‘Gems of Heaven’

Recent Research on Engraved

Gemstones in Late Antiquity,
c. AD 200–600

Edited by Chris Entwistle and Noël Adams

British Museum Research Publication
Number 177

Publishers
The British Museum
Great Russell Street
London WC1B 3DG

Series Editor Josephine Turquet

Assistant Production Editor: Anna Cunnane

Distributors
The British Museum Press
38 Russell Square
London WC1B 3QQ

‘Gems of Heaven’
Recent Research on Engraved Gemstones in Late Antiquity,
c. AD 200–600
Edited by Chris Entwistle and Noël Adams

Front Cover: Sardonyx cameo of Julia Domna as the Dea Caelestis.

British Museum, GR 1956,0517.1. See, Marsden, Pl. 2, p. 164

ISBN 978-086159-177-0
ISSN 1747-3640

© The Trustees of the British Museum 2011

Second printing 2012

Printed and bound in the UK by 4edge Ltd, Hockley

Papers used in this book are recyclable products made from wood
grown in well-managed forests and other controlled sources. The
manufacturing processes conform to the environmental regulations of
the country of origin.
Contents

Foreword v
List of Contributors vi
Illustration Acknowledgements viii

Non-destructive Gemmological Tests for the Identification of Ancient Gems 1

Çiğdem Lüle

A Case Study on Gemstone Origins: Chrysothrix, a Group of Roman Magical Gems 4

Lisbet Thoresen

The Garnet Millennium: the Role of Seal Stones in Garnet Studies 10

Noël Adams

Lithóis Indikois: Preliminary Characterisation of Garnet Seal Stones from 25

Central and South Asia
Noël Adams, Çiğdem Lüle and Emma Passmore

Archaeologies of Magical Gems 39

Richard Gordon

Text, Image and Medium: the Evolution of Graeco-Roman Magical Gemstones 50

Chris Faraone

The Colours of Magical Gems 62

Attilio Mastrocinque

Magic and Medicine: Gems and the Power of Seals 69

Véronique Dasen

Magical Gems and Classical Archaeology 75

Árpád M. Nagy

Studies on Magical Amulets in the British Museum 82

Simone Michel-von Dungern

‘Grylloi’ 88
Ken Lapatin

Engraved Gems from Sites with a Military Presence in Roman Palestine: 99

the Cases of Legio and Aelia Capitolina
Orit Peleg-Barkat and Yotam Tepper

Selected Antique Gems from Israel: Excavated Glyptics from Roman-Byzantine Tombs 105
Shua Amorai-Stark and Malka Hershkovitz

Christian Gems from Portugal in Context 114

Graça Cravinho and Shua Amorai-Stark

Intaglios and Cameos from Gaul in the 3rd and 4th Centuries AD 127
Hélène Guiraud
Late Roman Gems from Tilurium in Croatia 130
Bruna Nardelli

Three Degrees of Separation: Detail Reworking, Type Updating and Identity. 135
Transformation in Roman Imperial Glyptic Portraits in the Round
Elisabetta Gagetti

Gem Portraits of Soldier-Emperors 149

Erika Zwierlein-Diehl

Gods or Mortals – Images on Imperial Portrait Gems, Medallions and 163

Coins in the 3rd Century AD
Adrian Marsden

Love and Passion: Personal Cameos in Late Antiquity from the Content Collection 179
Helen Molesworth and Martin Henig

The Belgrade Cameo 186

Antje Krug

Late Antique and Early Christian Gems: Some Unpublished Examples 193
Jeffrey Spier

The Argument from Silence: Iconographic Statements of 1981 208

on Faked Gems Reconsidered
Josef Engemann

The Constanza Carnelian and the Development of 214

Crucifixion Iconography in Late Antiquity
Felicity Harley-McGowan

Seals in Transition: their Change of Function and Value in Late Antiquity 221
Gertrud Platz-Horster

Myth Revisited: the Re-use of Mythological Cameos and Intaglios in 229

Late Antiquity and the Early Middle Ages
Gemma Sena Chiesa

Inscriptions on Portrait Gems and Discs in Late Antiquity (4th to 6th centuries AD): 239
Between Epigraphical Tradition and Numismatic Particularism
Sébastien Aubry

Roman Intaglios Oddly Set: the Transformative Power of the Metalwork Mount 248
Genevra Kornbluth

The Re-use and Re-interpretation of Gemstones in Medieval Hungary 257

Tamás Gesztelyi

Reflections on Gems Depicting the Contest of Athena and Poseidon 263

Hadrien Rambach
Foreword
The papers in this volume derive from a conference held at the
British Museum on May 28 through to May 31 2009. Organised
under the auspices of the British Museum Byzantine Seminar,
this conference was the sixth in an ongoing series of annual
conferences held at the British Museum since 2003 devoted to
the material culture of Late Antiquity and Byzantium. All but
four of the 34 papers given at the conference in 2009 are
published here.
For their assistance with photographic enquiries I would
like to thank respectively Dr Irma Wehgartner of the Martin-
von-Wagner-Museum, University of Würzburg, and Drs Matteo
and Maria Campagnolo of the Musée d’art et d’histoire,
Geneva. My gratitude is also due to Saul Peckham, our
departmental photographer, Claudio Mari, Digital Image
Coordinator in Collections Services (Photography and
Imaging), in particular for his work on the magical gems, and
finally to Steve Crummy our departmental illustrator. Anna
Cunnane a voluntary intern at British Museum Press
generously gave a considerable amount of her time to the
initial lay-out and design. As ever, the volume would not have
been possible without both the energy and forebearance of
Josephine Turquet, Series Editor of British Museum Research
Publications.
Finally, given the inordinate expense in hosting
international conferences in this day and age, I would like to
thank the following who gave so generously towards the costs
of putting on the conference in 2009:
Dr and Mrs Jonathan P. Rosen (The Joseph Rosen
Foundation)
Phoenix Ancient Art S.A., New York
Derek J. Content Esq.
John W. Rassweiler Esq.
Chris Entwistle
Department of Prehistory and Europe
The British Museum
‘Gems of Heaven’ | v
 Dr Richard Gordon
List of Contributors Universität Erfurt
Postfach 90 0221
Dr Noël Adams 99105 Erfurt
c/o Department of Prehistory and Europe Germany
British Museum gordon.erfurt@gmx.org
London WC1B 3DG
UK Professor Hélène Guiraud
dnagranat@msn.com BAL 59
2 boulevard d'Arcole
Professor Shua Amorai-Stark 31000 Toulouse
Kaye College of Education France
Beer-Sheva guiraud.helene@numericable.fr
Israel
shua@macam.ac.il Dr Felicity Harley-McGowan
School of Culture and Communication
Dr Sébastien Aubry The University of Melbourne
Rue Jardinière 91 Victoria 3010
CH-2300 La Chaux-de-Fonds (NE) Australia
Switzerland fharley@unimelb.edu.au
sebastien.aubry@unine.ch
Dr Martin Henig
Dr Graça Cravinho Institute of Archaeology
Instituto de História da Arte 36 Beaumont Street
Universidade Nova de Lisboa Oxford, OX1 2PG
Av. de Berna, 26-C UK
P 1069-061 Lisboa martin.henig@arch.ox.ac.uk
Portugal
graca.silvaster@gmail.com Dr Malka Hershkovitz
Institute of Archaeology
Professor Dr Véronique Dasen Hebrew University
Institut des Sciences de l’Antiquité et du monde byzantin Mount Scopos, 91905
Séminaire d'archéologie classique Jerusalem
Université de Fribourg Israel
16 rue Pierre-Aeby mhershkovitz@huc.edu
1700 Fribourg
Switzerland Dr Genevra Kornbluth
veronique.dasen@unifr.ch 10508 Forestgate Place
Glenn Dale
Professor Dr Josef Engemann MD 20769
Moosstraße 145a USA
5020 Salzburg contact@KornbluthPhoto.com
Austria
ngmann-weil@aon.at Dr Antje Krug
c/o Deutsches Archäologisches Institut
Chris Entwistle Podbielskiallee 69-71
Department of Prehistory and Europe D - 14195 Berlin
British Museum Germany
London WC1B 3DG krugantje@arcor.de
UK
centwistle@thebritishmuseum.ac.uk Dr Ken Lapatin
Department of Antiquities
Professor Chris Faraone The J. Paul Getty Museum
Department of Classics 1200 Getty Center Drive
University of Chicago Suite 1000V
1115 E. 58th St Los Angeles, CA 90049-1745
Chicago, IL 60637 USA
USA KLapatin@getty.edu
cf12@midway.uchicago.edu
Dr Çiğdem Lüle
Dr Elisabetta Gagetti 1679 Henley Court
Masaryk University Wheeling
Faculty of Arts Illinois, 60090
Department of Archaeology and Museology USA
Arna Novaka, 1 cigdemlule@yahoo.com
60200 Brno
Czech Republic Dr Adrian Marsden
elisabetta.gagetti@gmail.com Norfolk Landscape Archaeology
Shirehall
Dr Tamás Gesztelyi Market Avenue
Institue of Classical Philology Norwich
University of Debrecen NR1 3JQ 
H-4032 Debrecen adrian.marsden@norfolk.gov.uk
Hungary
gesztelyi.tamas@arts.unideb.hu

vi | ‘Gems of Heaven’
 List of Contributors

Professor Attilio Mastrocinque Dr Gertrud Platz-Horster

Dipartimento di Arte, Archeologia, Storia e Società c/o Antikensammlung
Universitá di Verona Staatliche Museen zu Berlin
Via dell’Artigliere 8 Bodestrasse 1-3
37129 Verona D – 19178 Berlin
Italy Germany
attilio.mastrocinque@univr.it gertudplatz@gmx.de

Dr Simone Michel-von Dungern Hadrien Rambach

Museum Malerwinkelhaus 34 Campden Hill Towers
Stadt Marktbreit 112 Notting Hill Gate
Marktstraße 4 London W11 3QW
97340 Marktbreit UK
Germany coinadvisor@yahoo.co.uk
drsimonemichel@web.de
Professor Gemma Sena Chiesa
Helen Molesworth Dipartimento Di Scienze Dell’Antichità’
Avenue du Mail 25 University of Milan
Geneva 1205 via Festa del Perdono
Switzerland 20122 Milan
helen.molesworth@gmail.com Italy
gemma.chiesa@unimi.it
Dr Árpád M. Nagy
Classical Collection Dr Jeffrey Spier
Museum of Fine Arts 13316 E. Placita El Algodon
Dózsa György út 41 Tucson
1146 Budapest Arizona 85749 
Hungary USA
amnagy@szepmuveszeti.hu jbspier@gmail.com

Dr Bruna Nardelli Dr Yotam Tepper

Santa Croce 2333 Israel Antiquities Authority
30135 Venezia POB 35
Italy Nahalal 10600
brunanar@libero.it Israel
yotam@israntique.org.il
Dr Emma Passmore
Department of Conservation and Scientific Research Lisbet Thoresen
British Museum PO Box 1587
London WC1B 3DG Beverly Hills, CA 90213
epassmore@thebritishmuseum.ac.uk USA
lisbet@lthoresen.com
Dr Orit Peleg-Barkat
The Institute of Archaeology Professor Dr Erika Zwierlein-Diehl
Hebrew University of Jerusalem Institut für Kunstgeschichte und Archäologie
Mount Scopus, 91005 Universität Bonn
Israel Regina-Pacis-Weg 3
orit.peleg@mail.huji.ac.il 53113 Bonn
Germany
zwierlein-diehl@uni-bonn.de

‘Gems of Heaven’ | vii

Illustration Acknowledgements
N. Adams: The Garnet Millennium: the Role of Seal Stones in Garnet
Studies
Pls 1,2,3,4,5,6,7, 9,10 – author; Pl. 8 – Kubaba Photography, New York.
N. Adams, Ç. Lüle and E. Passmore: Lithóis Indikois: Preliminary
Characterisation of Garnet Seal Stones from Central and South Asia
Pl. 1 – Emma Passmore; Pls Group I, 1–8, Group II, 1–11 – Noël Adams
S. Amorai-Stark and M. Hershkovitz: Selected Antique Gems from
Israel: Excavated Glyptics from Roman-Byzantine Tombs
Pls 1–10, 12–36 – courtesy of the Israel Antiquities Authority and the
Israel Exploration Journal; Pl. 11 – London, British Museum.
S. Aubry: Inscriptions on Portrait Gems in Late Antiquity (4th to 6th
centuries AD): Between Epigraphical Tradition and Numismatic
Particularism
Pl. 1 – author; Pls 2,3,6,7,8,9,10,13,15,16,18,19,20,21,22,23,26,28 – after
J. Spier, Late Antique and Early Christian Gems, Weisbaden, 2007, nos
20,18,76,60,52,23,43,19,42,25,45, 74,39,1,44,17: my thanks to Jeffrey
Spier; Pl. 4 – © Numismatica Genevensis, Auction 2 (18 November 2002),
no. 142: thanks to L. Baglione, www.ngsa.ch; Pl. 5 – after J. Spier, ‘Some
unconventional Early Byzantine Rings’, in C. Entwistle and N. Adams,
‘Intelligible Beauty’: Recent Research on Byzantine Jewellery, London,
2010, pls 3a,3c; Pl. 11 – © Soprintendenza Speciale per i Beni
Archeologici di Napoli e Pompei, no. inv. 26054/216: thanks to P.G.
Guzzo, ssba-na@beniculturali.it; Pl. 12 – after M. Henig, The Content
Cameos, Oxford, 1990, no. 45; Pls 14,29,32 – after M.-L. Vollenweider
and M. Avisseau-Broustet, Intailles et camées II. Les portraits romains
du Cabinet des médailles, Paris, 2003, nos 228, 162, 135; Pl. 17 – after
M.-L. Vollenweider, Die Porträtgemmen der römischen Republik, Mainz
am Rhein, 1972–74, pl. 106/8; Pl. 24 – after A.M. McCann, The portraits
of Septimius Severus (MAAR XXX), Roma, 1968, 183 (j), pl. XCII; Pls
25,30 – E. Zwielein-Diehl, Glaspasten in Martin-von-Wagner Museum
der Universität Würzburg, Munich, 1986, nos 801,819; Pls 27,31 – after E.
Zwielein-Diehl, Antike Gemmen des Kunsthistorischen Museums in
Wien, vol. 3, Vienna, 1991, nos 1741, 1729; Pl. 33 – after E. Spagnoli and
M.C. Molinari, ‘Le monete’, in  A. Salvioni (ed.), Il tesoro di Via
Alessandrina, Rome, 1990, 94, no. 17; Pl. 34 – after L. Pirzio Biroli
Stefanelli, ‘I gioielli’, in A. Salvioni (ed.), Il tesoro di Via Alessandrina,
Rome, 1990, 45, no. 5.
G. Cravinho and S. Amorai-Stark: Christian Gems from Portugal in
Context
Pl.1 – © Luis Fraga da Silva; Pl. 2 – © Pedro Cravinho; Pls 3,4 – after
H.H. Hofstätter and H. Pixa, História universal comparada, III, Lisbon,
1985, 113; Pl. 5 – © Braga, Museu D. Diogo de Sousa; Pls 6,28 – © Lisbon,
Museu Nacional de Arquelogia; Pl. 7 – © Conimbriga, Museu de
Conimbriga and author; Pls 8,9,10,11,12,13,26,33,34,35,36,37,38,39 –
author; Pls 14,15 – after C.A. Ferreira de Almeida, ‘Arte paleo-cristã da
época das Invasões’, História da Arte em Portugal, II, Lisbon 1986, 14
and 10; Pl. 16 – after V.H. Correia, Conímbriga – Guia das Ruínas,
Lisbon, 2003, 12; Pl. 17 – after A.M. Alarcão, F. Mayet and J. Nolen,
‘Ruínas de Coimbra’, Roteiros da Arqueologia Portuguesa 2 (1989), 83;
Pl. 18 – after E. França, ‘Anéis, braceletes e brincos de Conimbriga’,
Conimbriga VIII (1969), 61; Pls 19,20,21,22,24,25 – © Conimbriga,
Museu Monográfico de Conimbriga; Pl. 23 – © Conimbriga, Museu de
Conimbriga; Pls 27,30,31 – after M. Fabião, M. Dias and M. Cunha, SIT
TIBI TERRA LEVIS – Rituais Funerários Romanos e Paleocristãos em
Portugal (Catálogo de Exposição), Museu Nacional de Arqueologia,
Lisboa, Lisbon, 2008, 50–1; Pl. 29 – © Museu de Arqueologia e
Numismática de Vila Real and author; Pl. 32 – © Arquivo Centro de
Arqueologia de Almadal; Pl. 33 – after F. Almeida, ‘Antiguidades da
Egitânia – alguns achados dignos de nota’, Arqueologia e História, 8ª
série, 11 (1965), pl. III, no. 1.
V. Dasen: Magic and Medicine: Gems and the Power of Seals
Pl. 1 – after L’oeil dans l’antiquité romaine, Lons-le-Saunier, 1994, fig. on
27; Pls 2,6,7,8,9 – London, British Museum; Pl. 3 – Paris, Cabinet des
médailles: photo A. Mastrocinque; Pl. 4 – after S. Michel, Bunte Steine –
Dunkle Bildern: ‘Magische Gemmen’, Munich, 2001, pl. 24; Pl. 5 – I.
Welner, ‘Aeskulapius és Hygieiát ábrázoló gemma Lenyomatával
díszített edény Aquincumból (un vase orné de l’empreinte d’une
gemme représentant Esculape et Hygie trouvé à Aquincum)’,
Archaeologiai Értesító 92 (1965), 42–4, fig. 1.
viii | ‘Gems of Heaven’
J. Engemann: The Argument from Silence: Iconographic Statements of
1981 on Faked Gems Reconsidered
Pls 1,5,6,7,8,11 – London, British Museum; Pl. 2 – after A. Mastrocinque,
‘Orpheos Bakchikos’, Zeitschrift für Papyrologie und Epigraphik 97
(1993), fig. 1; Pl. 3 – Rome, German Archaeological Institute; Pl. 4 –
after R. Garrucci, Storia dell’arte Cristiana, Prato, 1872, VI, pl. 479, n.
15; Pl. 9 – after J. Spier, Late Antique and Early Christian Gems,
Weisbaden, 2007, no. 432; Pl. 10 – after F.J. Dölger, ICHTHYS: Das Fisch-
Symbol in frühchristlicher Zeit, Rome, 1910, 334, fig. 50; Pl. 12 – after J.
Dresken-Weiland, Repertorium der christlich-antiken Sarkophage, Vol.
2: Italien mit einem Nachtrag Rom und Ostia, Dalmatien, Museen der
Welt, Mainz, 1998, 83–4, no. 242, pl. 80,2; Pls 13,14,19 – after C.
Christern-Briesenick, Repertorium der christlich-antiken Sarkophage,
Vol. 3: Frankreich, Algerien, Tunesien, Mainz, 2003, 23–5, no. 38; Pls
15,16,17,18 – after H. Brandenburg, Repertorium der christlich-antiken
Sarkophage, Vol. 1: Rom und Ostia, Wiesbaden, 1967, 271–2, no. 674 and
no. 43; Pl. 20 – after P. Angiolini Martinelli and P. Robino (eds), La
basilica di San Vitale a Ravenna 2, Modena, 1997, 210, fig. 411; Pl. 21 –
after M. Marcenaro, Il battistero di Albenga, Recco, 1994, fig. on 26; Pl.
22 – after C. Cecchelli, J. Furlani and M. Salmi, The Rabbula Gospels,
1959, Olten-Lausanne, pl. 9.4.
C. Faraone: Text, Image and Medium: the Evolution of Graeco-Roman
Magical Gemstones
Pls 1–7,10–12,14,15,17–19 – London, British Museum; Pl. 8 – after C.
Lenormant, Revue Archéologique (1846), 510; Pl. 9 – after E. Kunze,
Archaische Schildbänder, Berlin, 1950, IIIc, pl. 14; IVg, pl. 19; XLIIg, pl.
66; Pl. 13 – Washington DC, Dumbarton Oaks Collection; Pl. 16 – after
O. Tempkin, Soranus’ Gynecology, Baltimore, 1955, 9, fig. 1.
E. Gagetti: Three Degrees of Separation: Detail Reworking, Type
Updating and Identity. Transformation in Roman Imperial Glyptic
Portraits in the Round
Tables 1,2,3 – author; Pl. 1 – after H. Jucker, ‘Trajanstudien zu einem
Chalzedonbüstchen im Antikenmuseum’, Jahrbuch der Berliner
Museen 26 (1984), pl. 23i; Pl. 2 – after E. Gagetti, Preziose sculture di età
ellenistica e romana (Il Filarete. Collana di studi e testi. Università degli
Studi di Milano. Pubblicazioni della Facoltà di Lettere e Filosofia, 240),
Milan, 2006, pl. XI, A23; Pls 3,18 – after D. Boschung, ‘Die Bildnistypen
der iulisch-claudischen Kaiser-familie: ein kritischer Forschungs-
bericht’, Journal of Roman Archaeology 6 (1993), figs 30–5, 56–8; Pls 4,5
– after K. Fittschen and P. Zanker, Katalog der römischen Porträts in den
Capitolinischen Museen und den anderen kommunalen Sammlungen der
Stadt Rom, I, Kaiser- und Prinzenbildnisse (Beiträge zur Erschliessung),
Mainz am Rhein, 19942, pl. 12a; Pl. 6 – after G. Traversari, ‘Nuovo
ritratto di Cleopatra VII Philopator e rivisitazione critica
dell’iconografia dell’ultima regina d’Egitto’, Rivista di Archeologia 21
(1998), pls 18–20; Pl. 7 – after S. Walker and P. Higgs (eds), Cleopatra
regina d’Egitto (exh. cat., Rome), Milan, no. III.53a; Pl. 8 – after G.
Arbore Popescu (ed.), Traiano. Ai confini dell’impero (exh. cat.,
Ancona), Milan, 1998, no. 325 (front); A. Giuliano, I Cammei della
Collezione Medicea del Museo Archeologico di Firenze, Rome–Milan,
1989, pl. 181 (profile to right); A. Carandini, Vibia Sabina. Funzione
politica, iconografia e il problema del classicismo adrianeo (Accademia
Toscana di Scienze e Lettere ‘La Colombaria’. Studi, 13), Firenze, 1969,
pls 215, 217 (profile to left and reverse); Pls 9,10,11 – after A.M. Reggiani
(ed.), Adriano. Le Memorie al femminile (exh. cat., Tivoli), Milan, 2004,
105–6; 78; 139; Pls 12,13,14 – after F. Baratte, ‘Un portrait féminin des
collections du Louvre’, Revue Archéologique 1984 (2), pls. 1–4, 5–7, 9–11;
Pl. 15 – after Dr. Busso Peus Nachfolger, auction catalogue, 400 (April
22nd 2010), lot 653; Pl. 16 – after H. von Heintze, ‘Ein spätantikes
Mädchenporträt in Bonn. Zur stilistischen Entwicklung des Frauen-
bildnisses im 4. und 5. Jahrhundert’, Jahrbuch für Antike und
Christentum 14 (1971), pls 6b,d; Pl. 17 – London, British Museum; Pl. 19 –
after D. Salzmann, ‘Beobachtungen zu Münzprägung und
Ikonographie des Claudius’, Archäologischer Anzeiger (1976), pls 9–10;
Pl. 20 – after J. Meischner, ‘Studien zur spätantike Kaiserikonographie’,
JdI (1995), pl. 11; Pl. 21 – after E. Coche de la Ferté, Le camée Rothschild.
Un chef d’oeuvre du IVe siècle après J.-C., Paris, 1957, pls 2, 6; Pl. 23 – after
M.-L. Vollenweider and M. Avisseau-Broustet, Camées et intailles. Tome
II. Les portraits romains du Cabinet des médailles, Paris, 2003, pls 134–5;
Pl. 24 – after Le trésor de la Sainte-Chapelle (exh. cat., Paris), Paris,
2001, no. 56, pl. 2.

T. Gesztelyi: The Re-use and Re-interpretation of Gemstones in
Medieval Hungary
Pl. 1 – after A. Tocik, Altmagyarische Gräberfelder in der
Südwestslowakei, Bratislava, 1968, Taf. LV/17; Pl. 2 – after M. Hlatky, A
magyar gyűrű (The Hungarian Ring), Budapest, 1938, 48; Pls
3,4,5,8,13,14 – National Archives of Hungary (Magyar Országos
Levéltár), Budapest; Pl. 6 – http://www.corvina.oszk.hu/images/
CoatofArms/cimerhunyadi02.jpg& imgrefurl; Pl. 7 – Archives of
County Hajdú-Bihar, Debrecen; Pl. 9 – after E. Brandt and E. Schmidt
(eds), Antike Gemmen in deutschen Sammlungen I 2, Staatliche
Münzsammlung München, Munich, 1970, no. 1467; Pl. 10 – after J.
Jerney, Magyar Történelmi Tár 2, Budapest, 1855, 155, fig. 17; Pl. 11 –
Debrecen, Déri Múzeum; Pl. 12 – after E. Zwierlein-Diehl, Antike
Gemmen und ihr Nachleben, Berlin–New York, 2007, pl. 899; Pl. 15 –
Budapest, St Stephen Basilica, http://bin.sulinet.hu/ikep/2004/05/
sztjobb.jpg&imgrefurl; Pl. 16 – Komárom, Klapka György Múzeum.
R. Gordon: Archaeologies of Magical Gems
Pls 1–11,13–14 – London, British Museum; Pl. 12 – after Preisendanz
PGM II 166 (Gordon, n. 59).
H. Guiraud: Intaglios and Cameos from Gaul in the 3rd and 4th
Centuries AD
Pl. 1 – Y. Deslandes; Pl. 2 – Badisches Landesmuseum Karlsruhe; Pl. 3 –
S. Prost, service archéologique municipal; Pls 4–7 – author.
F. Harley-McGowan: The Constanza Carnelian and the Development of
Crucifixion Iconography in Late Antiquity
Pls 1,6,8 – London, British Museum; Pl. 2 – Rome, German
Archaeological Institute; Pl. 3 – author; Pl. 4 – after R. Garrucci, Storia
della arte Cristiana nei primi otto secoli della chiesa, 6 vols., Prato, 1880,
vol. 6, tav. 483; Pl. 5 – photo by Robin Jensen; Pl. 7 – Rome, Museo
Palatino, Inv. 381403; Pl. 9 – after J. Spier, Picturing the Bible; the
Earliest Christian Art, New York and London, 2008, 227, fig. 1.
G. Kornbluth: Roman Intaglios Oddly Set: the Transformative Power of
the Metalwork Mount
Pls 1,2,3,4,6,7,8,11,12,13,14,15,16,18,19,20,21 © Author; Pl. 5 – Abbey of
St-Maurice d’Agaune; Pls 9,10 –Foto Marburg; Pl. 17 – Oxford,
Ashmolean Museum; Pl. 22 – after C. Boulanger, Le Cimetière
FrancoMérovingien et Carolingien de Marchélepot (Somme): Étude sur
l’Origine de l’Art Barbare, Paris 1909; Pl. 23,24 – after W. Veeck, Die
Alamannen in Württemberg (Germanische Denkmäler der
Völkerwanderungszeit.1), Berlin 1931, pl. G8.
A. Krug: The Belgrade Cameo
Pls 1,5 – Belgrade, Narodni Muzej u Beograd; Pls 2,3,8 – plaster cast,
Akademisches Kunstmuseum, Bonn; photos © author; digital
processing: Hans Rupprecht Goette; Pl. 4 –drawing Gisela Höhn,
Bonn; digital processing: Hans Rupprecht Goette; Pls 6,7 – photos ©
author; digital processing: Hans Rupprecht Goette; Pls 9,11 – digital
processing: Hans Rupprecht Goette; Pl. 10 – London, British Museum.
K. Lapatin: ‘Grylloi’
Pls 1,12–16,22–35,37, Addendum – author after historical sources noted
in the captions; Pls 2,4,9,10,18–21,36 – London, British Museum; Pls
3,5–8 – Courtesy of the Michael C. Carlos Museum of Emory University:
photos by Bruce M. White, 2010; Pl. 11 – courtesy of the Derek J. Content
Collection; Pl. 17 – courtesy of the Beazley Archive, Oxford.
Ç. Lüle: Non-destructive Gemmological Tests for the Identification of
Ancient Gems
Pl. 1 – Elina Ratcheva; Pl. 2 – Stuart Robertson; Pl. 3 – Lisbet Thoresen;
Pl. 4 – author.
A. Marsden: Gods or Mortals – Images on Imperial Portrait Gems,
Medallions and Coins in the 3rd Century AD
Pls 1,2,3,12,17,21,27,28,32,35,39,45 – London, British Museum; Pl. 4 –
private collection; Pl. 5 – Colchester Castle Museum; Pl. 6 – private
collection; Pl. 7 – Berlin, Antikensammlung; Pl. 8 – after L. Endrizzi
and F. Marzatico (eds), Ori delle Alpi, Trento, 1997, no. 1183; Pl. 9 – after
O. Neverov, Antique Intaglios in the Hermitage Collection, Leningrad,
1976, 79, no. 141; Pls 10,20 – after J. Tassie and E. Raspe, A Descriptive
Catalogue of a General Collection of Ancient and Modern Engraved Gems,
Cameos and Intaglios, Taken from the Most Celebrated Cabinets in
Europe; and Cast in Coloured Pastes, White Enamel, and Sulphur,
London, 1791, nos 12067, 12081; Pls 11,26 – after G.M.A. Richter,
Illustration Acknowledgements
Engraved Gems of the Romans, London, 1971, no. 586, 589; Pl. 13 – after
M. Schlüter, G. Platz-Horster and P. Zazoff, Antiken Gemmen in
Deutschen Sammlungen, Band IV, Hannover Kestner-Museum, Hamburg
Museum für Kunst und Gewerbe, Weisbaden, 1975, no. 1599; Pl. 14 – after
E. Zwierlein-Diehl, Die Antiken Gemmen des Kunsthistorischen
Museums in Wien III, Munich, 1991, no. 1730; Pl. 15 – courtesy of Drs
Matteo and Maria Campagnolo, Musée d’art et d’histoire, Geneva; Pls
16,18,23 – after E. Zwierlein-Diehl, Glaspasten im Martin-von-Wagner
Museum der Universität Würzburg, Munich, 1986, nos 13, 794, 793; Pl.
19 – Naples, Museo Archeologico Nazionale; Pl. 22 – private collection;
Pls 24,31 – after M.-L. Vollenweider and M. Avisseau-Broustet, Camées
et intailles. Tome II : Les portraits romains du Cabinet des Médailles:
catalogue raisonné, Paris, 2003, no. 206; Pl. 25 – © Christie’s Images/
The Bridgeman Art Library; Pls 29,30 – after R. Delbrueck, Antike
Porphyrwerke, Berlin and Leipzig, 1932, pls 57b,58a; Pls 33,34,36,37 –
after J.M.C. Toynbee, Roman Medallions, New York, 1944, pls XV,5–6,
XLIV,1 and XLVI,5; 64; Pl. 38 – after A.B. Marsden, ‘Some sing of
Alexander and some of Hercules: artistic echoes of Hercules and
Alexander the Great on coins and medallions, ad 260–269’, in L.
Gilmour (ed.), Pagans and Christians – from Antiquity to the Middle
Ages, BAR International Series 1610, 2007, 66; Pl. 40 – Paris,
Bibliothèque nationale; Pls 41,42,43 – Oxford, Ashmolean Museum; Pl.
44 – after In Pursuit of the Absolute. Art of the Ancient World from the
George Ortiz Collection, Royal Academy of Arts, London, 20 January–6
April 1994, London, 1994, no. 238; Pl. 46 – Internet.
A. Mastrocinque: The Colours of Magical Gems
Pl. 1 – Rome, private collection; photo: author; Pls 2,10 – author; Pl. 3 –
courtesy Civic Museum of Verona; photo: author; Pls 4,5,8,12 – London,
British Museum; Pls 6,9,11 – courtesy of Cabinet des médailles, Paris;
photos: author; Pl. 7 – after: medicalimages.allrefer.com.
S. Michel-von Dungern: Studies on Magical Amulets in the British
Museum
Pl. 1 – Hamburg, Collection W. Skoluda; photo: author; Pls
2,3,4,5,6,9,14 – London, British Museum; Pl. 7 – Diagram ‘Dodekaoros’
after Teukros and the ‘Daressy Zodiac’: author; Pls 8,10 – Hamburg,
Collection E. Sossidi; photo: author; Pls 11,12 – Malibu, The J.P. Getty
Museum; Pl. 13 – Ann Arbor, Kelsey Museum; photo: author; Pl. 15 –
Kansas City, Linda Hall Library (For copyright: PDF).
H. Molesworth and M. Henig: Love and Passion: Personal Cameos in
Late Antiquity
Pls 1–27 – courtesy Derek J. Content Collection
Á.M. Nagy: Magical Gems and Classical Archaeology
Pl. 1 – St Petersburg, The State Hermitage Museum; photo: V.
Terebenin, L. Kheifets, Y. Molodkovets; Pls 2,3,5 – London, British
Museum; Pls 4,8, – St Petersburg, The State Hermitage Museum;
photos: A. Rázsó; Pl. 6 – after M. Maaskant-Kleibrink, Catalogue of the
Engraved Gems in the Royal Coin Cabinet, The Hague, The Hague, 1978,
355, no. 1119; Pl. 7 – courtesy of M. Avisseau-Broustet, Cabinet des
médailles, Bibliothèque Nationale, Paris; photo: A. Mastrocinque
(d.r.); Pl. 11:1 – courtesy of B. Shipman, Taubman Medical Library, Ann
Arbor; photo: B. Shipman; Pl. 11:2 – Budapest, Museum of Fine Arts;
photo: L. Mátyus; Pl. 11:3 – after S. Michel, Bunte Steine – Dunkle Bildern:
’Mágische Gemmen’, Munich, 2001, pl. 15,89; Pl. 12 – courtesy of M.
Torbágyi, Hungarian National Museum, Budapest; photo: A. Dabasi.
B. Nardelli: Late Roman Gems from Tilurium in Croatia
Pls 1–7,9,11–13 – © T. Sesel, Archaeological Museum of Split; Pls 8,10 –
© I. Prpa Stojanac, Archaeological Museum of Split.
O. Peleg-Barkat and Y. Tepper: Engraved Gems from Sites with a
Military Presence in Roman Palestine: the Cases of Legio and Aelia
Capitolina
Pls 1–13 – © Yotam Tepper
G. Platz-Horster: Seals in Transition: their Change of Function and
Value in Late Antiquity
Pl. 1 – Bonn, Rheinisches Landesmuseum, photo: S. Taubmann; Pl. 2 –
author; Pl. 3 – Stuttgart, Landesmuseum Württemberg: after L.
Wamser, Die Römer zwischen Alpen und Nordmeer, Mainz, 2000, cat.
no. 141; Pl. 4 – London, British Museum; Pl. 5 – The J. Paul Getty
Museum, Malibu, California, Inv. no. 83.AM.228.1-7; Pl. 6 – Krefeld,
Museum Burg Linn: after R. Pirling, Römer und Franken am
Niederrhein. Burg Linn, Krefeld, Mainz, 1986, pl. 133.
‘Gems of Heaven’ | ix
Illustration Acknowledgements
H. Rambach: Reflections on Gems Depicting the Contest of Athena and
Poseidon
Pls 1,2 – Soprintendenza per i Beni Archeologici delle Province di
Napoli e Caserta; Pl. 3 – after H. Brunn, Denkmäler griechischer und
römischer Skulptur); Pl. 4 – © Gemini auctions; Pls 5,6,26 – © Classical
Numismatic Group; Pl. 7 – © LHS Numismatik; Pls 8,9,38 – © Beazley
Archives; Pls 10,11,13,23,35 – London, British Museum; Pl. 12 – ©
Hanover, Kestner Museum; Pl. 14 – © Gisela Richter; Pls 17,31,33 – ©
Paris, Bibliothèque nationale; Pls 18,32,37 – © Vienna,
Kunsthistorisches Museum; Pl. 19 – © Erika Zwierlein-Diehl; Pls 20,21
– © Geldmuseum, Utrecht; Pl. 22 – © RGZM T 74/2452–2453 (Ernst
Künzl); Pl. 24 – © Heritage Auctions; Pl. 25 – © Bonn, Rheinische
Landesmuseum; Pls 26,27 – © Fritz Rudolf Künker GmbH & Co. KG; Pl.
28 – © Gorny & Mosch; Pl. 29 – © St Petersburg, Hermitage; Pl. 30– ©
Jeffrey Spier; Pl. 34 – after N. Dacos, Il tesoro di Lorenzo il Magnifico. Le
gemme, Florence, 1973, pl. 81; Pl. 36 – © Genevra Kornbluth/Hadrien
Rambach; Pl. 39 – © Diana Scarisbrick, courtesy of Claudia Wagner.
G. Sena Chiesa: Myth Revisited: The Re-use of Mythological Cameos
and Intaglios in Late Antiquity and the Early Middle Ages
Pl. 1 – after E. Zwierlein-Diehl, Magie der Steine. Die antike Prunk-
kameen im Kunsthistorischen Museum, Vienna, 2008, pl. 39; Pls
2,3,4,5,7,10,12,21, – after G. Sena Chiesa (ed.), Gemme. Dalla corte
imperiale alla corte celeste, Milan, 2002, folding page, pl. 1 on 19, pl. 3 on
23, pl. 3 on 31, pl. 49 on 201, pl. 2 on 44, pl. 36 on 210, pl. 37 on 210; Pl. 6 –
after Le trésor de Conques (exh. cat. Paris), Paris, 2001, pl. 15; Pl. 8 – after
M.-L. Vollenweider and M. Avisseau-Broustet, Camées et intailles, Tome
II, Les Portraits romains du Cabinet des Médailles. Catalogue raisonné,
Paris, 2003, pl. 132; Pl. 9 – after B. Nardelli, I cammei del Museo
Archeologico Nazionale di Venezia (Collezioni e Musei Archeologici del
Veneto, 43), Rome, 1999, no. 16; Pls 11,18 – London, British Museum; Pl.
13 – after http://museoarcheologiconazionale.campaniabeniculturali.
it/itinerari-tematici/galleria-di-immagini; Pl. 14 – after http://
commons.wikimedia.org/wiki/File:Affrescoromano-_eracle_ed_
onfale_-_area_vesuviana.JPG; Pls 15,16 – after M.-L. Vollenweider, Die
Steinschneidekunst und ihre Künstler in spätrepublikanischer und
augusteischer Zeit, Baden-Baden, 1966, pls 37.5, 68.7; Pl. 17 – after P.
Zanker and B.C. Ewald, Mit Mythen leben. Die Bildenwelt der römischen
Sarkophagen, Munich, 2004, pl. 125; Pl. 19 – after C. Rizzardi, L.
Martini, C. Muscolino and E. Cristoferi (eds), Avori bizantini e
medioevali nel Museo Nazionale di Ravenna, Ravenna, pl. 1; Pl. 20 – after
J. Spier, Late Antique and Early Christian Gems, Weisbaden, 2007, pl.
572; Pls 22,23 – after E. Zwierlein-Diehl, Die Gemmen und Kameen des
Dreikönigenschreines. Der Dreikönigenschrein im Kölner Dom, Band I.1
(Denkmäler Deutscher Kunst. Die großen Reliquienschreine des
Mittelalters, Studien zum Kölner Dom, 5), eds A. Wolff and R. Lauer,
Cologne, 1998, pls 8, 250; Pls 24,25,26,29,30,31 – after R. Nanni and
M.C. Monaco, Leda, storia di un mito dalle origini a Leonardo, Firenze,
2007, pls 10,14,16,22,53,59; Pl. 27 – after D. Willers and L. Raselli-
x | ‘Gems of Heaven’
Nydegger (eds), Im Glanz der Götter und Heroen, Meisterwerke Antiker
Glyptik aus der Stiftung Leo Merz, Mainz am Rhein, 2003, pl. 59; Pl. 28 –
after M. Maaskant-Kleibrink, Catalogue of the Engraved Gems in the
Royal Coin Cabinet, The Hague, The Hague, 1978, pl. 351; Pl. 32 – after
http://www.settemuse.it/pittori_scultori_ europei/ matisse/1944
J. Spier: Late Antique and Early Christian Gems: Some Unpublished
Examples
Pl. 1 – after Gorny & Mosch; Pls 2–8,11–14,16,17,19,21–26,28,31,32,34–
37,41,50 – courtesy C.S. Collection, Munich; Pl. 9 – after Harlan J. Birk
Ltd; Pls 10,18,20,27,30,39,42,49,51,52 – private collection; Pls 15,29,43–
46,53 – courtesy Derek. J. Content Collection; Pl. 33 – after
Münzenhandlung Gerhard Hirsch; Pl. 38 – courtesy of Penelope Rogers
and the Anglo-Saxon Laboratory; Pl. 40 – courtesy of Erika Zweirlein-
Diehl; Pl. 47 – © Bengt Lundberg; Pl. 48 – after F. Althaus and M.
Sutcliffe (eds), The Road to Byzantium, London, 2006, 165, no. 102; Pls
54–56 – German eBay.
L. Thoresen: A Case Study on Gemstone Origins: Chrysothrix, a Group
of Roman Magical Gems
Pls 1,2,4,8 – photomicrographs: John Koivula; Pl. 3 – author; Pl. 5 –
Herbert Wiegandt; Pls 6,7 – Harold and Erica Van Pelt.
E. Zwierlein-Diehl: Gem Portraits of Soldier-Emperors
Pls 1,2,42 – Vienna, Kunsthistorisches Museum; Pls 3,6,25 – Martin-
von-Wagner-Museum der Universität Würzburg, photos: Isolde
Luckert; Pls 4,5,33,43 – Munich, Staatliche Münzsammlung; Pls 7,12 –
St Petersburg, Hermitage (after O. Neverov, ‘Concordia Augustorum’,
Wissenschaftliche Zeitschrift der Universität Rostock 19 (1970), 605–12,
pl. 28, 6 and O. Neverov, Antique Intaglios in the Hermitage Collection,
Leningrad, 1976, no. 141); Pls 8,9 – Berlin, Staatliche Museen,
Antikensammlung, photos: Isolde Luckert (Pl. 8), Johannes Laurentius
(Pl. 9, Lippert2 II 1767, 847); Pls 10,11 – Międzyrzecz (Poland), Museum
(after M. Ruxerówa, ‘Gemma Międrzyzecka’, Fontes archaeologici
Posnanienses 8–9 (1957–8), 443–7, figs 2–3); Pl. 13 – London, Victoria &
Albert Museum, Tassie & Raspe No. 12081, photo: Claudia Wagner; Pls
14,15,31,32 – Cologne, Römisch-Germanisches Museum, photos:
author (Pl. 14), Jutta Schubert (Pl. 15), Museum (Pls 31,32); Pls 16,19,20
– Collection of Derek J. Content; Pls 17,24,37,38,41 – Bonn,
Akademisches Kunstmuseum, Cades Auswahl Bonn cl. 13 III 31, 407,
414, 423, 424, 432, photos: Jutta Schubert; Pls 18, 27,40 – Paris, Cabinet
des médailles (Pl. 40 after M.-L. Vollenweider and M. Avisseau-
Broustet, Camées et intailles. II Les Portraits romains du Cabinet des
médailles, Paris, 2003, no. 258, pl. 133); Pls 21,22,28,29,30,34,35,36,37 –
London, The British Museum, photos: Genevra Kornbluth (Pls
21,28,30,34,36), Jutta Schubert (Pls 22,29,35); Pl. 23 – Baltimore,
Walters Art Museum; Pl. 26 – Naples, Museo Archeologico Nazionale
(after U. Pannuti, La Collezione Glittica II, Rome, 1994, no. 214); Pl. 39 –
Copenhagen, Thorvaldsens Museum.
Non-destructive Gemmological Tests for the
Identification of Ancient Gems
Çiğdem Lüle
This paper aims to provide a brief discussion of the new
research field known as archaeogemmology; a relatively
unknown discipline that applies a multidisciplinary approach
to investigate the identity and origin of gem artefacts of the
ancient world.
Archaeogemmology defined
Archaeogemmology combines aspects of gemmology,
archaeology and geology in order to identify ancient gems and
determine their mineralogical and geographical origin. The
relevance of the discipline is to contribute gem-origin
information to aid the investigation of social patterns such as
migration and the trade routes of ancient cultures and
civilisations. Gemmology itself, as an offshoot of mineralogy, is
a relatively young discipline which emerged about 100 years
ago as a separate subject because the modern jewellery trade
needed a more scientific and practical methodology, as well as
a greater precision in terminology for the identification of
gems. It is generally accepted that this first occurred in the
European market to address emergent concerns posed by the
production of synthetic rubies.1
Gemmological testing is the first step in the archaeo-
gemmological investigation of the material of an ancient gem.
The initial gem testing provides a non-destructive and quick
identification process which can be performed anywhere from
an excavation site to a museum. Gem testing instruments are
easy to use, relatively inexpensive and portable. Basic
gemmological instruments are: a 10x loupe, standard
gemmological microscope, refractometer, polariscope, hand-
held spectroscope, dichroscope and light source (Pl. 1). The
majority of gem materials can be identified with these
instruments without the use of any destructive method and
regardless of whether the gem is mounted or unmounted.
Plate 1 Basic
gemmological
instruments:
gemmological
refractometer, light
source, polariscope,
dichroscope, hand-
held spectroscope,
Chelsea Colour
Filter, 10X loupe
and tweezers.
Gemmology is also studied in conjunction with other
scientific subjects such as mineralogy, physics and chemistry.
Although any skilled gemmologist should be able to identify
the majority of gems, detecting treatments and separating
natural stones from their synthetic counterparts might require
further testing in an advanced gem-testing laboratory. These
advanced tests are mainly non-destructive, but minimally
invasive tests may also be occasionally necessary.
Archaeologists frequently have the challenge of identifying
gem materials and interpreting the identifying terms used by
other related disciplines. Archaeogemmology offers several
vital tools for archaeologists, beginning with a lexicon of
recognised scientific terminology for the classification of gem
materials. Furthermore, the application of advanced
mineralogical and geochemical identification methods will
provide information on the geological and geographical origin
of gem materials.
The role of the archaeogemmologist is enhanced by an
understanding of the history of gem research. Although the
surviving body of literature is limited, it is sufficient to indicate
that precious metals and stones have been the focus of study for
thousands of years. The earliest text known to date is
Theophrastus’ treatise Peri Lithon (‘On Stones’) written in the
4th century bc,2 followed in the 1st century ad by Pliny the
Elder’s Naturalis Historia, XXXVII.3 Another significant text for
stones and their treatments is the Papyrus Graecus Holmiensis
(known as the ‘Stockholm Papyrus’) copied in the 4th century
ad in Greek-speaking Egypt.4
Problems of nomenclature
There is a discrepancy in the nomenclature used for gem
materials between archaeologists, gemmologists and
mineralogists. It is interesting to observe that the old terms
‘precious’ and ‘semi-precious’ for gem materials are still used in
archaeology. These concepts are considered to be invalid in
modern gemmological terminology. The vast majority of gems
are natural minerals and rare. In the past this rarity led to the
use of the expression ‘precious stones’. According to culture
and country, this term typically refers to diamond, ruby,
sapphire, and emerald. For a given gem, only a few fine
examples are highly priced, and the rest do not really deserve
the term ‘precious’ in lower grades. Gemmologists do not
recommend using the term ‘semi-precious stones’, which in
many cases is meaningless. A typical example of this is
amethyst. This transparent purple variety of quartz has been
considered as ‘semi-precious’ for the last two centuries due to
its plentiful supply from sources in Brazil. However, before the
discovery of Brazilian amethyst mines, it was a rare gem and
generally not seen in large quantities in archaeological
contexts.5 Yet it is not unusual to still hear scholars referring to
amethyst as ‘semi-precious’. In addition, not all ‘stones’ are
‘‘Gems of Heaven’ | 1
Lüle
Plate 2 A polished slice of a
quartz geode (diam. c. 25cm).
The outer rim of the geode
with brown and black bands is
constructed of microscopic
quartz grains and represents
micro-crystalline quartz. The
grain size and the translucent
structure of this layer indicate
that this is an example of
chalcedony. Quartz crystals
are visible to the naked eye
towards the centre.
minerals as there are very important gem materials of organic
origin such as pearls, amber and mother of pearl.6
Quartz group gems illustrate another nomenclature issue.
Due to their availability, durability, variety of transparency and
colour, quartz species are more suitable for engraving than
harder and rarer gems such as corundum. Archaeologists are
more familiar with quartz and chalcedony varieties as these
stones were commonly used over many millennia. The
difficulty in identifying the different species of the Quartz
Group has led to significant problems in the consistent use of
nomenclature between scholars in the archaeological world.
For example, it is usual for archaeologists to use the vague term
‘prase’ instead of translucent green chalcedony or green jasper.
Mineralogically, however, these are two different species of
Quartz Group gems. The same applies to the differentiation
between onyx, sardonyx and agate. This issue has been
addressed in the past by researchers from both archaeological
and gemmological perspectives, yet it has not been
disseminated sufficiently for widespread application.7
From a gemmologist’s point of view, the issue is simply
resolved by recognising the difference between
microcrystalline (quartz grains which are only visible under
magnification i.e. chalcedony and jasper) (Pl. 2) and
macrocrystalline (quartz grains which are visible to the naked
eye, i.e. aventurine quartz and quartzite). The next step is
separating the samples by transparency and colour. Jasper, for
instance, is opaque and green chalcedony is translucent. By
simple transparency and grain size inspection with a 10x loupe,
problems over terms like ‘prase’ can be resolved.
Identification of sources
It is understandable that ancient writers had a limited
understanding of different gem species and their sources.
Although the scale of mining and transportation of gems was
limited in comparison to the modern world, some early sources
are still known to us, such as the city of Alabanda in western
Turkey mentioned by both Theophrastus and Pliny as a source
of red garnet. The Carian city Alabanda was considered to be
the type locality of the almandine garnet named ‘Alabandicus’
by Pliny.8 In recent excavations the city walls of Alabanda were
found to have been built with local migmatitic rocks containing
red garnets, but large samples of garnet gems are not yet
known in this area.9 It is probable that the town was the trading
point for garnets brought from other sources.
There are also examples of archaeological gems whose
origin cannot be located today. Problems in relating the ancient
material to a known modern source can arise due to the
widespread availability of the stone or the fact that many
2 | ‘Gems of Heaven’
ancient sources are unknown or played out. Quartz group
species are the best examples of the problem: quartz is the
second most abundant mineral in the earth’s continental crust
and its species are found almost everywhere. However, the
majority of Quartz Group species lack significant trace
elements to aid in pinpointing their locality.
Chromian chalcedony
The material known as chromian chalcedony provides a good
example of the application of archaeogemmological
techniques.1o Furtwängler was the first to point out that a
particular translucent green chalcedony was different from the
common green chalcedony or ‘prase’.11 His observations were
based upon the extensive collection of gems in the
Antikensammlung in Berlin and in 1900 he named the material
‘kleine Praser’. Later this same material was identified as a rich
bluish-green translucent chalcedony with black inclusions and
published by, inter alia, Henig, Hutchinson, Guiraud, Hoey-
Middleton, Thoresen and Platz.12 The earliest Roman intaglios
carved from chromian chalcedony have been dated to the 1st
century bc (Pl. 3), but by about the 3rd century ad, it had
vanished from lapidary workshops.13 Although the material
was widely distributed throughout the Roman Empire, it
appears to have originated from a single source.
A handful of localities for the rare yet commercially
insignificant ‘chromian chalcedony’ have been reported in
recent times. The first discovery was in Mtoroshanga,
Zimbabwe, in 1956, published by C.C. Smith in 1967.14 Likewise,
J. Hrysl published the occurrence of ‘chromium bearing
chalcedony’ in Bolivia in 1999,15 and in 2003 M. Willing and
S.M. Stocklmayer reported the discovery of a chromian
chalcedony deposit in Western Australia.16 Although an African
source could have been known in the Roman period, none of
the modern sources of chromian chalcedony can be associated
with the material used in antiquity, either geographically or
mineralogically. Several researchers have suggested that the
material exploited in the Roman period may have originated
from the vicinity of the chromium mines in Turkey or Cyprus.17
To investigate this theory, the author has collected rough gem
material from Sivrihisar in Eskişehir province in northwest
Anatolia. The chromian chalcedony of this region has
demonstrated strongly similar characteristics through
gemmological examination.18
Plate 3 Chromian chalcedony intaglios of the Roman Empire, 1st–2nd century
AD. Berlin, Antikensammlung, Staatliche Museen zu Berlin, (clockwise from
top left) inv. nos 32, 237, 226, 249, 202
 Non-destructive Gemmological Tests for the Identification of Ancient Gems
Plate 4 Euhedral
(fully formed),
prismatic, black and
opaque chromite
crystal inclusions in
chromian
chalcedony from
Eskişehir, Turkey.
Magnification 10X
Besides the standard gem testing methods and detailed
microscopy analyses, whole rock ICP analyses on the Anatolian
material and ED-XRF analyses were performed on selected
ancient intaglios. The results showed high chromium and
nickel content ratios, which was significantly higher than
mtorolite (the Zimbabwean material) or the Bolivian material
sampled. The ancient samples came from different collections,
i.e. the Antikensammlung, Staatliche Museen, Berlin and the
Michael Shubin Collection in California.19 As indicated above,
they all can be stylistically dated to around the 1st century ad.20
Chromian chalcedony from Bolivia, Zimbabwe and Turkey
and Roman intaglios from various collections were compared
using microscopy and SEM analyses. The Anatolian material
and the Roman intaglios were significantly similar in general
observation, basic gemmological properties as well as
microscopy and chemical analyses. Under high magnification
both have the similar amount and distribution of euhedral
chromite inclusions in matrix (Pl. 4). In contrast, the fibrous
body texture, a significantly lower concentration of chromium
and nickel ratio, as well as relatively few chromite inclusions
clearly differentiates Zimbabwean mtorolite from either the
ancient Roman or Anatolian rough material.21
At this stage of research, archaeogemmology provides a
scientific approach to resolving the problem without
speculation. Both the modern and ancient examples of
chromian chalcedony are still being compared with detailed
and careful analyses in order to prove their origin and the final
results are to be published in the near future. The non-
destructive testing techniques used in gemmology pose
virtually no risk to valueable archaeological objects and serve
as initial tests. Additional advanced mineralogical analyses
provide further information regarding the source of the
material.  14 C.C. Smith, ‘A Preliminary account of Rhodesia’s new gemstone-
Conclusion
Archaeogemmology is a multidisciplinary study contributing a
new perspective to the interpretation of classical sciences.
Through it, archaeologists will be able to work with more
scientific and positive data in order to identify gem materials. It
should be considered as a subdivision of archaeometry – the
application of scientific techniques to the analysis of archaeo-
logical materials. As a new discipline, archaeogemmology aims
to identify the geological origin and geographical source of the
gem materials insofar as it is possible. However challenging
this might be, the results can provide critical information to
archaeologists. Through the application of practical
gemmological tests and reliable mineralogical analyses, the
archaeogemmologist can contribute invaluable information to
assist archaeologists to better understand otherwise elusive
activities such as the ancient gem trade.
Notes
1 The first successful production of synthetic rubies in commercial
quantities was achieved in 1891 by the French chemist Auguste
Verneuil. Their identification required scientific knowledge rather
than experience by eye. This event led to the formation of the
Gemmological Association of Great Britain in 1908 established by
the Education Committee of the Association of Goldsmiths in order
to educate qualified gemmologists. See P.G. Read, Gemmology,
Oxford, 2006 (3rd edn).
2 Theophrastus, Peri Lithon - On Stones (trans. E.R. Caley and J.F.C.
Richards), Colombus, Ohio, 1956.
3 Pliny, Naturalis Historia (trans. H. Rackham, W.H.S. Jones and
D.E. Eichholz), 10 vols, Loeb Classical Library, Cambridge, 1938–
62.
4 O. Lagercrantz (trans.), Papyrus Graecus Holmiensis, Uppsala,
1913.
5 But for the use of amethyst beads in the 6th and 7th centuries ad in
the West, see: J. Drauschke, ‘Byzantine Jewellery? Amethyst
Beads in East and West during the Early Byzantine Period’, in C.
Entwistle and N. Adams (eds), ‘Intelligible Beauty’: Recent Research
on Byzantine Jewellery (British Museum Research Publication no.
178), London, 2010, 50–60.
6 E. Fritsch and B. Rondeau, ‘Gemology: The Developing Science of
Gems’, Elements 5, no. 3 (2009), 147–52.
7 M. Sax, ‘The recognition and nomenclature of quartz materials
with specific reference to engraved gemstones’, Jewellery Studies 7
(1996), 63–72.
8 See notes 2 and 3.
9 C. Lüle-Whipp, Mineralogical-Petrological and Geochemical
Investigation on Some Garnets from Volcanic Rocks of Gorece
Village-Cumaovasi, Izmir and Metamorphites of Menderes Massif
and their Possible Archaeogemological Connections, PhD. thesis,
Hacettepe University, Ankara, Turkey, 2006.
10 The identification of this specific chalcedony required a definitive
name. Mineralogically speaking, the term chrome chalcedony is
incorrect as the element chromium is not part of its chemical
structure. This term typically applies to single crystal gems such as
tourmaline or diopside when they have chromium as part of their
chemistry. Another commonly-used term, ‘chromium chalcedony’,
is not completely accurate either. The terms chromium-bearing or
chromian are the most definitive names from a mineralogical
point of view. The term ‘chromian’ chalcedony was first suggested
by Thoresen and is adopted by the author for accuracy as well as
practicality.
11 A. Furtwängler, Die antiken Gemmen, vol. III, Berlin/Leipzig, 1900,
309.
12 G. Platz-Horster, ‘Kleine Praser and Chromium-bearing
Chalcedonies. About a small group of engraved gems’, Pallas.
Revue d’Études Antiques 83 (2010), 179–202.
13 Ibid., 191–5.
chrome chalcedony’, Chamber of Mines Journal, December (1967),
31–4.
15 J. Hyrsl, ‘Chrome Chalcedony – a review’, Journal of Gemmology
26, no. 6 (1999), 364–70.
16 M.J. Willing and S.M. Stocklmayer, ‘A new chrome chalcedony
occurrence from Western Australia’, Journal of Gemmology 28, no.
5 (2003), 265–79.
17 Lisbet Thoresen, pers. comm.
18 C. Lüle-Whipp, ‘Chromium Chalcedony from Turkey and its
possible Archaeological Connections’, Proceedings of the 4th
International Gemological Symposium & GIA Gemological Research
Conference. Gems & Gemology v. 42, no. 3, Carlsbad, 2006, 115.
19 L. Thoresen and C. Lüle, unpublished.
20 G. Platz-Horster, pers. comm.
21 See Lüle-Whipp (n. 18).
‘Gems of Heaven’ | 3
A Case Study on Gemstone Origins
Chrysothrix, a Group of Roman Magical Gems
Lisbet Thoresen
Introduction
The gemstone sources that supplied raw materials to ancient
gem cutters are known to us today mostly from ancient
inscriptions and texts. Many sources are described, and clearly
more varieties of gems were known than have been identified
positively or have survived to the present day. Together, texts
and archaeological work sometimes have revealed associations
between extant carved gems and gem-producing localities,
despite the paucity of direct evidence of gem-mining activity.
In addition, archaeology has rectified written accounts that
sometimes confuse centres of gem industry with gems at their
sources. It also has provided illuminating details about gem
production and the circulation of gems throughout the ancient
world – whether raw materials, partially worked blanks, or
finished, fashioned gems discovered in workshops, trading
centres, or settlement and burial contexts.
Where archaeology is lacking and an ancient writer’s
meaning is elusive, geology and its sub-specialty, mineralogy,
can at least provide useful references for evaluating the
prospective associations between the gemstones and sources
described in the ancient and archaeological literature. The
application of gemmological and analytical techniques to gems
of the ancient world, or archaeogemmology, is a specialised
field of study that has been profitably applied to correctly
identifying gemstones, recognising treatment or enhancement
techniques applied by ancient gem cutters, and geographic
provenancing. The instruments and techniques used in
gemmology are non-destructive and in general, non-invasive
or minimally invasive.1
The application of gemmology to geographic provenancing
relies upon a combination of properties that includes
observable features, optical properties, physical constants,
crystalline structure, and chemistry. A brief summary is given
below of analytical techniques and microscopy used in
gemmology that are useful for geographic provenancing with
several studies on Roman emeralds cited as examples. Because
the variety of gemstones and the geographic range of their
prospective origins in the later Roman Empire is a subject of
great breadth and complexity, a single engraved gem, the
chrysothrix, or rutilated quartz, will be used here to illustrate
the aspects of ancient gemstone origin that gemmology can
illuminate. The focus will be on the visually observable
characteristics of this gemstone and what it may reveal about
itself in relation first, to its origin, second, to other gemstones
with which it may share a common origin, and third, to the
interpretation of gems described in ancient texts, here with
reference to Pliny the Elder’s Naturalis Historia and the Orphic
Lithika. The latter two points may be of more practical interest
to gem scholars investigating the relationships between
individual gems as products of a particular artist, workshop, or
period and place. To a certain degree, such associations have
4 | ‘Gems of Heaven’
been recognised: the gem engraver Dexamenos’ predilection
for brecciated jaspers; Graeco-Persian seals carved in finely
textured translucent chalcedony of pale rose, blue, gray, and
white colour; Roman nicolos – onyx with a bluish-white layer
on a dark blue-black substrate; magical gems whose use is
related to colour and subject, such as the depiction of a
scorpion in yellow jasper (see the paper by Faraone this
volume, Pl. 14). The intent in the present study is to refine the
depth of observations that could be made through the
recognition of salient physical features which often are
overlooked. The principal tool is the microscope.
Geographic gemstone provenancing
The origins of gemstones of the ancient world can be discussed
in terms of a range of possibilities in relation to geologic
environments. Some materials such as quartz are too
ubiquitous and variable to be able to identify their sources.
Provenancing is especially challenging with rocks and
heterogeneous materials, for example, varieties of
microcrystalline quartz, as well as some polycrystalline
materials such as serpentinite, chlorite, and steatite. Rare
materials of limited occurrence also may be difficult to localise
to their sources. Lapis lazuli, for example, which is a very rare
gem, but one that is well-represented in early dated burial and
workshop contexts, has been the subject of numerous
analytical studies. The properties of materials from some
different localities overlap, so the results have not been
unequivocal.2
Analytical studies
Analytical techniques useful for provenance studies may be
broadly categorised according to bulk chemical analysis, trace
element analysis, stable isotopic analysis, inclusion chemistry,
and infrared and spectral fingerprinting. Since about the mid-
1970s, analytical provenance studies employing a variety of
techniques have been applied successfully to manufactured
glass and a variety of natural gem materials used in ancient
glyptic, including emeralds, garnets, lapis lazuli, and obsidian.
Several studies on ancient emeralds are summarised below to
illustrate the application of various techniques to geographic
provenancing of ancient gemstones.
Since the 1990s a variety of different analytical methods
has been applied to provenance studies on emeralds, several
focusing specifically on ancient and post-antique gems. Gaston
Giuliani and other researchers at the Centre for Research in
Petrography and Geochemistry (CRPG), Nancy, France, have
produced a series of provenance studies on gemstones. In 1998,
he and others published a study in which ion microprobe
oxygen isotopic analysis was used to associate emeralds from
different post-antique cultures with their geological origins.
Samples from Egypt, Pakistan, Austria, and Colombia were

included in the study. The criteria for the genetic geological
classification of the Egyptian emeralds in this study have been
questioned by other researchers whose investigations concur in
making a different assignment.3
Ion Beam techniques (IBA) – proton-induced X-ray
emission (PIXE) and proton-induced gamma emission (PIGE)
analyses – have been used in provenance studies on a range of
minerals, including emeralds. At Queens University, Kingston,
Canada, X.P. Ma and others used PIXE/PIGE analyses to
measure quantitatively light and trace elements in emeralds.
IBA techniques are also used to characterise museum objects at
the Centre de Recherche et de Restauration (CNRS) des
Musées de France. Using chemical analyses to determine light
and trace elements, data on emeralds mounted in Visigothic
jewellery (7th to 8th century ad) were compared against
emeralds from various localities, including Egypt, Austria,
Afghanistan, Pakistan, Central Urals, and India. Thomas
Calligaro argued that the analyses showed that the Visigothic
emeralds were the earliest attested use of emeralds from
Habachtal Valley.4
The Habachtal mines are situated in the Tyrolian Alps in
the south-western part of ancient Noricum (most of modern
Austria), a region invaded by Celtic tribes and later annexed
into the Rome Empire (c. 16/15 bc). Under the Romans,
Noricum became an important source of iron ore (the raw
material was processed at Aquileia). It has not been shown
through archaeological evidence that the Romans knew of the
Habachtal emerald mines; however, it would not be surprising
if an earlier dated association came to light between the
Habachtal emeralds and Roman glyptic or jewellery.
Another study on emeralds from a dated Roman context
was performed by an Italian team in Naples. Non-destructive
techniques, including electron microprobe analysis and
Fourier Transform Infrared Spectroscopy (microFTIR), were
used on the unfaceted emerald crystal beads strung on a gold
necklace found at Oplontis, which was destroyed in the
eruption of Vesuvius in ad 79.5 The authors considered
Habachtal as a possible source of the emeralds, but the
presence of mica and limonite inclusions inclined them to
favour an Egyptian origin. More methodologies applied to the
study of the Oplontis necklace (and more samples) might have
yielded less ambiguous results.
In provenance studies on ancient gems such as some of the
studies described above, determinations sometimes will be
qualified or tentative. In general, it is not feasible to make
definitive determinations if too few samples are available to
produce statistically meaningful results. Sometimes the range
of available tools and techniques are limited, or the specimens
preclude the use of some techniques due to the fragility of the
gemstone or because its mount is too large or obstructive for
the instrument. Certainly, small studies can provide useful
results, even if the results cannot pinpoint an origin with a high
degree of certainty. Data that can be compared directly with
other data independently of operating conditions help to
develop the wider field of general knowledge in gem studies
over the long-term; however, the importance of having access
to suitable reference standards and mineral libraries cannot be
overstated. Few museum laboratories have access to suitable
reference material or specialised expertise in the analysis of
minerals, even if the instruments are available. Partnering
A Case Study on Gemstone Origins
with institutions that have experience and facilities for
routinely performing gemstone analysis is a logical solution.
Visual observation – microscopy
Geographic provenancing relies on optical and physical
properties such as refractive index and density, and especially
observable features such as clarity (diaphaneity or
transparency), colour (including colour zoning), cleavage and
fracture, and crystal growth patterns. Much useful information
about a gemstone’s identity and its genetic origin is lost after a
stone has been cut and fashioned. Crystal size, shape, and
condition reflect the gem’s rock-forming environment and
sometimes deposition, whether, for instance, it was a surficial
find plucked from an alluvial placer in a streambed or mined
from a pegmatite pocket.
A gem’s internal features or inclusions retain physical
evidence of their paragenesis, which sometimes can be
observed in a light microscope. Inclusion paragenesis refers to
the formation of guest minerals within a host gem crystal: it
provides evidence of the geologic environment and phase in
which it and the gem host grew. The identity, form, and
condition a guest inclusion exhibits sometimes may be
diagnostic of geographic origin. The usefulness of inclusion
identity, paragenesis, and chemistry to geographic provenance
research has been elucidated in the pioneering work of
gemmologists Eduard Gübelin and John I. Koivula.6
Similar inclusions may be found in gems of the same
species from different localities, but sometimes their presence
and/or morphology are highly characteristic for a given
locality, and sometimes they are uniquely identifiable with a
specific locality.7 For example, the internal ladder-like fracture
pattern resembling a centipede in a moonstone is virtually
diagnostic for a Sri Lankan origin (Pl. 1).8 Likewise, a dense
arrangement of rounded grains of apatite and a treacly or
roiled body structure are highly characteristic features of
hessonite, the brownish-orange variety of grossular, from the
gem gravels of Sri Lanka (Pls 2–3).9 Guest inclusions of
metamict zircons are strongly, although not exclusively
correlated to a gem’s Sri Lankan origin (metamict minerals
have an amorphous structure due to the radioactive decay of
the elements uranium and thorium, which transform their
original crystalline structure) (Pl. 4).10
Plate 1 Moonstones from Sri Lanka sometimes exhibit distinctive inclusions
resembling centipedes, because exsolution of albite produces a pattern of
lateral fissures; centipede or ladder-like inclusions in this variety of orthoclase
feldspar, as seen here, are diagnostic for a Sri Lankan origin
‘Gems of Heaven’ | 5
Thoresen
Plate 2 The dense concentration of rounded protogenetic apatite crystallites
in hessonite from Sri Lanka is a characteristic feature (30 x mag)
Plate 3 This gem, engraved with a gutus or jar, is hessonite; identified by
Çiğdem Lüle, it is heavily included with rounded apatite crystallites and fine
rutile needles. The inclusions as seen in this stone are virtually diagnostic of a
Sri Lankan origin. End of 1st to beginning of 2nd century ad, grossular, var.
hessonite, 8.5 x 6.7 x 2.9mm. Berlin, Antikensammlung, Staatliche Museen,
misc. inv no. 7074,32
Plate 4 Evoking the gem Pliny the Elder calls ‘anthracites’ or ‘carbuncle stone’,
which ‘appears to have sparks running in different directions through it’ (Nat.
Hist. XXXVII.190), here, a Hellenistic almandine-pyrope garnet is host to a
metamict zircon crystal in which the radioactive elements contained in its
crystal lattice caused it to rupture; the inclusion is seen against a background
of acicular crystals of rutile that are oriented in two directions (field of view
7.0mm)
6 |’Gems of Heaven’
Cohabitation of multiple associated guest minerals in a gem
host also may help to isolate salient locality-specific features,
especially when gems from different localities otherwise
exhibit some overlapping characteristics. For example, a
distinctive suite of inclusions, rather than a single unique
inclusion helps to differentiate emeralds from Habachtal,
Austria; the Central Urals, Russia; Ajmer, Rajasthan State,
India; Swat District, NWFP, Pakistan; Lake Manyara,
Tanzania.11 Even visually distinctive inclusions may require
confirmation by other techniques. Chemical analyses
sometimes are needed to confirm the identity of the mineral or
fluid inclusions within the gem host, in addition to determining
the gem host’s chemistry.12
What’s in a name? Chrysothrix and rutilated quartz
Chrysothrix – a magical gem
The Orphic Lithika, putatively a 4th century ad work, describes
the symbolic and physical properties of about 30 gemstones,13
nearly all of which are recognisable in Pliny’s Naturalis
Historia.14 Chrysothrix (‘golden hair’), however, is a gem term
the Orphic Lithika introduces for the first time in ancient
literature. It was recognised by Nathaniel F. Moore in 1834 and
more clearly elucidated by Ruslan I. Kostov as rutilated quartz,
a crystalline quartz containing guest mineral inclusions of
golden-coloured acicular (needle-like) crystals of rutile.15
The author identified a rutilated smoky quartz intaglio in
the collection of the J. Paul Getty Museum, Malibu, in 1992,
and although appreciating that the material was rare, perhaps
exceptional in ancient glyptic, did not recognise that this
magical gem might have a special appellation until recently.16
Independently, Jeffrey Spier examined the stone, also noting
the rarity of the material in ancient glyptic.17 He has identified
three additional magical gems engraved in rutilated quartz,
which together with the gem in Malibu all date between the
2nd and 3rd centuries ad (Pl. 5).18 He observed that these gems
manifestly had been selected by the gem cutter specifically for
the inclusions to complement the subject of the engravings they
all share in common: the sun god Helios or Helioros (Helios-
Horus) or a lion, representing the sun.19
Plate 5 Biconvex oval ringstone depicting frontal radiate figure (Helios/Sol?)
holding a phiale and whip. 2nd century ad, rutilated quartz with faceted back
(modern addition), mounted in a 17th century setting, 14 x 10 x 4mm.
Baltimore, Walters Art Museum, acc. no. 42.1157, 33

Each of the magical gems in the group Spier identified are
colourless quartz or have a very slightly smoky brown tone, and
all contain a random arrangement of eye-visible rutile needles,
whose golden sub-metallic to metallic luster resembles the glint
of the sun’s rays when the stone is tilted back and forth. The
gem in Malibu, a pale smoky quartz, is characteristic of the
group: the shape is a biconvex oval; on the obverse Helios is
depicted radiate, with hand upraised and standing in a
quadriga facing frontally, with an anchor in the field below the
horses’ hooves; on the reverse is a three-letter inscription, iaw
(Iao) (Pls 6–7). A magical gem described by Sliwa is perhaps
another chrysothrix – it depicts Helios/Harpokrates and bears
the inscription IAW engraved in rock crystal quartz containing
‘golden inclusions’.20
Rutilated quartz – the gemstone and its sources
The mineral rutile, titanium dioxide (Ti02), is said to be
widespread in the host rock of the localities where it occurs,
and it also appears in various forms as a guest inclusion in
many gemstones from many different localities (Pls 4, 5, 7 and
8).21 When present as acicular (needle-like) or fibrous crystals
in crystalline quartz it is called rutilated quartz, and while not
rare in nature, neither is it a common gemstone. The needles
typically exhibit adamantine metallic luster and have a golden
to reddish colour. Rutilated quartz is sometimes called by
several different descriptive names, including Venus hair
stone. The rutile crystals may be straight or curving, oriented
or arranged in dense waving swatches resembling lockets of
golden or red to reddish-brown hair.22 The crystals can appear
A Case Study on Gemstone Origins
Plates 6–7 Front and back views of a
magical gem with Helios (obverse,
left) and the inscription IAW (Iao)
(reverse, right), 2nd–3rd century AD,
rutilated smoky quartz, 13.0 x 10.8 x
5.3mm. Gift of Stanley Ungar, The
J. Paul Getty Museum, Villa
Collection, Malibu, California, acc.
no. 82.AN.162.76
Plate 8 Detail of above. A 2-phase inclusion that consists of a cavity containing
a liquid and a gas vesicle hovering above acicular (needle-like) inclusions of
golden rutile
very fine to coarse and when oriented in a reticulated or net-
like structure it is termed sagenitic rutile.23 Rutilated quartz is
found in Karagandy Province, Kazakhstan; Balochistan and
North-West Frontier Province (NWFP), Pakistan; the Central
Urals, Russia; and in Europe: Viseu District, Portugal; Brittany,
France; Ticino and Uri in the Swiss Alps; several localities in
Austria, including Carinthia, and several sites in the environs
of Salzburg, including Habachtal Valley, and also Styria and
the Austrian Tyrol. Specimens have been collected from the
cliff faces of a quarry in Wadebridge District, Cornwall,
England.24
Discussion
The gemstone chrysothrix is first mentioned in ancient
literature in the 4th century ad, in the Orphic Lithika. Ruslan I.
Kostov deduced that the stone implied in the text is rutilated
quartz, a crystalline quartz containing guest inclusions of
golden-coloured acicular or fibrous crystals of the mineral
rutile. Today, this gem is also known as Venus hair stone, a
term derived from Pliny’s veneris crinis.25
It does not appear that the chrysothrix of the Lithika has
been identified with any extant ancient engraved gems
previously. Golden inclusions in quartz which complement the
sun god motifs of Helios/Helioros have been noted by Sliwa
and perhaps others, but these distinctive features were
recognised in the early 1990s for their rarity in ancient glyptic
in the Malibu gem, which disposed Spier to notice similar types
related first by the material – rutilated quartz – and then the
shape of the stone, motifs, and inscriptions. The dates for his
group ranges from the 2nd to 3rd century ad, suggesting that
they were probably products of a tradition more than a
workshop. Previously, Campbell Bonner observed that rock
crystal quartz – not specifically rutilated quartz – is a relatively
rare material for magical gems, and he speculated that the lion
motif seen on these special stones might belong to a group
produced in the same workshop.26 Spier’s group of quartzes
engraved with the related sun god motif of Helios/Helioros
may help to reinforce Bonner’s hypothesis.
‘Gems of Heaven’ | 7
Thoresen
Pinpointing the origin of the magical gems carved from
rutilated quartz, although rare in ancient glyptic, is
problematic. Rock crystal quartz (SiO2) is ubiquitous in nature.
Rutile may occur as an inclusion in quartz anywhere rutile
(also not a rare mineral) is present in the local host rock. Even
so, reported sources of rutilated quartz are relatively few
compared against the widespread occurrence of its two
principal constituents. In most localities, whether mining is
small, large or commercial in scale, rutilated quartz is likely to
be a minor or anomalous occurrence and not the predominant
mineral.27 Reported (and especially unpublished) occurrences
of rutilated quartz in Cornwall, England; Brittany, France;
Viseu, Portugal or other localities may be minor or virtually
negligible, which perhaps is similar to the circumstances as
found in antiquity. A serendipitous find or several may explain
the appearance of so few rutilated quartzes (bearing in mind
that the more common rock crystal is also not very prevalent as
a material for magical gems).
Conclusion
Moore and more recently Kostov enabled an association to be
made between chrysothrix, an obscure gemstone named in an
ancient poem, and an extant group of rare gemstones in
Roman glyptic. In the light of characteristics of the magical
gems demonstrably identified by the term chrysothrix, it would
be helpful now not to confuse or conflate it with other gems
Moore and Kostov proposed previously.28 Chrysolite is a
deprecated and confusing historical term applied to
chrysoberyl and also to the gem olivine, peridot. Chrysoberyl
has not been shown to be a gem of the ancient world. Sunstone,
a phenomenal variety of gem feldspar, likewise is a gem not yet
identified in the ancient gem cutter’s repertoire. The
interchangeable use of the terms rutilated quartz and sagenitic
quartz is unhelpful, because ‘sagenitic’ connotes reticulated
rutile, or rutile having a network structure.29 They are very
different in appearance and only the acicular and fibrous (Pls
6,7,8) morphology has been identified. Cupid’s arrows is
another unhelpful descriptive term. At least Venus hair stone
has a clear derivation – Pliny’s veneris crinis. Rutilated quartz is
correct and unambiguous.
It is the subject for another study to investigate the
typological connection between all the magical gems carved
from crystallus and chrysothrix or to hypothesise further about
their prospective association with a common workshop or
other groups of engraved gems. The material of the ancient
chrysothrix, in itself is not likely to reveal its geographic origin
on the basis of visual observation alone or even with the help of
currently available analytical techniques. However,
developments have been dramatic in the past 15 years, and the
next frontier in mineralogy promises that useful developments
are forthcoming rapidly, especially with the availability of
portable FTIR and portable Raman instruments, isotopic tools
and micro-Raman spectometry (used to identify mineral and
fluid inclusions in gemstones for provenancing) (Pl. 8).30
In future studies of the gems described in this essay and its
related types, suitable techniques useful for geographic
provenance determination should include analyses not only on
the rutilated specimens, but on all of the quartzes. Some of
them may have been obtained from the same source. Reference
standards for comparison from the localities enumerated
8 |’Gems of Heaven’
above, especially the European sources, and others should be
included. Also, quartz specimens from archaeological contexts,
especially workshop wasters or debitage would be excellent
controls for comparison, for example, the ancient workshops at
the lapidary centres of Aquileia and Roman Ammaia,
Lusitania, especially the latter site where a local source of
quartz has been identified.31 (As noted previously, pinpointing
geographic origin may not be possible, but perhaps a useful
and obtainable goal may be to relate more securely this group
of engraved gems to each on the basis of chemistry.) In the
meanwhile, techniques available to us now may be
constructively employed to extend our knowledge – the simple
act of making critical observation using a microscope as
described in this essay did not establish a single quantifiable
proof or fact, but it concentrated attention in a slightly different
and useful way. It enabled the identification of a related group
of gemstones on the basis of a distinctive material, which in
turn validated Kostov’s interpretation of rutilated quartz as the
ancient chrysothrix.
Acknowledgements
The author would like to thank Jeffrey Spier who generously
contributed his notes, references, and many useful comments on the
special group of magical gems that otherwise might not have been
identified as the ancient chrysothrix; thanks to John I. Koivula who
examined and photographed the Getty gem, produced all the photo-
micrographs used in this essay, and read the manuscript; thanks to
James A. Harrell who read the manuscript and made many useful
suggestions; thanks to Erica and Harold Van Pelt, who photographed
the Getty gem; and thanks also to Herbert Wiegandt, who
photographed the Walters gem.
Notes
1 Mineralogy employs a wider range of techniques, including
invasive and destructive analyses, many of which are extremely
informative, but cannot be applied to minerals having pecuniary
or cultural value.
2 For a chemical analysis of lapis lazuli artifacts compared against
reference standards from sources including Badakhshan Province,
Afghanistan; Pamir Mountains, Tajikistan; and Chagai Hills,
Balochistan, Pakistan; Lake Baikal, Siberia, see, A.B. Delmas and
M. Casanova, ‘The lapis lazuli sources in the ancient Near East’, in
M. Taddei (ed.), South Asian Archaeology 1987 – Proceedings of the
9th International Conference of the Association of South Asian
Archaeologists in Western Europe, Rome, 1990, 493–505;
M. Casanova, ‘The Sources of the lapis-lazuli found in Iran’, in
C. Jarrige (ed.), South Asian Archaeology 1989 – Proceedings of the
Tenth International Conference of the Association of South Asian
Archaeologists in Western Europe, Musée nationale des Arts
asiatiques, Paris, 3–7 July 1989, Madison, 1992, 49–56; J. Zöldföldi
and S. Kasztovszky, ‘Provenance study of lapis lazuli. Non-
destructive Prompt Gamma Activation Analysis (PGAA)’, in
Y. Maniatis (ed.), ASMOSIA VII (Proceedings of the 7th
International Conference of Association for the Study of Marble
and Other Stones in Antiquity), Athens, 2009, 677–91; A. Lo
Giudice, A. Re, S. Calusi, L. Giuntini, M. Massi, P. Olivero,
G. Pratesi, M. Albonico and E. Conz, ‘Multitechnique
characterization of lapis lazuli for provenance study’, Analytical
and Bioanalytical Chemistry 395(7) (2009), 2211–17.
3 G. Giuliani, C. France-Lanord, P. Coget, D. Schwarz, A. Cheilletz,
Y. Branquet, D. Giard, A. Martin-Izard, P. Alexandrov and
D.H. Piat, ‘Oxygen isotope systematics of emerald-relevance for its
origin and geological significance’, Mineralium Deposita 33 (1998),
513–19; other researchers disagree with the authors’
characterisation of the Wadi Sikait emeralds as Type II deposits,
but classified them as Type I: see, J.A. Harrell, ‘Archaeological
geology of the world’s first emerald mine’, Geoscience Canada 31(2)
(2004), 69–76; see also Zwann’s study, especially his conclusions
regarding the traditional genetic classification for emeralds
originating in schist-type deposits, for which he proposed the need

for a revised scheme, in: J.C. Zwaan, ‘Gemmology, geology and
origin of the Sandawana emerald deposits, Zimbabwe’, Scripta
Geologica 131 (2006), 161–3; and also, G. Grundmann and G.
Morteani, ‘Multi-stage emerald formation during Pan-African
regional metamorphism: The Zabara, Sikait, Umm Kabo deposits,
South Eastern desert of Egypt’, Journal of African Earth Sciences 50
(2007), 168–87, especially 185–6.
4 T. Calligaro, ‘Origin of Ancient Gemstones Revealed by PIXE,
PIGE, and µ-Spectrometry’, in U. Masayuki, G. Demortier and I.
Nakai (eds), X-Rays for Archaeology, New York, 2005, 101–12; see
also, T. Calligaro, J.C. Dran, J.P. Poirot, G. Querré, J. Salomon and
J.C. Zwaan, ‘PIXE/PIGE characterization of emeralds using an
external micro-beam’, Nuclear Instruments and Method in Physics
Research B161–3 (2000), 769–74.
5 C. Aurisicchio, A. Corami, S. Ehrman, G. Graziani and S.N. Cesaro,
‘The emerald and gold necklace from Oplontis, Vesuvian Area,
Naples, Italy’, Journal of Archaeological Science 20 (2005), 1–10; see
also Zwaan (n. 3), 159.
6 See, E. Gübelin and J. Koivula, ‘II. Thematic section: inclusions
and the microscope’, in Photoatlas of Inclusions in Gemstones
Volume 2, Basel, 2006, 30–64 (hereafter Photoatlas 2); E. Roedder,
‘Fluid inclusions in gemstones’, in E. Gübelin and J. Koivula,
Photoatlas of Inclusions in Gemstones Volume 1, Basel, 2004 (4th
edn), 62–87 (hereafter Photoatlas 1); and see also, J.I. Koivula,
‘Photomicrography for Gemologists’, Gems & Gemology 39(1)
(2002), 4–23.
7 See, E. Gübelin and J. Koivula, ‘III. Characteristics of inclusions:
diagnostic inclusions, diagnostic colours of mineral inclusions,
diagnostic morphology of mineral inclusions, fluid inclusions,
geological correlation’, in Photoatlas 2, 110–282.
8 Photoatlas 1, 276–7; Photoatlas 2, 125, 409–10.
9 Photoatlas 2, 128, 471.
10 Photoatlas 1, 197–9; Photoatlas 2, 446.
11 See, E. Gübelin and J. Koivula, ‘Inclusions in Emeralds’, in
Photoatlas of Inclusions in Gemstones Volume 3, Basel, 2009, 354–
441.
12 See, ‘II. Thematic section: inclusion analysis’, in Photoatlas 2,
65–94.
13 J. Spier suggests that chrysothrix not chrysotrix is the proper
transliteration of the Greek thrix meaning hair. E. Abel (ed.),
Orphei lithica. Accedit Damigeron de lapidibus, Berlin, 1881, 290–
302; for an updated translation and commentary see R. Halleux
and J. Schamp, Les Lapidaires grecs, Paris, 2001, 98–9 and 308; the
authors speculate on the identity of the inclusions being pyrite,
probably for the metallic luster. Rutile is a guest mineral inclusion
in many gemstones, frequently exhibited as short or long acicular
(needle-like) crystals and sometimes with a golden metallic luster.
14 Pliny the Elder, Naturalis Historia, Books XXXVI–XXXVII, trans.
D.E. Eichholz (Loeb Classical Library), Cambridge, MA, 1962.
15 N.F. Moore, Ancient Mineralogy: Or, an Inquiry Respecting Mineral
Substances Mentioned by the Ancients, New York, 1834, 142, where
the author cites in the same passage chrysolite (peridot, but
wrongly translated as our topaz), also containing ‘bright rays
resembling hairs’, peridot from some localities may contain fibrous
inclusions of ludwigite-vonsenite, asbestsos, or chrysotile (see
Photoatlas 2, 513, 518, 519, 521, 530, 531); G.F. Kunz, The Magic of
Jewels and Charms, Philadelphia, 1915, 29–30; R.I. Kostov, ‘Orphic
Lithica as a source of late antiquity mineralogical knowledge’,
Annual of the University of Mining and Geology ‘St. Ivan Rilski’, vol.
5(1), Geology and Geophysics, Sofia, 2008, 109–14; also: http://
www.mgu.bg/sessions/08/1/kostovri3.pdf. Kostov also provides
an overview of ancient texts on gemstones and also a survey of
various translations of the gemstones of the Orphic Lithica and its
dating.
16 S. Michel, Die Magischen Gemmen. Zu Bildern und Zauberformeln
auf geschnittenen Steinen der Antike und Neuzeit. Studien aus dem
Warburg-Haus, Band 7, Berlin, 2004, 118, notes 627, 628, gift of
Stanley Ungar, The J. Paul Getty Museum, Villa Collection,
Malibu, California, acc. no. 82.AN.162.76, rutilated smoky quartz,
13.0 x 10.8 x 5.3mm.
A Case Study on Gemstone Origins
17 J. Spier examined the gem in Malibu about the same time as the
author, but did not discuss the stone until 2009.
18 J. Spier kindly provided the following references for the three
gems: 1) a cabochon with convex back (faceted in modern times) in
which Helios stands facing, holding whip and phiale, formerly
Marlborough and Arundel collections, Walters Art Museum,
Baltimore, acc. no. 42.1157, in J. Boardman, The Marlborough
Gems, Oxford, 2009, 83, no. 128; and 2) an intaglio of a standing
figure of a lion-headed man (Helioros) holding a globe and whip,
Museum of Fine Arts, Boston, acc. no. 01.7556, in: Romans and
Barbarians (exh. cat., Boston), Boston, 1985, 57, no. 78, and
C. Bonner, Studies in Magical Amulets, Ann Arbor, 1950, 19–20 and
292, no. 234, pl. 11,2; and 3) an intaglio similar to the previous,
described as ‘tinged with brownish yellow’, so probably smoky
quartz, Ruthven collection (present whereabouts unknown), in
Bonner ibid., 293, no. 236. (The author has not personally
examined these gems.)
19 J. Spier points out that chrysothrix may be the same stone Pliny
called solis gemma: ‘Solis gemma candida est, ad speciem sideris in
orbem fulgentis spargens radios’ (Nat. Hist. XXXVII.181;
mentioned also in Les Lapidaires Grecs [n. 13]); chrysothrix is
mentioned also in a less well-known text, Pseudo-
Hippocrates Med. Έρμηνεία περὶ ἐνεργῶν λίθων, 35, which
appears to be a version of the Orphic Lithica.
20 Cited in Michel (n. 16), 273: the entry reads: ‘Bergkristall, goldene
Einschlüsse. SLIWA 84, pl. 23, 116: n.r. Helio/Harpokrates mit
Strahelenkranz rudert; Rs.; IAW’. (The author has not personally
examined this gem.)
21 R. Webster, Gems: Their Sources, Descriptions and Identification,
rev. P.G. Read, Boston, 1994 (5th edn), 223, 365–6.
22 Photoatlas 1, 188, 193, 313; Photoatlas 2, 627–9.
23 Photoatlas 2, 29, 630–1.
24 Rutilated Quartz: Rutilated Quartz mineral information and data
n.d., [online] Available at: http://www.mindat.org/min-3485.
html [Accessed 29 December 2010].
25 Nat. Hist. XXXVII.184. Veneris crinis is described as a ‘very dark,
brilliant stone, which has an inclusion resembling a lock of red
hair’; Pliny does not give a locality.
26 In Michel (n. 16), 77.
27 Brazil is a notable exception; today it is a significant producer of
high-quality, large pieces of rutilated quartz that are widely
available in the gem market.
28 See, n. 15.
29 Kostov (n. 15), 111–12.
30 For a general overview of developments and analytical tools useful
in gemstone provenancing see: C.M. Breeding, A.H. Shen,
S. Eaton-Magaña, G.R.R. Rossman, J.E. Shigley and A. Gilbertson,
‘Developments in gemstone analysis techniques and
instrumentation during the first decade of the 2000s’, Gems &
Gemology 46(3) (2010), 241–57; G.R.R. Rossman, ‘The geo-
chemistry of gems and its relevance to gemology: different traces,
different prices’, Elements 5(3) (2009), 159–62; M.A. Ziemann, ‘In
situ micro-Raman spectroscopy on minerals on-site in the Grotto
Hall of the New Palace, Park Sanssouci, in Potsdam’, Journal of
Raman Spectroscopy 37 (2006), 1019–25; H.A. Gilg and N. Gast are
collaborating on Raman studies on unusual fluid inclusions in
amethyst from the ancient mine in Wadi El-Hudi, Egypt,
unpublished, pers. comm. from H.A. Gilg to the author, 20 January
2011. 
31 The author is grateful to Graça Cravinho for providing the
references to the site at Roman Ammaia; G. Sena Chiesa, Gemme
del Museo Nazionale di Aquileia, Padua, 1966; G. Cravinho and S.
Amorai-Stark, ‘A Jewish intaglio from Roman Ammaia, Lusitania’,
Liber Annuus 56 (2006), 533–43; G. Cravinho, ‘Some engraved
gems from Ammaia’, Pallas. Revue d’études antiques 83 (2010),
14–15, notes 5–7.
32 See, C. Weiß, Die antiken Gemmen der Sammlung Heinrich Dressel
in der Antikensammlung, Berlin, Würzburg, 2007, 303, pl. 79.604.
33 See, Boardman (n. 18).
‘Gems of Heaven’ | 9
The Garnet Millennium
The Role of Seal Stones in Garnet Studies
Noël Adams
Garnets, famous for their magnificent colours and light
refractive properties, are now the most intensively studied of
all the gemstones used for glyptics and jewellery in the ancient
world. Scientific analysis of ancient garnets to date has
focussed on identifying garnet sources, a topic addressed with
equal enthusiasm in the first ancient texts on gemstones. As
with other precious stones, factors such as the difficulty of
extraction or acquisition, the distance stones must travel from
source to destination and the reliability of their supply, in
combination with their beauty, determine their value. In the
modern period pinpointing sources contributes to our
understanding of cross-cultural contacts and trading patterns
in the ancient world, and current research, concentrated
almost exclusively on garnets set in cloisonné made in the Early
Medieval period in Europe, has established links with sources
in India and Europe (Pl. 1).
Aspects of this research on garnets are explored further
below and in the following paper, but as an introduction to the
subject it is important to realise that garnet is one of the most
common gemstones found on the face of the earth. Modern
science recognises 24 species of this complex neosilicate
mineral, produced by a range of geologic processes which
result in crystals of different chemical compositions, size,
colour and translucency. Initial identification of garnet species
can be determined by gemmological examination but a
complete chemical analysis requires scientific examination in a
laboratory environment. Testing of ancient garnets has
established that the vast majority of stones used in the Classical
and Early Medieval periods fall in the red/purple colour
Plate 1 Ternary diagrams of the end members of the pyralspite series, plotted with distribution of Early Medieval garnet plates, Calligaro’s Types I–V (left) and
garnet specimens from India, Sri Lanka and Bohemia (right), after Calligaro et al. (n. 63), pl. III
10 | ‘Gems of Heaven’
spectrum and belong in the pyralspite series. These garnets
crystalise in continuous solid solution with one another along a
spectrum whose end members are known as pyrope,
almandine and spessartine (Pl. 1).1
This essay provides an introduction to the following paper
by Adams, Lüle and Passmore which presents the results of the
pilot phase of a project at the British Museum to scientifically
analyse the garnet seal stones in the museum’s collections. This
project, called Garnets: Classical, Eastern and Medieval
(hereafter GCEM), was initiated in 2009 and aims to create a
reference database of the chemical composition of garnets
across the millennium in which this gemstone was intensively
used.
Garnet intaglios date from the late 4th/early 3rd centuries
bc to the 6th century ad in the Classical West (the Etruscan,
Greek, Roman and Early Byzantine periods) and from the 2nd
century ad to the 6th or 7th century ad in the East (the Kushan,
Sasanian and Hunnic periods). The British Museum is
particularly well-placed to carry out this research as its various
departments hold major glyptic collections from all these
periods.
Documenting and testing the many hundreds of ornaments
incorporating garnets from this 1,000 year period would be a
major task. Garnet seal stones, however, constitute a discrete
body of material for study which is easily classified and
controlled. As we shall see below, the percentages of garnets
within larger collections of engraved ring stones can be used to
demonstrate the ebb and flow of garnet usage over a 1,000
years. Although the determination of sources, or at any rate the

characterisation of host rocks, remains a primary interest, the
GCEM project aims to broaden the parameters of current
scientific research into garnet. Our preliminary explorations
suggest that a combination of mineralogical and stylistic
analysis will allow new and closer definitions of workshop
traditions in the ancient world. Groups of mineralogically
similar stones, for example, can be compared to art historical
classifications of intaglios based upon style (and in some cases
epigraphy), allowing us to determine whether garnets of
varying compositions (and thus possibly sources) were used
within individual workshops. A comparison of stones from
eastern and western provenances may disclose similarities and
dissimilarities. In addition this extensive seal stone database
can be compared with the existing database of tested garnets
from Early Medieval ornaments which may reveal whether
similar sources were available over extended periods of time.
This paper is divided into three sections. In Part 1 the
overall pattern of usage of garnet seal stones from c. 300 bc to
ad 600 is surveyed, including a discussion of some of the
problems of identification and nomenclature with regard to
garnets. In Part 2 the issues of nomenclature and sources in
ancient texts is reviewed and set against the current scientific
thinking. Part 3 discusses the aspect of crystal size, bringing
together some garnet seal stones of similar size and date for
which a research database would be useful for examining
questions of stone sourcing and workshop practice.
Part 1
A survey of intaglios in the ‘garnet millennium’
A survey of the overall pattern of garnet usage from the
Hellenistic to the Early Medieval periods is a topic which has
never been addressed in the scholarly literature. The
archaeological record suggests that garnet was used
sporadically from c. 3000 bc onwards but then more or less
continuously from c. 300 bc–ad 700. Garnet was shaped into
cabochons and cameos for personal ornaments, engraved as
ring stones and, at the end of this time frame, polished into flat
plates to be assembled in cloisonné cellwork. After this
millennium of intensive exploitation, although garnets
continued to be used and re-used, notably on Early Medieval
liturgical objects in the West (see Sena Chiesa and Kornbluth in
this volume), they are not a dominant feature of gemstone
jewellery again until the upsurge of production of pyrope
garnet from Bohemian mines in the 18th century. We do not
know whether their diminished presence towards the end of
the 1st millennium ad was the result of changing tastes,
interruptions in supply, exhaustion of sources or combinations
of these factors.
The GCEM project began by assembling lists of garnet seal
stones for testing from British Museum printed catalogues.
Even those compiled nearly a century ago remain useful
compendiums of large numbers of gems which have been
grouped and approximately dated, primarily on the basis of
style. The initial comparisons of lists with actual stones
revealed two ongoing problems: the consistent
misidentification of gem materials and the inconsistent
application of terminology. In Walter’s 1926 catalogue raisonné
of the engraved gems in the British Museum, for example, some
garnet gemstones are either misidentified as amethyst or sard,2
The Garnet Millennium
Plate 2 Herakles to left, beating
Hydra with club, carnelian; intaglio:
4th century BC, ring: 2nd century BC;
L. (bezel) 13mm, D. (ring) 32mm.
London, British Museum, GR
1814,0704.1292; Townley Collection
or referred to using older terminology, such as ‘hyacinthine’ or
‘amethystine’, which does not correspond to modern scientific
terminology.3 Other terms used for garnets, such as ‘jacinth’,
describe different stones altogether in modern mineralogical
classifications (see below), or describe stones of a particular
size and cut: i.e., ‘carbuncle’.4 The problem persists throughout
the 20th century in catalogues of both western (Graeco-
Roman) and eastern (Central and South Asian) engraved gems,
where similar terms appear and garnets have been
misidentified as sard, agate or chalcedony.5 In some modern
catalogues, authors apparently have reserved judgement and
refrained from identifying stones they were not confident of,6
while even conscientious contemporary authors have mis-
identified some of the garnet species.7
Such misidentifications, however understandable, have
knock-on effects. Two examples illustrate the nature of the
problems that can arise. One of the more famous gemstones in
the Greek and Roman Department, at the British Museum,
formerly in the Townley Collection, has, since its acquisition in
1814, been identified as garnet. The suggested date for the
carving in the 4th century bc, the early Hellenistic period,
places it amongst the earliest garnet intaglios (Pl. 2).8 My initial
examination of the stone questioned this identification and,
when subsequently examined in the British Museum Research
Laboratory using Raman spectrography, it was shown to be a
carnelian.
Another misidentification has resulted in a different
inaccuracy. Prior to this study, the British Museum online
catalogue copied the original 19th-century catalogue entry
giving ‘jacinth’ for the scarab illustrated in Pl. 3; the material
was then entered by the programmer as ‘zircon’, another
gemstone altogether which shares the same name.9 An
exception to the problems reviewed above is the catalogue of
Plate 3 Bellerophon riding Pegasus
with spear, Chimera below, within
rope border, Etruscan, 3rd century
BC, garnet, 17 x 14 x c. 7mm.
London, British Museum GR
1865,0712.111
‘Gems of Heaven’ | 11
Adams
Sasanian seal stones by David Bivar, in which the garnets were
identified as almandines by Mavis Bimson in the British
Museum Research Laboratory.10 Hopefully with the publication
of a new garnet database for the British Museum, established
with the cooperation of the Research Laboratory, problems of
nomenclature and misidentification will be a thing of the past.
Hellenistic and Roman period intaglios
Leaving aside some early Chalcolithic beads in India, a few rare
garnets identified as grossulars amongst ancient Near Eastern
seal stones and the occasional employment of garnet in early
Egyptian jewellery,11 a marked escalation in the use of garnet
gems began in the Hellenistic period (c. 323–30 bc, from the
death of Alexander to the fall of the Ptolemaic kingdom in
Egypt). This is commonly attributed to the eastern advances of
the Greek armies under Alexander who invaded India via
Central Asia in 326 bc.12 The establishment of the Graeco-
Bactrian kingdom in the region of modern Tajikistan,
Afghanistan and Pakistan (c. 250–175 bc) and the successor
Indo-Greek kingdom which expanded south of the Hindu Kush
into India (c. 180–10 bc), may also have opened access to
eastern localities, or, perhaps more significantly, contributed to
a taste for gold ornaments set with colourful translucent
gemstones. Whatever the reason, an increase in garnet usage
in the Hellenistic period can be demonstrated by comparing
the relative percentages of garnet intaglios in museum
collections.
For example, despite the issues of identification, a large
collection such as that held in the Greek and Roman
Department at the British Museum (today over 4,000 intaglios)
is statistically significant for the interpretation of patterns of
usage.13 In the 1926 catalogue raisonné of the collection Walters
identified no engraved garnets in what he termed the Archaic
and Finest Greek periods, nor do they appear amongst the Italic
gems of the Roman Republican period.14 As we have seen above
the single intaglio he included amongst the 45 stones in his Late
Greek Period (4th century bc) has now been shown to be
carnelian rather than garnet. From the 3rd century bc Walters
classified two Late Etruscan scarabs as ‘jacinth’. Amongst the
Hellenistic gems, however, he identified 18 garnets and two
‘jacinths’ amongst 64 intaglios. Garnets thus represent almost
30% of the total gemstones from the period in the British
Museum collection. A not dissimilar pattern can be detected in
the large collection in the Antikensammlung in Munich, with
63 garnets amongst 272 Hellenistic stones, or 23.16% of the
total.15 These preliminary surveys suggest that in the
Hellenistic period garnets may have constituted a quarter to a
third of the stones being engraved.16
In the Roman period, garnets are for the most part smaller
in size and represent a smaller percentage of the total. Walters,
for example, lists only 34 garnets amongst a total of 1,475 stones
assembled under the prudent heading ‘Graeco-Roman’ (cat.
nos 1241–2716); these represent 0.024% of the total. The picture
is similar in the Munich collection noted above where a single
garnet catalogued in the Republican and Imperial periods
constitutes 0.0037% of the total of 273 and the 7 Late Roman
garnet stones representing only 0.0146% of the 478 gemstones
from that period.
How should we interpret this information in the light of the
opening of the sea trade routes in the 1st century ad,
12 | ‘Gems of Heaven’
demonstrated by the flow of Roman gold coinage into India
and often cited as evidence for an influx of precious
commodities into the Roman Empire?17 If Pliny is correct, the
Romans certainly valued deep red gemstones (which we
assume to be primarily garnet, see below) and garnet
continued to be used as cabochons on jewellery. Perhaps a
decrease in the demand for garnets coincided with an increase
in the volume of engraved gemstones in other materials.
Microcrystalline quartz varieties (carnelian, sard, jasper,
agate, onyx, etc.) dominate in the Roman period, and their
availability, relative ease of carving, flexibility of size and, of
course, affordability must be taken into account. It is also
interesting that the reduction in the numbers of garnets being
engraved is paralleled by a decrease in the variety of cabochon
garnet stones being produced for jewellery settings, a
phenomenon which may be suggestive of more industrially-
scaled production18 or perhaps bulk imports of pre-prepared
gem ‘blanks’. Pliny’s tastes were those of the Imperial period
and at some level it must be concluded that the matter of taste,
impossible to quantify, must have influenced usage.
Sasanian and Byzantine period intaglios
In the course of the 3rd century ad there is a well-documented
decline in the numbers of intaglios being engraved in the West
(see Plazt-Horster, Marsden and Spier in this volume). Many
factors may have contributed to this, including disruptions to
eastern trade routes in the transition from the Parthian to the
Sasanian periods, breaks in continuity in the gem-working
traditions in workshops across the Empire and the shift to new
forms of sealing practices in the West. Garnets, however, were
commonly engraved in the East in the Sasanian period, from
the 3rd to 5th century ad, and in the Early Byzantine period
from the late 5th to the 6th century ad. The expanding
Sasanian Empire absorbed the eastern provinces of the Roman
Empire from the mid-3rd century ad onwards and the
continuity of the ring stone tradition in the East suggests that
some glyptic workshops in the eastern Empire simply began
engraving new subject matter for new clients.
In these periods garnets again represent a substantial
percentage of the overall intaglios, ranging from 7.7 to 11.7% in
three collections of Sasanian seal stones19 to 10% in a group of
Byzantine-period intaglios recently assembled by Spier.20 He
suggests that the distinctive shape of Byzantine-period garnet
intaglios – all rounded or conical cabochons with a flat table
top – may derive from Sasanian ring stones (Pl. 4).21 It remains
Plate 4 Striding lion to left, front paw on bull-head; Sasanian-style intaglio in
an Early Byzantine ring, L. 23.5mm. London, British Museum, PE 1930,1107.1

Plate 5 Dagger hilt and detached garnet inlay from hilt, Tomb 3, Armaziskhevi,
Republic of Georgia, mid-3rd century AD, L. 10.5cm
to be determined whether these stones were being supplied to
both regions already fashioned to this shape or whether this
particular cut was just the prevailing fashion. In any case,
garnet cabochons of this shape, many with unambiguous
Christian iconography (crosses, doves, etc.), appear to have
been engraved largely in the later 5th and early 6th centuries
ad (see Spier, this volume, pls 32–7, 53). But a considerable
number of unengraved stones of this cut were in circulation as
they are appear on 6th- and 7th-century ad objects found in
Ostrogothic Italy, Visigothic Spain and Anglo-Saxon England.
Some of these were drilled with a ring and dot motif and, in a
few cases, they are incorporated within a cloisonné pattern.22
Sasanian intaglios and garnet cloisonné
Production in the Sasanian period overlaps chronologically
with the rise of garnet cloisonné ornaments from the second
half of the 3rd century ad onwards. Subsequently there was a
flood of small garnets onto the market to satisfy the taste for
this new style. Many hundreds of surviving garnet cloisonné
objects testify to the continued circulation of gem-quality
garnet from the 5th to the 7th centuries ad . With a few notable
exceptions, however, the majority of these ornaments
incorporate small (less than 10mm) thin flat plates of garnet
stone, only just adequate for intaglio engraving, but capable of
being ground into different shapes to fit into increasingly
complex cloisonné cellwork.
The largest and earliest garnet plate inlays for cloisonné are
comparable in size, shape and preparation to flat ring stones
(shape F or bezel D).23 Garnet intaglios of this shape are used as
early as the 1st century ad; by the 3rd to 5th centuries ad
virtually all of the flat ringstone inlays engraved for Sasanian
clients in the Persian Empire were made using garnet. In 1991
The Garnet Millennium
I demonstrated a statistical correlation between the sizes of flat
Sasanian garnet intaglios and the sizes of Roman intaglios (in
all materials), which confirms continuity between these two
traditions.24 However, a statistical correlation can also be made
between Sasanian intaglios and the first stones used in garnet
cloisonné which illuminates the new directions workshops
were taking.25
The transition in lapidary production from the Classical to
Early Medieval periods is evident on 3rd-century ad objects
found in ancient Iberia (the eastern regions of the modern
Republic of Georgia) (Pl. 5.1–2) and in Hatra, Iraq, which
incorporated flat geometric garnet inlays.26 These oval, square
and circular stones with bevelled edges are essentially ring
stone blanks which, rather than being engraved, were used as
inlays. It seems possible that, as the demand for lapidary
engraving declined, a new use was found for existing stocks of
pre-cut ringstones. Garnets were still traded and some
lapidaries accustomed to preparing gem intaglios may then
have turned their skills to the production of inlays alone,
requiring ever smaller stones. It remains to be determined
whether there is any mineralogical connection between the
garnet stones used in these two periods, although it may be
noted that the greatest dimensions of both Sasanian and Early
Medieval stones in one sample taken by the author peak at
around 12mm. This suggests that larger stones, above 15mm,
were scarcer and were selected and reserved for higher-status
ornaments and intaglios for the elite.
Central and South Asian intaglios
One further group of garnets in the British Museum is of
importance to this survey – seal stones found in Central and
South Asia in the territories of ancient Bactria, Gandhāra,
northwestern India and western China. The iconography and
inscriptions on these gems suggests they were produced
initially in a Graeco-Roman tradition for the Kushan Dynasty
which succeeded the Indo-Greeks in Central Asia (see Group I
in Adams, Lüle and Passmore below), and subsequently in a
Sasanian mode for the Hunnic clans who controlled the regions
of modern Afghanistan, Pakistan and northwest India, first
under Sasanian suzerainty and then independently (Group II
below).27
In the most recent catalogue of 208 stones from this region,
assembled by Pierfrancesco Callieri, 23 are garnets, a
proportion of the whole which is consistent with that in the
Sasanian and Byzantine periods.28 As in the Hellenistic period
in the West, a number of these intaglios were carved on gems of
a consistently large size, whose greatest dimensions are larger
than 15mm (cf. cat. nos I.1–5 in Adams, Lüle and Passmore
below). Given their proximity to good gem garnet deposits, it is
probable that these eastern stones, spanning the period from
the 1st or 2nd to the 6th or 7th century ad, utilised local
sources. The scientific examination of these stones will
establish parameters for comparison with rock formations in
Central and South Asia.
It is generally assumed that technical glyptic expertise and
indeed the use of hard gemstones for sealing was introduced
into Bactria by lapidaries from the West, perhaps even prior to
the period of Alexander as finds of Graeco-Persian seal stones
demonstrate that seals were in circulation, if not necessarily in
daily use.29 Garnet was used for beads at the time of the arrival
‘Gems of Heaven’ | 13
Adams
of the Greeks, but not apparently in any quantity, even into the
1st millennium ad. The scarcity of garnet may be attributed in
part to its relative hardness (6.5–7.5 on the MOH scale) which
was a factor in even the earliest times.30 Garnet beads, for
example, were excavated from 4th and 3rd century bc levels at
Taxila in present-day Pakistan, but their numbers were small in
comparison to microcrystalline quartz varieties.31 Similarly,
modern excavations at Sonkh near Mathurā, the Kushan capital
in northwest India, produced 765 beads, 595 of which were
clearly stratified; more than half of these belong to the Kushan
levels and there was no garnet.32 Garnet beads have been
reported at sites in Central and South Asia but always in lesser
quantities than the quartz group.33 At the trading port site of
Arikamedu on the southeast coast of India near modern
Pondicherry, there is clear evidence of garnet bead
manufacture; uncertain stratigraphy in many levels, however,
does not permit a close dating for these finds.34 Whether these
were made using local garnet sources as Peter Francis suggests
or with stones imported from India remains to be investigated.
Similarly scientific analysis of beads found in Oman with
examples from Sri Lanka established a correspondence but
could not rule out an Indian source.35
We are thus confronted with a paradoxical situation in
which we hypothesise that ‘India’ or ‘Bactria’ supplied stones to
the Hellenistic Greeks in the West, yet the local population
appears to have used garnet themselves only rarely, or at least
not extensively until the 1st millennium ad. At present, the
earliest garnet gems from excavated contexts suggest that
garnet may have been used primarily as a decorative inlay. The
garnet inlay in a 1st century bc reliquary deposit in a stupa at
Bhir Mound, the oldest of the excavated ruins at Taxila in
modern Pakistan, was flattened and notched in the manner of
garnet plates found on Kushan and Early Medieval garnet
cloisonné.36 In fact, the use of flat garnet stones used as
jewellery inlays appears simultaneously in the archaeological
record in Bactria and at Pompeii,37 and it seems possible that
not only techniques like foiling behind stones originated in
India, but, as I have discussed elsewhere,38 quite possibly
garnet cloisonné as well.
The earliest garnet seal stones in the East were excavated in
the treasury of the Hellenistic palace in the great Graeco-
Persian city of Aï Khanum. Two of these are large stones, close
to 20–30mm in greatest dimensions.39 All save one are too
fragmentary to assess the iconography. The most complete
intaglio depicts the lower portion of a female bust; Henri
Francfort, the excavator of the piece, compared the engraving
to Ptolemaic portraits, implying it may have been of western
manufacture. Aï Khanum, founded in the 4th century bc, is
generally thought to have fallen to the Yuehzhi from Inner Asia
in the decades after 145 bc. This suggests a general but
certainly not absolute terminus post quem for these finds as
some occupation continued in the city. Sited at the confluence
of the Amu Darya (Oxus) and the Kokcha Rivers on
Afghanistan’s northern border, the city controlled major trade
routes, including those from the mountains of Badakshan, the
source of lapis lazuli, as well as other precious minerals such as
garnet.
Garnet seal stones excavated in 1st century ad contexts
show the influence of Graeco-Roman iconography. A garnet
gem engraved with a profile female head was amongst the
14 | ‘Gems of Heaven’
grave goods in Grave VI in the Saka or Yuehzhi cemetery at
Tillya- tepe, currently dated to the second quarter of the 1st
century ad.40 The intaglio is in the Republican style typical of
the 1st century bc while the gold ring is also probably of this
date. It is uncertain whether the nine garnet seals, found at
Taxila in Pakistan in a jar in a hoard excavated in a 1st-century
ad context, are of Roman or local manufacture.41. Three
represent busts and six depict full-length figures drawn from
Roman mythology (Athena, Nike); these are smaller than the
gems published in the catalogue below and hollowed at the
rear. A gold ring with a garnet intaglio depicting a running
horse was excavated at Valabhipur (Vadodara/Baroda) in
Gujerat; it is certainly evidence of trading contact, either with
the Roman West or with the north of the Indian subcontinent.42
In addition, stones described as ‘intaglio blanks’ were found by
the excavators at the capital city of Anuradhapura; these were
all dated to Period F, c. ad 300–600.43 The greatest dimensions
of these are all less than 10mm and some would not be
adequate for engraving, so it is possible that some were
intended for use as inlays.
Further information is needed to determine whether the
glyptic skill to engrave garnet seal stones developed under the
Graeco-Bactrians, Indo-Greeks or Indo-Parthians but current
evidence suggests that consistent local production of garnet
seals began around the time of the expansion of the Silk Route
trade under the Saka and Yuehzhi clans whose descendents
established the Kushan Dynasty.
Investigations have begun into the mineralogy of the
garnet gems used for seal stones, beads and inlays in the East.
Claude Rapin first suggested that the beads and inlays found at
Aï Khanum were imports from India and a recent analysis
using X-ray fluorescence of mounted garnet cabochons on the
jewellery from Tillya-tepe suggests they may correspond to
sources in Rajasthan.44 The following paper provides the first
gemmological and mineralogical characterisation using
Raman spectroscopy of eastern garnet seal stones; when these
investigations are complete it may someday be possible to link
individual stones to a particular locality.
Following this historical overview of garnet usage across a
broad geographic and spatial continuum, it is useful to review
and compare the primary Greek and Roman texts with the
modern understanding of nomenclature, identification and
sources of the complex mineral called garnet.
Part 2
Ancient nomenclature
The identification of red stones described in early
‘mineralogical’ texts (as opposed to astronomical, magical and
medical texts) with the modern stone we call garnet is covered
in many sources.45 It is useful to remind ourselves, however,
that problems with identification and terminology have been
features of garnet study from the beginning and that the
picture presented by ancient texts is not always as clear-cut as
it is presented.46 While there can be no question that garnet was
recognised as a distinct gemstone, at the same time it is clear
that throughout Classical antiquity and the Medieval period
commentators recognised the difficulty of identifying and
assigning names to the range of red gemstones. Pliny himself
(Nat. Hist. XXXVII.26) expressed it the best in his descriptions

Plate 6 ?Meleager, holding
spears, standing to left before
altar, two hounds on groundline,
Hellenistic, garnet, 19 x 15mm.
London, British Museum, GR
1872, 0604.1215; ex-collection
Ferenc Pulszky, Alessandro
Castellani; from Palermo
of carbuncles: ‘Nothing is harder than the attempt to
distinguish the varieties of this stone…’.
The term which is generally assumed was used for garnet
in the Greek period was anthrax (ανθραξ) (Theophrastus, Peri
Lithōn, 18–19, 33–5), a word which has as its root ‘...θραξ’,
meaning to glow, as in glowing embers. A related word
άνθρακες (anthrakes) was applied to charcoal or soft coal
(lignite) and Theophrastus places the two side by side in the
text to highlight the paradox of two rocks called by similar
names, one of which burns in a common fire and the other of
which glows, but is incombustible in fire.47 One of the Roman
words used for garnet in Pliny (Nat. Hist. XXXVII.25–26),
carbunculus, likewise meant ‘little coal’ but was always used to
describe a hard gemstone. These two terms are generally, but
not universally, accepted to define the gemstone we call
garnet.48 Some of the best support for this is actually negative
evidence – modern archaeogemmological investigations of
Greek and Roman jewellery have turned up very little evidence
of the two other most common red gemstones – ruby and
spinel.41 This corresponds to the written evidence where
specific terms for rubies and spinels are not found until the
Islamic mineralogical literature of the 9th to 11th centuries.49
The terms anthrax and carbunculus survive in many texts
through the centuries, as do the terms alabandicus (Nat. Hist.
XXXVII.33) and hyancinthus (Nat. Hist. XXXVII.30. Another
word, lychnis (Nat. Hist. XXXVII.29), derived from λύχνος
(lamp) and used by Pliny to describe a pale carbunculus, also
appears frequently in Hellenistic lapidary texts.50 Other terms
used by Pliny for red stones such as phlogitis and anthracites
(Nat. Hist. XXXVII.73) do not appear elsewhere, although their
description suggests they may well have described garnets.
Inclusions, ancient and modern
Perhaps of greater interest than the multiplicity of terms and
even disputes which arose over what to call the various red
stones circulating in the Imperial period (cf. Nat. Hist.
XXXVII.25) is what Pliny recorded with regard to the
appearance of these stones to the naked eye. He notes that
‘male’ Carthaginian stones are ‘...sub caelo flammeos, contra
radios solis scintillare...’ (‘flaming under heaven but sparkling
against the rays of the sun’), while other ‘males burned inside
with a star’ (‘...maribus stellam intus ardere...’) (Nat. Hist.
XXXXVII.25). The males are deeper in colour and clearer; the
best are violet amethyst in colour and the next best have a
feathery bright radiance (‘...pinnato fulgore radiantes’). The
The Garnet Millennium
Plate 7 Young satyr with
panther skin and lagobolon
(throwing-stick) across his
shoulders, Hellenistic, garnet,
3rd-2nd century BC, 16.4 x 12.7
x 5.6mm. London, British
Museum, GR 1890,0601.86;
George James Howard, 9th Earl
of Carlisle
passage using the term anthracites (Nat. Hist. XXXVII.73),
seems again to describe the internal inclusions common to
garnet: ‘in phlogitide intus flamma videtur ardere, quae non
exeat, in anthracitide scinillae discurrere’ (‘In the phlogitis, a
flame seems to burn within which is not released; in the
anthracitis sparks run in different directions’).51
Such passages demonstrate that not only did the Romans
observe the internal appearance of garnet but also that they
were aware of the diagnostic potential of these phenomena for
classification. The modern identification of crystal inclusions
in garnet is an advanced science relying upon microscopy (see
Lüle and Thoresen this volume) and examination using
scientific instruments (Adams, Lüle and Passmore below).52 In
some cases individual crystal inclusions within garnets can be
mineralogically identified (i.e., rutile, zircon, apatite, ilmenite,
etc., to name the most common ones) while in others the effect
of these inclusions are described using terms such as silk, veils
and fingerprints. The intaglios in Pls 6–7 give some indication
of the kinds of inclusions the Greeks and Romans would have
been able to see with the naked eye. A prominent crystal to the
side of the head of Meleager on Pl. 6, for example, can easily
been seen when held against the light; it has been identified
using Raman spectroscopy as zircon (ZrSiO4). Of course, then,
as now, it would have been helpful to be myopic, which many
lapidaries almost certainly were; specialists and connoisseurs
may also have used crystals for magnification.53
In addition to being densely filled with ‘silk’ and with
prismatic inclusions (perhaps apatite), the stone shown in Pl. 7
exhibits asterism (an optical star) in two zones on the proper
left side, at the bottom and the top. The author has also
identified a large asteriated garnet set on a gold mount of the
later 5th century ad in the Early Medieval collections at the
Morgan Library and Museum in New York (Pl. 8).54 The optical
Plate 8 One of a pair of quatrefoil
mounts, Central Caucasus, second
half of 5th century AD, gold, garnet,
silver, H: 38mm; W: 36mm. New
York, The Pierpont Morgan Library,
2011.23:33b; Thaw Collection
‘Gems of Heaven’ | 15
Adams
effect of asterism in garnets has recently been scientifically
proven to be the result of bundles of elongated inclusions
(generally rutile) often in combination with elongated voids
which cause streaking of light; these are orientated along the
axes of the crystal, resulting in diagonal and right-angled
crossings.56 There are very few known sources of asteriated
garnets worldwide and only three of these could have been
known in the ancient world – India, Sri Lanka and Tanzania in
central East Africa.57
Ancient perceptions of sources
Discussions of the sources of garnet cited in the ancient
literature accompany most garnet studies; these are generally
taken at face value and form the basis of modern scientific
study. However, if classical authors did not apply consistent
terminology to the gemstones they handled, can we assume
that the sources they cite are necessarily accurate? The ancient
confusion between sites which were sources as opposed to
suppliers of stones through trade is well known, but a brief
review of the source information remains of interest.
In the late 4th century bc Theophrastus’ account of sources
for anthrax point specifically to Asia Minor (Miletus in western
Turkey) and Africa (Peri Lithōn, 18–19).58 He states that anthrax
is brought from Carthage and Massalia, the latter often taken
to be Marseilles in southern France. As Theophrastus cites only
African sources for gems and the term appears again in his
discussion of anthrakion as ‘the country around Massalia’ (Peri
Lithōn, 34–5), on balance this is probably a reference to
Masaesylia in the west of the Berber kingdom of Numidia
(between Algeria and Tunisia in Africa) (cf. Strabo Geographica
XVII.iii.6). Strabo (63–4 bc–c. ad 24; Geographica XVII.iii.19)
also notes that the precious stones are brought to Carthage
from the land of the Garamantes,59 the Saharan desert in the
province of Fazzān in modern southwestern Libya, still
occupied by oasis farmers and pastoral nomads.
By the later 1st century ad Pliny (Nat. Hist. XXXVII.25)
makes it clear that India had become the first of the two
primary sources of carbunculi. The Garamantic stones follow
the Indian and ‘to these are added the Æthiopian and the
Alabandic stones, the latter of which are found at Orthosia in
Caria, but are cut and polished at Alabanda’. Pliny clearly
knows nothing specific of the Indian sources and the
attribution to Caria seems incorrect on the basis of modern
knowledge but, thanks to an earlier Roman expedition, he
seems to have a closer knowledge of the African sources. In
Book V (Nat. Hist. V.35–38) he gives an account of the spoils,
proudly labelled and displayed, which were brought back from
Libya in the campaign of the African proconsul, L. Cornelius
Balbus, against the Garamantes in 19 bc; the final one in the list
bears the name of ‘Mount Gyri, which was preceded by an
inscription stating that this was the place where precious
stones were produced’.
The geologist Jean-Philippe Lefranc has suggested that
Mount Gyri should be identified with the rock formations at
Jabal al Hasāwnah or Jabal Fazzān.60 This is in south central
Libya, south of Tripoli, west of the modern Jabal as Sawdā,
which he identifies as the ancient Mons Niger. Following his
expeditions into the Precambrian outcrops in the southern
region of the massif, he reported that granites and schists were
traversed by pegmatite veins rich in tourmaline, mica and
16 | ‘Gems of Heaven’
garnets; further north hydrothermal zones were present at the
contact between eroded Cambrian-Ordovian sandstones and
volcanic phonolites. He also saw caved-in sumps which he took
to represent either ancient evidence of mining or water
exploration. Thus, although modern archaeologists have
proposed Pliny’s carbuncles were carnelian on the basis of
field-walking in this region of Libya which produced carnelian
beads in different states of preparation,61 it seems to this author
that we should also accept the likelihood of Garamantic
sources of garnet. It is, after all, clear that Pliny and his
contemporaries made a distinction between carnelian and
garnet.
Pliny’s only reference to Massalia is inserted later within a
list copying Theophrastus’ list of stones called anthrax; he
indicates no current knowledge on this source, but may have
assumed this referred to Massalia (Marseille) as he followed
this with a report of carbunculi, extracted with great difficulty
from the ‘argillaceous’ soil at Olispo (Lisbon) (Nat. Hist.
XXXVII.25). Of the alpine and central European sources of
garnet he knows or mentions nothing.
In summary, essentially Pliny records what modern
geological knowledge and research into garnet cloisonné
confirms today – that garnets can be found in India, Africa,
Asia Minor and Europe. Thus the broad orientation of the
ancient texts on mineralogical sources cannot be completely
rejected and, as with their discriminating perception of the
qualities of stones, still offers valuable information. Many of
these, in particular the North and East African sources, remain
to be explored and compared with ancient garnets.
Modern scientific source studies
The original impetus behind the scientific study of ancient
garnet was the desire to determine whether western
(Bohemian) or eastern (Indian and Sri Lankan) deposits
supplied garnets during the centuries when garnet cloisonné
flourished in the West.62 Thus the ever-expanding scientific
literature on the subject has to date focussed on the red garnet
inlaid into gold cloisonné used for personal ornaments and
made in Early Medieval Europe, Scandinavia and Anglo-Saxon
England.63 From a geological and scientific perspective, a more
daunting task could hardly have been imagined. Garnet
cloisonné uses small flat plates of garnet cut to the shape of the
cloisons, then cold set using backing foils and pastes. Intact
objects cannot be deconstructed and standard gemmological
methods often cannot be used successfully to determine the
species of the stone. Moreover, garnet cloisonné marks the
final phase in a millennium of steady use of garnet as a
decorative inlay and sealing stone. It is to be expected that by
this time stones from many different sources were in
circulation and that damaged intaglios, for example, might be
ground down and re-used. Such evidence need not uncover
Early Medieval sourcing patterns alone, but also re-use of an
existing pool of gemstones.
At the present time the use of different testing methods and
equipment by different researchers over a period of almost 50
years has produced scientific results which, in many cases, are
not, strictly speaking, comparable. Nonetheless, remarkable
results have been achieved at the Centre de Recherche et de
Restauration des musées de France (C2RMF) in Paris using
advanced non-destructive techniques capable of detecting

trace elements in the chemical composition of the stones.64
Museum specimens of garnet from India and Sri Lanka have
been shown to be possible candidates for particular clusters of
Early Medieval period garnet plates with a high almandine
component and a broader pyrope-almandine composition,
respectively (Pl. 1).65 Subsequent investigations at C2RMF and
in laboratories in Germany have revealed that the bulk
chemistry of some igneous pyropes corresponds to specimens
from the České Středohoří mountains in the central Czech
Republic while others can be matched by garnets from Monte
Suimo, Portugal.66 With only a few exceptions, examples of the
former appear on 7th and 8th-century ad ornaments, i.e. at the
very end or indeed, after, the garnet millennium defined
above. Gilg, Gast and Calligaro have also recently pointed out
that the chemistry of some chromium-poor, titanium-rich
pyropes are closer to gemstones from the Jos and Biu plateaus
in northern Nigeria than to other comparable sources such as
Scotland (the Elie Ness ‘rubies’), Mt. Carmel in Israel and
Shavaryn Tsaram in Mongolia.67
To date the intensive laboratory research in characterising
mounted garnet plates in an effort to pinpoint their sources has
far outpaced fieldwork to explore garnet localities. Calligaro’s
identification of radioactive metamict zircon inclusions in
some garnet plates is definitive for Precambrian paragenesis,
but this only narrows the picture to include the vast areas of
the Indian shield (roughly two thirds of the Indian sub-
continent), as well as the Precambrian rocks of Africa.68 But for
the first time fieldwork has been initiated to identify further
potential rock sources of garnet gems. Some calcium-rich
almandines used in the Early Medieval period have been
shown to be consistent with rock types in Scandinavia,69
although this is not clear evidence of their ever having been
utilised as Birgit Arrhenius has hypothesised. In addition the
discovery by Çiğdem Lüle of small, transparent and nearly
inclusion free garnets in the rocks of the Alabanda region of
Turkey establishes the possibility that gem quality almandine
garnets existed at this locality which might have been used
during the Hellenistic and Roman periods.70
The complexity of garnet chemistry, the vastness and
variability of many garnet deposits and the difficulty of proper
field investigation in key regions will continue to present
obstacles to the goal of wholly verifiable matching of garnet
stones to specific localities. One further critical factor must be
taken into account – the possibility, if not probability, that key
deposits were exhausted or ‘played-out’ in antiquity. We must
be prepared, given the large numbers of potential sources in
Africa, India, Afghanistan and Pakistan (all regions with
ongoing political conflicts), that it may take future generations
to carry out comprehensive work in this field.
In contrast to the burgeoning flow of garnet research in the
Early Medieval period, scientific research on garnets from the
classical world awaits publication. Some 30 garnet intaglios
from the Greek and Roman periods in museum and private
collections in America and Europe were recorded and
organised in colour groupings by Lisbet Thoresen in the 1990s
in association with a Getty Museum project. The chemistry of
nine of these held in the collection of the Getty Museum in Los
Angeles was investigated using electron microprobe analysis;
the results showed some correlation with the mineralogical
findings in Early Medieval garnet studies.71
The Garnet Millennium
Research into garnet seal stones of the Greek and Roman
and Sasanian periods in the British Museum is ongoing and the
following paper presents the results of initial gemmological
and mineralogical examination of seal stones from Central and
South Asia. This has defined groups of stones with
characteristic inclusions and consistent chemistry which allow
us to investigate issues of workshop production which both
support and transcend stylistic identifications. This suggests
that scientific investigation has a much broader role to play
than simply to allocate stones to sources and in fact offers a
new way forward in glyptic studies.
Part 3
The significance of size
Many topics for future exploration remain to be explored, but
one which is of particular interest to the GCEM project
concerns the size of ancient garnet gemstones. The Romans
were aware of extremely large garnet stones – of dimensions
large enough to hollow out a vessel,72 but, as noted above,
garnet seal stones above 15mm are relatively unusual in the
ancient world and modern gem quality garnet specimens of
sizes over 10mm remain relatively rare. The chemistry of
different species of garnet is, in some instances, closely linked
to the size of the stones themselves. The growth of any mineral
crystal is determined by time and space. Garnets, like many
other crystals, grow very slowly; recent geochemical
calculations suggest that garnet crystals 10mm wide could take
anywhere from 8–10 to 32 million years to form.73 Crystals
grow from seeds or nuclei. When the basic nutrients (elements)
are abundant and the temperature and pressure conditions are
favourable, crystallisation starts from the nucleus. Garnets
have what is known as high nucleation energy – i.e. they tend
to form many nuclei on which many small crystals grow.
Therefore, although it is considered to be a common mineral, it
is not very common to see very large crystals of gem quality
garnets. The aspect of size therefore is a factor in the
characterisation of garnets and their potential source localities.
Many significant deposits of gem-quality garnet crystals
whose chemical composition falls in the pyrope range are
relatively small. In the famous Bohemian pyrope mines, for
example, for every 2 tons of garnet mined in the 19th century,
only one stone capable of producing a five carat cut was
reported; in every 220 pounds a stone of 2.5 carats occurred.74
Carat is a measurement of weight (1 carat = 1/5 of a gram); this
can be affected by density, so it is not always strictly related to
size, but to put this information in the perspective of the
dimensions used in this study: a five carat stone would
correspond roughly to dimensions of less than 10mm. Thus,
although the bulk chemistry of some Early Medieval garnet
plates can be shown to be similar to the chemistry of Bohemian
garnets, it is unlikely that sources of pyrope garnet in Bohemia
supplied the larger stones of around 15 to 21mm+ used in the
Hellenistic period and in 5th-century Early Medieval garnet
cloisonné. Along these same lines, it has recently been shown
that what was deemed to be the largest known ‘pyrope’, set on
a medallion of the order of the Golden Fleece housed in
Dresden (measuring 35 x 27 x 15mm and weighing 46.75 carats)
was in fact an almandine, whose source is as yet
undetermined.75 In the same study the largest Bohemian
‘Gems of Heaven’ | 17
Adams
pyrope garnet in the garnet museum of Třebenice, Czech
Republic, measured 7 x 5mm, as would be expected.
Garnets rich in the almandine molecule are typically
metamorphic in origin, stable at high pressures compatible
with depths greater than 20km in the earth. They also occur in
lower crustal plutonic igneous rocks. Almandine-rich crystals
which formed in mantle rocks such as metamorphosed
granulites are capable of growing to very large sizes. The gem-
bearing pegmatites of northeast Afghanistan, Pakistan and
northwestern India are possible sources of some larger garnet
gemstones.76 Large almandine garnets, often with a high
spessartine component, are also reported from sites such as the
Umba Valley in Tanzania.77
Size is of particular significance to three groups of garnets
covered in the discussion above: 1) Hellenistic period intaglios,
2) Central and South Asian intaglios and 3) cabochon stones
and some plate garnets set on cloisonné made in the 4th and
5th centuries ad in the West. The second group is discussed
further in Adams, Lüle and Passmore below and this author has
explored elsewhere the importance of the size of garnets with
regard to the large garnet plates and cabochons used in the
Early Medieval period.78 Researchers in Paris have also pointed
to the potential significance of gemstones of larger size in the
scientific characterisation of their Type I and II garnets,79 while
other recent work has reiterated the small size (6 x 6mm) of
many Bohemian pyrope stones.80 A brief look at the Hellenistic
group outlines some possible directions for investigation.
The British Museum holdings include a number of
Hellenistic intaglios of long oval shape set on massive gold
rings with stepped bezels, a Ptolemaic form popular in the
second half of the 3rd and 2nd century bc (Pls 9a–c).81 Three
examples of these in the British Museum have been shown by
Raman spectroscopy to be closely similar in composition.82
While none of these are as large as the magnificent Ptolemaic
garnet intaglio measuring 39mm in length published by
Gertrud Platz-Horster,83 garnet seal stones of these proportions
are relatively rare in the archaeological record.
One question to consider is how to interpret the consistent
length (generally c. 20 to 23mm) and narrow width (c. 6.5 to
12mm) of these particular garnets. It could be argued for
example that such stones were cut to particular sizes because
the shape was fashionable or because it suited a ring setting.
But if these distinctive stones prove to be mineralogically
18 | ‘Gems of Heaven’
Plates 9a-c Muse standing to left,
reading from scroll, column behind,
garnet, gold, 21 x 10mm (GR 1772,
0314.1); Apollo or hermaphrodite with
thrysos, leaning on column, 20.2 x
6.5mm (GR 1872,0604.1191);
Dionysius holding lyre, leaning on
column, 20 x 8.3mm (GR 1872,
0604.1183). All London, British
Museum
consistent, then their shape may have been determined in part
by the available rough garnet, which in turn might prove to be
relevant to their source locality. Garnet specimens assembled
by museums and modern collectors tend to be perfectly formed
crystals, and, remarkably, natural garnet crystals were found
at Aï Khanum,84 but many ancient stones will have been
alluvially sourced or picked up as detritus.85 Plates 10a–c
shows some characteristic examples of rough garnet from the
Natural History Museum in London; these are from Africa, but
specimens of red garnet from other localities around the world
will be visually similar. Weathered or washed out of their host
rocks and roughly tumbled into shape, any consistency in size
they present will to some extent reflect their genesis.
Long and much narrower garnets than those shown in Pl. 9
appear amongst the jewellery inlays used in the Hellenistic
period and slender cabochon ‘bars’ are a distinguishing feature
of high-status cloisonné sword guards made in the 5th century
ad in workshops around the Black Sea. Many of the latter
stones consistently measure around 23mm in length. Do these
stones from different periods point towards access to a
particular gem source or simply represent the revival of a
specific gem cut? Future mineralogical investigations allowing
the comparison of garnets across many centuries may hold
some answers to such questions.
In fact an example of garnet sources influencing workshop
production may be adduced in the Early Medieval period. As
Bimson proposed many years ago with respect to cloisonné
plates, there is the possibility that some of the flatter garnet
intaglios derived from schistose garnet porphyroblasts in
metamorphosed rocks, predisposed to shatter into sheets with
parallel sides.86 Although this author has demonstrated that
some of the first garnet cloisonné inlaying in the West was
closely associated with ring stone production (see above, p. 13),
there is no reason why access to a particular gem source,
particularly if it was inexpensive, could not have stimulated the
expansion in the production of garnet cloisonné ornaments.
Review and conclusions
Having surveyed garnet intaglios and examined some aspects
of identification, sourcing and size, I would like to attempt a
chronological review of garnet production with an eye to the
East/West trade. Of course, any summary of the complex
patterns of usage across the garnet millennium can only be

Plates 10a–c Rough garnet
specimens from East Africa. Left:
Umba River Valley Area, N-E
Tanzania, gift of Dr J. Saul
1979.36915; centre: Madagascar, gift
of E.H. Florens, 1920,180; right:
Lisenfeld, Tanzania, gift of Fred
Marquordt, 1905,59–60. All London,
Natural History Museum
preliminary and, given the nature of the archaeological record,
imperfect in detail. Two primary strands of evidence have been
traced here: that of engraved seal stones and that of inlaying
with flat plates. This paper has attempted to synchronise some
of the evidence from these two strands. At present scientific
characterisation of ancient garnets with the goal of
provenancing has concentrated on evidence from the later end
of the cloisonné strand in the West. Given the probability that
many garnets came from eastern sources, it is essential to take
into account evidence from both the East and the West, as my
previous outline of the development of garnet cloisonné
attempted to do.87 The pilot phase of the GCEM project
published in the following paper begins the scientific
exploration of the strand of garnet seal stones from the East
with gemstones which are very likely to have been locally
sourced. Although we are still very far from scientifically
matching stones to localities, modelling the evidence for these
two strands together will provide a deeper foundation for
understanding patterns of trade and usage.
At present there is no evidence that garnet was used
consistently in the West before the end of the 4th or beginning
of the 3rd century bc, a generation or more after the death of
Alexander the Great (356–323 bc). A garnet-set ring found with
a coin of his successor Lysimachus (r. 306–281 bc )88 and a
superb garnet cameo of a satyr on a brooch excavated in a
barrow on the Taman peninsula dated to the first quarter of the
3rd century bc are amongst the earliest dated finds.89
Theophrastus’ treatise on stones is generally taken to have
been written near the end of the 4th century bc; he died c. 287
bc at the age of 85, so lived in the period when garnet was
beginning to be carved. The high value he gives to a single
anthrax stone (40 staters) underlines its rarity at this time.
Platz-Horster has recently argued that one exceptionally large
seal stone may date to the first half of the 3rd century bc (p. 18
above),9 and garnet intaglios first appear consistently on rings
of 3rd century bc type (Pl. 9.1).91 Much of the dating of
Hellenistic intaglios, however, remains internal and stylistic
and a comprehensive analysis of seal stones in this period
firmly dated by archaeological context or coin associations
remains to be undertaken.
Despite Alexander’s campaigns in the East, Theophrastus
was seemingly unaware of Bactrian/Indian sources of garnet
although his ‘sources’ for anthrax such as Miletus may
The Garnet Millennium
correspond to overland trade with the East. The appearance of
Carthage and probably Masaesylia in Numidia as gemstone
sources in his text and those of Strabo and Pliny suggest that
for this period garnet localities in Africa should be investigated
alongside those from Bactria and northern India. In the East
fragmentary garnet seal stones with classicising iconography
found at Aï Khanum in Bactria may date to the 3rd or first half
of the 2nd century bc, but it is not yet clear whether these were
imported from the West or produced using local sources.
The increased numbers of garnet seal stones produced in
the course of the Hellenistic period coincide with the rise of the
polychrome style of decoration on jewellery, combining
garnets with other coloured stones, pearls and inlays;
excavated finds in this style from the Bosphorus and Kuban
Basin region north-east of the Black Sea date to the 2nd and 1st
centuries bc.92 The concentration of finds in this region,
although biased by funerary practices, suggests that some
garnets may have been supplied via the northernmost trade
routes running from Transoxiana along the northern coast of
the Caspian Sea and across the Lower Volga region to the
northern Caucasus. Coin finds from the Black Sea suggests that
activity along this somewhat elusive trade route increased at
the end of the 2nd century bc after the fall of the Graeco-
Bactrian kingdom.93 This coincides with the movements of
Sarmatian clans from Central Asia into southern Russia and of
the nomadic confederations we know as the Saka and Yuehzhi
into Central Asia.
Pliny’s statement vis-à-vis Roman trends that ‘it was the
victory of Pompey over Mithradates (VI) that made fashion
veer to pearls and gemstones’ (Nat. Hist. XXXVII.6) fits well
with this evidence. The Roman triumph over this ruler of the
Kingdom of Pontus in the 60s bc extended the Empire’s Asian
protectorates to the Black Sea and the Caucasus. Much trade
with the East, however, remained funnelled through the
sprawling Parthian Empire and further attempts to exert
control over the Caspian trade routes, the ‘Way of the Golden
Fleece’, lay at the heart of the disputes between Rome and
Parthia over the region of Armenia.
Following the incorporation of Egypt into the Roman
Empire under Augustus, maritime trade along the seaways of
the India Ocean expanded dramatically, as the Greek mariner’s
handbook, the Periplus Maris Erythraei, written in Egypt in the
mid-1st century ad, attests.94 Finds of Roman or Roman-
‘Gems of Heaven’ | 19
Adams
influenced garnet seal stones in 1st century ad archaeological
contexts in Bactria at Tillya-tepe and in Gandhāra in the Indo-
Parthian levels at Taxila, may also document overland contact
with the West; again it is not yet clear whether these finds
represent imports or, as Callieri has suggested, evidence for
western gem engravers arriving in the East during the period
of Parthian rule.95 Further evidence of cross-cultural contact
appears in the form of garnet stones polished flat on both sides,
used both for intaglios and as jewellery inlays in the 1st century
ad in both the East and West.96
Somewhat surprisingly, in the period from the 1st to the
2nd and early 3rd century ad in the Roman West, preliminary
surveys suggest that overall less garnet was used for seal stones
than in the Hellenistic period. Jewellery inlays and beads in
garnet were still popular but in a narrower range of uniform
shapes, perhaps imported pre-cut. Expanded trade along the
sea routes to India and Sri Lanka may have supplied such gems,
but garnet seals were apparently not favoured by as many
clients in the Roman period, either because the stone’s
hardness made it more expensive to engrave or possibly
because of some perceived eastern or ‘barbarian’ flavour.
By the later 1st and certainly the 2nd century ad in the East
garnet seal stones were being produced locally with
iconography derived from Graeco-Roman models. Garnet seals
from the period of the Great Kushans (Group I in Adams, Lüle
and Passmore below) were presumably sourced from localities
with reliable access to the Kabul and/or Swat valleys. Although
Kaniška I extended the Kushan Empire into India, the absence
of garnet in excavations near Mathurā, the southern capital of
the Kushans, may suggest that seal-cutting expertise was
concentrated in workshops in Bactria and/or Gandhāra. Garnet
was undoubtedly in use across the Indian subcontinent as
jewellery inlays and beads and it seems that industrialised
garnet bead-making at the site of Arikamedu on the east coast
of India began in this period.
With the exception of the Indo-Parthian finds noted above,
the evidence for garnet engraving (or indeed hardstone
engraving at all) from the Parthian period is sparse.97 Garnet
jewellery inlays are abundant, however, and from the mid-2nd
to the mid-3rd century ad examples of cellwork garnet
cloisonné appear in the archaeological record on the eastern
borders of the Roman Empire, notably in trade route cities
taken from the Parthians by the Sasanians and in
Transcaucasia (Iberia and Armenia) where eastern Roman/
Parthian and Sasanian influence remained strong.98 Garnet
cloisonné can also be documented in Gandhāra and Bactria in
the 2nd to early 3rd century ad.99 Although cellwork cloisonné
in the Hellenistic tradition continued uninterrupted into the
first millennium,100 at present the surviving examples of such
work incorporate only glass and softer stone inlays. It is
possible that both the exemplars and expertise to produce
intricately-shaped interlocking garnet inlays were introduced
from the East to the West in the later Parthian period.
One outcome of the decline in seal stone engraving in the
West in the course of the 3rd century ad was the use of
unengraved seal stone blanks as inlays. These offer evidence of
a transitional period in lapidary workshops between the
production of intaglios and the first purpose-cut plates for
garnet cloisonné. Some lapidaries working in the eastern
Mediterranean found new clients amongst Sasanian patrons
20 | ‘Gems of Heaven’
and throughout the 4th and 5th centuries ad garnet enjoyed a
renewed popularity for engraved seals set in rings in the
classical manner. Garnets from Sasanian-controlled territories
in Bactria, Gandhāra and northern India, as well as Sri Lanka
and East Africa are all potential localities for these gems.
Sasanian seal stones in the British Museum have already been
identified as almandine; the GCEM project plans to undertake
a more complete mineralogical characterisation of the garnets
held in the Department of the Middle East.
In the 4th to 5th centuries ad in the East exceptionally fine
garnet seal stones were carved for the Hunnic clans ruling (in
many cases under Sasanian authority) in Central and South
Asia. The subject matter of these is no longer classicising, but
reflects Indian and, in a series of distinctive portrait gems,
Sasanian influence. The seal stones defined by the GCEM
project as Group II below are gemmologically and
mineralogically distinct from the earlier Group I stones,
suggesting different sources were being drawn upon. Many of
these are stones of violet hue, a feature of the gemstones
described in modern times from the mines at Sarwar and
Keloria in the garnet-mica schist deposits in the Aravalli
formations between Jaipur and Udaipur in Rajasthan.101 At
Anuradhapura in Sri Lanka, evidence of a flourishing trade in
garnet is provided by debitage from garnet bead working
concentrated in levels dating from 300–600 ad.
The final fascination with garnet in the West begins in the
second half of the 4th century ad with the arrival in southern
Russia of the Huns, steppe nomads with cultural links across
Inner Asia. In the 5th century ad garnet cabochons of good
colour and large size were available to workshops in the Pontus
and Danubian regions, together with a steady supply of small,
relatively inexpensive stones to supply the taste for garnet
inlaid ornaments favoured initially by the Hunnic/Iranian/
Germanic confederacies and eventually by the eastern Roman
military establishment as well. With the northern trade routes
under the broad control of Hunnic tribes, deposits in Bactria,
Gandhāra and northern India must be considered likely
sources, particularly for the large garnet plates and cabochons
still available in the decades after the death of Attila in ad 454.
Throughout the 5th, 6th and into the 7th century ad in the East
garnet seal stones of large size were produced for the
confederacies of Hunnic/Iranian/Turkic groups ruling in
Central Asia and northern India (Group II below).
In the second half of the 5th and early 6th centuries ad
there was a consistent supply of garnet gems, cut to a constant
shape, for engraving in Byzantine workshops. Mid-6th century
ad sources such as the Christiana Topographia, written by the
Greek seaman and monk, Cosmas Indicopleustes, record that
garnet (hyacinth) could be found at trading emporiums in Sri
Lanka.102 Few Sasanian-period seal stones in garnet can be
stylistically attributed to the 6th century ad and in the course
of the 6th century ad in the West, although cloisonné inlaying
flourished, garnet plates and cabochons of sizes over 10mm
diminish. Preliminary scientific analysis of garnet cloisonné
plates suggests that the stones being used in Europe in the late
5th and 6th centuries ad may have came from many sources,
including northern India (Rajasthan), eastern India (Orissa)
and Sri Lanka; by the 7th and 8th centuries ad European
sources of garnet in Bohemia, Portugal and possibly even
Scandinavia were increasingly tapped in order to satisfy the

continuing demand of the early Germanic kingdoms for garnet
cloisonné ornaments.
Much work remains to be done to align the archaeological
information with the few ancient texts on which we base many
of our assumptions and theories regarding sources and trading
patterns. Many garnet deposits remain untested and the
definitive assignment of garnet plates and seal stones to precise
localities is not yet possible. The creation of an ancient
database of garnet seal stones, combined with existing
research on garnet plates, may someday allow comparison of
stones across centuries in order to investigate issues of
sourcing, supply and demand in a truly meaningful way. Until
then it remains uncertain how closely we can link local
patterns of use to historical events which may have affected
trade.103
A lucrative trade in small and valuable luxuries may
continue under different political masters, but only if the
demand warrants it. It seems possible that, however important
stones from European and African sources may prove to be in
the overall picture, to a large extent the story of garnets in the
West may be that of the penetration of aesthetic preferences
resulting from contact with the East: a garnet millennium
initiated by Greek adventures in the Hindu Kush, expanded
under the Saka/Sarmatian/Parthian empires stretching across
Inner and Central Asia, Persia and southern Russia and
culminating with the integration of the Huns into the Sasanian
and eastern Roman Empires. From this perspective, the
importance of the northern trade routes should not be
underestimated.
Addendum
After this paper was completed, I learned from Albert Gilg that
he has also initiated studies of garnet intaglios in Bavarian
collections, beginning with inclusion characterisation and
Raman spectroscopy. Furthermore, Gilg and his PhD student,
Norbert Gast have completed an initial phase of fieldwork in
Rajasthan, which will be a major step to establishing a sound
basis for future comparative scientific studies.
Acknowledgements
The pilot phase of the GCEM project could not have been accomplished
without the generous input of numerous colleagues, foremost among
them Chris Entwistle who has offered support and liaised with the
curatorial departments in the British Museum.
Michael Willis in the Department of Asia made it possible for me to
begin this study of the Central and South Asian garnet intaglios and
Clarissa von Spee arranged for garnet seals from the Stein collection to
be gathered together and taken off display. I am grateful to Lesley
Fitton and Ian Jenkins for allowing access to the gems in the
Department of Greece and Rome and to Alex Reid and the other
museum assistants in that department who patiently opened many
dozens of drawers, helping me bring unpublished stones to light. Liz
Errington in Coins and Medals assisted with the Charles Masson
collection gems. Emma Passmore in the Research Laboratory at the
British Museum performed the Raman spectroscopy on the gems and
interpreted the results. I would also like to thank Catherine Higgitt and
Janet Ambers in the Research Laboratory at the British Museum.
At the Engraved Gemstones conference I met Dr. Çigdem Lüle
whose doctorate in mineralogy in Ankara investigated garnet sources
in the Menderes Massif in Turkey. As a qualified gemmologist she
undertook the initial phase of gemmological examination, and also
amended some of my mineralogical passages in the text. After the
conference Lisbet Thorsen generously shared her research on garnet
seal stones with me.
The Garnet Millennium
Harry Falk at the Institut für indische Philologie und
Kunstgeschichte, Freie Universitat, Berlin kindly translated the Indian
scripts on unpublished gemstones; Nicholas Sims-Williams updated
his transcriptions of published gems with Graeco-Bactrian inscriptions
and read the one stone inscribed in Middle Persian. Alan Hart at the
Natural History Museum in London gathered together many
specimens of rough and cut garnet for our preliminary enquiries.
Csanád Bálint offered his thoughts on an early version of this paper
and Angela Evans, as always, read my texts and patiently listened to
my ideas.
Notes
1 The classic source on garnet mineralogy remains W.A. Deer,
R.A. Howie and J. Zussman, Rock-Forming Minerals, Vol. 1A
Orthosilicates, London and New York, 1982 (2nd edn), 467–698.
Introductory overviews in English include: J.D. Rouse, Garnet,
London, 1987, and, most recently H.A. Gilg, D. Hile, S. Liebetrau,
P. Modreski, G. Neumeier and G. Staebler (eds), Garnet, Great Balls
of Fire, extralapis 9, East Hampton, Conn., 2008.
2 H.B. Walters, Catalogue of Engraved Gems & Cameos, Greek,
Etruscan & Roman in the British Museum, London, 1926, 162, no.
1462; also 202, no. 1905 (as sard). Pl. 6 in this catalogue
(GR1872,0604.1215) was still entered in the online Merlin records
as amethyst.
3 Amethystine garnet: Walters (n. 2), 189, no. 1763; 210, no. 1997 and
213, no. 2038; the term reappears in modern catalogues, cf.
M. Henig with D. Scarisbrick and M. Whiting, Classical Gems,
Ancient and Modern Intaglios and Cameos in the Fitzwilliam
Museum, Cambridge, Cambridge, 1994, 48, no. 78. The distinction
between garnet and ‘hyacinth’ also appears in A. Furtwängler, Die
Antike Gemmen, Geschichte der Steinschneidekunst, 3 vols, Berlin,
1900, III, 150. It is my impression that Walters used the term
‘hyacinthine’ to refer to clear, pale garnets with a purplish cast, but
he also employed the term ‘jacinth’ for similar stones, see n. 9
below. The stone labelled ‘hyacinthine garnet’ (Walters [n. 2], 322,
no. 3424; GR 1799,0521.54) was identified by John Rouse ([n. 1],
129, fig. 5) as a rather rare grossular (var. hessonite).
4 Carbuncle: Walters (n. 2), 383 (nos 4070, 4080, both re-used in later
Medieval settings) = O.M. Dalton, Franks Bequest. Catalogue of the
Finger Rings. Early Christian, Byzantine, Teutonic, Medieval and
Later, Bequeathed by Sir Augustus Wollaston Franks, London, 1912,
nos 658 and 222. The term also had biblical authority, cf. the King
James’ translation of the Hebrew bareketh (‘flashing’ stone) as
carbuncle (Exodus 28:17 and 39:10). In the United Kingdom in
particular, the term carbuncle is still used in the trade to refer to a
large cabochon in red garnet; the British Museum has garnets on
display in the Hotung gallery identified as carbuncles. Jacinth:
Walters (n. 2), 91, no. 741; 92, no. 750; 133, no. 1165; 134, no. 1179; no.
1854; 157, no. 1400; 160, no. 1438; 175, no. 1608; 254, no. 2576.
5 Inter alia: hyacinth: R. Göbl, Dokumente zur Geschichte der
Iranischen Hunnen in Baktrien und Indien, Wiesbaden, 1967, 222,
G3; jacinth: Sir A. Cunningham, ‘Coins of the Kushans or Great
Yeuh–ti’, Numismatic Chronicle, 3rd series, vol. xii (1892), 40–82
and 98–159, pl. XXII.18, and 116; J. Marshall, Taxila, vols I–III,
Archaeological Survey of India, Cambridge, 1951, vol. I, 160; vol. II,
650, vol. III, pl. 207a–i; O. Neverov, Antique Intaglios in the
Hermitage Collection, Leningrad, 1976, nos 58, 66–8; J. Boardman
and M.-L. Vollenweider, Catalogue of the Engraved Gems and Finger
Rings I. Greek and Etruscan, Ashmolean Museum, Oxford, 1978, 82,
no. 290; 84–5, no. 296; 108, no. 370, 111–2, no. 38; agate: Göbl ibid.,
223–4, G5, pl. 85. The 19th-century tickets accompanying many of
the gems in the Department of Asia, the British Museum, regularly
misidentified garnets as sard and prior to this study many of the
Masson gems (i.e. nos 1880.3559, 1880.3553, 1880.3554, 1880.3576,
1880.3577, 1880.3578, 1880.3596, 1880.3597, 1880.3598, 1880.3599,
1880.3729, 1880.3988) were identified as either ‘sard’ or ‘wine-red
chalcedony’. These have now been corrected on the British
Museum online database.
6 The catalogue of the gems in the Museo Nazionale di Aquileia lists
no garnet amongst the 1573 stones catalogued, although the colour
plates would seem to suggest that the collection does include
garnets (cf. G. Sena Chiesa, Gemme del Museo Nazionale di
Aquileia, Padua, 1966, pls B, nos 1182, 681, 645 and D, no. 1408, 176).
7 Callieri (P. Callieri, ‘Seals from Gandhara, Foreign Imports and
Local Production’ in M.-F. Boussac and A. Invernizzi (eds), Archives
et Sceaux du Monde Hellénistique, Torino, 1996, 413–22, at 414, pl.
‘Gems of Heaven’ | 21
Adams
93, fig. 2 and P. Callieri, Seals and Sealings from the North-West of
the Indian Subcontinent and Afghanistan (4th Century bc –11th
Century ad): Local, Indian, Sasanian, Graeco-Persian, Sogdian,
Roman, Naples, IsIAO, 1997) used the term ‘wine-red chalcedony’
for many garnets in his catalogue; see also n. 3 above.
8 F.H. Marshall, Catalogue of the Finger-Rings, Greek, Etruscan and
Roman, in the Departments of Antiquities, British Museum, London,
1907, no. 351; Walters (n. 2), 73, no. 603 (GR 1814,0704.1292,
Townley collection). The exhibition history and modern
publications of this famous ring may be found on the British
Museum’s online database.
9 The scarab was omitted by Walters. In modern mineralogy jacinth
(from the Greek hyacinth) refers to a variety of zircon (ZrSiO4 =
zirconium silicate, a neosilicate like garnet); it has a wide colour
range from colourless to yellow, brown, red, pink, blue and green
to black. Although Walters states in his introduction (n. 2, xiii) that
‘jacinth is an orange-coloured stone’, in comparing the actual
stones in the collection to his catalogue it is clear that he used the
term rather indiscriminately to describe stones of both orange and
pale violet hues.
10 A.D.H. Bivar, Catalogue of the Western Asiatic Seals in the British
Museum, Stamp Seals, II, The Sassanian Dynasty, London, 1969.
11 P.R.S. Moorey, Ancient Mesopotamian Materials and Industries: the
Archaeological Evidence, Oxford, 1994, 83.
12 Cf. Furtwängler (n. 3), III, 150; Walters (n. 2), xiii.
13 D.N. Adams, Late Antique, Migration Period and Early Byzantine
Garnet Cloisonné Ornaments, (unpublished D. Phil. thesis,
University College, London), 1991, Appendix III, 292–3, 301, table 1.
14 Walters (n. 2), 437–69 (nos 437–564); 110–31 (nos 948–1142).
Likewise J. Boardman, Greek Gems and Finger Rings, Early Bronze
Age to Late Classical, London, 2001, 373–7, includes no garnet but
takes Theophrastus’ text as evidence that it was in use.
15 E. Brandt, Antike Gemmen in deutschen Sammlungen (hereafter
AGDS) I, Staatliche Münzsammlung, München, 1. Griechische
Gemmen von minoischer Zeit bis zum späten Hellenismus, Munich,
1968; E. Brandt and E. Schmidt, AGDS I, Staatliche
Münzsammlung, München, 2, Italische Gemmen, etruskisch bis
römisch-republikanische, Italische Glaspasten vorkaiserzeitliche,
Munich, 1970; E. Brandt, A. Krug et al., AGDS I, Staatliche
Münzsammlung, München, 3, Gemmen und Glaspasten der
römischen Kaiserziet sowie Nachträge, Munich, 1972.
16 It must also be noted that some major museums have collected
remarkably few engraved garnets. For example only six garnets
were recorded amongst a total of 564 intaglios ranging from the
Minoan, Greek, and Etruscan through the Roman period in Berlin
(E. Zwierlein-Diehl, AGDS II, Staatliche Museen, Preussischer
Kulturbesitz Berlin, Munich, 1969) and 12 garnets amongst 1,739
intaglios in Hanover (M. Schlüter, G. Platz-Horster and P. Zazoff,
AGDS IV, Hannover, Kestner-Museum, Hamburg, Museum für Kunst
und Gewerbe, Wiesbaden, 1975). Whether this reflects the personal
taste of curators or the availability of particular stones at the time
the collections were assembled is difficult to assess.
17 E.H. Warmington, The Commerce between the Roman Empire and
India, London, 1974, 121–4, 394 h and I, passim.
18 Adams (n. 13), Appendix 1, 281–4.
19 Adams (n. 13), 293–4, Table II. Garnet seals constitute 11.59% of the
total (92 of 794 stones in all materials, excluding glass and metal)
and 7.7% of the collection in the Hermitage Museum in St
Petersburg (57 of 804 stones).
20 J. Spier, Late Antique and Early Christian Gems, Wiesbaden, 2007,
25, 87–94.
21 Ibid., 87, pl. 138, fig. 6; these are essentially deep versions of Bivar’s
Bezel A (Bivar [n. 10], 21, 142).
22 Spier (n. 20), 87; A. MacGregor, ‘A seventh-century pectoral cross
from Holderness, East Yorkshire’, Medieval Archaeology 44 (2000),
217–22. Treasure Annual Report 2000, DCMS, London, 2002, 44–5,
no. 61.
23 Bezel D: Bivar (n. 10), 21, 142; ringstone F: M. Maaskant-Kleibrink,
Catalogue of the Engraved Gems in the Royal Coin Cabinet, The
Hague. The Greek, Etruscan and Roman Collections, The Hague,
1987, 60, fig. 2, F1, F2 and F5.
24 Adams (n. 13), Appendix III, 294–7, Graphs 1–7.
25 Ibid., 294–7, Graphs 9–11; N. Adams, ‘The Development of Early
Garnet Inlaid Ornaments’, in Cs. Bálint (ed.), Kontakte zwischen
Iran, Byzanz und der Steppe in 6.–7. Jh. (Varia Archaeologica
Hungarica 10), Budapest-Naples-Rome, 2000, 13–70, at 22–3, pl. IV.
22 | ‘Gems of Heaven’
26 N. Adams, ‘Garnet Inlays in the Light of the Armaziskhevi Dagger
Hilt’, Medieval Archaeology 47 (2003), 167–75.
27 Göbl (n. 5); Callieri (n. 7). Further references in Adams, Lüle and
Passmore below.
28 Callieri (n. 7), 217.
29 J. Boardman, The Diffusion of Classical Art in Antiquity,
Washington DC, 1994, 118–19 Callieri (n. 7), 84–90, 235–7, cat. nos
4.1–17.
30 B. Barthélemy de Saizieu and J. Rodière, ‘Bead-Drilling: A look
from Mehrgarh and Nausharo. Preliminary Results of Micro-trace
Analyses’, in South Asian Archaeology 2003 (hereafter SAA),
Proceedings of the Seventeenth International Conference of the
European Association of South Asian Archaeologists (hereafter
EurASEAA) (7–11 July 2003, Bonn) (U. Franke-Vogt and H-J.
Weisshaar, eds), Aachen, 2005, 39–48, at 43, 45, fig. 5.2 have
demonstrated that perforations in garnet beads which first appear
at Mehrgarh in the Indus Valley in Pakistan in the Early
Chalcolithic Period III (c. 4500–4000 bc) could only be
accomplished by pecking, as opposed to drilling.
31 H.C. Beck, ‘The Beads from Taxila, an examination of 950 selected
beads 700 bc – 500 ad ’, in J. Marshall (ed.), Memoirs of the
Archaeological Society of India 65 (1941), 1–66, at 65, nos 1–68).
32 H. Härtel, Excavations at Sonkh, 2500 years of a Town in Mathura
District, Berlin, 1993, 298–300; Adams (n. 13), 60, 69.
33 Afghanistan: H. P. Francfort, Fouilles d’Aï Khanoum, III, Le
Sanctuaire de Temple à Niches Indentées, 2. Les Trouvailles, Paris,
1984; C. Rapin, Fouilles d’Aï Khanoum, VIII, La Trésorie du Palais
Hellénistique d’ Aï Khanoum, Paris, 1992, 132–3, 178, 340, pls 79 and
117; India: N. Lahiri, The Archaeology of Indian Trade Routes up to c.
200 bc: Resource Use, Resource Access and Lines of Communication,
Delhi, 1992, Upper Gangetic Plain: Charts 47a-b (Piprahwa and
Kausambi); Central India: 51b (Nagda); Gujarat: 52 (Nagara) and
53b (Bhokraden); northern Karnataka: 54b (Maski). Sri Lanka:
R. Coningham, Anuradhapura, The British-Sri Lankan Excavations
at Anuradhapura Salgaha Watta 2, vol II: The Artefacts, BAR
International Series 1508, 2006, 387–90; P. Francis Jr, ‘Beads and
Selected Small Finds from the 1989–92 Excavations’, in V. Begley et
al., The Ancient Port of Arikamedu, New Excavations and Researches
1989–1992, Vol. Two, Mémoires Archéologiques 22.2, Paris, 2004,
447–604, at 480–81, 491–7, a total of 168 garnet beads.
34 Francis (n. 33), 492–5.
35 C. Rösche, R. Hock, U. Schüssler, P. Yule and A. Hannibal, ‘Electron
microprobe analysis and X-ray diffraction methods in
archaeometry: Investigations on ancient beads from the Sultanate
of Oman and from Sri Lanka’, European Journal of Mineralogy 9
(1997), 763–83; U. Schüssler, C. Rösche and R. Hock, ‘Beads from
Ancient Sri Lanka – first results of a systematic material analysis’,
in H.-J. Weisshaar, H. Roth and W. Wijeyapala (eds), Ancient
Ruhuna. Sri-Lankan-German Archaeological Project in the
Southern Province 1, Mainz, 2001, 227–42.
36 Adams (n. 13), 64, Appendix II, cat. no. 173, pl. 21.2,3; cf. Beck (n.
31), 16, pl. vi., no. 33.
37 Adams (n. 25), 17–20, pl. I, 1–5 and pl. II; P. Cambon and J.-F.
Jarrige (eds), Afghanistan: les trésors retrouvés, collection du musée
national de Kabul, Paris, 2006, 169, nos 47–8.
38 Ibid. and N. Adams, ‘Back to Front: Observations on the
development and production of decorated backing foils for garnet
cloisonné’, Historical Metallurgy 40/1 (2006), 12–26, at 12. Cf. Pliny
(Nat. Hist. XXXVII.27.99).
39 Francfort (n. 33), 75, pl. 27, no. 825 and pl. XXXIII, no. 45; O.
Guillaume and A. Rougeulle, Fouilles d’Aï Khanoum, VIII, La
Trésories du Palais Hellénistique d’ Aï Khanoum, VII, Les Petits
Objects, Paris, 1987, 55, pl. XIV.12, no. 1004; Rapin (n. 33), nos O78
and O23.1
40 ­­Cambon and Jarrige (n. 37), 210, 282, no. 139 (MK 04.40.45).
41 Marshall (n. 5), vol. I, 160; vol. II, 650, vol. III, pl. 207a–i, Stratum
I-II, Sirkap Block E, indentified as ‘jacinth’, see above, p. 11; Callieri
(n. 7, 1996), p. 417, pl. 95, figs 16–17.
42 R.N. Mehta, ‘Urban Centres of Western India and the Western
World’, in J. Deppert (ed.), India and the West, Proceedings of a
Seminar Dedicated to the Memory of Hermann Goetz, New Delhi,
1983, 137–48, at 145, fig. 7.
43 Coningham (n. 33), 389.
44 T. Calligaro, ‘Analyse des matériaux: Tillia tepe, etude des
incrustations et de l’or’, in Cambon and Jarrige (n. 37), 292–7, at
291–2.

45 Inter alia, from an archaeological perspective: H. Roth,
‘Almandinhandel und -verarbeitung im Bereich des Mittelmeeres.
Zum archäologischen Befund und der schriftlichen Uberlieferung
in der Spätantike und im frühen Mittelalter’, Allgemeine und
Vergleichende Archäologie Beiträge 2 (1980), 309–35; from a
jewellery perspective: J. Ogden, Jewellery of the Ancient World,
London, 1982, 97–9; from a mineralogical perspective: Rouse (n.
1), 1–19.
46 A recent analysis of some of the differing and possible
identifications of ancient terms for stones in: A. Mottano and M.
Napolitano, ‘Il libro “Sulle pietre” di Teofrasto, Prima traduzione
italiana con un vocabolario dei termini mineralogici’, Rend. Ris.
Acc. Lincei, ser. 9, 8 (1997), 151–234, at 181–3; a further exploration
of the Hellenistic lapidary and Christian terminology in N. Adams,
‘The garnet on the Narses cross’, Dumbarton Oaks Bulletin,
forthcoming.
47 The juxtaposition of anthrax and anthrakion has caused some
authors to suggest that anthrax was a word for corundum which is
incombustible with a blowpipe, while garnet is not. The first
reference to άνθραξ as a gem appears in Aristotle (d. c. 322 bc)
who wrote that: ‘the seal-stone called anthrax is the least affected
by fire of all the stones’ (Meteorologica IV.ix, 387b18).
48 See D.J. Mattingly (ed.), The Archaeology of Fazzān, Volume 1,
Synthesis, Society for Libyan Studies, London and Department of
Antiquities Tripoli, Hertford, 2003, 356–7, who argues that the
archaeological evidence suggests that Carthaginian ‘carbuncle’
was carnelian.
49 Ogden (n. 45), 95 and 111.
50 Adams forthcoming, ‘Garnet in Aristotelian and Orphic
Traditions’.
51 R. Halleux and J. Schamp (eds and trans.), Les Lapidaires Grecs,
Paris, 1985, Orphei Lithica, 96, no. 8, and Orphei Lithica Kerygmata,
150, no. 7, notes on 306; both of these texts probably date to the
second half of the 2nd century ad although they survive only in
Late Medieval manuscripts.
52 I have followed the Latin quite closely here in this and the
preceding passages, diverging somewhat from the standard Loeb
edition, Pliny the Elder, Naturalis Historia, Books XXXVI–XXXVII,
(D.E. Eichholz (trans.), Loeb Classical Library, Cambridge, MA,
1962.
53 E.J. Gübelin and J.I. Koivula, Photoatlas of Inclusions in Gemstones,
Vol. 1, Basel, 1986, 287–97; idem, Photoatlas of Inclusions in
Gemstones, Vol. 2, Basel, 2005, 430–85; garnet inclusions in Early
Medieval gemstones in H. Gilg, N. Gast and T. Calligaro, ‘Vom
Karfunkelstein’, in L. Wamser (ed.), Karfunkelstein und Seide. Neue
Schätze aus Bayerns Frühzeit, Munich, 2010, 87–100.
54 R. Temple, The Crystal Sun, Rediscovering a Lost Technology of the
Ancient World, London, 2000, with a catalogue of actual and
potential ancient crystal magnifiers.
55 N. Adams, Bright Lights in the Dark Ages, The Eugene Thaw
Collection, New York, forthcoming 2012. I am grateful to the
J. Pierpont Morgan Library and Museum for permission to publish
this piece from the Thaw collection here. Chatoyancy resulting
from dense clusters of rutile needles has also been observed by this
author on other medieval stones.
56 M.J.-F. Guinel and M.G. Norton ‘The origin of asterism in
almandine-pyrope garnets from Idaho’, Journal of Materials
Science 40 (2006), 719–25.
57 The famous modern sources in Idaho, USA are of course excluded
(M.E. Gunter, ‘Idaho’s “Star” Garnet’, in Gilg et al. [n. 1], 30–3);
asterism has also been identified in an andradite garnet from
Namibia in southwestern Africa (E.A. Fritz, J.I. Koivula, B. Laurs,
M.E. Gunter and C.J. Johnston, ‘Inclusions in andradite from
Namibia’, Gems and Gemology 43/4 (2007), 370–3).
58 E.R. Caley and J.F.C. Richards (trans.) Theophrastus On Stones.
Introduction, Greek Text, English Translation and Commentary,
Columbus, Ohio, 1956.
59 ‘Above the coast-line which extends from Carthage to Cephalae
and to the land of Masaesylians lies the land of the Libo-
Phoenicians, which extends to the mountainous country of the
Gaetulians, where Libya begins. The land above the Gaetulians is
that of the Garamantes, which lies parallel to the former and is the
land whence the Carthaginian stones are brought’.
60 J.-P. Lefranc, ‘La geologie, Pliné l’Ancien et l’histoire de Cornelius
Balbus (20 avant J.C.). Nouvelles identifications’, Histoire et
archéologie de l’Afrique du nord 3 (1986), 303–16, at 309–10.
The Garnet Millennium
61 See n. 48 above.
62 Although the first examination of medieval garnet took place in
the 19th century in Bavaria (Gilg, Gast and Calligaro [n. 53]),
modern work was initiated by Birgit Arhennius in conjunction with
her seminal study of garnet cloisonné from the Merovingian
period in Europe: Granatschmuck und Gemmen aus nordischen
Funden des frühen Mittelalters, Acta universitatus
Stockholmiensis, Stockholm, 1971. Arguing that much of the
garnet used on ornaments of this period originated from
Bohemian sources, she arranged for Mellis (O. Mellis,
‘Mineralogische Untersuchungen an Granaten aus in Schweden
gefundenen Schmuckgegenstanden der Merowinger- und
Karolingerzeit’, Archiv for Mineralogi och Geologi 3/15 (1963), 297–
362) in Stockholm to perform basic gemmological tests on a
number of flat garnet plates. The later English language version of
her 1971 text included further analyses by Diego Carlström:
Merovingian Garnet Jewellery, Emergence and Social Implications,
Kungl. Vitterhets Historie och Antikvitets Akademien, Stockholm,
1985, 26–34.
63 Mellis (n. 62); P. Lundström, ‘Almandingranaten von Paviken auf
Gotland’, Early Medieval Studies 6 (1973), 67–77; L. Löfgren, ‘Die
mineralogische Untersuchungen der Granaten von Paviken auf
Gotland’, Early Medieval Studies 6 (1973), 78–96; M. Bimson, S.
La Niece and M.N. Leese, ‘The Characterisation of Mounted
Garnets’, Archaeometry 24 (1982), 51–8; S. Greiff,
‘Naturwissenschaftliche Untersuchungen zur Frage der
Rohsteinquellen für frühmittelalterlichen
Almandingranatschmuck rheinfränkischer Provenienz’, Jahrbuch
des Römisch-Germanischen Zentralmuseums Mainz 45/2 (1998),
599–646; F. Farges, ‘Mineralogy of the Louvres Merovingian
garnet cloisonné jewelery: Origins of the gems of the first kings of
France’, American Mineralogist 83 (1998), 323–30; D. Quast,
‘Mineralogische Untersuchungen zur Herkunft der Granate
merowingerzeitlicher Cloisonnéarbeiten’, Germania 78/1 (2000),
75–96; T. Calligaro, P. Perin, F. Vallet and J.-P. Poirot, ‘Contribution
à l’étude des grenats mérovingiens (Basilique de Saint-Denis et
autres collections du musée d’Archéologie nationale, diverses
collections publiques et objets de fouilles récent)’, Antiquités
Nationales 38 (2006/7), 111–44; M. Mannerstrand and L.
Lundqvist, ‘Garnet Chemistry from the Slöinge Excavation,
Halland and Additional Swedish and Danish Excavations –
Comparisons with Garnet Occurring in a Rock Context’, Journal of
Archaeological Science 30 (2003), 169–83; P. Périn with T. Calligaro
and C. Sudres, ‘A propos du “trésors de grenats de Carthage,
attribué à l’époque vandale”’, Antiquités nationales 40 (2009)
(2010), 155–65; H.A. Gilg, ‘Anthrax, Carbunculus, and Granatus:
Garnet in Ancient and Medieval Times’, in H.A. Gilg et al. (n. 1),
12–18; Gilg, Gast and Calligaro (n. 53), 87–100.
64 Calligaro, Perin, Vallet and Poirot (n. 63). Chemical analyses of
garnet plates at C2RMF uses the PIXE method with particle
acceleration AGLAE, micro-spectrometry RAMAN and micro-
madding in the PIXE mode.
65 Greiff (n. 63), tab. 2 and abb. 2, performed microprobe analysis on
ten garnet specimens from Indian localities and one from
Afghanistan; Calligaro, Perin, Vallet and Poirot (n. 63), Type I
which is equivalent to Gilg, Gast and Calligaro (n. 53), Cluster B.
66 Bohemia: Farges (n. 63); Quast (n. 63); Calligaro, Perin, Vallet and
Poirot (n. 63), 126–7, Type IV, equivalent to Cluster D in Gilg,
Norbert and Calligaro (n. 53); Portugal: Calligaro, Perin, Vallet
and Poirot (n. 63), Type V equivalent to Gilg, Norbert and Calligaro
(n. 53), Cluster E.
67 Gilg, Norbert and Calligaro (n. 53), xx.
68 Cf. Calligaro, Perin, Vallet and Poirot (n. 63), 125, fig. 8. The author
has been informed by modern gem dealers in Hatton Garden in
London that gem-quality stones from East Africa are today
regularly shipped to India for preparation where the skills and
labour force exceed those of the native source.
69 Löfgren and Mannerstrand and Lundqvist (both n. 63).
70 C. Lüle­­-Whipp, Mineralogical-Petrographical and Geochemical
Investigation on some Garnets from Volcanic Rocks of Görece Village
– Cumaovasi İzmir and Metamorphites of Menderes Massif and their
possible Archeogemmological Connections, unpublished PhD thesis,
Hacettepe University, Ankara, Turkey, 2006.
71 I met Lisbet Thoresen, who directed these studies, for the first time
at the 2009 Engraved Gems conference. I am grateful to her for
sharing her unpublished work with me which dovetails so well
‘Gems of Heaven’ | 23
Adams
with the GCEM project; her results are forthcoming in: L. Thoresen
(ed.), On Gemstones: Gemological and Analytical Studies of Ancient
Intaglios and Cameos.
72 Nat. Hist. XXXVII.25: ‘The Carchedonian stones, they say, are of
much smaller size than the others; but those of India admit of
being hollowed out, and making vessels that will hold as much as
one sextarius.’
73 J.N. Christensen, J. Selverstone, J.L. Rosenfeld and D.J. DePaolo,
‘Correlation by Rb-Sr geochronology of garnet growth histories
from different structural levels within the Tauern window,
Eastern Alps’, Contributions to Mineralogy and Petrology 118
(1994), 1–12.
74 Rouse (n. 1), 42.
75 J. Hyršl, ‘New gemmological study of large garnets of supposedly
Czech origin’, Deutschen Gemmologie, Zeitschrift der Deutschen
Gemmologischen Gesellschaft 50/1 (2001), 37–42.
76 Recently crystals of different species measuring from 30mm to
70mm have been recorded in Pakistan and Afghanistan (D.
Blauwet, ‘Pakistan and Afghanistan: Garnet from the Roof of the
World’, in Gilg et al. (n. 1), 70–6).
77 Commercial examples of these: http://customgemstonestudio.
com/custom-gemstones/worlds-largest-umba-valley-red-garnet.
The largest stone on this particular site weighs 59ct, diam. 24mm
and depth 14mm; it has been faceted.
78 Adams (n. 13), Appendix V, 337–9; Adams (n. 25), 38–41.
79 Calligaro, Perin, Vallet and Poirot (n. 63), 125.
80 Hyršl (n. 75), 41.
81 J. Boardman, Intaglios and Rings, Greek, Etruscan and Roman from
a private collection, London, 1975, 92, no. 56; J. Boardman and D.
Scarisbrick, The Ralph Harari Collection of Finger Rings, London,
1977, 24, no. 28; Boardman and Vollenweider (n. 5), 100, no. 342.
82 Unpublished results from tests performed by Emma Passmore.
83 G. Platz-Horster, ‘Königliche ‘Artemis? Eine neue Granatgemme im
Kestner-Museum zu Hannover’, Niederdeutsche Beiträge zur
Kunstgeschichte 34 (1995), 9–26; D. Plantzos, Hellenistic Engraved
Gems, Oxford, 1999, 127, no. 399.
84 Rapin (n. 33), pl. 68, M23.1-3
85 L.A.N. Iyer, ‘Indian Precious Stones’, Bulletins of the Geological
Survey of India, Series A, Economic Geology, no. 18 (1961), 50–4,
summarises 19th and 20th century records of garnet, recording
both mining and panning for garnet from river sands.
86 Bimson, La Niece and Leese (n. 63), 52, note 1; M. Bimson, ‘Dark-
age garnet cutting’, Anglo-Saxon Studies in Archaeology and
History 4 (1985), 125–8, at 125, figs. 2–3; Adams (n. 26), 167–8, n. 3.
87 Adams (n. 25), 17–20.
88 Marshall 1907 (n. 8), 118, no. 707 (garnet-set ring found with a coin
of Lysimachus (r. 306–281 bc).
89 M. Treister, ‘Late Hellenistic Bosphoran Polychrome Style and its
Relation to the Jewellery of Roman Syria (Kuban Brooches and
24 | ‘Gems of Heaven’
Related Forms)’, Silk Road Art and Archaeology 8 (2002), 29–72, at
31 (Zelenskoi barrow, Taman peninsula).
90 Platz-Horster (n. 83).
91 Inter alia, Marshall 1907 (n. 8), 67, pl. XI, S; Boardman (n. 81), 91,
no. 51.
92 Treister (n. 89); V. Mordvinceva and M. Treister, Toreutik und
Schmuck in nördlichen Schwarzmeergebiet, 2. Jh. v. Chr. –2. Jh. n.
Chr., 3 vols, Simferopol, Bonn, 2007.
93 M. Mielczarek, ‘Remarks on the numismatic evidence for the
northern Silk Route: the Sarmatians and the trade route linking
the northern Black Sea area with Central Asia’, in K. Tanabe,
J. Cribb and H. Wang (eds), Studies in Silk Road Coins and Culture,
Papers in honour of Professor Ikuo Hirayama on his 65th birthday,
Kamakura, 1997, 131–47, at 135. The author emphasises that trade
was probably ongoing before this time, using barter mechanisms
rather than coinage.
94 L. Casson (trans. and ed.), The Periplus Maris Erythraei, Princeton,
1989.
95 Callieri (n. 7), 266–7.
96 Adams (n. 25), 17–20, pls I.1–5, pl. II, pl. III.1–4.
97 Bivar (n. 10) 13–14, discusses the ‘baffling scarcity of evident
Parthian seals’. The garnet portrait of the Iberian eristavi Asparug,
excavated at Armazis-khevi in Tomb 1 (tpq coin date of ad 128), is of
the Parthian period but where it was carved is unknown (A.M.
Apakidze, G.T. Gobedžišvili, A.N. Kalanadze, and G.A.Lomatidze,
Mcxeta, Itogi Archeologičeskih Issledovanij I, Arxeologičeskie
Pamjatniki Armazis-xevi po raskopkam 1937–1946 gg., 1958, Tbilisi,
pl. 1,1, XLV, 1, fig. 4). Cf. also R. Gyselen, ‘La glyptique’, in
B. Overlaet and M. Ruyssinck (eds), Splendeur des Sassanides,
L’Empire perse entre Rome et la Chine (224–642 ad), Brussels, 1993,
123. On bullae impressions: A.D.H. Bivar, ‘Seal-Impressions of
Parthian Qūmis (Qūmis Commentaries no. 4)’, Iran 20 (1982), 161–
76; A.B. Nitkin, ‘Parthian Bullae from Nisa’, Silk Road Art and
Archaeology 3 (1993/4), 71–80.
98 Adams (n. 25), 21–2.
99 Ibid., 21, pl. III.5; 42, pl. XII.1–2. I now believe these pieces should
be dated to the 2nd or early 3rd century ad.
100 M. Triester, ‘Cloisonné- and champlevé-decoration in the gold
work of the Later Hellenistic-Early Imperial Period’, Acta
Archaeologica 75/2 (2004), 189–219.
101 Iyer (n. 85), 53–4.
102 Roth (n. 45).
103 E.g. U. von Freeden, ‘Das Ende engzelligen Cloisonnés und die
Eroberung Südarabiens durch die Sasaniden’, Germania 78/1
(2000), 97–124, who argued that a disruption in supplies of garnet
to Europe when the Sasanians assumed control of trade routes in
southern Arabia could account for a group of empty cloisonné disc
brooches (i.e. not inlaid with garnet plates) dating to the late 6th
and early 7th century ad.
Lithóis Indikois
Preliminary Characterisation of Garnet Seal Stones from Central and South Asia1
Noël Adams, Çiğdem Lüle and Emma Passmore,
with contributions by Harry Falk and Nicholas Sims-Williams
The GCEM project (Garnets: Classical, Eastern and Medieval)
This paper presents the initial findings of the GCEM project,
designed to create a scientific database of garnet intaglios held
in different departments in the British Museum. Of all the
ancient stones used as seal stones, garnets are perhaps the most
promising in terms of yielding a broad reference database.
Unlike diamond, the corundum series (ruby, sapphire) and
spinels which were used relatively rarely in antiquity,
substantial numbers of garnets were used steadily over many
centuries. And, unlike the inclusions in the quartz group whose
various members are ubiquitous throughout antiquity, the
distinctive inclusions within garnets, when compositionally
identified, may in some cases be used to match a particular
gemstone to a host rock.
As reviewed in the previous paper in this volume,
considerable scientific research into the mineralogy of ancient
garnets has been done on stones set on Early Medieval objects
with the goal of identifying source localities. Some of the
problems associated with the investigation of Early Medieval
period garnets – examining mounted garnets from the end of a
millennium of continuous usage – can be sidestepped by
shifting the focus to another corpus of garnet stones. The
GCEM project will generate reference data derived from
engraved seal stones used across the ‘garnet millennium’, from
c. 300 bc to c. 700 ad, ranging from the Hellenistic, Graeco­­-
Roman and Byzantine periods in the West, to the Sasanian
empire in greater Persia and finally to the Indo-Greek, Indo-
Parthian, Kushan and Hunnic dynasties in Central and South
Asia.
Unlike mounted garnet plates on Early Medieval objects,
many of which are held in settings which cannot be
deconstructed or subjected to vacuum pressure, most intaglios
can be easily tested and quantified. Mounted intaglios can in
some cases present obstacles to initial gemmological
identification, and of course the individual stability of each
gem must be considered before subjecting it to some scientific
tests. The following pilot study employed non-destructive
testing combining standard gemmological identification for
inclusion characterisation, followed by mineralogical
examination using Raman spectroscopy. The initial application
of this straightforward and non-destructive methodology to
engraved garnets from Central and South Asia in the British
Museum identified two distinct clusters of stones on the basis
of their gemmology, inclusions and mineralogy.
This preliminary work has been undertaken in the
knowledge that, at present, even advanced levels of garnet
identification do not necessarily result in information which is
source-specific. As Gübelin and Koivula noted many years ago
with respect to inclusions: ‘while these inclusions are
symptomatic for almandine garnet, they do not differ from
deposit to deposit, i.e. all over the world the inclusion scenes in
almandines are analogous’.2 In the next phase of the project it
may be possible to obtain further geochemical information
using LA-ICP-MS (Laser Ablation Inductively Coupled Plasma
Mass Spectrometry) to analyse trace elements within selected
garnets and their inclusions; these results will further refine
the parameters of clusters of stones which in turn may someday
be used to pinpoint a particular locality.
A published reference database giving a clear
gemmological and mineralogical picture of surviving garnet
intaglios, combined with the published data on garnet
cloisonné plates, will provide more closely defined parameters
for scientific researchers as to which of the many dozens of
viable garnet sources fieldwork should be directed towards. For
art-historical researchers the established groupings may
define, re-define or substantiate groups of garnet intaglios
classified on the basis of stylistic analysis alone (on some
problems of stylistic attribution see Marsden in this volume).
With an expanded body of analytical information from
intaglios we can begin to analyse patterns of garnet usage
within and across chronological time frames and geographical
regions and, for the first time, obtain an overview of garnet
usage in antiquity.
Significance of the collections
The garnet gems held in the Department of Asia in the British
Museum constitute the largest collection of eastern intaglios in
a western museum. Although lacking excavated provenances,
many of these were collected in the 19th century in the former
North-West Frontier Province in British India (hereafter
NWFP). Many are inscribed in eastern scripts (Graeco-
Bactrian, Kharoṣṭ hī and Brāhmī) and most exhibit distinctive
local iconography, making it probable that the majority were
engraved in Central or South Asia. Scientific investigations into
Early Medieval garnet cloisonné plates have already suggested
that India was a major source for many of the garnets used in
that period.3 In addition to the multiple sources reported in
India,4 gem-quality red garnets are also known from
Afghanistan and Pakistan (Map).5 It is therefore highly likely
that the British Museum seal stones were locally sourced in
these regions. As a consequence these particular garnets are of
the utmost importance to garnet studies as they provide a
control group of ancient garnet stones from the East against
which gemstones in the West, potentially gathered from many
different sources from the Hellenstic to the Early Medieval
periods, may be compared.
This paper presents two groups of garnets whose
gemmological and mineralogical characteristics suggest that
they came from compositionally similar sources. The two
clusters were initially identified on the basis of a gemmological
examination which was subsequently confirmed by Raman
spectroscopy; these were blind tests, undertaken with no
'Gems of Heaven’ | 25
Adams, Lüle and Passmore
Map of place names and
archaeological sites
noted in the text, with
some garnet localities
and mines recorded in
modern sources
foreknowledge of the date or style of the gemstones. Together
these two groups establish the first mineralogical fingerprint of
garnet stones used in the East from around the 1st to the 6th
century ad.
The brief discussion in the catalogue demonstrates how
even a basic scientific characterisation of garnet intaglios may
be useful for stylistic, workshop and historical analyses. Group
I, for example, constitutes a uniform set of seals whose style of
carving is classicising and whose iconography derives from
Graeco-Roman subject matter. Key stones in this group have
already been intensively studied and assigned to the period
when the Kushan dynasty was at its height in the 2nd and 3rd
centuries ad. Work is ongoing to characterise the optically
distinctive amphibole inclusions found in six of these stones,
which may be useful in determining the composition of the
host rock. Many of the stones in Group I are not only
compositionally similar but are related in terms of size,
preparation, iconography and date. This could suggest that one
locality was being drawn upon by one or more closely related
lapidaries for the majority of these seal stones.
Group II is more stylistically heterogeneous, comprising
stones that are diverse in terms of size, iconography and date.
Most of the stones exhibit the same inclusion set and three have
been identified as having distinctive inclusions whose chemical
composition remains unidentified. Some of the stones in this
group belong within the cultural and political sphere of the
26 | 'Gems of Heaven’
Hunnic successors to the Kushan empire, ruling from the later
4th to the mid-6th centuries ad. These overlap chronologically
with the Sasanian period and may someday provide valuable
comparative data, not only with Sasanian seal stones in the
British Museum previously identified as almandine by Bimson,6
but also with the range of garnets found on cloisonné which
first came into fashion in the Early Medieval period in the West
during the Hunnic period, i.e. the late 4th and first half of the
5th centuries ad.
Background to the collection
The collection of intaglios in the Department of Asia in the
British Museum includes seals collected on the ‘North-West
Frontier’ (comprising parts of modern Afghanistan, Pakistan
and northern India) by Charles Masson (alias James Lewis;
1800–53) and Major-General Sir Alexander Cunningham
(1814–93), as well as gems bought by Sir Aurel Stein (1862–
1943) on the western borders of China. The majority of these
gems were purchased, not excavated. Masson’s gems came
primarily from the bazaars in Kabul or were stray finds from
the urban site at Begrām, the ancient Kāpiśī, north of Kabul,
which flourished from the 2nd century bc under the Indo-
Greeks and their successors the Kushans until attacked by the
Sasanians in c. ad 241. The final excavated occupation levels at
this site date to the 5th century ad, but Hunnic confederacies
were still based there in the 6th century ad.7

Masson’s collection came to the British Museum in 1878
with the closure of the India Museum, run by the East India
Company where he was employed for much of his career.
Cunningham travelled widely across South Asia and
established the Archaeological Survey of India. In addition to
serving in Tibet, Kashmir, Burma and the NWFP, he explored
Buddhist monuments in central India. His collection was
purchased and donated to the British Museum by A.W.F. Franks
in 1892. The Stein collection gems were purchased in Yōtkān,
the ancient capital of Khotan in present-day Xinjiang, China.8
Cunningham himself published some of the gemstones in
his collection9 and in 1960 M.G. Dikshit published a partial
catalogue of the stones in the Department of Asia.10 David Bivar
worked on some of the inscribed stones, first in his unpublished
PhD dissertation and subsequently in publications for the
Corpus Inscriptionum Iranicarum.11 Robert Göbl’s four volumes
on Kushan and Hunnic coins published in 1967 also included
many of the Hunnic period seals in the collection, juxtaposed
against the coin issues of the region.12 Work is ongoing on the
Masson collection as part of a larger project within the British
Museum, led by Elizabeth Errington who has published some
of the Masson gems.13 Most recently, many, but by no means all,
of the Central and South Asian gemstones in the department
were published by Pierfrancesco Callieri in 1997.14 His analysis
of the gems resulted in groupings based upon stylistic
considerations, inscriptions and historical context. He also
published a few stones in the Stein collection, now being
investigated by the Stein project at the museum.15
Almost half the 80 garnets identified in the collection in the
Department of Asia are unpublished; eight of those are
published here. The registration numbers for the most part
reflect the date of acquisition of the collections although a
number of the seal stones acquired by the department in the
late 19th century were not assembled and formally registered
until 2005 and thus do not appear in Callieri’s 1997 catalogue.
With the exception of the first four gemstones in Group I
which are uniformly convex with flat backs, many of these
gems are quite irregular in shape in comparison to western
intaglios. Seals from Bactria and northwest India have been
classified according to both Graeco-Roman and Sasanian
conventions,16 and cat. nos I.1–5 and II.1 correspond to
western ring stone type B,17 but others of these engraved stones
vary so much from the Graeco-Roman tradition that it is not
clear to what extent this terminology is meaningful.
Accordingly at this stage we have made no attempt to
categorise the stone shapes. The unusual shape of many of
these seal stones, of course, must ultimately have a bearing on
whether or not they were fitted into rings or other settings.
All measurements were taken with dial callipers.
Photographs were taken by Noël Adams in the student’s room
in the Department of Asia and in the British Museum Research
Laboratory. Almandine garnets can be difficult to photograph
accurately for colour and while the colour may be seen to be
reasonably accurate, there has been no attempt to exert rigid
colour control in either the shooting or printing of the
catalogue photographs at this stage.
Following the scientific contributions by Çiğ dem Lüle (CL)
and Emma Passmore (EP), the catalogue text sets the gems in
their historical context and provides a preliminary analysis of
their iconography and style. This includes readings and
Lithóis Indikois
commentary on the scripts on the stones by Harry Falk (HF)
and Nicholas Sims-Williams (NS-W). It will be clear from this
discussion that much work remains to be done as we reconsider
how to structure our often subjective understanding of dating
and style vis-à-vis the mineralogical aspects of lapidary
production.
Gemmological remarks
Çiğdem Lüle
Garnet seal stones held in the Department of Asia in the British
Museum, London were tested gemmologically on the 24, 25 and
27 August 2010. Approximately 80 stones were tested and 72
samples were identified as red garnets. Two distinctive
gemmological groups were distinguished.
Garnets consist of a relatively large group of 24 related
species which crystallise in the cubic system. Only six of these –
pyrope, almandine, spessartine, andradite, grossular and
uvarovite – are important gems in the modern world.
Structurally similar, garnets differ from one another by
interchanging chemical compositions through solid solutions.
For the most part, it is primarily red garnets (pyrope,
almandine, spessartine, pyrope almandine, and rarely
grossular) which have been identified as gem materials on
classical objects.18
The terms used in the report follow modern gemmological
terminology which is simplified and based upon colour,
refractive index and spectrum of the material. Garnet group
minerals are considered to be complex structures in
mineralogy and pure species are extremely rare in nature. The
pure end members expressed in ternary diagrams are thus
largely hypothetical (see Pl. 1 Adams in this volume). On these
diagrams, moving along between a line drawn between the
points indicates a ratio change in the atoms of the elements
that constitute the chemical composition of the garnets. These
shifts in chemical structure affect the physical and optical
properties of the stones.
In the almandine-pyrope series, almandine is the term
used for iron aluminium silicate garnets when the iron in the
chemical composition exceeds that of magnesium and when
iron exceeds that of manganese in its solid solution series with
spessartine. The magnesium-rich end member of this solid-
solution series is pyrope, and the manganese-rich end member
is spessartine. There are many intermediate members within
this series as the structure of the crystal lattice can
accommodate different elements as long as they ‘fit’ in the
structure. Differences in composition result in different
refractive indices and density and without full chemical
analysis it is difficult to pinpoint precisely where any one stone
sits within the solid solution series. Many garnets fall between
two or three of the end members pyrope-almandine and
pyrope-spessartine, but they always have a small amount of a
third component.
Broadly speaking, almandine forms in metamorphic rocks
such as mica schist, amphibolite, granulite and gneiss, as well
as in igneous rocks such as granite and associated granitic
pegmatites. Pyrope on the other hand occurs in peridotite,
eclogite, in ultra high-pressure metasomatic rocks, in
magnesium-rich gabbros and as phenocrysts in rare alkali
basalt tuffs. In the first instance pyropes are chromium-rich
and in the latter, chromium-poor. Pure pyrope does not occur
'Gems of Heaven’ | 27
Adams, Lüle and Passmore
in nature but if it did, it would be colourless. Its hue is
commonly the result of variable amounts of the trace elements
of iron and chromium. Pyrope is also found in kimberlites
which may contain either peridotite or eclogite as xenoliths
and can be an indicator mineral for diamond.
Gem testing and identification process
The initial gem testing was performed with basic
gemmological equipment; this included a standard
gemmological refractometer, a utility lamp providing white
and monochromatic light, a standard gemmological binocular
microscope with maximum 60x magnification and a handheld
diffraction type gemmological spectroscope.
General observation was done under a daylight equivalent
fluorescent lamp. The Refractive Index (RI) of each stone was
taken with a monochromatic light source. The spectrum of a
stone was observed to the extent permitted by its transparency.
Microscopy provided valuable information on inclusions for
each specimen and the magnification range was from 10x to
40x, depending on the clarity of the stone. Good transparency
and the deep colour of the samples made the spectra of stones
generally easy to obtain with a pointed white light source
provided by the utility lamp.
Colour
The colour description given for each sample is based on
gemmological colour terminology used by many respected gem
testing laboratories and avoids trade terms. This colour
grading system identifies the body colour as hue, intensity of
the colour as saturation and lightness or darkness of the colour
as tone. Within this system brown and gray are not considered
to be colours but lower saturation. For example, brownish red
refers to low saturation of red versus bright red. Lower
saturations are typically observed in very light or very dark
tones of any hue. Red garnets tend to have dark tones and
lower saturations of red, purplish red or orangey red.
Colour observation of the samples was done under a
daylight equivalent fluorescent light source. It is important to
bear in mind that the depth of the stone impacts on how the
body colour is perceived and therefore the same samples may
be observed with different colours depending on the cut.
For example, a small and thin garnet sample would display
more vivid colours than a bigger and a deeper sample of the
same stone. Since most red garnets are very dark in tone and
lower in saturation, they tend to be cut with a hollow back in
order to display better colour, otherwise they might look
almost black.
Surface features
Garnet group minerals possess a high refractive index and
display a high vitreous polish lustre which is unmistakable to
the trained eye. The samples tested for this project show high
lustre on polished surfaces.
Garnets often contain fissures and fractures due to
processes occurring during their formation. These fissures
might be healed with secondary inclusions. The irregular
structure of these fissures and fractures also creates
weaknesses within the garnet and consequent conchoidal
breaks, chips or even deep cavities can occur easily during or
after the fashioning of garnets. The British Museum samples
28 | 'Gems of Heaven’
show typical conchoidal breaks and chips in and around the
carved surfaces.
Microscopy
Observation under magnification, typically 10X to 60X,
provides information about the stones’ internal structure. In
modern gemmology inclusions found within minerals are used
for the separation of natural from synthetic and imitation
gems. The many different types of inclusions in garnet crystals
include crystals of different minerals such as apatite, rutile and
zircon, as well as gases, fluids and voids. As noted above,
garnets grow in various genetic processes of the igneous and
metamorphic cycles, occurring in basic igneous rocks, granites,
pegmatites, schists, certain marbles and skarn zones, and in
each deposit they feed from various chemical sources and are
subjected to different temperatures and pressures. Although
particular inclusions do not provide sufficient evidence to
pinpoint the geographical origin of any particular garnet, they
can indicate the geological conditions under which the garnet
formed.
Almost all the samples tested for this project contained
individual inclusions and combinations of inclusions that are
typically seen in red garnets. It is important to note that these
inclusions are not specific to different species of garnets
General properties for all samples
Lustre: all samples displayed a high vitreous polish lustre.
Surface conditions: all samples displayed conchoidal
fractures, fissures, abrasions visible to the unaided eye. Tool
marks in the carvings were observed under high
magnification.
Optic character and pleochroism: due to the fact that
garnets crystallize in the cubic system, all samples are singly
refractive and display no pleochroism.
Spectrum: both almandine and pyrope-almandine display
the same absorption spectrum. Although the spectrum
distinguishes these two from all the other garnet species, they
cannot be separated from each other via spectrum. All the
samples tested displayed very distinct absorption spectra with
three strong bands at 504, 520 and 573nm and fainter lines at
423, 460, 610 and a cut off 680–690nm. All almandine garnets
have a refractive index that is over the limit of standard
gemmological refractometers. Therefore no RI reading is
provided for this group.
Raman Spectroscopic analysis
Emma Passmore
Previous work has demonstrated that the Raman spectra of
garnets bear a direct relationship to their chemical
composition.19 The work described here constitutes a
preliminary Raman investigation of a sub-set of the garnets
included in this paper, focussing on the chemical composition
of both the garnets themselves and of some of the inclusions
within them.
Raman spectra were collected for the garnet matrixes of 19
of the intaglios described in the catalogue, and for two
optically distinct groups of mineral inclusions within them (see
'Gemmological remarks', Çiğ dem Lüle above). Analyses were
performed in the Department of Conservation and Scientific
Research at the British Museum, using a Dilor Infinity Raman
 Lithóis Indikois

Plate 1 Raman spectra obtained from 19 garnet intaglios. The two spectra shown in grey (Pl. 1a) are reference spectra for the Mg-rich pyrope end member and
the Fe-rich almandine end member of the pyralspite solid solution series. The wave numbers of the highest intensity peaks are labelled in all figures for
comparison, and object registration numbers are shown next to their respective spectra. The spectra have been grouped according to the position of the highest
intensity peak (in the range 916–928cm-1), and the second highest intensity peak (in the range 342–370cm-1). Spectra in black came from garnets where no
inclusions were measured, the spectrum in red from a garnet that had a single quartz and several amphibole-type inclusions, spectra in blue came from garnets
containing amphibole-type inclusions, and spectra in green came from garnets containing an as yet unidentified inclusion.

spectroscope with a near infrared (785nm) laser, with a spot
size of around 3 microns and a maximum power of 20 mW at
the sample. Each spectrum was collected for 150 seconds. No
pretreatment of the objects was necessary and the process of
analysis was completely non-invasive. Two analyses were
obtained from the garnet matrixes of each object, using two
different locations on the surface chosen to be free of dirt and
grease and not adjacent to inclusions. As described previously,
the intaglios are of different shapes and sizes and have variable
surface morphologies due to the depth and style of the carving.
Analyses were made on the flat polished backs of the objects,
but where this was not feasible (for example for cat. no. I.4,
which is in a metal mount), the garnet was angled and secured
in place so that a flat area on the carved surface was
perpendicular to the laser beam. Prior to analysis, each intaglio
was inspected with a binocular microscope using a
combination of reflected and raking light, and the location of
inclusions to be analysed was noted. Two different types of
optically distinct inclusions were seen: six of the garnets
contained multiple dark inclusions, ranging in size from 0.1–
1mm, which appeared optically to all be of the same type; and
three of the garnets contained a small number of large
(>1mm), colourless inclusions. The remaining stones were
apparently inclusion-free, although as cat. no. I.4 is in a metal
mount, it was not possible to see clearly into the body of the
garnet, and this stone may also contain inclusions. Where
possible, the inclusions were analysed where they were
exposed on polished surfaces. For those below the surface the
confocal microscope of the spectrometer was adjusted so that
the inclusion was in focus rather than the surface of the garnet.
Due to time restrictions, and the difficulty in accessing
inclusions that occurred below the garnet surface, only one
measurement of each type of inclusion was made per object.
The Raman spectra obtained for the garnets were
compared to garnet reference spectra obtained from
experimentally-grown garnets with known compositions.20 In
this study all the garnets were measured twice at two different
locations on the garnet surface, and all produced identical first
and second spectra, suggesting that composition is
homogeneous across each gemstone and they are therefore
unlikely to be compositionally zoned with respect to major
elements. The spectra produced indicate that all the garnets
measured have compositions within the pyralspite group
(X32+Y23+(SiO4)3; where X=Mg, Fe2+, Mn, and Y=Al), in a solid-
solution between the Mg-rich pyrope end member, and the
Fe-rich almandine end member. No spectra were obtained that
indicated a spessartine composition (Mn end member).21 The
garnets have been broadly grouped in Pl. 1 according to the

'Gems of Heaven’ | 29
Adams, Lüle and Passmore
shifts in the highest intensity peaks in the Raman spectra
(wavenumbers between 916 and 928cm-1). The results indicate
that all the garnets measured are almandine-rich; Pl. 1a shows
the garnet with a composition closest to the almandine end
member, whilst Pls 1b–e demonstrate a gradual shift in
composition towards a higher pyrope content.
Inclusions were measured in nine of the 19 garnets. All
produced clear Raman spectra, which fell into three categories;
those which could be identified as quartz or amphibole and an
as yet unidentified mineral. Amphibole inclusions were found
in six of the garnets (indicated in blue and red in Pl. 1), and are
a close match to the RRUFF reference spectra for actinolite,22
within the tremolite – ferroactinolite ((Ca2)(Mg5)(Si8O22)(OH)2
– (Ca2)(Fe52+)(Si8O22)(OH)2) solid-solution series. One garnet,
cat. no. I.6, contained numerous amphibole-type inclusions
and a single quartz inclusion and was the only garnet
measured that contained more than one type of inclusion
(indicated in red on Pl. 1a). Note however that only one of the
multiple dark inclusions was analysed in each of the six garnets
that contained them, hence it is possible that the dark
inclusions may not all be amphibole. The as yet unidentified
type of mineral inclusion was found in three of the garnets
(indicated in green in Pl. 1). During the gemmological analysis
this mineral was found to be transparent and seemingly
colourless and was speculatively identified as apatite. Its
Raman spectrum shares many features with several in-house
British Museum reference spectra for apatite, and a peak at
~967cm–1, which is suggestive of a phosphate bond. A number
of additional peaks are also present, however, and the precise
identification of this mineral remains unclear.
Whilst the mineralogical findings are at a preliminary
stage, it is interesting to note several trends already apparent in
these data. Garnets that contain numerous large amphibole-
type inclusions tend to have compositions that are closer to the
almandine end member than garnets that contained the
unidentified mineral inclusion. Only one garnet contained
more than one type of mineral inclusion (quartz and
amphibole in the most almandine-rich garnet analysed), and
amphibole and the unidentified mineral inclusion seem never
to occur together in the same crystal. These findings are good
evidence that further compositional trends may be found
within this group of garnets, and that more geochemical
analyses are required to enhance these preliminary data.
30 | 'Gems of Heaven’
CATALOGUE
Noël Adams
Group I
This group is composed of almandine garnets, generally with
dark, brownish orangey red colours. Their gemmological
properties are between almandine and spessartine and they
are all distinguished by inclusions of dark opaque tabular
crystals. Raman spectroscopy revealed the composition of
these garnets to be close to the almandine end member and the
dark inclusions which were examined proved to a type of
amphibole (specifically within the tremolite – ferroactinolite
solid solution series).
The iconography of the group is strongly classicising,
representing either Graeco-Roman mythological figures
(Herakles, putto) or Kushan-period deities represented in a
western classical manner (Nanā, Hāritī, Mahāsena/Skanda).
Two are inscribed in Bactrian, two in Kharoṣṭ hī and one in
Middle Persian. The first five are consistent in size and cut and
are oval convex stones with flat backs.
Group I includes two well-known seals from the Kushan
period (cat. nos I.1 and I.2), both inscribed with compound
personal names in Bactrian. In the 2nd century bc the Graeco-
Bactrian kingdom in Central Asia fell to nomadic tribes named
in the Chinese sources as the Yuezhi. The descendents of one
dominant tribe amongst the Yuezhi confederacy, the Kushans,
established a dynasty which eventually controlled much of
Central Asia and India. They adopted the language of their
subjects, Bactrian, an eastern Iranian language with signs
adapted from the Greek alphabet, as their administrative
language.23 By the time of Kaniṣka I their kingdom extended
from Kashgar, Khotan and Yarkand in the Tarim Basin region
(modern Xinjiang) to northern India. The territory was
administered from two capitals: Purushapur (Peshawar in
northern Pakistan) and Mathura in northern India, with a
summer capital at Kāpiśī in Afghanistan.
As Callieri and others have pointed out, the female deity
seated on a lion on cat. no. I.1 resembles representations of
Nanā on Kushan coinage of Kaniṣ ka I and Huviṣ ka;24 this
suggests a date in the 2nd century ad but there is no way to be
certain that the subject matter was not present before the coins
were struck or how long it survived. In any case, the
significance of the imagery to the royal Kushans is made clear
by the famous inscription in the sanctuary at Rabatak, north of
Pul-i Khumri in Afghanistan, which states that Kaniṣ ka
‘…obtained the kingship from Nanā’.25 She holds a shallow bowl
in her right hand and a pronged staff in the other; a crescent
moon appears above her head and the ties of a diadem float
down behind the tied bun at the back of her head.
Cat. no. I.2 depicts a standing goddess with a cornucopia,
and with a small male child before her, which suggests she
represents Hāritī, the goddess responsible for protection from
childhood diseases. Bivar has suggested that the proliferation
of Hāritī images in Gandhāran art documents the rise of
smallpox in South Asia, which rapidly became a pandemic,
spreading to the West in the reign of Marcus Aurelius in ad 166.26
As he notes, the disease was the dark side of the flourishing Silk
Route trade which underpinned the stupendous gold coinage
issued by Huviṣ ka (c. ad 154/55–186/7) and indeed the cross-
cultural imagery reflected in the seals of the period.

If the image of a putto driving a griffin engraved on cat. no.
I.7 seems a fairly faithful rendition of the subject matter of
Imperial Rome, the Herakles depicted on cat. no. I.4, with a
sketchily-rendered lion skin, heavy club and jockey-cap helmet,
represents a local interpretation of this most popular god of the
Hellenistic East. Herakles with a skin and club appears on the
reverse of coins of Kujula Kadphises (ad 30–80), who united
the Yuehzhi tribal confederation and became the first Kushan
emperor.27 According to the Rabatak inscription he was the
grandfather of Kaniṣ ka, and coins of Kaniṣ ka’s successor
Huviṣ ka also featured the standing Herakles motif. A group of
clay sealings impressed with Herakles figures in similar poses
were excavated in stratified levels of the 2nd–3rd century ad at
Rajghat, near Varanasi (Benares), Uttar Pradesh.28 These
provide evidence of trade, most probably with Bactria, of local
goods such as ivory carvings and cloth;29 whether they represent
the presence of foreign merchants, as the excavators suggest,
or simply the internal movement of goods is less easy to
determine.
An elegantly executed fusion of the attributes of the Roman
gods Mars (shield and spear) and Mercury (cockerel) appears
on cat. no. I.3. The warrior stands in a contraposto pose attired
in a version of a Graeco-Roman lorica, worn with Iranian-style
leggings; his headband displays a cockade. Carter has argued
that in this region the militant deity associated with a cockerel
can be identified as the Kushan Maaseno (Mahāsena); this
figure also relates to the Zoroastrian god Sraoša (mentioned
amongst the deities in the Rabatak inscription, see above) and
to the Gandhāran god Skanda/Kumāra.30 A similar figure
engraved on a carnelian in the British Museum (1892,1103.176)
was the personal seal of ‘Yōl, son of Orl’, according to its
Bactrian inscription, while another in Sir John Marshall’s
collection in Peshawar has a personal name in Kharoṣṭ hī:
pahatigasa (‘of Prabhātika’).31
Cat. no. I.4 is engraved with a personal name in Kharoṣṭ hī,
the script used primarily for the Prakrit dialect of Gāndhārī, the
language of the ancient kingdom of Gandhāra; it was in use
from the 3rd century bc to the 3rd century ad in modern
northern Pakistan and eastern Afghanistan. It is structurally
related to Brāhmī, represented on the gemstones in Group II,
which eventually replaced it. Callieri identified cat. no. I.4 as a
Roman gem depicting Mars Gravidus, proposing that a later,
less skilful hand inserted the inscription in the 2nd or 3rd
century ad.32 The fact that a palm is depicted as well suggests
the image may also evoke Mars Quirinus, bringer of peace, the
complementary aspect of Gravidus (the marching god). The
plumed helmet with a wide flat brim copies the Alexandrian
bull-horn helmet first depicted on coinage of the Bactrian king
Eukratides (c. 171–139 bc) and his successors which still
featured on Indo-Greek coinage issued in the eastern Punjab in
the 1st century bc.33 The close mineralogical grouping
established by GCEM confirms that this stone was carved in the
East, possibly as early as the 1st century ad.
Cat. no. I.5 also bears an inscription in Kharoṣṭ hī, but the
characters cannot be resolved easily into a name in any
language. The inscriptions giving a personal name on cat. no.
I.6 are in Middle Persian script typical of the early Sasanian
period. The figure bears some comparison with other garnet
gems engraved with what Callieri termed ‘Kushan period
Gandhāran tutelary deities’,34 but whether the inscriptions
Lithóis Indikois
were added subsequently in the 3rd or 4th century ad to an
earlier seal or whether classicising imagery like this persisted
into the Kushano-Sasanian period is difficult to assess. The fact
that the preparation of the seal is completely different to the
other stones in this group might argue for the latter.
Callieri placed cat. nos I.1 and I.3 in two different classes
on stylistic grounds but in the context of Group I it seems clear
that they are stylistically related to one another and prepared
in an identical manner, with a fine bevel around the
circumference. Cat. nos I.1 and I.3 are also both chipped at the
top, perhaps suggesting they were removed from similar
settings. These, together with cat. no. I.7, display the most
expert carving of the gems in this first group. Cat. nos I.2 and
I.5 are blocked out and worked in much the same manner but
with considerably less detail and cat. no. I.6 is even more
sketchily rendered. This posits the question of whether gems
could be sold in different stages of preparation. Obviously the
latter, less refined examples would be less expensive than the
former. Alternatively we might imagine that lapidaries of
varying competence shared access to the same garnet sources.
In any case, it seems likely this source or sources were
exploited for some considerable time, before, during and after
the period of the Great Kushans.
1. Nanā seated on lion, facing left, Bactrian inscription
2nd–3rd century ad
19.7 x 17 x 5.5mm; low cabochon, flat back, bevelled, chipped top and
bottom.
Reg. no. 1892,1103.100; Major-General Sir Alexander Cunningham
collection; Franks Bequest.
Bibliography: Cunningham 1892 (n. 9), pl. XXII.18, and 116 (as jacinth;
he notes a ‘duplicate in 'red carnelian’ in the collection of General
Pearse’); idem, 1893 (n. 9), 128; A. Maricq, ‘La grande inscription de
Kaniṣka et l’étéo-tokharien, l’ancienne langue de la Bactriane’, Journal
Asiatique 246/4 (1958), 345–439, at 420, no. 14; Göbl (n. 12), vol. 1, 222–3,
G3 (as hyacinth), vol. III, pl. 85; J.M. Rosenfield, The Dynastic Art of the
Kushans, Berkeley-Los Angeles, 1967, 102, seal 4; Bivar 1968 (n. 11), pl.
I,2; B.N. Mukherjee, Nana on Lion. A Study in Kushāna Numismatic Art,
Calcutta, 1969, 23, 110–11, pl. IV.9 (as jacinth); V.A. Livšic, ‘K otkrytiju
baktijskix nadpisejna Kara-tepe’, Raskopki buddiijskie pešč ery Kara-
tepe v Starom Termeze. Osnovnye itogi rabot 1963–1964 gg, Moscow,
1969, 47–81, at 57–8; Göbl (n. 24), 153, pl. 177, no. 4; B. Ja. Staviskij,
La Bactriane sous les Kushans. Problèmes d’histoire et de culture, Paris,
1986, 145, n. 55; Callieri (n. 14), 197–8, 233, 310, cat. no. U 7.23, Class IX;
Ghose (n. 24), 100, fig. 10b.
Inscription: ФРЕІΧΟAΔHΟ/FreiXoadēo/Freykhwadew, a personal name
equivalent to Sogdian ‘friend of the lord’ (Sims-Williams in Callieri (n.
14), 310; see also G.D. Davary, Baktrisch. Ein Wörterbuch auf Grund der
Inschriften, Handschriften, Münzen und Siegelsteine, Heidelberg, 1982,
107, 190, Sig. 3 and N. Sims-Williams, Bactrian Personal Names,
Iranisches Personennamenbuch, Fasc. 7, R. Schmitt, H. Eichner,
B.G. Fragner and V. Sadovski (eds), Band II, Mitteliranische
Personennamen, Vienna, 2010 (hereafter IPNB II/7), 144 (no. 501).
Gemmology: almandine garnet. Inclusions: large tabular and prismatic
opaque inclusions. CL.
'Gems of Heaven’ | 31
Adams, Lüle and Passmore
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions are amphibole, in the tremolite-ferroactinolite
series. All inclusions identified as amphibole belong to this same
compositional series. EP.
2. Draped female deity, standing frontal, head to left, wearing polos,
cornucopia in left hand, male child before, Bactrian inscription
2nd–3rd century ad
17.3 x 14.3 x 3.7mm, low cabochon, flat back.
Reg. no. 1892,1103.173; Major-General Sir Alexander Cunningham
collection; Franks Bequest.
Bibliography: Cunningham 1892 (n. 9), 111–12, no. 61, pl. xxi.16 (as
agate); idem 1893 (n. 9), 127; Maricq (cat. I.1), 420, no. 10; Göbl (n. 12),
vol. I, 223–4, G5, vol. III, pl. 85 (as Ardoxso); Bivar 1968 (n. 11), pl. I.3;
Mukherjee (cat. I.1), 118, pl. IX.38 (as agate); Livšic (cat. I.1), 57–9;
Göbl (n. 24), 153, pl. 177, no. 6; Callieri (n. 14), 114, 310, U 7.22.
Inscription: ÞΟΟΓΑΟ/Šoogao/Shu-gaw, a compound personal name
(Sims-Williams in Callieri (n. 14), 310; Davary (cat. no. I.1), 108, Sig. 5;
IPNB II/7 (cat. no. I.1), 158, no. 559.
Gemmology: probably almandine garnet, medium dark strongly
brownish red; however, the spectrum is unclear and requires Raman.
Inclusions: large opaque rounded crystal inclusions with clusters of
very small colourless and transparent crystals which may be apatite.
CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The large, opaque, rounded crystals were confirmed as
amphibole, but the small, colourless transparent crystals were not
analysed. EP.
3. Cuirassed warrior standing frontal, head to right with spear in left
hand, shield with cockerel in his right
2nd–3rd century ad
20.5 x 16 x 4.5mm, low cabochon, flat back, bevelled, chipped at bottom
and proper left.
Reg. no. 1892,1103.170; Major-General Sir Alexander Cunningham
collection; Franks Bequest.
Bibliography: P. Callieri, ‘La glittica romana nel Gandhāra: presenze e
influssi’, Atti dell’Accademia Nazionale dei Lincei (1989), s. 8°,
Rendiconti, 44 (1991), 243–57, at 251, fig. 11; Callieri (n. 36), 418, pl. 96,
no. 21; Callieri (n. 14), 191, 234, 270–71, U 7.5, Class XI.
Gemmology: probably almandine garnet, dark orangey brown;
however, the spectrum is unclear and requires Raman. Inclusions:
clusters of tabular black and opaque crystals, many prismatic
transparent and colourless crystals, possibly apatite. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The tabular, black, opaque crystals were shown to be
amphibole, but the transparent, colourless crystals were not analysed.
EP.
32 | 'Gems of Heaven’
4. Helmeted warrior standing in profile to right, spear in right hand,
left arm upraised, palm behind, Kharoṣṭ hı̄ inscription
1st–2nd century ad
19 x 16.5 x 6.5mm (with bezel), cracked.
Reg. no. 1880.3730; Charles Masson collection.
Bibliography: Callieri (n. 14), 186, 282, U 5.1.
Inscription: Mastukasa: (‘[seal] of Mastaka’), equivalent to Sanskrit
mastaka (summit or peak) (Garbini in Callieri (n. 14), 282); Dastukasa:
(‘of Dastuka’) (H. Falk, pers. comm. August 2011).
Gemmology: back covered by metal bezel, so no spectroscope reading.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. EP.
5. Herakles with lion-skin and club, standing frontal, head turned to
right, helmeted, Kharoṣṭhı̄ inscription
1st–3rd century ad
18.8 x 16.1 x 5.2mm, cabochon, flat back; paper label ‘Punjab’35 on
reverse as with some Cunningham stones.
Reg. no. 2005,0815.4.
Inscription: damage to the surface obscures some letters and the
symbol above the right shoulder could be a glyph or an attribute.36 The
Kharoṣṭ hī letters can be read: haṭ akhohajaḍ a-sa, but the language
(?Scythian) and meaning of the name are unknown. HF.
Gemmology: almandine garnet. Inclusions: large tabular and prismatic
opaque inclusions. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions were shown to be amphibole. EP.
6. A figure in profile to right, ?Herakles with ?spears and ?lion-skin
1st–3rd century ad
10.6 x 11.6 x 2.7mm, broken, low cabochon, flat back.
Reg. no. 2005,0815.16.
Gemmology: almandine garnet, medium slightly orangey red.
Inclusions: clusters of large opaque dark crystals, mostly tabular;
groups of prismatic colourless transparent crystals and very small
needles in a cloud. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. This garnet had the most almandine-rich composition of all the
garnets analysed. The large, opaque dark inclusions were shown to be
amphibole, and a single colourless crystal was quartz. This stone was
the only one where two different types of inclusion were measured in
the same stone. EP.

7. A putto harnessing a griffin to right
1st–2nd century ad
12.5 x 11.9 x 3.2mm, originally larger, low cabochon, deeply hollowed
back.
Reg. no. 1880.3576; IM.Gems.41; Charles Masson collection.
Bibliography: Callieri (n. 14), 52–3, cat. 1.25.
Gemmology: almandine garnet, medium slightly brownish orangey
red. Inclusions: almost fibrous/silk cloud with black opaque tabular
crystals, occasional small transparent colourless crystals which may
be apatite. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions were shown to be amphibole. EP.
8. Draped female figure, standing frontal, head turned to left, left
hand upraised, wearing a fillet/diadem, Middle Persian inscriptions
2nd–4th century ad
13.8 x 9 x 2.8mm, low cabochon, flat back.
Reg. no. 2005,0815.7.
Inscriptions: Šḥpwh. ry d'lyk (Shāhpuhr-dārīg), personal name. NS-W.
Gemmology: possibly almandine garnet but the spectrum is unclear
and requires Raman. Inclusions: fingerprint37 inclusions with dark
opaque tabular crystals, no other inclusions. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions were shown to be amphibole. EP.
Group II
This group has been gemmologically identified as almandine
garnet of purple and reddish purple colours. Raman
spectroscopic analyses show that the composition of these
garnets is in the pyrope-almandine range, but with a greater
pyrope component (more Mg) than Group I. All the garnets in
Group II have the same set of inclusions which in some cases
included crystals of distinct structure and yellowish colour.
Spectra produced by analysing the inclusions using Raman
spectroscopy have not yet been matched to reference spectra,
and these remain as yet unidentified.
These intaglios are stylistically and iconographically
heterogenous, with subject matter ranging from animal
representations, narrative human subjects, a tamgha,
inscriptions alone and individualised portraits. Bactrian and
Brāhmī scripts are represented. Although not all are easy to
date, the intaglios appear to span two centuries or more; this
could suggest that the locality or localities that produced these
stones provided gems over centuries to different workshops.
Depending upon their ultimate source, they could provide
evidence of regional trade in garnets between the North-West
frontier and India, but much more investigation needs to be
done before reaching any conclusions.
Lithóis Indikois
Kushan rule in Bactria came to an end not long after the
Persian dynasty of the Sasanians came to power. Having taken
Iran from the Parthians around ad 224, by c. ad 233 they had
conquered Bactria as well.38 The iconographic themes on cat.
nos II.1–3 appear on Persian intaglios dated to the 4th and 5th
centuries ad. Lions and humped bulls such as those on cat.
nos. II.1–2, for example, are common subjects on Sasanian
seals and sealings,39 cat. no. II.3 can be paralleled on Sasanian
gems depicting seated figures with their arms upraised40 and
the winged horse Pegasus (cat. no. II.4), perhaps because of its
Zoroastrian and astrological connections, was popular on
Sasanian glyptic.41 Mounted riders and bull and lion motifs,
however, appear earlier on coinage struck by Indo-Scythian
rulers and their satraps in northwest India from the late 1st
century bc and 1st century ad,42 so seals with such imagery may
also have drawn from indigenous objects. The powerful
depiction of two rearing confronted lions above their prey on
cat. no. II.1, for example, finds good parallels on Gandhāran
sculpture.43 Cat. nos II.2 and II.3 are stylistically related,
executed with broad wheel strokes, while cat. no. II.1 is
markedly more skilfully carved. Although the shape of cat. no.
II.1 recalls those of Group I.1–5, the extremely low, almost flat
profiles of cat. nos II. 2–3 and the round shape of II.4 can be
compared to some Sasanian period garnet intaglios.
In the decades after ad 350, Bactria, Gandhāra and
northwest India fell under the control of clans of tribes of
mixed Hunnic and Iranian ethnicity – known in the Roman
and Sasanian sources as the Chionites, Kidarites and
Hephthalites. Initially they ruled under Sasanian suzerainty
and, until c. ad 475, were loyal to the Sasanid kings. The
historical picture in the region, based upon complex,
overlapping coin issues, has recently undergone considerable
revision, although there is still no agreement about how the
issues of the various Hunnic rulers relates to their authority
and presence in the different regions of the North-West
frontier.44 The present picture sees the Huns taking over the
Sasanian mint in Kāpiśī/Kabul after ad 384. Kidāra, the ruler
who is the namesake of the Kidarites, is known from coinage
struck after Kushano-Sasanian examples; the Kidarites
maintained control in Sogdia, Bactria and Gandhāra into the
mid-5th century ad and Kashmir until the later 5th century
ad.45 Coinage issued in the late 4th century ad with the
Bactrian legend 'alχanno', gives the name to the tribal groups
now referred to as Alkhan or Alchon Huns. These clans spread
from the regions around Kabul east to Gandhāra, eventually
controlling northwest India. The group sometimes now
referred to as the ‘Genuine’ Hephthalites came to power in the
late 5th century ad. Whereas the power base of the Alchon
Huns was located south of the Hindu Kush, the Hephthalites
issued coins from mints centred in ancient Bactria, north of the
Hindu Kush. They are famous for having captured the Sasanian
king Peroz (d. ad 484) and after his death minted their own
coins in Balkh in imitation of his last issues. As Alram has
noted, the Hepthalites played a key role in trade between
Central Asia and China, ruling during the period when
maritime trade from Persia via Sri Lanka to south China was
also thriving.46
Returning to the gems, cat. no. II.5 is engraved with a
tamgha or tribal sign; another garnet in the collection
(2005,0815.23) is engraved with a closely similar sign. Both of
'Gems of Heaven’ | 33
Adams, Lüle and Passmore
these consist of a triangle whose side bars cross at the apex and
extend into asymmetrical terminals; a short bar separates the
triangular element from the footed stand or base. I have traced
no exact parallels for these tamghas. Triangular elements are
found on Sarmatian tamghas dating from the Hellenistic
period onwards47 and occasionally on Sasanian gems,48 while
the lower base is characteristic of Sasanian and Kushan period
tamghas.49 Tamghas appear in Bactria on coins of the late 4th
century ad issued by the Huns, first in the name of Varhrān III
or IV and Kidāra,50 and another well-known garnet gem in the
British Museum collection, from the relic deposit of Tope Kelan,
Hadda, is engraved with the not dissimilar Alchon Hun
tamgha,51 so perhaps it is appropriate to associate these gems
with this time period.
Group II includes three flattened and irregular garnet
stones engraved only with inscriptions in Brāhmī, the great
‘syllabic alphabet’ which formed the basis of many writing
systems throughout South, Southeast and East Asia. Cat. nos
II.6–7 are personal names, one analogous to ‘boar’, the avatar
of the Hindu Viṣṇu; the name ‘boar’ was also significant in
Sasanian culture. Cat. no. II.7 begins with a Zoroastrian fire
altar followed by a name related to others of royal status. Other
seals and impressions are known with fire altars, including a
lapis lazuli seal with a Brāhmī inscription (sangha-vala) which
formed part of the Oxus treasure and another of the same
material inscribed with the name Ripuśalya (‘a javelin for his
foes’).52 Callieri published other garnets bearing only Brāhmī
akṣ aras and three further examples (untranslated) are in the
collection of the Cabinet des médailles, Paris; two of these were
purchased in the bazaar in Peshawar.53 On epigraphical
grounds these inscriptions date from the 3rd to the 5th century
ad, with examples such as cat. no. II.7 perhaps reflecting the
spread of coinage with the fire altar motif imitating issues of
Sāpūr III (ad 383–88).54
Three of the remarkable portrait seals from the eastern
Hunnic period represented in the British Museum collection
belong in Group II. These follow on from royal and princely
seal stones carved under the Sasanians, the finest of which
date to the 3rd and 4th centuries ad,55 but diverge from the
hieratic treatment of those representations and recapture some
of the Hellenism of the Indo-Greeks. Closest to the Sasanian
portraits are examples like the seal stone of Tiroado, cat. no.
II.9, who wears a diadem with ties above his bowl-like cap of
hair; unlike conventional Sasanian representations, he is
clean-shaven with long mustachios. A number of seal
impressions from this general class of stone have survived56
and it is possible some were used in an official or quasi-official
capacity. Callieri placed cat. no. II.9 in his stylistic Class II,
which he suggested originated in Afghanistan at the end of the
4th and beginning of the 5th century ad. In his 1997 catalogue,
he referred to this group as ‘Kidarite’, and the following group,
Class IV, as ‘Hephthalite’, but was clear that these were not
absolute designations of ethnicity.57
The male and female portraits, cat. nos II.10–11, depict
local dignitaries with distinctive jewellery, hairstyles and
dress, identified with personal names in Brāhmī. Callieri
placed these in his Class IV ‘Hephthalite’ group and suggested a
date from the second half of the 5th to the end of the 6th
century ad.58 He suggested that the prevalence of Brāhmī
inscriptions on these Hunnic period stones perhaps indicates a
34 | 'Gems of Heaven’
North Indian provenance for the engravings.59 Falk has recently
recorded a number of Hunnic period seals with Indian names
which provide an indication of cultural assimilation in
Gandhāra.60
The portrait of Rostam (cat. no. II.10) remains quite close
to the Hunnic type seen on cat. no. II.9, although the feature of
the diadem has been replaced with three rows of puffy curls.
Göbl saw some parallels on coinage issued by the Nēzak king
Narendra, struck in the second half of the 6th century ad,61 but
for another stylistically similar garnet seal, formerly in the
Kevorkian and now in the Rosen collection in New York (21 x 18
x 5mm), Lerner has drawn parallels with 5th-century ad
Hephthalite coinage; it bears a Bactrian inscription reading
alχono.62 Paleographically the inscription on the British
Museum stone belongs to the late Gupta period, in this case
probably to the 6th century ad,63 and Grenet has explored the
possibility that this might be evidence for the cult of the great
hero Rustam amongst the Hephthalites.64 The Alchon lost their
power over Gandhāra in the second quarter of the 6th century
ad and groups of them returned to the regions of Kabul and
Ghazni where their coinage overlaps with that produced by
another Iranian clan, the Nēzak.65 Whether the contrasting
assessments of the Rosen and British Museum seals can be
reconciled by these historical circumstances remains to be
explored.
The bold portrait in cat. no. II.11 is one of a small number
of female representations, suggesting its owner must have been
of very high, if not royal, status. The engraving relates to gems
produced for the later Hunnic kings.66 The hairstyle of this
older woman depicts either a braid or a row of double curls on
the forehead with the hair then pulled back into a coiled bun;
this is apparently secured by a net, recalling Hellenistic
traditions. The oversized wheel earrings and necklace strung
with round beads or pearls finds parallels in Indian jewellery
fashions. The inscription ‘splendour of the lotus’ belongs in the
Buddhist domain, as do other inscriptions on related stones in
Callieri’s class IV.67
1. Two rearing lions above a zebu, Bactrian inscription above
2nd–4th century ad
23.6 x 20.1 x 6.3mm, cabochon.
Reg. no. 1907,1111.26; Sir Aurel Stein collection, purchased at Yōtkan
site, Yo.0008a.
Bibliography: Stein (n. 15), vol. I, 210–11, vol. II, pl. XLIX, Y.008.a; Göbl
(n. 12), vol. I, 227, G8, vol. III, pl. 85; Zwalf (n. 43).
Inscription: variant transcriptions of the Bactrian are gathered
together in Davary (cat. no. I.1), 109, 186, Sig. 8, but Nicholas Sims-
Williams (pers. comm. January 2011) is of the opinion that the
inscription is too unclear to be deciphered properly.
Gemmology: almandine garnet, dark, slightly reddish purple.
Inclusions: low relief fingerprints, zircon crystals with dark discoid
fractures, yellowish transparent crystals which may be apatite. CL.

Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions were not analysed. EP.
2. Lion and horned ?bull en face
3rd–4th century ad
15.8 x 12 x 3mm, low profile, almost flat, chipped along all edges.
Reg. no. MAS 212; Sir Aurel Stein collection, from Khotan.
Gemmology: almandine garnet, medium slightly reddish purple.
Inclusions: clouds (very fine silk?), zircon crystals with dark discoid
fractures, yellowish transparent crystals which may be apatite, no
needles. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusion analysed produced a Raman spectrum that has
not yet been matched to a reference spectrum, and remains
unidentified. EP.
3. Two seated figures, the smaller on lap with instrument
3rd–4th century ad
12 x 11 x 2.8mm, low cabochon, flat back, chipped, traces of whitish
paste in engraving.
Reg. no. 1892,1103.146; Major-General Sir Alexander Cunningham
collection (no. 39); Franks Bequest.
Bibliography: Dikshit (n. 10), 128, no. 38 (1st–2nd century ad).
Gemmology: almandine garnet, medium slightly reddish purple.
Inclusions: yellowish transparent prismatic crystals, fine fingerprints
and small scattered rutile needles. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions were not analysed. EP.
4. Pegasus to right
3rd–5th century ad
12 x 11.4 x 2.8mm, very low cabochon, flat back, high polish, traces of
white paste in engraving.
Reg. no. MAS 219; Sir Aurel Stein collection, from Khotan.
Gemmology: almandine garnet, medium dark reddish purple.
Inclusions: yellowish transparent prismatic crystals, zircon crystals
with dark discoid fractures and some small clouds of very fine silk. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions were not analysed. EP.
Lithóis Indikois
5. Tamgha
4th–5th century ad?
11.6 x 9.2 x 4mm, cabochon of irregular shape, slightly hollow back.
Reg. no. 2005,0815.14.
Gemmology: almandine garnet, medium dark reddish purple.
Inclusions: variously-sized yellowish transparent crystals and zircon
crystals with dark discoid fractures around them. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet, although this garnet had the joint largest pyrope component
(most Mg-rich) of all the garnets analysed. The inclusion analysed
produced a Raman spectrum that has not yet been matched to a
reference spectrum, and remains unidentified. EP.
6. Brāhmı̄ inscription: varāhula
3rd–5th century ad
17.3 x 12.6 x 3.2mm, low cabochon, flat back, bevelled.
Reg. no. 2005,0815.15.
Inscription: Varā hula is a hypochoristic name, abbreviated from a
compounded name which includes varā ha, the boar, often standing for
Viṣṇu. HF.
Gemmology: almandine garnet, medium slightly reddish purple.
Inclusions: low relief fingerprints, zircon crystals with dark discoid
fractures, yellowish transparent crystals, perhaps apatite, no needles.
CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions were not analysed. EP.
7. Brāhmı̄ inscription: fire altar rolādityah.
3rd–5th century ad
18.5 x 14 x 4.9mm, low cabochon hollow back, very worn, irregular,
chipped, broken and repaired.
Reg. no. 1903,1116.3; W.S. Talbot collection.
Inscription: Rolā dityaḥ is probably a local form of Lolāditya, which
compares to lolārka, ‘playful sun’. Similar names of kings in the
Rājataraṅgiṇ ī are Madanāditya and Lalitāditya. HF.
Gemmology: almandine garnet, medium reddish purple. Inclusions:
fingerprints, zircon crystals with dark discoid fractures. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions were not analysed. EP.
'Gems of Heaven’ | 35
Adams, Lüle and Passmore
8. Fragmentary Brāhmı̄ inscription: ///[ka?]vı̄ra
4th–5th century ad
11.5 x 11.4 x 3mm, low cabochon, hollowed back, badly chipped.
Reg. no. 2005,0815.17; originally stored with gems from Franks
Bequest.
Gemmology: almandine garnet, medium purple. Inclusions: prismatic
transparent colourless crystals which may be apatite. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions were not analysed. EP.
9. Male bust profile to right, diademed, mustachioed, Bactrian
inscription
Late 4th–first half of 5th century ad
18.7 x 14.4 x 4.1mm, cabochon with irregular bevel, flat back, all edges
chipped.
Reg. no. 1892,1103.171; Major-General Sir Alexander Cunningham
collection; Franks Bequest.
Bibliography: Göbl (n. 12), vol. I, 241–2, G33, vol. III, pl. 86; Bivar 1968
(n. 11), pl. II,7 (OA 28); Callieri (n. 14), 202, 311, cat. U 7.34, Class II;
Callieri 1999 (n. 57), 282–4, pl. 4. as Cat. U7.36, Class D.
Inscription: τιροαδο (Tiroado) Davary (cat. no. I.1), 112, Sig. 33; Sims-
Williams in IPNB II/7, 136–7 (no. 468) suggests: ‘Dedicated to Tir and
Wad’, a personal name combining the divine names T īr (τειρο) and
Wād ‘Wind’ (οαδο), both attested on Kushan coins.
Gemmology: almandine garnet, colour medium dark, slightly reddish
purple. Inclusions: a few fissures with brown FeO staining, various
sized zircon (?) crystals with dark discoid fractures around them. No
needles. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions were not analysed. EP.
10. Male bust, profile to right, earring, facing Brāhmı̄ inscription
6th–7th century ad
21 x 18.3 x 5.9mm, low cabochon, flat back, chipped at bottom.
Reg. no. 1880.3505; Charles Masson collection.
Bibliography: Callieri (n. 14), 202, 229–31, 285, 297, cat. no. U 7.36, Class
IV; Callieri 1999 (n. 57), 282–4, pl. 4 as 7.41; Grenet (n. 64).
Inscription: Rostama (Rustam), a personal name.
36 | 'Gems of Heaven’
Gemmology: almandine garnet, medium dark, slightly reddish purple.
Inclusions: low relief fingerprints, zircon crystals with dark discoid
fractures, yellowish transparent crystals which may be apatite, no
needles. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet. The inclusions were not analysed. EP.
11. Female bust profile to right, Brāhmı̄ inscription
6th–7th century ad
20.2 x 17.2 x 3.4mm, low cabochon, flat back.
Reg. no. 1892, 1103.121; Major-General Sir Alexander Cunningham
collection (no. 8); Franks Bequest.
Bibliography: Dikshit (n. 10), 125, pl. V,14; Göbl (n. 12), vol. I, 251–2, G55,
vol. III, pl. 87; Callieri (n. 14), 204, cat. no. U 7.42, Class IV; Callieri 1999
(n. 57), 282–4, pl. 5, cat. no. U 7.42, Class D.
Inscription: Patmaśrī, ‘splendour of the lotus’ (Garbini in Callieri (n.
14), 285 (6th–7th century ad).
Gemmology: almandine garnet, medium reddish purple. Inclusions:
large fingerprints, zircon crystals with dark discoid fractures,
yellowish transparent prismatic crystals (?apatite), no needles. CL.
Raman spectroscopic analysis: confirms almandine composition of the
garnet, although this garnet had the joint largest pyrope component
(most Mg-rich) of all the garnets analysed. The inclusion analysed
produced a Raman spectrum that has not yet been matched to a
reference spectrum, and remains unidentified. EP.
Notes
1. The term ‘λιθοις ινδικος’ was used by Priscus to describe the
precious stones brought to Attila by the eastern Roman diplomatic
mission of ad 449 (Prisci Fragmenti, in Historici Graeci Minores I
(ed. L. Dindorf), Leipzig, 1870, 290, 8.22. The definitions and
boundaries of Central and South Asia are variable. Many of these
stones can be associated with the modern regions of Afghanistan
and Pakistan which are included in both geopolitical regions in
different schemes.
2 E.J. Gübelin and J.I. Koivula, Photoatlas of Inclusions in Gemstones,
Vol. 1, Basel, 1986, 287–97; idem, Photoatlas of Inclusions in
Gemstones, Vol. 2, Basel, 2005, 430–85, at 439.
3 S. Greiff, ‘Naturwissenschaftliche Untersuchungen zur Frage der
Rohsteinquellen für frühmittelalterlichen Almandingranat-
schmuck rheinfränkischer Provenienz’, Jahrbuch des Römisch-
Germanischen Zentralmuseums, Mainz 45/2 (1998), 599–646. See
pp. 16–17 and n. 63 in Adams above.
4 V. Ball, A Manual of the Geology of India, Part 3, Economic Geology,
Calcutta, 1881, 521–8; L.A.N. Iyer, Bulletins GSI Series A – Economic
Geology, no. 18, Indian Precious Stones, Delhi, 1961, 50–4; R.V.
Karnath, Gems and Gem Industry in India, Memoir 45, Geological
Society of India, Bangalore, 2000, 204.
5 A. Shareq, V.M. Chmyriov, K.F. Stazhilo-Alekseev, V.I. Drunov,
P.J.G. Gannon, L.N. Rossovskiv, A. Kh. Kafarskiy and
E.P. Malyarov, Mineral Resources of Afghanistan, United Nations
Development Programme, Kabul, 1977, 202; G.W. Bowersox and
B. Chamberlain, Gemstones of Afghanistan, Tucson, 1996; http://
www.gems-afghan.com/NGO8thdraft.pdf; M. Qasin Jan and
D.R.C. Kempe, ‘Recent researches in the Geology of Northwest
Pakistan’, Geological Bulletin University of Peshawar 6/1 (1970),
1–32; S.K. Kakar, S.B. Mian and J. Khan, ‘The Geology of Jandul
Valley, Western Dir’, Geological Bulletin University of Peshawar
(1971), 54–73, at 71; D. Blauwet, ‘Pakistan and Afghanistan’, in
H.A. Gilg, D. Hile, S. Liebetrau, P. Modreski, G. Neumeier and
G. Staebler (eds), Garnet, Great Balls of Fire, extralapis 9, East

Hampton, Conn., 2008; A. Khan, ‘Investment opportunities in the
mineral sector of Pakistan’, Journal of Himalayan Earth Sciences 43
(2010), 42.
6 M. Bimson, ‘Notes on the Material of the Seals’, in A.D.H. Bivar,
Catalogue of the Western Asiatic Seals in the British Museum, Stamp
Seals, II, The Sassanian Dynasty, London, 1969, 35–6.
7 The site can be identified with the Indo-Greek Kāpiśī (M.L. Carter,
‘Begram’, Encyclopædia Iranica 1989 (www.iranica.com/articles/
begram), but recent research has also supported Masson’s
association of the site with the ancient city of Alexandria ad
Caucasum, founded by Alexander the Great (E. Errington, ‘Charles
Masson’, Encyclopædia Iranica 2004 (www.iranica.com/articles/
masson-charles); P. Bernard, ‘Alexandrie du Caucase ou
Alexandrie de l’Oxus’, Journal des Savants (1987), 217–42).
8 In M.A. Stein, ‘Archaeological Work about Khotan’, Journal of the
Royal Asiatic Society of Great Britain and Ireland (1901), 295–300,
he describes how gems, like coins, were extracted by villagers who
regularly mined and washed the alluvial soils for ‘treasure’; he
himself dispatched ‘professional treasure seekers’ on
reconnaissance missions to known sites.
9 Sir A. Cunningham, ‘Coins of the Kushans or Great Yeuh-ti’,
Numismatic Chronicle, 3rd series, xii (1892), 40–82 and 98–159;
idem, ‘Coins of the Later Indo-Scythian and Later Kushâns, Scytho-
Sasanians and Little Kushâns’, Numismatic Chronicle, 3rd series,
xiii (1893), 93–128 and 166–202.
10 M.G. Dikshit, ‘Cunningham Collection of Seals in the British
Museum’, Journal of the Numismatic Society of India 22 (1960), 123–30.
11 A.D.H. Bivar, The Kushano-Sassanian Episode, unpublished Ph.D.
diss., Oxford University, 1955; idem, Kushan and Kushano-
Sasanian Seals and Kusano-Sasanian Coins: Sasanian Seals in the
British Museum, Corpus Inscriptionum Iranicarum III, Vol. IV,
London, 1968.
12 R. Göbl, Dokumente zur Geschichte der Iranischen Hunnen in
Baktrien und Indien, Wiesbaden, 1967.
13 E. Errington, ‘Rediscovering the collections of Charles Masson’, in
M. Alram and D.E. Klimburg-Salter (eds), Coins, Art and
Chronology. Essays on the pre-Islamic History of the Indo-Iranian
Borderlands, Österreichischen Akademie der Wissenschaften, Vol.
280, Vienna, 1999, 207–37, at 212–28, pl. 7.3–4, 7.14, 7.18–19.
14 P. Callieri, Seals and Sealings from the North-West of the Indian
Subcontinent and Afghanistan (4th Century bc – 11th Century ad),
Local, Indian, Sasanian, Graeco-Persian, Sogdian, Roman, Naples,
IsIAO, 1997.
15 Sir A.M. Stein, Ancient Khotan, 2 vols, Oxford, 1907.
16 Callieri (n. 14), 34–5, pl. II; J. Lerner, ‘Some Central Asian Seals in
the Rosen Collection’, in Alram and Klimburg-Salter (n. 13), 265–
76, at 265.
17 Cf. M. Maaskant-Kleibrink, Catalogue of the Engraved Gems in the
Royal Coin Cabinet, The Hague. The Greek, Etruscan and Roman
Collections, The Hague, 1987, 60, fig. 2.
18 J.D. Rouse, Garnet, London, 1987, 130; M. Henig with D.
Scarisbrick and M. Whiting, Classical Gems, Ancient and Modern
Intaglios and Cameos in the Fitzwilliam Museum, Cambridge,
Cambridge, 1994, 428, no. 884 (reg. no: E.58.1982), identified as
‘grossular-melanite garnet’; Dr. Lüle has also identified a yellow
grossular seal in the Antikensammlung in Berlin (Thoresen, this
volume, Pl. 3).
19 B.A. Kolesov and C.A. Geiger, ‘Raman spectra of silicate garnets’,
Physics and Chemistry of Minerals 25 (1998), 142–51; D. Smith, ‘The
RAMANITA© method for non-destructive and in situ semi-
quantitative chemical analysis of mineral solid-solutions by multi-
dimensional calibration of Raman wavenumber shifts’, Spectro-
chimica Acta Part A: Molecular and Biomolecular Spectroscopy 61
(2005), 2299–2314; D. Bersani, S. Andò, P. Vignola, G. Moltifiori,
I.G. Marino, P.P. Lottici and V. Diella, ‘Micro-Raman spectroscopy
as a routine tool for garnet analysis’, Spectrochimica Acta Part A:
Molecular and Biomolecular Spectroscopy 73 (2009), 484–91.
20 Kolesov and Geiger (n. 19).
21 Ibid.
22 R.T. Downs, ‘The RRUFF Project: an integrated study of the
chemistry, crystallography, Raman and infrared spectroscopy of
minerals’, Program and Abstracts of the 19th General Meeting of the
International Mineralogical Association, Kobe, Japan, 2006, O03–13.
23 The use of Bactrian diminished with the rise of the Hunnic
confederations in the 4th and 5th centuries ad, but examples dated
as late as the 9th century ad have survived.
Lithóis Indikois
24 Callieri (n. 14), 198; cf. R. Göbl, System und Chronologie der
Münzprägung des Kušānreiches, Vienna, 1984, 43, pl. 167, Nana 4;
M. Ghose, ‘Nana: The “Original Goddess on the Lion”’, Journal of
Inner Asian Art and Archaeology 1 (2006), 97–112, at 100.
25 N. Sims-Williams, ‘A New Bactrian Inscription of Kanishka the
Great, pt. 1: “The Rabatak Inscription, Text and Commentary”’,
Silk Road Art and Archaeology 4 (1995–96), 77–97; J. Cribb, ‘The
Early Kushan Kings: New Evidence for Chronology, Evidence from
the Rabatak Inscriptions of Kanishka I’, in Alram and Klimburg-
Salter (n. 13), 177–205; see also in the same volume, R. Göbl, ‘The
Rabatak Inscription and the Date of Kanishka’, 151–76. H. Falk,
‘The Yuga of Sphujiddvaja and the Era of the Kuṣanas’, Silk Road
Art and Archaeology 7 (2001), 121–36, has analysed early Sanskrit
astronomical texts which suggest that the start date of the Kaniṣka
inscriptions should begin around ad 127/28.
26 Bivar summarises his arguments in A.D.H. Bivar, ‘Kushan Dynasty
i. Dynastic History’, Encyclopædia Iranica 2009 (www.iranica.
com/articles/ Kushan Dynasty i).
27 J. Cribb and O. Bopearachchi, ‘Greek hero-god, Heracles’, in
E. Errington and J. Cribb (eds), Crossroads of Asia, Transformation
in Image and Symbol (exh. cat., Fitzwilliam Museum), Cambridge,
1992, 79–82, nos 75–6.
28 K.K. Thaplyal, ‘Greek Devices on some Rajghat Sealings: a Review’,
Journal of the Numismatic Society of India 31 (1969), 135–9;
V.S. Agrawala, Varanasi Seals and Sealings, Varanasi, 1984, 16–18,
pl. XIX, nos 493–6.
29 Thaplyal, ibid., 137–9.
30 M.L. Carter, ‘Buddhist Aspects of the Imagery of Skanda in
Gandhāra and Central Asia’, in South Asian Archaeology 2003,
Proceedings of the Seventeenth International Conference of the
European Association of South Asian Archaeologists (EASAA) (7–11
July 2003, Bonn: U. Franke-Vogt and H.-J. Weisshaar [eds]),
Aachen, 2005, 399–404; M.L. Carter, ‘The Gods of the Rabatak
Inscription’, in South Asian Archaeology 1999, Proceedings of the
Fifteenth International Conference of the EASAA (5–9 July 1999,
Leiden), (E.M. Raven [ed.]), Groningen, 2008, 289–95. Callieri
(n. 14), 191, identified the figure as the Hindu god Kārttikeya whose
symbol was the peacock. On Gandhāran images of Kumāra and
their relationship to Palmyrene sculpture, see P. Pal, Indian
Sculpture, Vol. I, c. 500 bc–ad 700, A Catalogue of the Los Angeles
County Museum of Art Collection, Berkeley, Los Angeles and
London, 1986, 164, no. S41.
31 Callieri (n. 14), 106, no. 7.2 and 191, no. U 7.4; A. ur Rahman and H.
Falk, Seals, Sealings and Tokens from Gandhāra (Monographien
zur Indischen Archäologie, Kunst und Philologie, 21), Wiesbaden,
2011, 204, no. PM 07.03.01.
32 Ibid., 186.
33 Cf. O. Bopearachchi, Monnaies Gréco-Bactriénnes et Indo-Grécques,
Catalogue Raisonné, Bibliothèque nationale de France, Paris, 1991,
202–15, pls 16–22 (Eukratides I [c. 170–145 bc]) and his successors
in eastern Bactria and 368, pl. 68 (Apollophane [35–25 bc]).
34 Cf. Callieri (n. 14), 107, 7.5; 195–6, U 7.17.
35 ‘Punjab’ in this context refers to the Punjab Province in British
India. This covered a large area which comprised the present-day
provinces of Punjab and Khyber Pakhtunkhwa (formerly the
NWFP) as well as the Islamabad Capital Territory in Pakistan,
together with the states of Punjab, Himachal Pradesh and
Haryana, Delhi and the Chandigarh Union Territory in India.
36 Lighting-rod zig-zags appear above the shoulders of a Herakles
carved on another gem from the region (P. Callieri, 'Seals from
Gandhara, Foreign Imports and Local Production', in M.-F.
Boussac and A. Invernizzi (eds), Archives et Sceaux du Monde
Hellénistique, Archivi e Sigili nel Mondo Ellenistico, Turin, 1996,
413–22, at 417–18, pl. 96, fig. 18a–b.). Callieri suggests they derive
from the stylised ends of a Hellenistic ribbon or taenia as depicted
on Indo-Greek coins of the 1st century bc.
37 ‘Fingerprint’ is a generic term used in gemmology to describe
minute fluid and solid inclusions formed on a fissure or fracture as
secondary processes try to ‘heal’ the break.
38 A review of the evidence for the ad 233 date in M.L. Carter, ‘Notes
on Kuṣāṇa Chronology and the Bactrian Era’, Journal of Inner Asian
Art and Archaeology 1 (2006), 81–3.
39 Bivar (n. 6), 74, pl. 11, DJ1–6; R.N. Frye (ed.), Sasanian Remains
from Qasr-I Abu Nasr, Seals, Sealings and Coins, Cambridge (MA),
1973, nos D252–5, 257.
'Gems of Heaven’ | 37
Adams, Lüle and Passmore
40 Ibid., 64–5, pl. 8, CD1, 3, 5–6. The identical iconography appears on
a nicolo collected by Cunningham (1892,1103.147).
41 Bivar (n. 6), 77–9, pl. 12, nos ED1-13, EE1-9.
42 Cf. O. Bopearachchi, ‘Jihoṇika, a ruler in northwestern India
known to us from his coins and an inscription (1st cent. ce)’,
Encyclopædia Iranica 2008 (www.iranica.com/articles/Jihonka).
43 W. Zwalf, A Catalogue of the Gandhāra Sculpture in the British
Museum, London, 1996, vol. I, 299, vol. II, 265, no. 480, with
further references.
44 Cf. M. Alram, ‘Hunnic coinage’, Encyclopædia Iranica 2004 (www.
iranica.com/articles/Hunnic coinage); M. Alram, ‘Huns And
Western Turks In Central Asia And Northwest India’, A Survey of
Numismatic Research 2002–2007, International Association of
Professional Numismatists, Special Publication 15, Glasgow, 2009,
Section 3: Oriental and African Coinages, 516–19; M. Alram and
M. Pfisterer, ‘Alkhan and Hepththalite Coinage’, in M. Alram,
D. Klimburg-Salter, M. Inaba and M. Pfisterer (eds), Coins, Art and
Chronology II, the First Millennnium C.E. in the Indo-Iranian
Borderlands, Vienna, 2010, 13–38. In the same volume: J. Cribb,
‘The Kidarites, The Numismatic Evidence’, 91–146, and
E. Errington, ‘Differences in the Patterns of Kidarite and Alkhon
Coin Distribution at Begram and Kashmir Smast’, 147–68.
45 F. Grenet, ‘Kidarites’, Encyclopaedia Iranica 2005, online edition
available at: www.iranica.com/articles/kidarites.
46 Alram and Pfisterer (n. 44), 32.
47 E.I. Solomonik, Sarmatskie Enaki Severnogo Prichernomor’ja, Kiev,
1959, fig. Iv.1, nos 2–3, 5–6 (sculpture); 76–80 (metalwork), all
1st–3rd centuries ad.
48 Bivar (n. 6), 111–12, NB6, ND2 and ND4, all dated 4th century ad.
49 Ibid., 112, ND6-7; R. Göbl, Die Tonbullen vom Tacht-E Suleiman,
Berlin, 1976, 83–7, pls 48–9 (Sasanian tamghas on bullae).
50 Göbl (n. 12), vol. III, G16–17 and vol. IV, pls 15 and 17, tamgha S82.
Tamghas were still in use on coins dated as late as the 8th century ad.
51 Errington (n. 44), fig. 1l and figs 4–5.
52 E.J. Rapson, ‘Notes on Indian Coins and Seals, Part IV, Indian Seals
and Clay Impressions’, Journal of the Royal Asiatic Society of Great
Britain and Ireland (1901), 97–108, at 106–7; Göbl (n. 12), vol. III,
nos G51-2, pl. 87; Callieri (n. 14), U 7.49–50.
53 Callieri (n. 14), 3.20–22, U 7.47; G. Fussman, ‘Intailles et
Empreintes Indiennes du Cabinet des médailles de Paris’, Revue
38 | 'Gems of Heaven’
Numismatique, VIe series, 14 (1972), 21–48, at 47, nos 34–6.
54 Göbl (n. 12), vol. I, 47.
55 Bivar (n. 6), 44–8, pls I.AA1–10, II.AC1–14.
56 M. Alram, ‘Three Hunnic Bullae from Northwest India’, Bulletin of
the Asia Institute 17 (2003/2007), 177–84.
57 Callieri (n. 14), 227–8. In an earlier paper (P. Callieri, ‘Huns in
Afghanistan and the North-West of the Indian Subcontinent’, in
Alram and Klimburg-Salter (n. 13), 277–91), he places cat. nos.
II.9, II.10 and II.11 together in his Class D which he regarded as
Hephthalite. As is true of the Hunnic period in the West,
designations such as Alchon, Kidarite or Hephthalite should be
understood to refer to periods and not necessarily to specific
ethnic groups.
58 Callieri (n. 14), 229–31.
59 Callieri 1999 (n. 57), 285.
60 H. Falk, ‘Names and Titles from Kuṣāṇa Times to the Hūṇas’, in
Alram, Klimburg-Salter, Inaba and Pfisterer (n. 44), 73–90, at 81.
61 Göbl (n. 12), vol. I, G 55, 251–2; on balance, a date extending into the
7th century ad (Garbini in Callieri (n. 14), 285), might be too late.
As Lerner (n. 16), 268, has noted, the close resemblance between
coins and seals in this period suggests that if die-cutters did not
also engrave seals, then seal-cutters drew much of their
inspiration from coinage.
62 Lerner (n. 16), 268–9, fig. 19.
63 Falk (n. 60), 80–81; Garbini in Callieri (n. 14), 285, suggests the
script on this particular stone is late or post-Guptan, 6th–7th
century ad.
64 F. Grenet, ‘Regional Interaction in Central Asia and Northwest
India in the Kidarite and Hephthalite periods’, in N. Sims-Williams
(ed.), Indo-Iranian Languages and Peoples, Proceedings of the
British Academy 116, Oxford, 2002, 203–24, at 218–19, pl. 7b.
65 K. Vondrovec, ‘Coinage of the Nezak’, in Alram, Kilmburg-Salter,
Inaba and Pfisterer (n. 44), 169–90; also Grenet (n. 64), 214–18.
66 Göbl (n. 12), vol. I, 251, suggested this was carved in the same
workshop as a rock crystal (G54) inscribed devadata in Brāhmī. He
dated these to the 6th–7th century ad and proposed they were
carved either in India or Afghanistan (?Kabul).
67 Cf. Callieri (n. 14), 203, 285, 287, 298, cat. U 7.39 (1892,1103.140),
another garnet portrait of this class inscribed dharmadāsa
(‘servant of the dharma’).
Archaeologies of Magical Gems
Richard Gordon
It is a topos in the study of Graeco-Egyptian magical gems that,
after the lively interest shown in them during the early modern
period,1 the rise of an aestheticism of the sublime associated
with Winckelmann, above all in Germany, eventually led to
their almost total neglect. Winckelmann himself was
forthrightly dismissive:
Die geschnittenen Steine, welche man Abraxas nennt, sind jetzt
durchgehends für Gemächte der Gnostiker und Basilidianer aus
der ersten christlichen Zeiten erklärt und sind nicht würdig, in
Absicht der Kunst in Betracht gezogen zu werden.2
Art is art, religion – and especially the religion of what for
Winckelmann was a period of steep cultural decline –
something else.3 We may however remark that this exclusion
from the emergent realm of Classical archaeology was based
upon an attempted historicisation of the ‘Abraxases’ that
sought to break free of their earlier valuation as effective
amulets, just as it was founded upon their unreadability in
terms of the familiar stock of antiquarian knowledge. Everyone
knows the story of how Adolf Furtwängler, unwilling to include
the ‘oriental’ and magical gems in the Berlin collection in his
catalogue, tried to pass them out of the Antiquarium into the
charge of the Museums’ (Vorder-) Orientalische Abteilung.4 In
pursuit of this scheme, he persuaded Adolf Erman to store
them in the Ägyptisches Museum, where they remained
unsystematised and scarcely known until 1986.5 As late as 1983
Peter Zazoff could complain about the numbers of unpublished
magical gems in important collections, in Vienna, Florence,
Copenhagen and elsewhere.6
Since then, however, as a result of the expansion of
museum staffs and university departments of Classical
archaeology, and within the context of a heightened general
awareness of ancient Kleinkunst, there has been a great
increase in the publication of the relevant holdings in museum
and private collections (notably, apart from the Berlin amulets:
London, Vienna, the University of Cologne, Naples, Perugia,
the Skoluda Collection).7 Apart from the re-edition of (often
fanciful) images drawn from older publications, Attilio
Mastrocinque has published an almost complete catalogue of
such gems in Italian museums and revised the old catalogue of
the first-rate collection in the Cabinet des médailles by Armand
Delatte and Philippe Derchain, with its very poor images.8
Altered views about the use of impressions, which Campbell
Bonner defended, and the practice of enlarging photographs
have greatly increased legibility.9 The importance of colour
photos is now taken for granted.10 Given the doubts raised
already by Bonner about the feasibility of a complete corpus, in
view of the expense involved in conventional publication and
the repetitive nature of many types,11 internet publication
offers the obvious way forward – a start has already been made
on this by Árpád Nagy and Ildikó Csepregi at the Budapest
Museum of Fine Arts.12 Scientific exhibitions – despite the
technical difficulties in rendering such small objects legible –
have increased public awareness of these objects; we await
with impatience the promised exhibition in Budapest at the
end of 2011.13 Symposia such as the one held in the British
Museum, hitherto barely thinkable, offer an indispensable
opportunity for the direct exchange of information and ideas.14
Hannah Philipp, Erika Zwierlein-Diehl and Simone Michel
have improved the standard of commentary on individual
stones out of all recognition. The publication in recent years of
important studies of individual motifs, such as the uterine
amulets, the Phoenix and the ouroboros have contributed both
to the understanding of thematic groups and to improving
methodology.15 And with the publication of Simone Michel’s
Die Magischen Gemmen (2004) we now possess a synthetic
work that assembles the holdings of numerous museums, with
many original photos, and offers a catalogue of types that will
be of permanent value.16 In short, more is now known, and in
greater detail, about museum holdings and about individual
types than ever before.
On the other hand, this is a very small professional field,17
in which only a relatively small number of questions tend to get
asked. What type of stone? What parallels are there for the
image(s)? Is there an intelligible text or classifiable logos? What
are the implications of these for our judgement about the
intention of the amulet? Given the obscurity of much of the
iconography, such a focus is of course intelligible; one might
even hold: the only responsible procedure if we are to avoid
mere speculation. It goes without saying that these studies, as
well as the continuing work of cataloguing and publication, are
of primary importance to the field, have led to major advances
in our understanding, and need to be continued by those whose
skills and enthusiasm lie in this direction.18
Thirty years ago Gertrud Platz-Horster remarked of ancient
gem-studies in general, ‘Die Angleichung an der Stand der
Methodik in anderen Bereichen der Archäologie erfolgt spät,
oder ist noch im Gange’.19 Much has been achieved in the
meantime, but, Classical archaeology being itself subject to
constant theoretical development, the process of catching-up is
ceaseless. The basic issue is the need to find ways of
historicising the magical gems more adequately. I would
suggest there are at least four areas that require some brief
discussion in the spirit of Platz-Horster’s comment; none is
totally new, but all in my view need more explicit attention
than they currently receive.
1. The creation of new facts
A major problem, as with all Kleinkunst, is the sheer museality
of magical gems. It is enough to recall a few details. Most of the
c. 2,700 gems in the collection of the Royal Coin Cabinet in The
Hague were amassed between 1741 and 1822 from the earlier
collections of wealthy Dutch merchants, for whom provenances
‘Gems of Heaven’ | 39
Gordon
were of no concern.20 Not one of the 50 intaglios in the
collection of the University of Pavia, acquired by Pietro Vittorio
Aldini during his tenure of the chair of numismatics and
antiquities 1819–42, has a provenance.21 The list of cities where
in the mid- to late 19th century the collector and dealer
Constantine Schmidt-Ciążyński purchased his final gem and
cameo collection (now in the Museum of the Jagellonian
University in Cracow) includes: Paris, London, St Petersburg,
Moscow, The Hague, Vienna, Milan, Turin, Venice, Rome,
Florence, and in addition ‘thousands of other [dealers] from
whom I have purchased cameos’.22 In the particular case of
magical gems, Furtwängler’s distaste meant, and often
continues to mean, that they are generally poorly represented
in private collections.23 Museality, however, does not
necessarily prevent us from thinking of ways of re-presenting
them so as to yield new information.
The production of new facts is one major function of
archaeology, not in the sense of objects newly dug up, or even
site plans, but as it were secondary or constructed facts
produced from counting or otherwise synthesising already
existing data, and so representing them in a different mode.
Two now standard kinds of such constructed second-order
facts are distribution maps and statistical charts. Could a
distribution map of provenances of our objects be constructed?
At first sight, in view of the point I have just made about
museality, we would have to say it is impossible, at any rate if
we think in terms of trying to map the totality. But, as both
Zazoff and Philipp long ago pointed out, a certain number of
magical amulets do have attested provenances, notably two
dozen in Aquileia, one or two in Germany, one or two in
Britain, one or two from Carnuntum, items from Porolissum,
Micia and Celeia in Romania, in Bulgaria, the northern Black
Sea area, Greece and especially Thessaloniki, the west coast of
Asia Minor, Cyprus, Israel, Egypt.24 Additional information
could be derived from more or less reliable reports of the
provenance of individual collections or part collections.25 Even
if proveniences would be much more telling, mapping such
provenances – assuming the cases provide a random sample,
which of course needs arguing – would constitute a new sort of
fact of the kind I envisage.26 In the long run, the creation on the
internet of a unified virtual database of all or almost all
magical gems will make it possible to produce such a map by
just clicking the mouse;27 but in the meanwhile pencil and
paper will be a surer if much slower means of achieving the
same result. Further information about ancient recognition of
the category of magical gems would be forthcoming if such a
map could be compared with a map of all ascertainable gem-
provenances of the relevant period.
Gems are like relief sculpture, coins and pottery in that
they are relatively hard to destroy. They survive infinitely
better than papyrus or wax tablets.28 Quite apart from telling us
something about the distribution of demand, a map of
provenances would permit inferences about familiarity with
the ‘discourse’ implicit in such amulets, and so about the
extension into the (eastern and central) Mediterranean of
awareness of Graeco-Egyptian magical practice, which is
otherwise, given the total loss of relevant papyri and codices
outside Egypt, hardly measurable. A first impression, based on
the tiny number of magical gems found at Aquileia (1.5% of
1573 catalogued gems), Altino (VE) (0.75% of c. 400), Caesarea
40 | ‘Gems of Heaven’
Maritima (4.8% of 165 catalogued) and, more tentatively, at
Gadara in the Dekapolis (1.2% of the roughly 410 relevant items
in the Sa’d Collection), is that they were not very prominent.29
Such a finding would correlate with the evidence for the
distribution of defixiones and phylacteries displaying
knowledge of Graeco-Egyptian techniques.
Beyond that, it is safe to say that among the provenances
are also a small number of proveniences; and, as Martin Henig
and Antje Krug long ago recognised in relation to conventional
gems, an inventory of proveniences would also help to provide
a more reliable basis for what are at the moment impressions
about the contexts of deposition/loss, which in turn would
provide a more secure basis for arguments about Sitz-im-
Leben.30 As for statistics, it would now be possible on the basis of
Michel’s Die Magischen Gemmen, notwithstanding that her
collection is far from exhaustive, to construct graphs or other
types of abstract representation of type-frequencies, type-
deviations, perhaps even of type-deterioration (granted the
irreducible problems inherent in type-recognition). New
representations of this kind would again help to get away from
mere impressions, and provide the basis for testing existing
hypotheses about the development of the genre on the one
hand, and the invention and development of types on the
other.
2. Designers and workshops
Maps and graphs are already standard modes of
re-representation of complex materials in archaeology – here it
is a matter merely of seeing ways in which the amuletic
material can be re-presented, so as to establish a new basis for
questions and hypotheses that would not otherwise become
apparent. Mapping however will not solve a further problem,
namely the point or points of manufacture. Already in 1914
Delatte rejected the traditional term ‘amulettes gnostiques’:
‘ce sont simplement des amulettes qu’on doit attribuer à
l’époque de l’efflorescence des doctrines et des pratiques de la
magie gréco-égyptienne’.31 Bonner, Derchain and Philipp all
rightly insisted that, with some obvious exceptions, they are
Graeco-Egyptian and only interpretable on that basis. For
Philipp, this fact, together with the predominantly eastern
Mediterranean provenances, implied they were mainly created
in Alexandria, the obvious centre for such syncretistic
practice.32 Erika Zwierlein-Diehl has thrown her considerable
authority behind this view.33 Bonner himself however was
apparently undecided and seems never to have expressed a
firm opinion.34
There are three main objections to this hypothesis. The
first relates to the positive evidence in favour of Alexandria.
Although E.A. Burghart did collect a number of such gems in
Egypt in the early 19th century, it does not seem that they were
mainly acquired in Alexandria itself.35 Of a collection of 74
gems from the city and its environs, apparently made some
time in the 1830s and 1840s, which entered the Akademisches
Kunstmuseum in Bonn from the Museum rheinischer
Alterthümer in 1892/3, just one can on modern criteria be
counted ‘magical’ (1.35%).36 Thanks to the fellaheens’ awareness
of the potential value of such objects, the antiquities trade
flourished all along the Nile valley, not simply in Alexandria;
the provenanced finds seem to reflect this.37 The fact that so
few of the enormous numbers of engraved stones of the Roman

period found in Egypt are in the possession of the Cairo
Museum but rather in the hands of private individuals seems to
point in the same direction.38
The second consideration relates to the Greek-speaking
elites in Roman Egypt. At the time that the ‘Alexandrian
hypothesis’ was proposed, the socio-political role of the urban
elites, the ‘gymnasial class’, of the more important of the 42
metropoleis, such as Arsinoe, Herakleopolis, Oxyrhynchus,
Hermopolis, Ptolemaïs, Koptos, which formed the ‘vertebrae’ of
Roman Egypt, was hardly recognised. Although boulai were
only officially introduced in the Severan period, such local
elites were the profiteers of the incremental privatisation of
land in the chora throughout the Roman period.39 These urban
elites and sub-elites certainly had the resources to invest in
intaglios, including magical gems; although they were
relatively expensive, not something that was turned out every
day,40 it seems out of the question that this demand could be
satisfied only in Alexandria.41 Moreover, it is now recognised
that one of the major routes for the import of both stones and
incense from India and Arabia met the Nile at Hermopolis, so
that semi-precious stones were easily available far up-river.42
If we now consider the distribution of the relatively
specialised knowledge required to create such compositions, it
was widely, albeit differentially, disseminated among the
personnel in some sense attached to the Egyptian temples in
the chora, particularly if we assume that in the 2nd and 3rd
centuries ad it was to a large extent temple priests in their
unofficial capacities, and ‘freelance’ specialists outside the
temple, or with only a loose relationship to it, who wrote both
the receptaries and the activated texts (‘magical papyri’) that
have been found in the rubbish dumps of the metropoleis.43
These mostly very damaged texts sometimes refer to recipes in
the temple libraries, but, generally speaking, reflect a much
lower level of expertise than most of the famous Graeco-
Egyptian receptaries collected from unknown sources by
Giovanni Anastasi in the 1820s – which likewise have no
plausible link to Alexandria.44 The shift to Greek from Demotic,
especially in the case of the magical gems, which were visible,
and partly intelligible, to those who bought them, clearly
implies a Greek-speaking clientele throughout the chora able to
afford such ‘enhanced’ semi-precious stones. The shift to
writing receptaries, and a fortiori activated texts, in Greek,
whose contents were in principle inaccessible to clients, is
likewise suggestive of the linguistic culture of the major target-
group.45 In both cases, however, the sheer range in quality
between the finest and the most mediocre examples (compare
Pls 13, 14 with Pl. 3) implies a wide range of clients with very
different purchasing power.
A fourth consideration relates to the possibility that at least
some magical amulets were created outside Egypt altogether.
This is most plausibly the case with the Sacrifice of Isaac and
the Burning Bush;46 nor should we dismiss at least
Goodenough’s better examples of Jewish coding, even if his
exaggerated inclination to see Jewish influence almost
anywhere has largely discredited his work.47 But we may
perhaps go further. Whatever roles we assume for the Latin
Kingdom of Jerusalem (1099–1291),48 the personal fiefs on
Cyprus, and the Venetian Levantine trade,49 it is difficult to
believe that the great majority of the 3,500–5,000 surviving
(known) amulets, now mainly in west European and North
Archaeologies of Magical Gems
American museums, or, in the case of eastern Europe,
purchased from west European dealers, were imported in
early-modern and modern times from the Levant. Though it is
only an indirect pointer, the re-use of antique gems in the
Middle Ages, a practice more or less confined to western
Europe, seems to indicate that the gems in question were found
and re-set mainly in Italy, and to a lesser extent in France,
Spain and the German-speaking lands.50 Of course amuletic
gems – ‘piccoli materiali di facile dispersione’51 – were objects
that could easily have been traded in antiquity.
On the other hand, the wide spectrum of competence and
quality might be explained not simply as a function of
differential competence and pricing within Roman Egypt itself
but also as the result of the imitation of unfamiliar models by
workshops outside Egypt. To my mind, one plausible example is
the practice of adding simple magical texts, usually a single vox
magica, and often in an irregular form, to standard religious
iconography. Prof. Zwierlein-Diehl has recently drawn
attention to a series of amulets showing the Danubian Riders,
and suggested, in keeping with her general position, that
adherents of this cult came to Alexandria and there had these
gems cut.52 That is of course possible, but it seems much more
economical to suppose that simple magical formats of this type
might also have been composed in the provinces outside Egypt,
in this case in Dacia or Moesia. The same is true of magical
text, even images, intended to enhance the efficacy of Mithraic
scenes,53 of occasional gems with Dionysiac motifs, Hermes,
Asklepios, Nemesis and others,54 and those with Latin texts.55
Moreover, the discovery all over the Mediterranean basin
(outside Hispania) of phylacteries in precious and base metals,
more or less in the Graeco-Egyptian tradition, even if some
were transported from elsewhere, points to the existence of
practitioners outside Egypt who may have been capable of
creating their own designs and having them engraved on gem-
stones.56 The same inference can be made from the (much less
frequent) case of competent defixiones in the same tradition
discovered outside Egypt, for example in Rome, Athens,
Antioch, Carthage and Hadrumetum.57
The likeliest hypothesis in my view is suggested by a point
Árpád Nagy made in the Verona conference proceedings, that
very occasionally one finds mistakes in the amuletic texts
indicating that they were taken from written models, in this
case WC ΠΡΟΚΕΙΤΑ[I on a gem in Budapest.58 This in turn
suggests that there may have been analogous models for the
commoner designs, exactly as we find in the high-quality
magical papyri (Papyri graecae Magicae: hereafter PGM),59
such as the drawing of a scarab to illustrate the ritual required
at PGM II 152–61, or the drawing of Akephalos from a few lines
lower, PGM II 166, to illustrate the location of the prescribed
voces (see Pl. 12). If so, amulets could have been created by
those with access to the relevant materials almost anywhere in
the eastern Mediterranean world, as well as in Italy and, to a
much smaller extent, in the Latin-speaking provinces.60 The
name of this game would be (relative) routinisation through
the medium of the book.
The only way of assessing these possibilities would be to
adapt the pioneering work done by Marianne Maaskant-
Kleibrink on Republican gem engraving, which indicated the
shape of the drill-head and the angle of application as the
decisive differentia.61 Since this involves the use of a powerful
‘Gems of Heaven’ | 41
Gordon

Plate 1 Ringstone of dark green jasper, 12 x 16mm. London, British Museum, Plate 2 Ringstone of red-green jasper, 15.5 x 11mm. London, British Museum,
PE G522 PE G135

Plate 3 Ringstone of red jasper, 18 x 13mm. London, British Museum, PE Plate 4 Ringstone of mottled dark brown jasper, 14 x 11mm. London, British
1986,0501.81 Museum, PE G 443

Plate 5 Ringstone of haematite, 19 x 14mm. London, British Museum, PE G76 Plate 6 Ringstone of yellow jasper, 12 x 16mm. London, British Museum, PE
G415

Plate 7 Ringstone of haematite, 21 x 15mm. London, British Museum, PE Plate 8 Haematite amulet for cutaneous application, 44 x 19mm. London,
1986,0501.71 British Museum, PE G497

Plate 9 Siltstone neck amulet, 48 x 38mm. London, British Museum, PE G477 Plate 10 Ringstone of carnelian, 17.5 x 13.5mm. London, British Museum,
PE G109

42 | ‘Gems of Heaven’

Plate 11 Ringstone of lapis lazuli, 23 x 16mm. London, British
Museum, PE 1986,0501.67
Plate 12 Drawing of the Akephalos from col. IV of P. Berol. 5026
= PGM II 166, to indicate the design to be drawn on papyrus (l.
46f.) in order to obtain a direct revelatory vision
Plate 13 Ringstone of almost black haematite, 21 x 15mm. London, British
Museum, PE G 557
binocular microscope, and the creation of matrices for
comparison, it could only be done by sampling in the context of
a wider project devoted to the historicisation of the magical
gems. Moreover, if it is indeed the case in the Imperial period
that the three same kinds of drill, ‘Flachperl’, ‘Zylinderkopf’ and
‘Scharfkantiger’, were everywhere employed, and that specific
styles and techniques cannot be assigned to particular centres
but belong to a craft-koinê, the results might still be
inconclusive. A trial run, however, would still have probative
value and might create significant new facts.
3. Magical gems: between functionality and accessory
To judge from the surviving magical papyri of Ptolemaic date,
such as the Brooklyn papyrus (IV-IIIa), and the style of amulets,
including contemporary scarabs, much of the traditional
magical regime survived well into the period of Macedonian
rule in Egypt.62 But a relatively dramatic shift seems to have
occurred in the late Hellenistic-early Roman period. The
traditional genres of temple magic, particularly protective
magic against demons, snakes and scorpions, seem largely to
have disappeared, or at least are hardly attested,63 and new, or
hitherto rather rare, genres become dominant, for different
types of divination (direct visions of god, bowl divination,
dream sending, Homer oracles), malign magic, aggressive
(mainly erotic) magic, magic for personal success and
attractiveness; written phylacteries.64 This shift implies a
response on the part of practitioners not merely to the
changing status of the temple in the metropoleis and the
emergent socio-political dominance of local elites and sub-
elites in these centres but also to changing demand.
Archaeologies of Magical Gems
Plate 14 Ringstone of dark green jasper, 18 x13mm. London, British Museum,
PE G 449
Practitioners were open to experimentation inasmuch as they
were forced to compete with one another in a market
relationship with their Greek-speaking clients.
The shift from traditional amulets (which already show
typological changes in the Hellenistic period) to our ‘magical
amulets’ was surely likewise market driven.65 It implies
customers who, already familiar with the Hellenistic intaglio in
semi-precious stone, demanded a type of amulet analogous to
these but drawing (at least indirectly) upon a prestigious
tradition of temple learning (Pl. 1).66 It is also very likely that
the Hellenistic – essentially Babylonian – lore concerning the
properties of stones, associated with the names Sotakos,
Sudines, Zachalias of Babylon and others, played a role in the
forging of this new fashion.67 At all events, the market
relationship resulted in a sharp increase in the number of types
created, which could be amplified by appeal to Greek
mythology or astrological schemes as well as by the addition of
voces magicae and charakteres. The increase in types must be
connected both with the topographic dispersion of production
and with the variable competence of practitioner-designers and
their gem-cutters.
We can also turn the market argument on its head, and
consider the creation of new-felt needs, through specialisation
of types, with the specific intention of increasing sales. Many
amuletic gems have non-specific aims, appealing to the deity
invoked for general protection, sometimes explicit as in the
injunction διαφύλασσε, ‘Watch carefully over (the bearer)’ (Pl.
2).68 But a good number are intended to help against specific
dangers. We may adduce here the type of Herakles and the
Nemean lion to aid recovery from digestive ailments (Pl. 3),69
‘Gems of Heaven’ | 43
Gordon
or the invocation of melothesic schemes, such as the zodiacal
sign Scorpius to protect the genitalia (Pl. 4).70 Amulets to aid
childbirth and its attendant ills and dangers (Pl. 5) are another
very common specialised type.71 Mariangela Monaca has
proposed that yet another group may have been used as aids in
(astrological) divination (Pl. 6).72
We may assume specialisation even in the numerous cases
where the precise intention is to us obscure, as in the case of
so-called Pantheos figures (Pls 7, 11).73 Some non-ringstones
were evidently considered analogous to the terracottas of the
Isiac cycle that fill our museums,74 and used performatively (Pl.
8).75 Simone Michel has even revived the thought that certain
designs may have served as tokens in mystery initiations,
though this seems very speculative.76 More, perhaps many
more, were apparently intended eo ipso to evoke a deity or
group of deities for now unspecifiable purposes, thus
functioning in exactly the same manner as the voces barbarae
and the charakteres (Pl. 9).77 The magical papyri suggest that a
few may have been used as phylacteries in the course of
performing rituals.78 Except in cases of extreme routinisation,
it is wise to assume that a specific intention lies behind the
design, particularly when it is more or less unique (Pl. 10).79 We
may also assume that non-standard designs were more
expensive.
It would however be over simple to think only in terms of
the explicit intentions of designers. In my view, it is quite
implausible to suppose that all practitioners could have
provided the type of commentary to individual designs that
one finds in modern catalogues. The extent of routinisation
(Pl. 11),80 as well as the probable existence of receptaries for
common designs, surely excludes this. Routinisation also
implies that most amulets were not elaborately consecrated.81
Such considerations lead us on to considering magical gems as
an aspect of culturally specific consumption.
We usually think of the magical gems exclusively in
instrumental terms, as objects intended to cause changes in the
real world (or to prevent possible changes) by appeal to
imputed specialist knowledge of the ins and outs of the divine
world. Insofar as they are given a Sitz-im-Leben, we invoke
‘belief’ in their ‘power’. But in the wake of Appadurai’s notion of
the ‘social life of things’ we might also try to think of ways of
seeing them in relation to other contexts, thus challenging the
determinacy of imputed meanings.82 Just as clothes and
personal adornment in antiquity provided a material language
of social stratification and role-expression,83 just as the
adoption in the north-western provinces of terra sigillata and
even insignificant articles of use such as hair-tweezers for
depilation imply larger shifts in self-understanding and social
place,84 so perhaps the Graeco-Egyptian magical gems have
implications for self-understanding which could be teased out.
As purely elective items, magical gems offered benefits
beyond their overt instrumental value. One of these was
membership in a loosely defined imagined community of the
pious. From that point of view the amulet is a pre-paid votive
offering, amounting to an expectation. Within Egypt, these
gems mediated between two worlds, the metropolis and the
temple. The group of designer-practitioners collectively
referred its authority to an institution, the temple, which
served not merely as the guarantor of the age-old efficacy of
Egyptian religious culture but prided itself on its command
44 | ‘Gems of Heaven’
through texts of the relevant mythology in all its evolving
complexity (Pl. 12).85 To that extent the gems they designed
affirmed the resilience of the Egyptian heritage of the elites
and sub-elites of the metropoleis. Outside Egypt, such gems
evoked the authority of the oldest of the wise nations, but also,
like superior hair-tweezers as it were, served as an aspect of
Romanisation in the loose sense, an exotic commodity
available solely thanks to the ‘hodological space’ of the Empire,
and thus in themselves a social statement. Again, the relative
expense of such gems excluded the great majority of the
population; but at the same time their availability in all
qualities, from the fine (Pl. 13)86 to the dreadful (Pl. 8),
provided a range of options appropriate to one’s status. To the
extent that their quality was calibrated with purchasing power,
they can also be seen as objective affirmations of the ability of
the ‘theodicy of good fortune’ to cope with the threat of
misfortune.87 At the same time, these gems, whether worn as
ringstones or at the neck, were often openly visible and thus
capable of stimulating an aesthetic (Pl. 14)88 as well as a
religious response.89 As Georg Simmel pointed out:
Es ist das Wesen und der Sinn des Schmuckes, die Augen der
Anderen auf seine Träger zu lenken, und er ist insofern der
Antagonist des Geheimnisses, das sich aber auch seinerseits der
personal-akzentuierenden Funktion nicht entzog.90
The aesthetic response of the other is a graduated one
evoked both by the quality and value of the stone and by the
mastery of technique, the latter in rewarding tension to the
flouting of dominant Graeco-Roman iconographic norms
regarding sujet (‘Geheimnis’). Such jewellery is socially
interesting precisely because of the other’s involuntary
response of admiration-cum-envy. Routine work in the same
mode relies upon the existence of high quality exemplars for its
own more pallid effects.
4. The category ‘magical gem’
The final set of issues is complex; I can here only outline one or
two of the problems.91 The first concerns the sub-divisions or
sub-groups within the larger category of Graeco-Egyptian
magical gems. The compilation of every new catalogue in this
field brings the scholar up against the limits of the adequacy of
previous classificatory systems.92 One has only to compare the
major examples (London, Paris, Berlin, Vienna, The Hague, the
Italian collections [see n. 10]) to realise how little agreement
there is over the most appropriate ‘internal’ classification once
one leaves the area of medical applications. Is the category
‘regeneration’ useful? Is it appropriate to classify Harpocrates
gems under the larger heading ‘solar’ on the grounds that they
evoke (inter alia) the daily course of the sun? How helpful is it
to distinguish between ‘Egyptian images’ and ‘new’ ones? Are
we to include images of Egyptian deities, say the Isis group,
even when they lack magical text?
This lack of agreement, which is partly a function of the
thematic diversity of different collections but mainly a sign of
our irremediable ignorance about the practitioner-designers’
contextualisations, will need to be resolved at least
pragmatically as the unified virtual database on the Internet
takes shape over the next few years.93 Although it requires more
intensive discussion, however, it is not, I think, the major
problem, which is the question of the relation between the
group of magical gems ‘mainly Graeco-Egyptian’ (to employ

Campbell Bonner’s phrase) and the wider classes of gem
treated as magical – or at any rate as marvellous – by the texts
we can roughly group as ‘les lapidaires grecs’.94
In practice, we have settled for the easy solution of
iconographic difference: our ‘magical amulets’ are defined as
such primarily by their Egyptian motifs – explaining the
iconography has since Delatte’s day mainly involved appeal to
Egyptological expertise; in the penumbra are arranged, with or
without appeal to texts such as PGM IV 1722–45 or V 447–58, a
variety of other motifs, mostly from the Greek iconographic
repertoire, usually legitimated as ‘magical’ by a vox magica.
More recently, colour and stone types have been emphasised.
Once one invokes ‘magic’ and stone types however it is difficult
to claim that the ‘Abraxas’ gems are sufficiently different in
their Sitz-im-Leben from many other antique gems to warrant
being treated as a separate category, except for purely
pragmatic purposes, such as organising museum holdings. The
relation between the Graeco-Egyptian magical gems and the
ancient lithic tradition describing the natural properties of
numerous ‘magical’ stones and their effects is vexingly obscure.
Would we be justified in using, say, Damigeron-Evax 7.3 to
classify images of Hekate or Medusa on Graeco-Egyptian
amulets as thumokatochoi, to protect the wearer against the
anger of masters, adversus iras dominorum?95 Or a standing Isis
as intended to ensure success and vigour, according to the
prescription of Socrates and Dionysius?96
Our most important single source, Pliny the Elder (in Nat.
Hist. XXXVI–XXXVII), highlights the immanent power of
stones and says little or nothing about images or settings; but I
think we should infer, on the basis of e.g. Kyranides Bk 1, that
they are assumed – it is simply that they are irrelevant to his
topic, which is the inherent powers of the stones themselves.
The same is true of the lithic tradition, though it includes many
more references to image-stone combinations. Two of these
texts, Damigeron-Evax and Socrates and Dionysius, are in the
present connection the most important. The first Incipit to our
Latin text of Damigeron-Evax, with its repeated reference to
Egyptian lore, suggests that at least one version was written in
Greek in Egypt.97 Yet the text as we have it contains virtually no
reference to any type we classify as Graeco-Egyptian.98 In
Socrates and Dionysius however we do find clear references to
such types, for example Chnoubis on an onyx against stomach
problems, prescriptions for (admittedly rather feeble) voces,
and the occasional use of charakteres.99 This same text however
also prescribes Greek divinities, such as Poseidon or Artemis,
for various kinds of aid, in precisely the same way; and
translates into Greek iconographic terms what seem clearly to
be Egyptian conceptions.100 This implies that the absence of
‘hand-books’ of Graeco-Egyptian magical amulets is due not to
their being considered a genre unrelated to the interests of the
Orphic Lithika but rather to the vagaries of transmission.101
Against this background, the habitual separation between
‘Greek’ glyptics and Graeco-Egyptian magical gems seems to
owe rather more to the connoisseurship of a Furtwängler than
seems justifiable. How perceptible were such boundaries in
antiquity? And how are our own choices to be grounded?
These four topics for further research could easily be
extended, and others will have their own particular questions.
Although the work of composing and issuing catalogues must
continue, it is equally important to make use of the initiatives
Archaeologies of Magical Gems
issuing from conferences such as this in London, in Fribourg
and elsewhere, and the data-base project in Budapest, to
establish networks of co-operating scholars able to meet on a
fairly regular basis. Despite the rich holdings of museums, and
the intrinsic interest of the material, the study of magical gems
will remain a niche area – could one in good faith encourage a
PhD student to start a project in this area, given the
competition from other more central topics in Classical
archaeology? – until we succeed in creating such networks of
scholars willing to pool their expertise.
Notes
1 On the Nachleben, see now the brief but helpful discussion by E.
Zwierlein-Diehl, Antike Gemmen und ihr Nachleben, Berlin, 2007,
264–86; cf. eadem, ‘Antikisierende Gemmen des 16.-18. Jhdts.’, in
T. Hackens and G. Moucharte (eds), Technology and Analysis of
Ancient Gemstones. Proceedings of the European Workshop held at
Ravello, European University Centre for Cultural Heritage, 13–16
November 1987, Strasbourg, 1989, 373–403; eadem, Magische
Amulette und andere Gemmen des Instituts für Altertumskunde der
Universität zu Köln (Papyrologica Coloniensia 20), Opladen, 1992,
12–15. On Macarius/Chifletius and Capello, see briefly P. and H.
Zazoff, Gemmensammler und Gemmenforscher, Munich, 1983,
30–3; on the magical gems of the Praunsches Kabinett in
Nuremberg, which are now mainly or partly in the British
Museum, see S. Michel, ‘Nürnberg und die Glyptik.
Steinschneider, Sammler und die Gemmenkunde in 17. und 18.
Jhdt.’, Nürnberger Blätter zur Archäologie 16 (1999/2000), 65–90.
On Marchese Alessandro Capponi, see M.-L. Ubaldelli, Corpus
gemmarum: Da ctyliotheca capponiana. Collezionismo romano di
intagli e cammei nella prima metà del XVIII secolo (Bollettino di
Numismatica, Monografia 8.1), Rome, 2002.
2 J.J. Winckelmann, Geschichte der Kunst des Altertums, Dresden,
1764, repr. Darmstadt, 1993, 68. Thanks to his cataloguing of the
enormous von Stosch collection (see n. 4), Winckelmann was
thoroughly familiar with the sub-genre of magical gems:
Description des pierres gravées du Baron de Stosch, Florence, 1760,
1775². On his influence in Germany, see briefly S. Marchand, Down
from Olympus: Archaeology and Philhellenism in Germany, 1750–
1970, Princeton, 1996, 7–13.
3 ‘Die eigentliche bestimmte Zeit, in welcher der gänzliche Fall der
Kunst erfolgte, war vor dem Constantin, zur Zeit der großen
Verwirrung durch die dreißig Tyrannen’: Winckelmann 1764 (n.
2), 387.
4 A. Furtwängler, Die antiken Gemmen, I–III, Berlin, 1900, 363; cf.
H. Philipp, Mira et Magica: Gemmen im Ägyptischen Museum der
Staatlichen Museen ∙ Preußischer Kulturbesitz Berlin-
Charlottenburg, Mainz, 1986, 2f.; P. Zazoff, Die antiken Gemmen,
Munich, 1983, 352. Many of the magical gems came to Berlin
thanks to the purchase by Frederick the Great in 1764 of the 3,444
gems in the collection of Baron Philipp von Stosch; they were
included in the catalogue by E.H. Toelcken, Erklärendes
Verzeichniß der antiken vertieft geschnittenen Steine der Königlichen
Preußischen Gemmensammlung, Berlin, 1835.
5 For his part, Erman, who had been director of the Ägyptisches
Museum since 1885, considered the magical gems ‘un-Egyptian’ –
at any rate they represented a sharp break with the styles and
motifs of earlier Egyptian amulets, even of the Ptolemaic period.
To be fair, Furtwängler was fully aware that they required for their
interpretation very specialised knowledge, which he did not
possess; but he certainly added to confusion by, for example,
retaining one gem showing Horus with falcon-head in the
Antiquarium because it was more ‘Greek’ but removing others; and
retaining the Archaic Graeco-Phoenician scarabs in the Classical
collection: Philipp (n. 4), 3, n. 10. Furtwängler claimed to view
even Trajanic-Hadrianic gems as ‘decadent’; on such fastidious
connoisseurship, see briefly A. Snodgrass, ‘What is Classical
Archaeology? Greek Archaeology’, in S.E. Alcock and R. Osborne
(eds), Classical Archaeology (Blackwell Studies in Global
Archaeology), Malden, 2007, 13–29 (19–23). Other collections,
notably that of the British Museum, of course suffered equally
arbitrary treatment by opinionated curators.
6 Zazoff (n. 4), 352.
‘Gems of Heaven’ | 45
Gordon
7 London: S. Michel, Die Magischen Gemmen im Britischen Museum,
London, 2001a; Vienna: E. Zwierlein-Diehl et al., Die Antiken
Gemmen des Kunsthistorischen Museum in Wien, 3: Die Gemmen der
späteren römischen Kaiserzeit, Teil 2, Munich, 1991; Cologne:
Zwierlein-Diehl 1992 (n. 1); Naples: U. Pannuti, Cataloghi dei musei
e gallerie d’Italia. Museo Archeologico Nazionale di Napoli. La
collezione glittica, 2, Rome, 1994, nos 267–310; Guardabassi:
P. Vitellozzi, Gemme e magia dalle collezioni del Museo Archeologico
Nazionale dell’Umbria, Perugia, 2010; Skoluda: S. Michel, Bunte
Steine – Dunkle Bilder. «Magische Gemmen», Munich, 2001b.
Smaller groups have of course also been published in their
museological contexts, e.g. M. Henig, Classical Gems: Ancient and
Modern Intaglios and Cameos in the Fitzwilliam Museum,
Cambridge, 1994; Z. Kiss, ‘Amulettes magiques gréco-égyptiennes
au Musée National de Varsovie’, in K.M. Ciałowicz (ed.), Les
civilisations du Bassin Méditerranéen: Hommages à J. Śliwa
(Instytut archeologii), Cracow, 2000, 375–82; T. Gesztelyi, Antike
Gemmen im Ungarischen Nationalmuseum, Budapest, 2000, nos
253–7; F. Marco Simón, ‘New Magical Gems in Madrid’, in
A. Mastrocinque, Gemme gnostiche e cultura ellenistica, Bologna,
2002, 87–101. So far as I know, however, some important
collections, e.g. the c. 50 magical gems in the National Museum in
Copenhagen, remain unedited.
8 A. Mastrocinque (ed.), Sylloge gemmarum gnosticarum, 1–2
(Bollettino di Numismatica, Monografia 8.2.1 and 2), Rome, 2004–
7 [or 2008]. His revision of Delatte-Derchain (see n. 54) is due to be
published in the first half of 2012.
9 ‘Der Abdruck gibt nur ein Exzerpt des Originals’: E. Zwierlein-
Diehl, Antike Gemmen in Deutschen Sammlungen, Band II:
Staatliche Museen ∙ Preußischer Kulturbesitz, Antikenabteilung,
Berlin, Munich, 1969, 11; cf. C. Bonner, Studies in Magical Amulets,
chiefly Graeco-Egyptian, Ann Arbor, 1950, 251. Bonner was
habituated to the traditional means of copying gems for private
collections, cf. G.M. Facchini, ‘Riproduzione di gemme nel XVIII-
XIX secolo: la collezione di impronte di «Ennio Quirino Visconti» a
Novara’, in Hackens and Moucharte, (n. 1), 405–15.
10 S. Michel, Die Magischen Gemmen (Studien aus dem Warburg-
Haus 7), Berlin, 2004, 32. All the gems in A. Mastrocinque (ed.),
Sylloge gemmarum gnosticarum 2 (Bollettino di Numismatica,
Monografia 8.2.2), Rome, 2007 [or 2008], are presented both in
b/w and in colour plates at the end of the volume; Vitellozzi (n. 7),
provides magnificent colour enlargements in his Perugia
catalogue.
11 Bonner (n. 9), vii. Already in his review of vol. 1 of Karl
Preisendanz’ Papyri graecae magicae (1928) A.D. Nock had
expressed the hope that a comparable corpus of magical gems
might be created: Journal of Egyptian Archaeology 15 (1929), 219–
35, at 235 = idem, Essays in Religion and the Ancient World 1 (ed.
Z. Stewart), Oxford, 1972, 176–94, at 194; cf. A.A. Barb, review of
Delatte-Derchain (n. 54), Gnomon 41 (1969), 298–307, at 307.
12 See pro tem. the eight entries available under <http://www2.
szepmuveszeti. hu/campbell>, whose images however cannot be
enlarged. The Bollettino di Numismatica has generously made
Mastrocinque (n. 10) in its entirety available on its web-site.
13 Michel 2001b (n. 7) is the catalogue of an exhibition that was
shown between 2001 and 2003 in Hamburg, Hanau, Freiburg i/B,
Dresden, Künzing, Stendal, Hamm, Kassel and Hannover;
Vitellozzi (n. 7) is the catalogue of an exhibition in Perugia entitled
‘Mira et magica. Gemme e magia dalle collezioni del Museo
Archeologico Nazionale dell’Umbria, Perugia mostra 26
settembre–31 dicembre 2009’ [a summary version in Mastrocinque
(n. 10), 102–18]. The Budapest exhibition is scheduled for
December 2011, but may be delayed .
14 For the proceedings of the conference organised by Prof. Attilio
Mastrocinque on October 22–23 1999 in Verona, see Mastrocinque
(n. 7). Dr. Ken Lapatin organised a symposium on magical amulets
at the Getty Villa Museum in February 2010; another, arranged by
Prof. Véronique Dasen, was held at Fribourg in September 2010. So
far as I know, the first gem conference of any kind – though
magical gems were scarcely mentioned – was the interdisciplinary
effort organised by Tony Hackens, the then professor of Classical
archaeology and numismatics at Louvain, and held at the
European University Centre for Cultural Heritage at Ravello, 13–16
Nov. 1987: Hackens and Moucharte (n. 1), 7f.
15 Uterine: V. Dasen, ‘Métamorphoses de l’utérus d’Hippocrate à
Ambrise Paré’, Gesnerus 59 (2002), 167–86; eadem, ‘Représenter
46 | ‘Gems of Heaven’
l’invisible: La vie utérine et l’embryon sur les gemmes magiques’, in
V. Dasen (ed.), L’embryon humain à travers l’histoire. Images,
savoirs, rites, Gollion, 2007, 41–64; also her contribution to the
present volume; phoenix: Á.M. Nagy, ‘Le phénix et l’oiseau benw
sur les gemmes magiques: Trois notes sur le phénix gréco-
égyptien’, in S. Fabrizio-Costa (ed.), Phénix: Mythes et signes, Bern,
2001, 57–84 (I thank Dr. Nagy for sending me a copy of this paper);
ouroboros: M.G. Lancellotti, ‘Il serpente ouroboros nelle gemme
magiche’ in Mastrocinque (n. 7), 71–85.
16 Michel (n. 10); for some general comments see also my review in:
Journal of Roman Archaeology 21 (2008), 713–18.
17 ‘La bibliographie des gemmes magiques se réduisit-elle à quelques
articles perdus dans la vastitude des études classiques’: Á.M. Nagy,
‘Gemmae magicae selectae: Sept notes sur l’interprétation des
gemmes magiques’, in Mastrocinque (n. 7), 153–79, at 153.
18 See the comments on method by Nagy (n. 17), 168f.; also
C. Sfameni, ‘Culti egiziani e magia: il contributo delle gemme
magiche’, in E. Sanzi and C. Sfameni, Magia e culti orientali: Per la
storia religiosa della Tarda Antichità (Hierá: Collana di studi storio-
religiosi 11), Cosenza, 2009, 141–78.
19 Review of M. Maaskant-Kleibrink, The Classification of Ancient
Engraved Gems, in Gnomon 50 (1978), 494–8, at 495.
20 M. Maaskant-Kleibrink, Catalogue of the Engraved Gems in the
Royal Coin Cabinet, The Hague: The Greek, Etruscan and Roman
Collections, The Hague, 1978, vol. 1, 5.
21 C.M. Tomaselli, ‘La collezione ottocentesca di scarabei, gemme e
cammei del Museo di Archeologia dell’Università di Pavia’, in
Hackens and Moucharte (n. 1), 249–79, at 251f.
22 J. Śliwa, Egyptian Scarabs and Magical Gems from the Collection of
Constantine Schmidt-Ciążyński, Cracow, 1989, 27–9. The
collection, including Babylonian seals etc., amounted in 1886 to
some 2,517 items and is almost free of early-modern imitations and
plain forgeries. As was usual, Schmidt-Ciążyński also purchased a
number of existing private collections en bloc. Several of the
magical amulets in the British Museum collection were donated by
him: Śliwa, ibid., 30, n. 16; M.L. Bierbrier, ‘The Schmidt Collection
in the British Museum’, in Ciałowicz (n. 7), 49–52.
23 For example, the fine Yüksel Erimtan collection of rings and gems
from Asia Minor contains just two magical gems (nos 166–7), one
of which carries no text: K. Konuk and M. Arslan, Ancient Gems and
Finger-Rings from Asia Minor, Ankara, 2000 [original Turkish
version, 1992], 190f.
24 Zazoff (n. 4), 352–4; Philipp (n. 4), 8–11; Michel (n. 10), 2, n. 7.
Carnuntum: H. Jobst, ‘Syrische Kulte’ [sic], in W. Jobst (ed.),
Carnuntum. Das Erbe Roms and der Donau. Katalog der Ausstellung
... in Bad Deutsch Altenburg, Bad Deutsch Altenburg, 1992, 59–71, at
69f., nos 17–19; G. Demski, Die antiken Gemmen und Kameen aus
Carnuntum (Archäologischer Park Carnuntum, Neue
Forschungen 1), Vienna, 2005, 161–3, nos 1105–13. B. Nardelli,
‘Gemme magiche della Dalmazia’, in Mastrocinque (n. 7), 181–94,
observes in the case of the Dalmatian gems that the vicissitudes of
war, political turbulence and the fates of museums have caused an
almost total loss of precise information relating to provenance.
Nevertheless, we do know that these magical gems were found
either in tombs or in urban contexts in Pola and the Croatian
coastal area, which is more than we can say about virtually any of
the gems in the great museum collections.
25 E.g. Caesarea Maritima, where A. Hamburger and colleagues
collected 165 gems in the dunes prior to the Joint Expedition’s
excavations: A. Hamburger, ‘Gems from Caesarea Maritima’,
’Atiqot 8 (1968), 1–38, at 1; Gadara: M. Henig and M. Whiting,
Engraved Gems from Gadara in Jordan: The Sa’d Collection of
Intaglios and Cameos, Oxford, 1987, though the authors
unfortunately make no reference to the policies pursued in
creating the collection.
26 Since the 1940s it has become usual in the Anglo-American world
to observe the distinctions introduced by the New Archaeology of
that period: provenance = the general location of an
archaeological find; provenience = its precise location within an
assemblage. This distinction is not noted in e.g. W. Martini,
Sachwörterbuch der klassischen Archäologie, Stuttgart, 2003,
where we read s.v. ‘Provenienz’: ‘Herkunft eines Kunstwerkes aus
dem Umkreis eines Künstlers, einer Kunstrichtung oder einer
Region bzw. aus dem Besitz eines früheren Eigentümers’. There is
no entry ‘Herkunft’.
27 M.G. Lancellotti, ‘Problèmes méthodologiques dans la constitution

d’un corpus des gemmes magiques’, in A. Moreau and J.-C. Turpin
(eds), La magie, 2: La magie dans l’antiquité grecque tardive,
Montpellier, 2000, 153–66, at 161–3. Mastrocinque (n. 7), 7, rightly
advocates the creation of collective projects in view of the mass of
materials and the obscurity of their iconography. The co-operation
of Egyptologists, and particularly Demoticists competent in late
Egyptian religion, is of course a particular desideratum; cf. J.-F.
Quack, review of Michel 2001a (n. 7): Gnomon 76 (2004), 257–62.
28 L. Barkan, Unearthing the Past: Archaeology and Aesthetics in the
Making of Renaissance Culture, New Haven, 1999, 126.
29 Aquileia: Philipp (n. 4), 9, n. 18; Altino: Mastrocinque (n. 10), 11f.;
Caesarea: Hamburger (n. 25), 33, counted 13 + 1 gems as ‘magical’,
on modern criteria the number can be reduced to eight; Gadara:
Henig and Whiting (n. 25), nos 423–7.
30 M. Henig, A Corpus of Roman Engraved Gemstones from British
Sites (BAR 8.1), Oxford, 1974, 55–81; A. Krug, ‘Römische Gemmen
und Fingerringe im Museum für Vor- u. Frühgeschichte Frankfurt
a.M.’, Germania 53 (1975), 113–25 (includes several from
Heddernheim); eadem, ‘Römische Fundgemmen, 2’, Germania 55
(1977), 77–84; eadem, ‘Römische Fundgemmen, 3’, Germania 56
(1978), 476–503; eadem, ‘Römische Fundgemmen, 4’, Germania 58
(1980), 117–35; eadem, Antike Gemmen im Römisch-Germanischen
Museum Köln (Bericht der Römisch-Germanischen Kommission
61, 1980), Frankfurt am Main, 1981, 151–260. She rightly warned
however that dealers’ claims about provenance are often false. The
St Petersburg collection is exceptional in containing a number of
magical gems with proveniences (precise tomb numbers): O. Ya.
Névérov, ‘Gemmes, bagues et amulettes magiques du Sud de
l’URSS’, in M.B. de Boer and T.A. Edridge (eds), Hommages à M.J.
Vermaseren (Études Préliminaires aux Religions Orientales dans
l’Empire Romain 68.2), Leyden, 1978, 833–48: necropolis of Olbia
(nos 18, 28); of Chersonnesus (nos 27, 49); of Samtavo (no. 33); of
Armasiskhevi (no. 40).
31 A. Delatte, ‘Études sur la magie grecque, 4: Amulettes inédits des
Musées d’Athènes’, Le Musée belge 18 (1914), 21–96, at 21f.
32 Philipp (n. 4), 11f.
33 Zwierlein-Diehl et al. (n. 7), 18; eadem 1992 (n. 1), 15 (‘vermutlich
Alexandria’); eadem 2007 (n. 1), 210.
34 Bonner (n. 9), 26, did however draw attention to the occasional
similarity between amuletic designs and Alexandrian coin types:
cf. e.g. 262, no. 56 (Artemis). Michel (n. 10), 1–9, likewise avoids all
reference to a specific centre of production.
35 Of the almost 100 magical gems in the Kunsthistorisches Museum,
Vienna (nos 2178–2273), 11 were certainly acquired by Burghart
(but also by Ph. Agnello: see nos 2253, 2256), in Egypt prior to 1821;
in all, Burghart collected 27 gems and 27 glass pastes in Egypt: A.
Bernhard-Walcher, ‘Zur Geschichte der Gemmensammlung’, in
Zwierlein-Diehl et al. (n. 7), 28–38, at 34. Neither of the rather fine
items acquired by Agnello is inscribed. The Description de l’Égypte,
vol. 5, pls 85–7, which illustrates a good number of scarabs and
other amulets, includes no ‘magical gems’ in our sense.
36 U. Mandel-Elzinga, ‘Eine Gemmensammlung aus Alexandria im
Akademischen Kunstmuseum der Universität Bonn’, Bonner
Jahrbücher 185 (1985), 243–98. She counted no less than 12,
including a seal of Solomon and a John the Baptist, as ‘magical’;
but only one of the ten (no. 64 = Michel [n. 10], 310, §37.B.1.f), a
fragmentary lion-head with Chnoubis, is inscribed, and
incompetently at that. Despite part of it having been acquired by
the von Claer family, prominent in the Rhineland since the late
17th century, the collection as a whole is of rather poor quality.
37 Cf. A.M. El-Khachab, ‘Some Gem-Amulets depicting Harpokrates
seated on a Lotus-Flower’, Journal of Egyptian Archaeology 57
(1971), 132–45.
38 A.M. El-Khachab, ‘A Collection of Gems from Egypt in Private
Collections [sic]’, Journal of Egyptian Archaeology 49 (1963), 147–
56, at 147; cf. too the remarks of J. Thompson, Edward William
Lane: The Life of the Pioneering Egyptologist and Orientalist,
London, 2010, 335f., on the lively antiquities trade in the 1830s all
the way up the Nile. It is of course unknown what proportion of
these gems in private hands in Egypt is to be classed as ‘magical’.
39 L.E. Tacoma, Fragile Hierarchies: The Urban Élites of Third-Century
Egypt (Mnemosyne Supplements 271), Leyden, 2006, 21–152, is
fundamental.
40 Assuming the 3,500–5,000 surviving amulets of all qualities were
made over a period of 200 years, the annual production would
have been between 14.5 and 25 items. Of course we have no idea of
Archaeologies of Magical Gems
the actual numbers ever cut.
41 The significance in this connection of Tacoma’s ‘regional elite’, i.e.
the major families of Alexandria, who owned land throughout
Egypt, must remain conjectural.
42 This route ran from Qana’ (Bi’r ‘Ali in Yemen) to Berenike or Myos
Hormos (Quseir al Qadim) on the Egyptian Red Sea coast, and
then by camel across the mountain/desert to Hermopolis: A.
Sedov, ‘The Port of Qana’ and the Incense Trade’, in D. Pocock and
D. Williams (eds), Food for the Gods: New Light on the Ancient
Incense Trade, Oxford, 2007, 71–111; S. Gupta, ‘Frankincense in the
‘Triangular’ Indo-Arabian-Roman Aromatics Trade’, ibid., 112–21.
43 J.Z. Smith, ‘The Temple and the Magician’, in idem, Map is not
Territory, Chicago, 1993, 172–207 (orig. 1978); idem, ‘Trading
Places’, in M. Meyer and P. Mirecki (eds), Ancient Magic and Ritual
Power (Religions in the Graeco-Roman World 129), Leyden, 1995,
13–27, repr. in idem, Relating Religion: Essays in the Study of
Religion, Chicago, 2004, 215–29.
44 Cf. K. Preisendanz, ‘Die Überlieferung der griechischen
Zauberpapyri’, in Miscellanea critica: Festschrift zum 150-jährigen
Bestehen des Verlages B.G. Teubner, 1, Leipzig, 1964, 203–17.
45 One of the conclusions of J. Dieleman, Priests, Tongues and Rites:
The London-Leiden Magical Manuscripts and Translation in
Egyptian Ritual (100–300 ce) (Religions in the Graeco-Roman
World 153), Leyden, 2005, is that the London-Leyden Demotic
papyrus, whose constituent recipes were originally composed in
Greek in IIp, was in effect translated into Demotic in c. 200p so as to
confer upon the contents the authority of a sacred language. Such a
translation can at that date only have been performed by temple
priests. The marginalisation of Coptic, which is scarcely
represented in the PGM, is likewise instructive.
46 L. Robert, ‘Amulettes grecques, i’, Journal des Savants (1981), 3–27 =
Opera min. 7, Paris, 1990, 465–93 = Choix d’écrits (ed. D. Rousset),
Paris, 2008, 357–75.
47 E.R. Goodenough, Jewish Symbols in the Graeco-Roman Period, 2:
The Archaeological Evidence from the Diaspora, Toronto, 1953, 208–
95; Michel (n. 10), 113f.; see also M. Smith, ‘Goodenough’s Jewish
Symbols in Retrospect’, Journal of Biblical Literature 86 (1967), 67,
add. note 2. A modern version, that the anguipede is a sort of visual
rebus on the Hebrew root gbr, is argued by Á.M. Nagy, ‘Figuring out
the Anguipede ... and his relation to Judaism’, Journal of Roman
Archaeology 15 (2002), 159–72. G. Bohak, Ancient Jewish Magic,
Oxford, 2008, 197, n. 152 is unwarrantedly sceptical.
48 Cf. M. Henig, ‘Archbishop Hubert Walter’s Gems’, Journal of the
British Archaeological Association 136 (1983), 56–61.
49 On Venetian collectors in the 15th century, important both because
of Venice’s control over the Levantine trade long after 1453 and on
account of her possession of Crete and other islands, see briefly P.
Fortini Brown, Venice and Antiquity: The Venetian Sense of the Past,
New Haven, 1996, 77; on Cyriacus of Ancona, 83.
50 A. Krug, ‘Wiederverwendung und Gebrauch antiker Gemmen im
Mittelalter’, in Hackens and Moucharte (n. 1), 369 (abstract). As far
as I know, the project announced there was never completed; but
cf. e.g. the study of 15 re-used gems in eadem, ‘Antike Gemmen an
mittelalterlichen Goldschmiedarbeiten im Kunstgewerbemuseum
Berlin’, Jahrbuch der Berliner Museen n.f. 37 (1995), 103–19.
51 G. Sfameni Gasparro, ‘Religione e magia nel mondo tardo-antico:
il caso delle gemme magiche’, in Mastrocinque (n. 7), 243–69, at
251; Philipp (n. 4), 23.
52 E. Zwierlein-Diehl, ‘Donaureiter in Alexandria’, Kölner Jahrbuch
43 (2010), 847–57 (I thank Prof. Zwierlein-Diehl for kindly sending
me an off-print), cf. Michel (n. 10), 92, n. 478.
53 Among these are: M.J. Vermaseren, Corpus inscriptionum et
monumentorum religionis mithriacae, 2 vols, The Hague, 1956 and
1960, nos 2353, 2354 (2355 is a modern imitation), 2356, 2359, 2361,
2364, 2365; it must be admitted that 2359, which is in the Cairo
Museum, is of Egyptian provenance; on the theme in general:
A. Mastrocinque, Studi sul Mitraismo (il Mitraismo e la Magia),
Rome, 1998.
54 Dionysiac: e.g. Zwierlein-Diehl 1992 (n. 1), 100f., no. 29; Hermes:
Michel (n. 7), 40, nos 61–2; Asklepios: A. Delatte and P. Derchain,
Les intailles magiques gréco-égyptiennes, Paris, 1964, 178f., nos
234f.; Nemesis: ibid., 193, no. 256; Artemis of Ephesus: ibid., nos
239–40; Cybele (?), ibid., no. 291. We can also think of highly
summary, indeed largely unintelligible, representations such as
the gems listed by Michel (n. 7), 262–4, nos 418–23.
55 E.g. Zwierlein-Diehl 1992 (n. 1), 103f., no. 31.
‘Gems of Heaven’ | 47
Gordon
56 A rough idea of their distribution is provided by R. Kotansky, Greek
Magical Amulets. The Inscribed Gold, Silver, Copper and Bronze
Lamellae, 1: Published Texts of known Provenance. (Papyrologica
Coloniensia 22.1), Opladen, 1994. For a provisional list of
(Christian) phylacteries on lead, see S. Giannobile and D.R.
Jordan, ‘A Lead Phylactery from Colle san Basilio (Sicily)’, Greek,
Roman and Byzantine Studies 46 (2006), 73–86, at 81–4.
57 See the survey with translated texts by J.C. Gager, Curse Tablets
and Binding Spells from the Ancient World, New York, 1992, repr.
1999; the recent volume by M. Martin, Sois maudit! Malediction et
envoûtement dans l’Antiquité, Paris, 2010, mainly concerns
defixiones outside this tradition.
58 Nagy (n. 17), 162.
59 K. Preisendanz (ed.), Papyri graecae Magicae (ed. 2 A. Henrichs),
2 vols, Stuttgart, 1973–74; ed. 1, 3 vols, Leipzig and Berlin, 1928,
1931, [1941].
60 In her survey of the gems from the Lower Danube area, where
there were gem workshops at Novae, Ratiaria, Serdica and perhaps
elsewhere, A. Dimitrova-Milčeva, ‘Die Gemmen und Kameen vom
Unteren Donaulimes in Bulgarien’, Studien zu den Militärgrenzen
Roms, 2 (Beihefte der Bonner Jahrbücher 38), Cologne and Bonn,
1977, 282–7, mentions none that would now be classified as
‘magical’.
61 M. Maaskant-Kleibrink, ‘The Microscope and Roman Republican
Gem Engraving. Some Preliminary Remarks’, in Hackens and
Moucharte (n. 1), 189–204; no second edition of her Classification
of Ancient Engraved Gems (n. 19), where she promised to continue
this research, ever appeared; but cf. eadem, ‘Three Gem-engravers
at Work in a Jeweller’s Workshop in Norfolk: The Evidence of the
Roman Engraved Gems in the Jeweller’s Hoard found at
Snettisham’, Bulletin Antieke Beschaving 67 (1992), 151–67.
62 Brooklyn Museum inv. no. 47.218.156; illustrated in the pull-out in
S. Sauneron, Le papyrus magique illustré de Brooklyn (Wilbur
Monographs 3), Brooklyn NY, 1970, with the comments of E.M.
Ciampini, ‘Tradizione faraoniche e iconografie magiche’, in
Mastrocinque (n. 7), 27–40, at 34. On Egyptian temple magic, see
the authoritative account by J.-F. Quack, ‘La magie au temple’, in
Y. Koenig (ed.), La magie en Égypte. À la recherche d’une définition,
Paris, 2002, 41–68.
63 The range of Pharaonic magical texts can be judged from the
editions and translations by J.F. Borghouts, e.g. The Magical Texts
of Papyrus Leiden I. 348 (OMRL 51), Leyden, 1971; Ancient Egyptian
Magical Texts, Leyden, 1978; cf. idem, ‘Magical Texts’, in Textes et
langages de l’Égypte pharaonique: Hommage à J.-F. Champollion, 3,
Cairo, 1974, 7–19; idem, s.v. ‘Magie’, Lexikon der Ägyptologie 3
(1980), 1137–51.
64 W. Brashear, ‘Magical Papyri: Magic in Bookform’, in P. Ganz (ed.),
Das Buch als magisches und als Repräsentationsobjekt (Wolffen-
bütteler Mittelalter-Studien, 5), Wiesbaden, 1992, 25–57.
65 It is difficult to discover whether older types of Egyptian amulets
continued to be produced in the Roman period, i.e. co-existed with
the intaglios, which would again have implications for the type of
market served by ‘magical gems’. I know only of scattered items in
exhibition catalogues.
66 A typical example of this learned frame of reference is a green
jasper in the British Museum (Pl. 1). The obverse shows a papyrus
boat, with Horus-falcons at either end, carrying Harpokrates on
the lotus; the god is being adored by an ithyphallic hamadryad.
This motif evokes not just the complex of rituals that
commemorated the moment of creation but also the order
necessary to their continuation and their efficacy. The sun and
moon in the field reinforce this claim by suggesting the
immutability of this order. Cf. Philipp (n. 4), 76f., no. 100;
Zwierlein-Diehl 2007 (n. 1), 219, no. 782; Michel (n. 10), 272f., §19.3;
Ciampini (n. 62), 39.
67 Th. Hopfner, s.v. Λιθικά, Realencyclopädie der classischen
Altertumswissenschaft 13.1 (1926), 747–69, at 748f.; M. Wellmann,
‘Die Stein- und Gemmenbücher der Antike’, Quellen und Studien
zur Geschichte der Naturwissenschaften und der Medizin 4.4 (1935),
86–110; R. Halleux and J. Schamp (eds, trans.), Les lapidaires grecs
[Budé], Paris, 1985, xiii–xxxiv; E. Reiner, Astral Magic in Babylonia
(Transactions of the American Philosophical Society, 85.4),
Philadelphia, 1994.
68 Bust of Sarapis, surmounted by a kalathos and encircled by the
command addressed to the god (see Pl. 2); cf. Philipp (n. 4), 55f.,
nos 55f.; Mastrocinque (n. 10), 148, Si 3, also 132, Ro 4 (Asia Minor);
48 | ‘Gems of Heaven’
Michel (n. 10), 326, §45.3.c.
69 Red jasper with Herakles throttling the Nemean lion; his club in
the field. The reverse bears the charakteres: kkk hhh, against
‘colic’, cf. Zwierlein-Diehl 2007 (n. 1), 226, no. 799; Michel (n. 10),
280f., §23.1.a.
70 Brown jasper showing (obverse) a scorpion with an eight-pointed
star between its claws. The star suggests an allusion to Scorpius’
melothesic power over the genitalia, e.g. Manilius, Astronomica
2.462 = 4.707; Firmicus Maternus, Mathesis 2.24; the scheme went
back to ‘Nechepso-Petosiris’; cf. Philipp (n. 4), 86, no. 122 (reverse);
Zwierlein-Diehl 2007 (n. 1), 219, no. 798 (obverse); Mastrocinque
(n. 10), 94, Na 24; Michel (n. 10), 332, §51.1.b.
71 The obverse shows an uterus, with seven-tongued key and lateral
ligaments, protected by four deities, (from left) Isis, Chnoubis,
Anubis and Nephthys, all within an ouroboros, which serves to
confine the deities’ action within a specific space. Texts: around
rim: the Soroor-logos; in field: clockwise, the seven vowels;
vertically on the uterus: iaw; cf. Delatte-Derchain (n. 54), 252, no.
353; Philipp (n. 4), 112, no. 184; Zwierlein-Diehl 2007 (n. 1), 227f.,
no. 801; Michel (n. 10), 235, §54.2.c.; also the paper by Dasen, this
volume. The text on the reverse, opwpiovθ, is standard on gems in
this category; cf. A.A. Barb, ‘Diva matrix’, Journal of the Warburg
and Courtauld Institutes 16 (1953), 193–238, at 202.
72 Yellow jasper, obverse: crab = Cancer; on rim, the seven vowels
(?). Reverse: bapxai; cf. the paper by Michel, this volume, pl. 4;
Delatte-Derchain (n. 54), 270, no. 385; Philipp (n. 4), 83, no. 115
reverse; Mastrocinque (n. 10), 95, Na 26; Michel (n. 10), 303,
§33.1.a, with M. Monaca, ‘Gemme magiche e divinazione’, in
Mastrocinque (n. 7), 135–52.
73 Haematite (obverse): on a crocodile, a ‘pantheistic’ figure with
four ibis-heads, four wings, and four arms; between the four
necks, a pole bearing a star. In the lower pair of arms, the deity
holds a was-sceptre (left) and perhaps an ankh or a za-sign (right),
both poorly formed. Around this motif, the ablanathanalba-
akrammacharei-logos. No design on reverse. There are no very
close parallels, the nearest being Delatte-Derchain (n. 54), 51, no.
44 obverse (no wings, no neck-pole, on a lion; sun and moon
above); also Bonner (n. 9), 297, no. 265, reverse; 317, no. 378
obverse; Michel (n. 10), 320, §41.6.b. On ‘pantheistic’ figures see
now J.-F. Quack, ‘The so-called Pantheus. On Polymorphic Deities
in Late Egyptian Religion’, in H. Györl (ed.), Aegyptus et Pannonia
III. Acta Symposii Anno 2004, Budapest, 2006, 175–90.
74 E.g. L. Török, Hellenistic and Roman Terracottas from Egypt
(Biblioteca Archaeologica 15), Rome, 1995 (catalogue of the
holdings of the Museum of Fine Arts in Budapest).
75 This possibility is suggested by the fact that some amulets were not
made for rings, e.g. the elongated haematite in the British Museum
with (obverse): a lion-headed ‘pantheistic’ figure with six wings
holds a tabula ansata above its head inscribed ia[w]; in the field, on
either side, schematic palm-fronds and stars. Reverse: CT/OMA/XEΠ/
ΕΠ/ΤΕ, ‘Stomach, digest!’: Michel 2001a (n. 7), no. 180, cf. also nos
395–400; also Bonner (n. 9), 60f. (on digestive amulets in
haematite from Syria).
76 Michel (n. 10), 84–93.
77 Obverse: Isis-Hathor-Aphrodite, holding a fanciful sceptre
(conceivably a reference to Horus-Chnum-Agathodaimon)
embracing Osiris, who appears both mummiform and draped;
sun-disk and uraeus on his head. To left, Harpokrates with lotus-
bud on forehead, as a statue on base (no gesture, no flail); in the
field, stars. No close parallels; for a strange animal on Isis’ staff, cf.
Delatte-Derchain (n. 54), 86f., no. 108. For mini-Harpokrates in the
field, cf. Mastrocinque (n. 10), 175, Ve 2.
78 E.g. PGM IV 2877–90 (Hekate on magnetite); V 240–7 (scarab
engraved with Isis); 447–58 (seated Sarapis, with an onoma on the
reverse). These are all straightforward divine images; in general,
however, the magical papyri do not envisage stones as phylacteries
for the rituals they prescribe – metal lamellae are much more
common. The stone mnizouris is prescribed in Orac. Chald. frg. 149
des Places as a protection against terrestrial daemones, but there is
no mention of an image.
79 Carnelian (obverse): in the central oval (possibly, an ouroboros), a
phoenix; in the four quarters, from top right: Harpokrates on lotus,
[crocodile], Horus-falcon, winged Agathos-Daimon, i.e. four
aspects of Ra/Phre signifying the four quarters (the totality of the
cosmos). For two amuletic parallels, see A.A. Barb, ‘Abraxas-
Studien’, in Hommages à W. Deonna (Coll. Latomus 28), Brussels,

1957, 67–86, at 81–6); on the significance of the directions, cf. Chr.
Harrauer, Meliouchos (Wiener Studien, Beiheft 11), Vienna, 1987,
36–9 (both citing PGM II 105–15 and XII 87–9).
80 ‘Pantheos’ gems are often routinised, cf. Michel (n. 10), 317, §41.1.
Pl. 11 shows one example (obverse): on an ouroboros-cartouche, a
standing Bes-Pantheos with four wings, four arms and a bird’s tail.
In the lower left hand, a was-sceptre, in the right, a nhaha
‘ ‘
-flagellum; on the head, a clumsy atef-crown on top of a kalathos.
Reverse: a now largely indecipherable text of 13 lines: Michel 2001a
(n. 7), no. 161; cf. Bonner (n. 9), 295, no. 254; Delatte-Derchain (n.
54), 132f., no. 168.
81 The semantic cluster τελεῖν does occur in this context however, e.g.
PGM IV 1596–1715 passim, 1744; V 242 etc.; cf. also the frequent use
of consecratio/consecratus in Damigeron-Evax 3.6, 5.5, 6.4, 7.7 etc.,
and the (fanciful) ritual in Orph. Lith., 366–80. In my view
elaborate consecration was an ideal associated with ‘high’ ritual-
magical practice, a theoretical norm that was usually not followed
(cf. Damigeron-Evax 46.3: ‘hunc autem de qua vis consecratione
consecras’; Socrates and Dionysius 39.7, where ‘consecration’
involves simply the act of engraving and fitting into a ring). I
intend to return elsewhere to this question.
82 A. Appadurai, The Social Life of Things: Commodities in Cultural
Perspective, Cambridge, 1996; T. Dant, Material Culture in the
Social World: Values, Activities, Lifestyle, Buckingham, 1999,
40–59.
83 G. Woolf, Becoming Roman: The Origins of Provincial Civilization in
Gaul, Cambridge, 1998; A. Gardner, ‘Seeking a Material Turn: The
Artefactuality of the Roman Empire’, Proceedings of the 12th
Annual Theoretical Roman Archaeology Conference, University of
Canterbury (hereafter TRAC), Oxford, 2002, 1–13.
84 C. Martins, ‘Being Consumers: Looking beyond Wealth as an
Explanation for Villa Variability’, TRAC (n. 83), 84–100; cf. G.
Monteil, ‘Samian and Consumer Choice’, TRAC 13 (2003), 1–15.
85 Drawing of the Akephalos from col. IV of P. Berol. 5026 = PGM II
166, to indicate to the practitioner the design to be drawn on
papyrus (l.47f.) in order to obtain a direct revelatory vision.
86 Haematite (obverse): within an ouroboros, a carefully aligned
series of five triads of the (variable list of) creatures that symbolise
different phases of Ra/Phre, here: scarab, Horus-falcon, buck-
goat, crocodile, uraeus. In the field, two sets of vowels listed in
descending (left) and ascending (right) order. Reverse: the
chabrach-logos. Since all of these are standard in the iconography
of Harpokrates, his presence here is implied; cf. Delatte-Derchain
(n. 54), 124, nos 163f.; Philipp (n. 4), 78, no. 105; Michel (n. 10), 317,
§19.4a.
87 By the concept ‘theodicy of good fortune’ Weber meant the world
view that enables elites and the socially successful to believe in the
legitimacy of their good fortune, cf. G. Küenzlen, Die Religions-
soziologie Max Webers. Eine Darstellung, Berlin, 1980, 81.
88 Green jasper (obverse): the cock-headed anguipede; within the
shield iaw; round the shield: the seven vowels; upper border:
mamaraouth-logos; lower border: arponchnouphi-logos.
Reverse: chabrach-logos: Michel 2001a (n. 7), no. 181; cf. Michel
(n. 10), 242, §3.A.1.j.
Archaeologies of Magical Gems
89 As a ringstone: e.g. Socrates and Dionysius 27.1, 39.7, 50.2; hung
around the neck: idem, 48.1; as a brooch: Damigeron-Evax 6.5–6.
The expression καί ἄνευ γλυφῆς (‘even without an engraved
image’) at Socrates and Dionysius 28.1 seems to suggest that the
image serves to enhance or focus the inherent power of the stone.
Certain types of amulet were of course not intended for display,
such as those bound to the body (e.g. Damigeron-Evax 8.7; 9.9, 10.5,
28.6, 44.1 etc.; Dioscorides 5.142), clutched in the hand
(Damigeron-Evax 42.2), kept inside the zona (ibid. 55.4) or
wrapped in the skin of a sacrificed animal (Pliny, Nat. Hist.,
XXXVI.151). An unknown percentage of surviving stones has been
re-cut in order to fit modern ring settings, and may not originally
have been so used.
90 G. Simmel, ‘Das Geheimnis und die geheime Gesellschaft’, in his
Soziologie: Untersuchungen über die Formen der Vergesellschaftung,
Berlin, 1908, 256–304, at 304.
91 See also the remarks of Á.M. Nagy, this volume, 75–81.
92 Platz-Horster (n. 19), 496.
93 Compare the progress made in creating electronic data-bases in
the area of Latin epigraphy: F. Feraudi-Gruénais, ‘An Inventory of
the Main Electronic Archives of Latin Epigraphy’, in eadem (ed.),
Latin on Stone: Epigraphic Research and Electronic Archives,
Lanham MD, 2010, 157–60.
94 The title of Halleux-Schamp (see n. 67). The best brief surveys of
the ancient texts remain Hopfner (n. 67), 747–61; A. Hermann, s.v.
Edelsteine, Reallexikon für Antike und Christentum 4 (1958), 505–
52.
95 The recipe prescribes coral; but there are no such images on what
we would classify as coral.
96 Socrates and Dionysius 26.5 (on an emerald) = Damigeron-Evax
6.6–7 = PGM V 238–43. For a series of alternative speculations
about the possible referential contexts, see Halleux-Schamp (n.
67), 166, n. 4.
97 ‘Hoc enim mysterium ceteris egyptiis litteratis neque allophylis ne
traderis nec cuiquam alii ne ad sterilitatem huius scientiae
Aegyptus deveniat ...’, Incipit 3–4, 230 Halleux-Schamp (n. 67); cf.
e.g. 34.1, ‘dicunt magi et Egyptii synciten’. The manuscript tradition
of Damigeron-Evax, which, as a ‘living text’, is particularly
complex, and offers several different astrological-occult schemes,
is analysed by Halleux-Schamp (n. 67), at 193–215.
98 Only Isis (6.5) and Hathor (‘Latona’) + Harpokrates (37.5).
99 Socrates and Dionysius 35.3 (not named as such); 36.3 (‘with three
heads’ = possibly the triple Chnoubis sign); first decan of Cancer:
ibid. 32.2; vox magica: 39.7 (iaxw); 50.2 (apam); charakteres: 36.6,
43.4, 50.2.
100 Greek deities invoked: Poseidon: Socrates and Dionysius 27.1;
Athena with erodios (evoking the Doloneia), ibid., 29.3; Artemis,
ibid., 30.6. Egyptian themes hellenised: Ram + Athena with a
heart, 31.4 (Isis with Anubis? = Michel (n. 10), 30.3.d, but cf.
Halleux-Schamp [n. 67], 169, n. 3); 34.2 (‘an egg’ = ouroboros?);
35.3; 50.2 (bent man holding a δίκελλα, a mattock, but no doubt =
the ‘reaper’?, cf. Bonner [n. 9], 72–4).
101 One point of contact will have been the quotient of the miraculous.
‘Gems of Heaven’ | 49
Text, Image and Medium
The Evolution of Graeco-Roman Magical Gemstones
Christopher A. Faraone
Scholars generally agree that in the Classical and Hellenistic
periods only a scant handful of images, for example, the head
of the Medusa, the phallus, or the frontal eye were used as
traditional amulets to ward off evil or disease,1 but they find it
difficult to identify many other protective or curative amulets
or to distinguish them from ornaments used for aesthetic or
other purposes. The world of the Roman Empire, on the other
hand, seems marked by an explosion of images and texts on
metal lamellae, papyri, mosaics and gemstones – all designed
to protect or to heal. By some scholarly accounts this
proliferation of readily identifiable amulets is best explained by
a sudden need for magic or a precipitous rise in superstition in
this period.2 In my wider work on Greek amulets, I have taken
to calling this historical argument or assumption the ‘big-bang
theory’ of creation for Graeco-Roman amulets and it is a
problematic one. First and foremost, I argue that it is unlikely
that the diverse populations of such a large geographic area
(from Great Britain to Afghanistan and from the banks of the
Danube to the edge of the Sahara) all ‘got superstition’ within a
few generations. I suggest, in fact, that these amulets were not
invented out of whole cloth in this period, but rather that they
only become visible to us in the archaeological record, because
of the increased epigraphic habit of the Roman Imperial
period. Or to put it another way: it is only when magical texts
appear side by side with traditional pairings of media and
images that we become aware that these media and images
themselves were thought to have innate magical powers to
protect or heal.
Such an argument is, of course, much larger than I can
attempt in this essay, and so I limit myself to a handful of test
cases, all of them engraved gemstones that have in the past
been described as ‘Gnostic’, ‘Graeco-Egyptian’ or simply
‘magical’.3 These gems generally date to the Roman period and
they comprise by far the largest corpus of extant Greek
amulets.4 They usually have three important components:
medium, image and text. In the past scholars have focused
primarily on the images and texts and have often ignored the
specific medium or colour, despite the fact that it is logically
more probable that the type of stone on which various images
and texts were inscribed was of basic or even original
importance.5
In what follows I examine five popular types of gems –
haematite, lapis lazuli and three differently coloured kinds of
jasper – and argue that prior to the Roman period simpler
versions of these same gems (with images only) were used as
amulets and that the sorcerers and stonecutters of the Roman
period innovated primarily by adding text to the stone. In each
case I shall define a basic typology (defined by a consistent
correlation between medium, image and text) and then trace
the evolution of this type, in which the final stage is the
addition of text.6 The elaborately inscribed gems of the Roman
50 | ‘Gems of Heaven’
period, in short, provide excellent evidence for the
scribalisation of amulets, not for their invention or recent
arrival from the East. This argument relies, in part, on the fact
that prior to the Roman period inscriptions of any sort are rare
on gems and limited almost exclusively to the single names of
their owners or occasionally the artists who carved them.7 We
shall see, moreover, that in each of my five studies, although
various literary sources – for example, medical books,
lapidaries and natural histories – describe or prescribe the
making of various gem amulets, they almost never mention the
inscription or text.8
Translucent green or white jasper and the radiant serpent
I begin with a series of light-green jasper gems used to cure
stomach ache and heartburn, a type discussed in some detail
by the Greek physician Galen:9
The testimony of some authorities attributes to certain stones a
peculiar quality, which is actually possessed by the light green10
jasper. Worn as an amulet, it benefits the stomach and the
oesophagus. Some set it in a ring and engrave on it a radiate
serpent, just as King Nechepso prescribed in his fourteenth book.
Of this stone I, too, have personal experience. I made a string of
small stones of this type and hung it from my neck at such a length
that the stones touched the oesophagus. They seemed no less
beneficial, even though they had not the design that Nechepso
prescribed.
Galen lived in the Roman Imperial period, but the book he cites
was written several centuries earlier in the Ptolemaic period,
probably by a hellenised Egyptian pretending to be Nechepso,
one of the last native pharaohs.11 A few centuries after Galen
made his experiment, Marcellus of Bordeaux recorded a
similar recipe in Latin: ‘This is a powerful remedy for pain of
the stomach: on a jasper stone carve a radiant serpent, so that it
has seven rays; enclose it in gold and employ it on the neck’.12
He does not say, however, what colour the jasper should be. We
have, then, three different recipes to guide us, each from a
different era: Nechepso’s Hellenistic version (green jasper and
radiant serpent set in a finger-ring); Galen’s Roman-Imperial
one (a string of plain jasper stones hung over the oesophagus);
and Marcellus’ Late Antique one (jasper and radiant serpent set
in gold and hung around the neck). All, however, aim at curing
pains in the digestive tract by using the same gem and – in two
cases – the same image, but none make any mention of
inscribed text.
Numerous examples of this type of light green jasper have,
in fact, been discovered in the Mediterranean basin and Europe
and they are conventionally dated to the Roman Imperial
period in which Galen lived.13 There is some variation in the
type of stone, running the gamut from translucent jaspers of
various shades of lighter green to more transparent gems of the
same hue, such as olivine.14 A good example of the type is a
green jasper in the British Museum (Pls 1a–d): on the obverse

Plates 1a–d Jasper with Chnoubis, 15 x 12 x 3mm. London, British Museum, PE G 397
side of this gem we find the radiant lion-headed serpent that
the two medical writers describe. This gem has on its reverse a
pair of inscriptions that almost always appear on such stones:
the name Chnoumis or Chnoubis in Greek and a symbol formed
by three Ss with a horizontal line running through them.
A treatise on gems by Socrates and Dionysus, roughly
contemporaneous with Galen, confirms some of his
description, but recommends a transparent or translucent
white stone, rather than a green one:15
Another onyx stone, white and entirely transparent, just like the
air. It is a type of onyx. Engrave on it, then, a serpent coil with the
upper part or head of a lion and rays. If worn this stone completely
prevents pain in the stomach; rather you will easily digest however
many foods you make use of. Let the wearer not set it aside.
A number of these translucent stones have survived from
antiquity, for example, one in the British Museum.16 Here we
see the serpent with seven rays around its head sitting between
a small crescent moon and a six-pointed star (Pls 2a–b); on the
reverse the name Chnoubis with stars above and below (Pl. 2c).
The word Chnoubis or Chnoumis seems to be the Greek
rendition of the name of the Egyptian god Khnum, who in
earlier native iconography was depicted as a ram-headed man.
He appears independently, however, as a decan in Ptolemaic
astrological tables, where among other things he is connected
with an ‘amulet for the stomach’.17 This astrological feature is
also probably reflected in the stars and crescent moons that we
see on the example in Pl. 2.
Three other kinds of inscriptions, also in Greek letters,
appear frequently on these Chnoubis gems: nabis biennouth,
for which both Egyptian and Hebrew interpretations have been
proposed,18 and two other words which seem to be Greek
epithets in the vocative case, for gigantorêkta (‘breaker of
giants’) and perhaps barophita (‘crusher of snakes’),19 as on
the gray-green jasper gem in Pl. 3c. On the back of this stone
we see the name Chnoubis and the -SSS- symbol surrounded by
the epithet ‘breaker of giants’.
Plates 2 a–c Rock crystal with Chnoubis, 21 x 7mm. London, British Museum,
PE 1986,0501.19
Text, Image and Medium
The decipherable Greek inscriptions on these green and
white gems, then, all seem to be names or epithets used to
invoke the radiant serpent-god Chnoubis.20 This idea is
occasionally spelled out in brief prayers inscribed, for example,
on the back of a Chnoubis gem of greenish black jasper in Paris
(‘Keep the stomach of Proclus healthy!’) or a light green gem in
Berkeley (‘Avert all tension, all indigestion and all pain from
the stomach of Julian, whom Nonna bore’).21 A badly chipped
greenish yellow jasper in the British Museum has a different
prayer inscribed around the head of the serpent (‘[Chnou]bis,
digest, dig[est]!’), one in which Chnoubis himself is apparently
imagined to enter the stomach and help digest the food.22
Scholars have, then, made some sense of the combination
of Greek text and image on these stones, but Galen makes it
clear that the medium was crucial to their efficacy. Such power
also seems to be assumed in the final command of Socrates and
Dionysus: ‘let the wearer not set it aside’ – the jasper stones had
to be in contact with the patient’s body to be effective. The
image of the divine solar serpent, on the other hand, seems to
indicate the protective presence of the god, like a statue in a
temple or a house.23 The names and epithets (sometimes in the
vocative) and especially the prayers support this idea, because
they all seem to invoke Chnoubis to stop the stomach ache or
help digest the food. Their absence in the literary sources may
mean that they were perhaps originally spoken aloud to the
image of the serpent and only later inscribed on the gemstone.
The -SSS- symbol is, of course, another story, as it presumably
cannot be spoken aloud; like the image of Chnoubis himself, it
probably derives from a Ptolemaic Egyptian tradition of
representing the decan Khnum, for example at Edfu and
Dendera.24 In a section of De medicamentis on remedies for
pleurisy and pains in the sides, moreover, Marcellus reveals
that this symbol could magically heal by itself (24.7):
If the letters noted below [that is: -SSS-] will have been carved into
a jasper stone of the Phrygian translucent type, and the stone worn
suspended from the neck of the sufferer, it will work marvelously.25
Plates 3a–c Jasper with Chnoubis, 16 x 13 x 5mm. London, British Museum,
PE G 23
‘Gems of Heaven’ | 51
Faraone

Plate 4 Black-figure vase with Heracles and the lion, 520 BC. H. 43.18cm. Plates 5a–d Jasper with Heracles and the lion , 12 x 10 x 3mm. London, British
London, British Museum, GR 1843,1103.67 Museum, PE 1986,0501.80

Here we see that the -SSS- alone heals pains in the lungs and
sides, if inscribed on a whitish translucent stone.26 Very few
gems of this type have actually survived27 and Marcellus is the
only author to report the independent power of the -SSS-,
which appears on the reverse of most Chnoubis amulets, but is
never mentioned by any other writers. It is unwise, finally, to
dismiss the astrological features on some of these amulets:
both the name and image of Chnoubis and the -SSS- symbol
appear in astrological texts in the Hellenistic period and in
these texts the god is connected with the healing of the
stomach. This suggests that astrology probably always played a
role in the Chnoubis amulets. We run into a different problem
below when the discussion turns to scorpion amulets, where
the image of the eight-legged scorpion is much older and its
role in the astrological healing system differs significantly from
the amuletic. There, as we shall see, the protective nature of
the image clearly precedes its later astrological application.
Plates 6a–d Jasper with Heracles and the lion, 15 x 11 x 3mm. London, British
Red jasper and Heracles strangling the lion Museum, PE G 224
My second case is a series of opaque red stones with images of
Heracles and the lion that were designed to heal colic. This is a
more complicated assemblage, because unlike the image of the
Chnoubis serpent, this scene is not a recent Hellenistic
invention, but a very old icon, already popular in Archaic
Greece, as can be seen on a vase in the British Museum (Pl. 4).
This scene of Heracles’ famous labour was, however, thought to
have curative power: in his chapter ‘On the colicky condition’
Alexander of Tralles, a 6th-century ad Greek physician,
prescribes the following treatment for colic, a painful disease
of the lower intestine: ‘On a Median stone engrave Heracles
standing upright and throttling a lion. Set it in a gold ring and
give it to the patient to wear’ (2.579).28 There is some confusion
about the precise identity of the ‘Median stone’ in this passage,
which may have been a form of haematite or magnetite29 but
Alexander’s description coincides well with a popular series of
amulets that consist of an opaque red stone (almost always
jasper) engraved with the wrestling scene that he describes; for
example, three gemstones in the British Museum (Pls 5a, 6a, Plates 7a–d Jasper with Heracles and the lion, 18 x 13 x 3mm. London, British
7a).30 Museum, PE 1986,0501.81

52 | ‘Gems of Heaven’

Plate 8 Jasper with Heracles and the lion, and Plate 9 Archaic bronze shield-band reliefs, with Heracles and the lion, 600–550 BC. Olympia, Olympia
Hecate. Paris, Cabinet des médailles Museum
Alexander tells us that his colic-amulet has only two parts,
a special stone and a special image, but the extant gemstones
almost always include a third feature: three kappas inscribed
on the reverse of the gem (Pls 5–7b and d). Another red jasper
gem in the Bibliothèque nationale, Paris, (Pl. 8) confirms that
these gems were, in fact, used to cure colic several centuries
before Alexander wrote down his recipe: Heracles and the lion
are partially surrounded by the command: ‘Withdraw colic! the
divine one pursues you’.31 On this gem the three kappas (along
with an eight-pointed star) lie beneath the feet of the struggling
pair, and on the reverse is another common magical image: the
tri-form Hecate with the magical names Iao above and Abrasax
below.
The use of Heracles and the lion on the Paris gem is not,
perhaps, difficult to decipher: the image of the dangerous lion
strangled by a powerful hero is designed to frighten the colic
and cause it to flee from the person wearing the ring. The flee-
formula, however, appears on this gem alone and is part of a
much older Greek tradition of protective incantations.32 It is
more likely, in fact, that on most of these gems the sharp-
clawed lion itself somehow represented the sharp abdominal
pain of colic, which by the process of persuasive or sympathetic
magic is strangled and extinguished, just as Heracles strangles
the lion. A similar kind of operation is, in fact, imagined in a
recipe for weed-killer in a late Greek agricultural handbook
(Geoponica 2.42.2):33
If you wish that this plant [= ‘lion-pulse’] in no way appears [i.e. in
your field], take five potsherds and draw on them in chalk or in
another kind of white [a picture of] Heracles strangling the lion.
Place these in the four corners [i.e. of the field] and in the middle.
The power of these amulets is presumably triggered by the
analogy between the annoying weed, ‘lion-pulse’ (so named,
perhaps, because its grassy leaves resembled the mane of a
lion) and the lion. Here, too, although the author of this recipe
carefully specifies the image and the two media to be used
(white chalk on potsherds), he says nothing – just as Alexander
says nothing – about inscribing Greek words or letters on the
potsherds.
In the case of Heracles strangling the lion there are
numerous pre-Roman examples on gems that date as early as
the Archaic period. In the version of the scene on the colic
amulets, Heracles always stands, rather than kneels or
crouches; this is a less common version, but a very early one: it
shows up, for example, on Archaic bronze shield-band reliefs in
the Olympia Museum (Pl. 9), two of which show the hero’s
weapons in the background. Note also the alternate version of
the encounter in the right-hand example, where the hero stabs
the beast, rather than chokes it.34There are, moreover, a
Text, Image and Medium
number of examples of this scene on Archaic scarabs, often
carved in more translucent red gems, like carnelian or
sardonyx. If we use Alexander’s two criteria for a colic amulet,
these gems could presumably serve as amulets, because they
have the correct medium (a red stone) and the right image
(Heracles and the lion). But according to modern scholarship,
they are merely ornaments, because they lack the triple kappas
on the reverse side.
But how does one begin to argue that these Archaic or
Classical examples of the type might have also been used as
amulets? The first thing to point out is that the image of
Heracles and the lion is, in fact, a special one: it is by far the
most frequently depicted of all his labours and it is the only
episode in the cycle that regularly appears on gemstones from
the Archaic to the Roman period. It seems, in short, to be
popular beyond its place in mythic narrative. There is no doubt,
of course, that the image was used widely for ornamental
purposes on pots, shields, armour and elsewhere, but at what
point can we tell that it is being used as an amulet?
One lucky bit of evidence for the pre-Roman amuletic
power of this image was discovered in an early 5th-century bc
grave in Cyprus (Pl. 10): a red carnelian scarab that shows a
young, beardless Heracles standing and attacking a lion with a
sword as he did on one of the shield-band reliefs discussed
earlier. This is one of the earliest versions of the scene and the
method of killing, stabbing rather than strangulation, seems to
have been borrowed originally from Mesopotamia, where one
often finds images of various heroes and even the Assyrian
king dispatching a standing lion in this fashion.35 In the field of
the gem, however, behind the hero we see that something has
been added to the traditional design: two stylised eyes that are,
in fact, the so-called ‘eyes of Horus’ or the udjat-eyes. These
symbols are easily the most common device on Egyptian
Plate 10 Carnelian scarab with Heracles and the lion. 15 x 12mm. London,
British Museum, GR 1894,1101.458
‘Gems of Heaven’ | 53
Faraone

Plates 11a–c Lapis with Aphrodite Anadyomene, 12 x 9 x 2mm. London, British Plates 12a-c Lapis with Aphrodite Anadyomene, 14 x 10 x 3mm.
Museum, PE 1986,0501.141 London, British Museum, PE G 194

protective amulets throughout the Pharaonic period and down Imperial date, but preserving more ancient traditions:43
into Roman times.36 On this Cypriot gem, then, we have all In the euanthes stone, the multi-coloured one, Aphrodite is
three of the features that modern scholars, at least, expect to engraved binding up her hair and the locks on her head and the
root of the plant (i.e. arugula) and the tongue of the nightingale
find on a ‘magical’ amulet: a potent medium (a red gem); a
are put underneath. Set the stone and wear it and you will be loved
powerful image (Heracles and the lion) and a magical symbol by everyone and will be well known and seem sweet-voiced not
or text (the udjat-eyes, instead of KKK) that somehow gives it only to human beings but also to gods and demons (Kyranides
added curative or protective power. Do these eyes of Horus, 1.5.27–31).
then, reveal the innate amuletic power of all of the other If someone bears the right eye [i.e. of a bird called the wryneck]
under stainless lapis lazuli, in which Aphrodite is engraved, the
Archaic red-carnelian scarabs inscribed with Heracles and the
bearer will be charming, he will be famous amongst people and he
lion? Or do they simply reflect a single idiosyncratic impulse will gain every lawsuit. The left eye has the same effect if borne by
that, in just this one example, recasts an ornamental gem into women (Kyranides 1.10.39–42).
the role of an amulet?
As in the other medical or lapidarian descriptions of amulets,
Lapis lazuli and Aphrodite Anadyomene there is no command here to inscribe magical names or
My third category is a series of blue gemstones, usually of lapis symbols on these bluish gems (opal and lapis lazuli), although
lazuli or blue glass,37 that show the famous image of Aphrodite these recipes do add another feature that is invisible to us in
Anadyomene (‘Rising from the Sea’), for instance on two lapis the archaeological record: plant or animal matter placed
lazuli gemstones in London (Pls 11–12).38 On the back of the beneath the gem, presumably between the gemstone and its
stone we find the magical word arôriphrasis which is typical gold or silver setting.44
of these amulets, as we can see from one of the lapis stones in
the British Museum (Pl. 12c). Scholars, myself included,
sometimes repeat the claim that arôriphrasis transliterates an
Egyptian title of the goddess Hathor as ‘The Lady of the Blue
Stone’; the epithet exists, but apparently bears no phonetic
resemblance to the word arôriphrasis on the gemstone.39
We do find, however, on one lapis gem of this type the
name ‘<H>athor’ inscribed before arôriphrasis,40 suggesting
that the word – whatever it means – was, indeed, an epithet of
some sort, and since Hathor and Aphrodite are assimilated in
Graeco-Roman Egypt, there is some logic to the appearance of
the name and Aphrodite on these blue stones. As in the case of
the other types, a rare inscribed prayer tells us much about the
perceived power of this kind of amulet: on a greenish lapis gem
of another type (with Aphrodite and Ares) we read:
‘Arôriphrasis, give your charm to the bearer’.41 It would seem,
then, that like the name Chnoubis, arôriphrasis identifies a
divinity – probably Aphrodite/Hathor – who is thereby invoked
to produce charm and beauty in the person who wears the gem.
We find this same focus on ‘charm’ or ‘grace’ on two other
stones that depict Aphrodite Anadyomene: a circular lapis gem
in the Ashmolean Museum, Oxford, which has on its back,
instead of the expected epithet, the Greek noun for ‘charm’
(charis); and a recently discovered, but unpublished, blue-glass
version with a command along the bevel: ‘Give power
(dynamis) and charm (charis) to the one who wears it!’42
This crossover between power and charm is illustrated well
by the authors of the Kyranides, an encyclopaedic work of Plate 13 Necklace pendant. Washington DC, Dumbarton Oaks Collection

54 | ‘Gems of Heaven’

It is difficult to see any obvious evolution in the case of
these blue amulets. Given the focus on erotic charm, classicists
and classical archaeologists are apt to see the image of
Aphrodite as the most important and original focus for this
series,45 but here, too, I suspect that the magical name, the
possible connection with Hathor and the type of stone all point
to an earlier Egyptian practice, since lapis lazuli was popular in
Egypt for many centuries before the Greeks discovered it.46 It is
also interesting to note that the use of this traditional amulet
seems to have continued in Egypt down to Byzantine times, as
we can see in a gold and lapis lazuli necklace in Dumbarton
Oaks said to come from Egypt (Pl. 13).47 Here Aphrodite is
rendered in gold and strikes her familiar pose against a
background of lapis lazuli. There are no discernible words or
magic symbols on this pendant, but I suspect that I would not
be too far off the mark, if I were to suggest that its owner may
have known about the charm-giving power of this image and
this attractive blue stone.
Yellow jasper and the scorpion
A fourth type of popular amulet in the Roman world was a
yellow jasper gemstone inscribed with an eight-legged scorpion
(Pls 14a–b). This stone has the yet-to-be-deciphered magical
name or word ôrthmenchiniambôn inscribed on the reverse
(Pl. 14c), as do many in this series. Museum catalogues are apt
to file a gem like this under the rubric ‘magical’, but if a
scorpion is inscribed on a yellow jasper with no magical word,
it is more liable to be categorised as an ‘animal-type’.48 In
modern eyes, then, text is once again the crucial feature that
makes a scorpion gem ‘magical’.
The fact that Scorpio is one of the signs of the zodiac and
that these yellow scorpion-gems sometimes depict stars,
crescent moons and even the scales of Libra further
complicates matters – indeed some authorities regularly
categorise scorpion gems as ‘astrological’.49 As in the case of the
Chnoubis amulets discussed earlier, one cannot dismiss these
astrological elements, because over time an elaborate
connection between astrology and healing clearly did evolve,
according to which each zodiac sign governed a special part of
the human body; and since the region assigned to Scorpio was
the genital organs, it is reasonable to suggest that some of these
amulets may indeed have been designed to heal that part of the
body.50 Nonetheless, it seems that here (unlike the case of the
Chnoubis amulets) the astrological use of these gems
constitutes a later and novel adaptation or interpretation
generated by the appearance of the scorpion in both systems.51
This argument is supported by, among other things, the
images of scorpions that often appear on Egyptian amulets of
all periods, where they clearly work by the logic of like-
banning-like.52 Here, too, a simple inscribed prayer reveals this
Text, Image and Medium
protective function: one of the yellow jasper gems mentioned
above has on its reverse a magical word, aiaêilegreô
(presumably the name of a protective entity), followed by the
command ‘Protect!’53 Yellow jasper was also useful by itself as a
painkiller for a scorpion-sting. Here once again the lapidary of
Socrates and Dionysus is crucially helpful (39.1–4):54
Agate stone: The agates have the greatest power. They are from
Hermes. The one that has a colour similar to the pelt of a lion is
powerful, if set upon those who have been stung by a scorpion or if
ground up and applied with water. For immediately it makes the
victim painless. It is also suitable for those bitten by vipers, ground
up and applied to the bite or even drunk with wine.
The innate power of this tawny gem is deployed in various
ways: as a gemstone placed directly on the patient (presumably
on the sting or bite), when ground up with water and applied as
an ointment, or when drunk with wine.55 The authors make no
mention of an image of the scorpion or any magical letters in
these recipes: the power lies in the stone itself.
To date scholars have offered no compelling explanation
for this close connection between yellow jasper and
scorpions.56 As it turns out of the several species of scorpion
found in the eastern Mediterranean, two were especially
feared: the familiar large black scorpion with its arched tail
and a much smaller species called the ‘yellow Palestinian
scorpion’, which (because of its yellow or light brown colour)
blends in more easily with soil and sand. The yellow scorpion
is, in fact, the most dangerous because of its camouflage and
the power of its venom: it can kill a small child in a matter of
hours and completely disable a grown man for a day or two.57
I suggest, then, that the consistent yellow colour of these
scorpion gems is as important as the image: a yellow stone
inscribed with a scorpion is designed to keep yellow scorpions
away by a two-fold strategy of like-banning-like. The yellow
Palestinian scorpion is, moreover, found all around the
Mediterranean basin, but not in northern Europe, so it is
interesting to note that only a single yellow scorpion gem has
been found in northern European soil,58 in sharp contrast to
other gem types used for colic, stomach ache or gynaecological
problems. This insight can, of course, be overturned by future
discoveries, but it suggests that yellow scorpion amulets were
limited to the areas where yellow scorpions dwelt. Thus it
would seem that both the image and the yellow medium work
together to ban the dangerous yellow scorpion. And indeed we
do have some evidence that the combination of the eight-legged
scorpion and a yellow stone continues a very old Egyptian
tradition: a yellowish brown stone, in this case a glazed
steatite, set in a ring discovered on the left hand of the mummy
of a noblewoman named Hanofer, who apparently served in
the court of Queen Hatshepsut.59
Plates 14a–c Jasper with scorpion, 15 x 11 x 2mm. London,
British Museum, PE G 180
‘Gems of Heaven’ | 55
Faraone
Plate 15 Haematite with womb, 18 x 13 x 2mm. London, British Museum,
PE G 540
The bloodstone, the womb and the key
My fifth example is the use of bloodstone (haematite) for a
variety of gynaecological amulets (Pl. 15). This gem is a good
example of the typical reddish brown colour and slightly
metallic texture of the haematite stone, which is in fact a
mineral form of iron oxide. The name of this gem means
‘bloodstone’ in Greek and explains its medical usefulness: one
of its primary functions seems to have been to control the flow
of blood, usually from a woman’s uterus, although these same
gems could also be used to prevent premature birth, to stop the
womb from wandering and even to heal intestinal diseases,
which presumably are also marked by internal bleeding.60
The imagery on the simplest version of the series (see
above, Pl. 15) is entirely Greek. The idea that a woman’s womb
resembled an upside-down jug appears in the earliest Greek
medical writers,61 and is nicely illustrated in a late manuscript
of the gynaecologist Soranus (Pl. 16).62 The small device below
the womb on the gemstone in Pl. 15 – it looks like a modern
toothbrush with a short and contorted handle or crank – is in
fact an ancient Greek door-key that in this scheme controls the
opening and closing of the patient’s womb.63 On the other side
of these gems, we most often find the single word orôriouth
(Pl. 17c).64 There is no consensus on the meaning of orôriouth:
it is not Greek and although a number of Egyptian and Hebrew
phrases and words have been adduced to explain it, none have
been convincing.65 The expansive and unique title on the back
of a gem from Cairo – ‘Lord of the womb of women, orôriouth
aubax’ – suggests a male demon, god or angel, but that is all
one can say.66 On the other hand, the commands found on other
stones point in another direction: three contain a command
composed as a full iambic trimeter: ‘Contract, womb, lest
Typhon sieze you!’67 The other examples are truncated, but
always retain the initial imperative and sometimes its direct
object as well.68 In every case, however, the womb is invoked,
not a god, and since the womb is indeed depicted on the gem,
perhaps orôriouth is a word or epithet for ‘womb’.
Plate 16 Drawing of a womb from a manuscript of Soranus, AD 900
56 | ‘Gems of Heaven’
Plates17a–c Haematite with womb, 16 x 9 x 3mm. London, British Museum,
PE G 371
In a slightly more complex version of this series, the uterus
and key are surrounded by an ouroboros, the well-known
Egyptian device of a snake eating its own tail (Pls 18a–b).69 In
this case the snake either protects the womb from outside harm
or prevents it from moving.70 On many gems of this type we
also find Egyptian gods standing or sitting on top of the womb
as, for example, Chnoubis flanked by Isis and Nephthys (Pls
19a–b).71 Because of the large number of complicated images
like this one, it is commonplace to label all of these womb-
amulets ‘Graeco-Egyptian’ and assert, for example, that
‘Egyptian ideas and practices exercised the strongest influence
upon the makers and users of Greek magical amulets’.72 This is
certainly the case in this, the most complex version of the
uterus gem, but theorising Egyptian influence as the ‘strongest’
makes it difficult to explain the existence of the simplest
designs, which show Greek images of the womb and the key
and the magic word orôriouth, but no Egyptian iconography at
all.
The evidence suggests, in fact, that like the other series
discussed in this essay, these haematite gems underwent some
period of evolution from a simple form to gradually more
complex ones. I suggest that uninscribed haematite was first
used alone in the Greek world as an amulet for gynaecological
bleeding and other complaints. This first stage gains
considerable support from the fact that already in the Classical
period Greek doctors were prescribing the use of ground-up
haematite in drinks and ointments for similar complaints.73 The
second step seems to have been the addition of the womb and
key: both images are Greek and seem to engage, as the image
of Heracles and the lion does, in a simple form of sympathetic
magic designed to lock up the contents of the womb and make
the bleeding stop. We know, moreover, that the womb-and-key
device lies at the heart of the magical operation, because it
appears on almost all of the nearly 200 amulets of this type; the
Egyptian gods, on the other hand, appear on only two-thirds of
them, usually in the same peripheral position on top of the
Plates 18a–c Haematite with womb, 17 x 13 x 2mm. London, British Museum,
PE G 320

Plates 19a–c Haematite with Egyptian gods and womb, 18 x 12 x 3mm.
London, British Museum, PE G 77
womb and having little to do with the key and the magical
action of closure or containment.74 The Egyptian accretions, in
other words, apparently enhance the power of the original
womb-and-key device, but they are not central to its mission.
Conclusions
I have, then, traced the evolution of some of these popular
magical gems, beginning with the idea that a powerful, but
unadorned, stone has some innate curative, painkilling or
other power. We know from Greek medical texts and lapidaries
that haematite and yellow jasper were thought to have such
natural powers when ground up and mixed in with liquids for
drinking or application; Galen’s experiment with green jasper,
moreover, proved to his satisfaction, at least, that this
gemstone when placed near the site of his heartburn was also
effective without any image engraved upon it. Socrates and
Dionysus likewise recommend that we place an unadorned
yellow stone on a scorpion sting to alleviate the pain. The
power to kill pain may, in fact, be an important feature of
jasper gems more generally: the green or white ones were
thought to heal stomach ache, the sky-blue ones pleurisy; the
red ones intestinal pain and the yellow ones snakebite and
scorpion-sting. The Aphrodite amulets are anomalous in this
regard: they do not obviously fit the category of a curative or
protective amulet75 and we have no independent evidence that
lapis lazuli had any magic power of its own, beyond its status
(in the Greek mind) as a rare and exotic stone connected with
Egypt and the Near East.
At the second stage it seems that the sorcerers or
stonecutters added images. In the case of the scorpion or
Heracles and the lion these images were very old, imported
from the East and probably thought to have inherent powers of
their own. The scorpion appears alone on a yellowish gem from
the time of Hatshepsut and the 5th-century bc Cypriot gem
with Heracles, the lion and the two ‘eyes of Horus’ suggests
that a number of the other pre-Roman red stones that carry the
scene may have also been used as an amulet. The three other
images discussed here are, however, relatively novel.
Aphrodite Anadyomene was, for example, the late 4th-century
bc invention of the Greek painter Apelles and Chnoubis in the
form of a radiant lion-headed snake, is first attested in
Hellenistic Egypt. And although we have no direct icono-
graphic evidence, there is literary evidence from the late
Classical period for the metaphor of the womb as an inverted
jug. All this suggests that the Hellenistic period may have been
a time when some of the most popular images were beginning
Text, Image and Medium
to be added to already powerful stones: Greek Aphrodite to
Egyptian lapis lazuli; Egyptian Chnoubis to green jasper and
the Greek womb and key to haematite.
We have seen, too, that the creators of these gems often
distinguished a central and ever present image from the
peripheral and more transient additions. Chnoubis and the
scorpion appear centrally on every gem of their type, whereas
the stars and crescent moons, which appear on a minority of
examples and around the edges of the central figures, seem in
the case of the scorpion amulets to be later accretions designed
to adapt the traditional central image to new astrological uses.
And although uterine amulets sometimes deploy along their
periphery an old Egyptian image like the ouroboros-serpent or
place tiny Egyptian gods on top of the womb, the central and
ever present womb-and-key device deploys only native Greek
images and ideas that are attested no earlier than the Classical
period. A motif’s position in the visual field of the gem, in short,
can tell us something important about the evolution of the type
from a simple core image to which peripheral or encircling
designs are later added.
Text, then, is clearly the last to arrive of the three important
features of magical gems. This is perfectly obvious in the
medical and lapidary sources, which occasionally describe
images (e.g. the Chnoubis serpent or Heracles and the lion) and
even a symbol, like the -SSS-, but only rarely tell us to add
Greek words to the amulet. Scholars speculate that these
recipes, preserved in later encyclopaedic works (like the
Kyranides or the lapidary of Socrates and Dionysus) or in even
later medical writers (like Marcellus of Bordeaux or Alexander
of Tralles), often preserve material of Hellenistic or earlier
date, and Galen proves this is so, when he (unlike the others)
actually cites a Hellenistic source (Nechepso) for his
description of the green jasper Chnoubis amulet. Thus the pre-
Roman recipes seem to be copied or translated into Latin in a
closed bookish manner, apparently without any detailed
examination of contemporary Roman practices, although their
ignorance can perhaps be partially explained by the fact that
on some of these gems (e.g. the Heracles and lion or the
scorpion) the inscriptions usually appear only on the reverse
and thus would be invisible to the observer.
The title of this essay begins with the words ‘Text, image
and medium’ and, in addition to being a gently rising tricolon,
it reflects perfectly my own philological training and prejudice:
I tend to begin with texts, add in a few images if necessary and
only rarely think about media. In the course of my study of the
magical gemstones I am, however, more and more convinced
that we should, perhaps, acknowledge their importance in
precisely the reversed order and that this view reflects their
evolution over time. It is, I suggest, only in the Roman period
when the written text takes on any importance at all. Indeed,
as we have seen, there were probably a number of powerful
stones and powerful images circulating independently around
the Mediterranean basin long before the advent of the Roman
Empire. The ‘big-bang’ approach to magical gems, therefore,
misreads the texts on these gems as evidence for an historic rise
of superstition or magic, when we would do better to see these
texts as additional evidence for the rise in Roman epigraphic
habits and the increasing scribalisation of magic.
‘Gems of Heaven’ | 57
Faraone
Notes
1 They are all usually thought to guard against the evil eye; see O.
Jahn, ‘Über den Aberglauben des bösen Blicks bei den Alten’,
Berichte über die Verhandlungen der königlich sächsischen
Gesellschaft der Wissenschaften zu Leipzig, Philologisch-Historische
Classe 7 (1855), 28–110, at 59–60 (gorgoneia), 63–6 (frontal eyes)
76–80 (phalloi); for a recent survey and bibliography for the
Roman world, see A. Alvar Nuño, El mal de ojo en el occidente
Romano: Materiales de Italia, norte de África, Península Ibérica y
Galia (doctoral diss.), Madrid, 2009, passim, especially for the use
of the phallus first on the Italian peninsula in the 2nd century bc
and then spreading to the provinces. For the gorgoneion in earlier
Greek culture, see M. Halm-Tisserant, ‘Le gorgonéion, emblème
d’Athéna: introduction du motif sur le bouclier et l’égide’, Revue
Archéologique (1986), 245–78; G.H. Clarke, The Shield Devices of the
Greeks in Art and Literature, Cambridge, Mass., 1902, 50–2,
discusses frontal eyes and gorgoneia on Archaic and Classical
Greek shields.
2 Regarding the gemstones, see, C. Bonner, Studies in Magical
Amulets, Chiefly Graeco-Egyptian (University of Michigan Studies,
Humanistic Series 4), Ann Arbor, 1950, 22: ‘The very great number
of magical gemstones ... from the first few centuries of the
Christian era suggests that in some way magic had got a stronger
hold upon people of those times than ever before’. A.A. Barb, ‘The
survival of the magical arts’, in A. Momigliano (ed.), The Conflict
between Paganism and Christianity in the Fourth Century, Oxford,
1963, 100–25, speaking generally about magic provides the classic
formulation, e.g., in a discussion of Roman Imperial laws against
sorcery, ‘when the syncretistic, rotting refuse-heap of the dead and
dying religions of the ancient world grew to a monstrous height’ (at
104) or ‘magic-ridden centuries’ (at 105).
3 For recent reviews of the history of these terms, see, M.G.
Lancellotti, ‘Médicine et religion dans les gemmes magiques’,
Revue de l’Histoire des Religions 218 (2001), 427–56, and R. Gordon,
rev. of S. Michel, Die magischen Gemmen im Britischen Museum, 2
vols, London, 2001, in Journal of Roman Archaeology 15 (2002),
666–70. A. Delatte, ‘Etudes sur la magie grecque IV. Amulettes
inédites des Musées d’Athènes’, Musée Belge 18 (1914), 21–96, at
21–22, seems to be the first to call them ‘magical gems’, which was
then made popular by Bonner (n. 2), 1–2. A. Nagy, ‘Daktylios
pharmakites: magical healing gems and rings in the Graeco-
Roman world’, in C. Burnett and I. Csepregi-Vardabasso (eds),
Ritual Healing in Antiquity and the Middle Ages (The Warburg
Institute Colloquia Series), London, 2010 (in press), agrees that
‘magical gems’ is a convenient archaeological designation for a
phenomenon of the Roman period ‘with three defining
iconographic motifs: magical names (voces magicae), magical
signs (charaktêres) and non-standard iconographical types’. But
he openly acknowledges that magically protective rings and gems
existed as early as the Classical period, but since these do not carry
any of the three criteria listed above, he dubs them ‘talismanic’ to
avoid confusion.
4 S. Michel’s catalogue (Die magischen Gemmen: eine Studie zu
Zauberformeln und magischen Bilderen auf geschnitten Steinen der
Antike und Neuzeit, Geissen, 2004) lists around 2600 examples,
but scholars rightly estimate that the number could be twice as
large; see, M. Smith, ‘Relations between magical papyri and
magical gems’, Papyrologica Bruxellensia 18 (1979), 129–36, at 131,
or R. Gordon, rev. of Michel ibid., in Journal of Roman Archaeology
21 (2008), 713–18, at 713, n. 3. This is a huge number when
compared with the 68 amulets inscribed on metal foil collected in
R. Kotansky, Greek Magical Amulets, vol. 1 (Papyrologica
Coloniensia 22.1), Opladen, 1994, or the 36 papyrus amulets
collected in the first volume of R. Daniel and F. Maltomini,
Supplementum Magicum, 2 vols (Papyrologica Coloniensia 16.1 and
2), Opladen, 1990–1.
5 Bonner (n. 2), 3, for example, begins his study by acknowledging
the inherent power of stones like amethyst and galactite and
cautiously avers that ‘the presumption that classical ringstones
had quasi-amuletic value is quite strong’ (at 6), but then he notes a
‘marked change’ in the 1st century ad when ‘rings and pendants of
semi-precious stones ... show that they are magical, either by
designs of so particular a character as to admit to no other
classification or by the unmistakable evidence of the inscriptions’.
Here his distinction between ‘amuletic’ and ‘magical’ begs the
question.
58 | ‘Gems of Heaven’
6 This process is similar to that adopted in A.M. Nagy, ‘Gemmae
magicae selectae. Sept notes sur l’interprétation des gemmes
magiques’, in A. Mastrocinque (ed.), Atti dell’incontro di studio
‘Gemme gnostiche e cultura ellenistica’, Verona, 22–23 ottobre 1999,
Bologna, 2002, 153–79, at 157–62, for a series of gemstones
depicting a gorgoneion on the obverse, a tri-form Hecate on the
reverse, and some form of the magical word or name
gomandarêarêgoro. I extend his method in two ways: (i) I add
medium as a significant feature of the ‘type’ (of the 12 examples he
discusses eight are of red jasper or a similar red-orange stone); and
(ii) where he sees different schools of magicians and creative
individuals simultaneously making different versions of the same
archetype (he suggests that ‘it is not possible to establish relative
chronologies between the original and its variations’), I will argue
that some of these differences reflect historical change and can be
organised in an evolutionary scheme.
7 Bonner (n. 2), 5.
8 The disparities between the lapidaries and the extant stones has
been noted before. Bonner (n. 2), 41, for example, while discussing
a recipe in a lapidary for the positive effects of carrying a topaz
engraved with an image of Poseidon, says that without an extant
example of this kind of stone with magical words or charaktêres we
cannot know if the stone was used in that fashion; Lancellotti (n.
3), 446–7, responds, rightly in my view, that the lapidary in this
case is ‘un témoignage eloquent des pouvoirs magiques attributés
à cette iconographie même sans l’addition d’inscriptions’.
9 De simpl. 10.19 (XII 207 = C.G. Kühn, ed. and Latin trans., Galeni
opera omnia, 20 vols, Leipzig, 1821–1833). I follow the translation of
Bonner (n. 2), 54.
10 The adjective chlôros usually refers to lighter, paler or yellowish
shades of green.
11 Bonner (n. 2), 54 dates Nechepso’s book to about 150 bc.
12 Marcellus, De medicamentis 20.98 (M. Niedermann (ed.), Marcelli:
De medicamentis liber [Corpus Medicorum Latinorum 5], Berlin,
1916).
13 These amulets were popular; Michel (n. 4), 255–62, lists over 150.
Most of these gems were chance finds and ended up in private
collections, but recently some have appeared in controlled
archaeological excavations, for instance: a green jasper with
serpent, name and-SSS- from an industrial installation with
several large kilns or vats at Tel Dor in Palestine; see E. Stern and
I. Sharon, ‘Tel Dor, 1993: a preliminary report’, Israel Exploration
Journal 45 (1995), 26–36, at 35, fig. 35; G. Bohak, ‘A note on the
Chnoubis gem from Tel Dor’, Israel Exploration Journal 47 (1997),
255–6, corrects their reading of the inscription and notes that the
Chnoubis serpent is explicitly mentioned several times in Rabbinic
literature.
14 In the collection of the British Museum, for example, we find two
green jaspers (Michel [n. 3], nos 305 and 317), three greenish
chalcedonies (nos 306 and 308–9) and two olivines (nos 307 and
310; see the colour photos in Michel [n. 4], ‘Farbtafeln IV’) and a
grayish green chalcedony (no. 329; also in Michel [n. 4]). The
nomenclature for gems is fuzzy both among the ancient Greeks
and modern scholars. M. Sax , ‘Recognition and nomenclature of
quartz materials with specific reference to engraved gemstones’,
Jewellery Studies 7 (1996), 63–72, for instance, divides (at 63, fig. 1)
microcrystalline quartz (a category into which most magical gems
fall) into two categories: (i) those that are ‘commonly translucent
with a fibrous microstructure’ and commonly called
‘chalcedonies’, with subgroups such as ‘chalcedony’, ‘carnelian’,
‘agate’ and (ii) those that are ‘commonly opaque with a granular
microstructure’ and commonly called ‘cherts’ or ‘jaspers’, the latter
having subgroups such as ‘jasper’, ‘plasma’ and ‘banded jasper’.
One can see immediately the problem that ‘chalcedony’ and
‘jasper’ appear as both categories and subcategories. Since it is
nearly impossible with the naked eye to tell a granular
microstructure from a fibrous one, the distinction can only be
made roughly in terms of opaque vs. translucent. The green
digestive gems under discussion are always called ‘jaspers’, but
their translucence should give one pause.
15 Socrates and Dionysius no. 35 = R. Halleux and J. Schamp (eds and
trans.), Les lapidaires grecs, Paris, 2003, 170 = Ch. Mély-Ruelle,
Lapidaires grecs, Paris, 1898, vol. 2, 177. Since the authors make no
mention of bands of colour, we should probably understand that a
light-coloured chalcedony or jasper is used here.
16 It is a rock crystal as is Michel (n. 3), no. 326; there are also three

white agates in the collection (nos 323–4 and 327). A chipped gem
of this type (described as ‘crystal glass’) was discovered in Corinth
(it has the serpent and the name, but the reverse is inscribed with
the three Greek letters znx and then a symbol that looks like a
sideways turned psi); see, G. Davidson, Corinth XII: The Minor
Objects, Princeton, 1952, 225, no. 1777.
17 Chnoubis appears as a decan of Leo and Cancer in astrology, and
thus also cures by an elaborate coordination of zodiac sign and
body part; see Bonner (n. 2), 25 and 54–5, and especially
Lancellotti (n. 3), 449–51, and Michel (n. 4), 165–70.
18 W.M. Brashear, ‘The Greek magical papyri: an introduction and
survey; annotated bibliography’, Aufstieg und Niedergang der
römischen Welt II, 18.5 (1995), 3593, cites two possible Hebrew
interpretations (‘bound by chains’ or ‘bound by incantations’) and
one Egyptian: ‘le tour à potier + âme de dieu’.
19 Bonner (n. 2), 168–9, who also discusses some variants for the
former, e.g. gigantopantorêkta (‘breaker of all giants’); or
gigantopniktorêkta (‘throttler-breaker of giants’). Michel (n. 4),
258–9, lists one example of gigantopantoplêkta (‘striker of all
giants’).
20 See Nagy (n. 6), 164–5, for some PGM recipes for gems that are
likewise engraved with the gods’ images and their names (PGM =
K. Preisendanz [and A. Henrichs], Papyri Graecae Magicae: Die
Griechischen Zauberpapyri2, 2 vols, Stuttgart, 1973–4).
21 A. Delatte and P. Derchain, Les intailles magiques gréco-
égyptiennes de la Bibliothèque Nationale, Paris, 1964, no. 80 (a
greenish black jasper in Paris) and C. Bonner, ‘A miscellany of
engraved stones’, Hesperia 23 (1954), 138–57, pls 34–6, at no. 36 (a
light green stone – possibly jadeite – which now resides in the
collection of the Department of Near Eastern Languages at the
University of California at Berkeley. The second prayer uses a
plural imperative, suggesting that the author imagined Chnoubis
and at least one other addressee. Michel (n. 4), 259–60, describes a
haematite gem in the Skoluda collection, on the back of which
appears the Chnoubis serpent sitting on the womb-and-key design
(see below) beneath which lies the -SSS- symbol. Round the edge is
the prayer: ‘Chnoubis, stop the pain of the stomach, Abrasax!’
22 Michel (n. 3), no. 338. See the paper by Mastrocinque, this volume,
pl. 12. For other examples, see Bonner (n. 2), no. 83, a bluish
chalcedony that has the serpent, -SSS- and the name chnoubis on
the front and on the reverse ‘Digest, digest!’; Bonner (n. 2), 59, also
mentions a ‘brownish chalcedony’ Chnoubis amulet in the Museo
Borgiano that has ‘Digest!’ inscribed three times. He also observes,
ibid. ad no. 83, that ‘many Chnoubis stones are strongly convex’;
I would add that many of them are quite small and do not seem to
have been designed for a ring or pendant setting: if Chnoubis is
called on to ‘digest’ and if he is called the ‘crusher of snakes’, might
it be the case that some of these smaller or seed shaped convex
gems were actually swallowed so they could enter the stomach and
allow the god to do his work directly? Recall how Galen insisted
that the string of green jasper gems be placed over the oesophagus
to cure pain in the oesophagus. Direct contact with the painful site,
in short, seems to have been important for the efficacy of these
amulets.
23 For the idea of a gemstone as a miniature statue, see Theophrastus,
Charact. 21.10, who describes a man who cleaned his Asclepius ring
daily and then oiled it and crowned it with a wreath: see Nagy (n.
3) for a brief discussion.
24 E.g. Bonner (n. 2), 25, and Michel (n. 4), 168, n. 859.
25 Marcellus, De med. 24.7 (n. 12). For the preservation of the -SSS- in
the manuscripts, see R. Heim, Incantamenta Magica Graeca-Latina
(Jahrbücher für classische Philologie suppl. 10), Leipzig, 1893, 480,
n. 3.
26 The Latin word aerizusa renders the Greek aerizousa, a participle
of the verb aerizô, ‘to resemble the air, to be pure as the air’. See C.
Lewis and C. Short (eds), A Latin Dictionary, Oxford, 1879, s.v., who
define the verb as ‘to be light as air’ and ‘to be light blue’ (perhaps
‘grey, cloudy’). Bonner (n. 2), 60, suggests that aerizusa means ‘of
gray or bluish colour’, but when Socrates and Dionysus describe a
variation of the Chnoubis amulet (discussed above) as an ‘onyx
stone ... white and transparent just like air’, I suspect they are
describing the same type of crystalline or translucent white stone.
27 For example: M. Buora and F. Prenc, Gemme Romane da Aquileia,
Udine, 1996, no. 206 (chalcedony; -SSS- between an epsilon and a
sigma and the magical word abramaôth); and S. Amorai-Stark,
Engraved Gems and Seals from Two Collections in Jerusalem,
Text, Image and Medium
Jerusalem, 1993, no. 137, in a Jerusalem collection (‘white and light
orange opal, opaque’ with an ouroboros serpent with -SSS-
(missing the last S apparently) on top of its head on the obverse and
on the reverse a badly corrupted inscription with ‘breaker of
giants, crusher of snakes’).
28 T. Puschmann, Alexander von Tralles, 2 vols, Vienna, 1879.
29 Bonner (n. 2), 62–3.
30 Michel (n. 4), 178–9, discusses the type: on 280–1, she lists 37
examples, all red jasper, except for a single green jasper, three
carnelians and one obsidian. She lists only one example without
the KKK (see Pls 6–7), which instead has on its reverse Zeus with
an eagle, for which see eadem, ‘Medizinisch-magische Amulett-
gemmen’, Antike Welt 26 (1995), 379–87, at 383, pls 10a and b.
31 Delatte and Derchain (n. 21), no. 280; the drawing is after the
unnumbered figure in C. Lenormant, ‘Lettre à M. A. de
Longpérier’, Revue Archéologique (1846), 510.
32 C. Faraone, ‘Hipponax frag. 128W: epic parody or expulsive
incantation?’, Classical Antiquity 23 (2004), 209–45.
33 The Geoponica is a compilation of Byzantine date that preserves
selections from a number of earlier Greek agricultural handbooks:
H. Beckh (ed.), Geoponica: sive Cassiani Bassi Scholastici De Re
Rustica Eclogae (Bibliotheca scriptorum Graecorum et
Romanorum Teubneriana), Leipzig, 1895.
34 Drawings after E. Kunze, Archaische Shildbänder, Berlin, 1950, IIIc,
pl. 14 (600–575 bc); IVg, pl. 19 (575–550 bc) and XLIIg, pl. 66 (575–
550 bc).
35 J. Carter, ‘The beginning of narrative art in the Greek geometric
period’, Annual of the British School at Athens 67 (1972), 25–58;
T.H. Carpenter, Art and Myth in Ancient Greece, London, 1991, 120.
36 See, W.M. Flinders Petrie, Amulets, Illustrated by the Egyptian
Collection in University College London, London, 1914, 32–4, nos
138–43.
37 As the list in Michel (n. 4), 250–1, shows, there is a clear correlation
between the blue stone, the image and the word arôriphrasis (for
which see below): of the 26 gems that have the image of Aphrodite
Anadyomene alone and the word, only seven of the stones are not
blue: three red stones, two green, one haematite and one
chalcedony whose colour is not given. On the other hand on the 19
gems on which the Anadyomene figure appears with others (e.g.
standing on a lion or in combination with a dog-headed figure)
there is only one blue gem and this is the only one that also has the
word arôriphrasis (= Michel [n. 3], no. 85). Likewise in Michel’s list
ibid. of the 13 gems showing Ares and Aphrodite, there are only
two blue stones (one lapis and another glass), neither of which
have the word arôriphrasis.
38 All of these gems imitate the pose of Aphrodite Anadyomene in a
famous late-4th century bc painting of Apelles in the sanctuary of
Asclepius on Cos that depicted the moment when the goddess first
stepped out of the sea fully grown; see Pliny, Nat. Hist. XXXV.91
and Lexicon Iconographicum Mythologiae Classicae, ‘Aphrodite’,
nos 423–54. Scholars of these gems disagree about the pose of the
goddess: Bonner (n. 2), 196, thinks she is drying out her hair, while
Delatte (n. 3), 44–5, suggests she is tying up her hair, which seems
to be how the gem-cutters understood it: as M. Waegeman,
‘ΑΡΩΡΙΦΡΑΣΙΣ: Aphrodite’s secret name’, L’Antiquité Classique 61
(1992), 237–42, at 237–8, points out, the Kyranides (in a passage
quoted and discussed below) refers to this type as ‘Aphrodite ...
binding up the hair and the locks of her head’.
39 E.g. A.A. Barb, rev. of Delatte and Derchain (n. 21), Gnomon 41
(1969), 298–307, at 306, P. Zazoff, Antike Gemmen in deutschen
Sammlungen, III: Braunschweig, Göttingen, Kassel, Wiesbaden,
1970, 244, ad no. 185, U. Mandel-Elzinga, ‘Eine Gemmensammlung
aus Alexandria im Akademischen Kunstmuseum der Universität
Bonn’, Bonner Jahrbücher 185 (1985), 243–98, at 290; M.
Waegeman, Amulet and Alphabet: Magical Amulets in the First Book
of the Kyranides, Amsterdam, 1987, or Michel (n. 4), 211, who all
ultimately depend on the interpretation of W. Helck, ‘Zu den
ägyptischen Sinai-Inschriften’, Orientalistische Literaturzeitung
53 (1958), 423, of the Sinai Inscription. My Egyptologist colleague
R. Ritner has (pers. comm.), however, recently studied the
inscription and Helck’s article and although he agrees that ‘the
titles he mentions for Hathor are correct: ‘Lady of Turquoise, Lapis
or Amethyst’, none of these seem to match up phonetically with
arôriphrasis. Brashear (n. 18), 3580, cites another possible
Egyptian interpretation (‘fac me gaudere, redde me hilarem’) and
Waegeman (n. 38), 239, notes that the name is often divided on the
‘Gems of Heaven’ | 59
Faraone
stones into two parts (arôri and phrasi) and then suggests (at
242) that the second half preserves a purposely scrambled version
of the Greek transliteration of Hebrew saphir (‘lapis lazuli’; the
Greek word is sappheiros).
40 E. Drioton, ‘Notes diverses’, Annales du Service des Antiquités de
l’Egypte 45 (1947), 82–3, no. 12, who remarks: ‘Il est remarkable que,
pour illustrer la vielle formule qui débute par le nom Hathor, le
gravieur n’ait pas reproduit une déesse égyptienne’.
41 Delatte and Derchain (n. 21), no. 333. See, G. Bevilacqua, ‘Ares e
Afrodite sulle gemme magiche’, in Mastrocinque (n. 6), 13–25.
42 Ashmolean: M. Henig and A. MacGregor, Catalogue of the Engraved
Gems and Finger-Rings in the Ashmolean Museum. II: Roman (BAR
International Series 1332), London, 2004, no. 13.20.
43 On the date, see Lancellotti (n. 3), 437. I use the translation of
Waegeman (n. 38), 41 and 79, with one change in each passage: in
the first section the Kyranides describes the euanthos stone as ‘all
golden’ (pagchrusos), but Waegeman (at 43–4) argues persuasively
that this must be a mistake for the very similarly spelled adjective
‘all colourful (pagchroos)’, a kind of opal with bluish tinge. In the
second recipe the gem is called ‘sapphire’ (sappheiros), but, as
Waegeman (n. 38), 1, and C. Oldershaw, Gems of the World,
Richmond Hill, 2008, 216, explain, this was a common way of
referring to lapis lazuli in the ancient world.
44 We find a similar combination of Aphrodite’s magical name and
vegetal matter in the following recipe (PGM IV 2231–40): ‘For (i.e.
erotic) fetching spells: burn roses and sumac, then write a
different spell on myrtle leaves and put them under the tablet. The
spell: ‘Steneriô arrôriphrasis yyyy i i i i, fetch her, so-and-so, for
him, so-and-so! Wear it on a woollen cord’.
45 E.g. Delatte and Derchain (n. 21), 183–7, and Michel (n. 4), 250–1.
46 The name lapis lazuli combines Latin lapis (‘stone’) with the
Persian word lazhuward, which means ‘sky’ or ‘heaven’ and was
added to the Latin language as lazulum. It is a form of
metamorphosed limestone, that has been mined in Afghanistan
for more than 6000 years (it is mentioned in the epic of
Gilgamesh). It was very popular in Mesopotamia and Persia and
used extensively in religious ceremonies by the Egyptians: see
Oldershaw (n. 43), 217–18. It is also interesting to note M.J.
Treister, ‘Bronze matrices in the Georg Ortiz collection’, in
A. Calinescu (ed.), Ancient jewelry and archaeology, Bloomington,
1996, 178: ‘Aphrodite Anadyomene was so popular in Hellenistic
and Roman Egypt, where it was reproduced in bronze, marble,
bone, terracotta and faience, that some scholars believe the type
originated in Alexandria’. Cf. LIMC (n. 38), s.v. ‘Aphrodite (in
peripheria orientali)’, nos 40–66.
47 M.C. Ross, Catalogue of the Byzantine and Early Mediaeval
Antiquities in the Dumbarton Oaks Collection, Volume 2: Jewelry,
Enamels and Art of the Migration Period (With an Addendum by S.A.
Boyd and S.R. Zwirn), Washington DC, 2005 (2nd edn), 18–19, no.
12 and LIMC (n. 38), ‘Aphrodite’, no. 89.
48 See e.g. G. Sena Chiesa, Gemme di Luni, Rome, 1978, who places
under the general category of animals gems no. 149 (eight legs;
sardonyx; no inscriptions; 1st century ad), 150 (yellow jasper with
no inscriptions), and 151 (red carnelian with no inscriptions), but
under magical gem no. 173 (red jasper with different and unique
inscriptions), or Amorai-Stark (n. 27) who under animals
catalogues no. 63 (a sardonyx with no inscription).
49 Delatte and Derchain (n. 21), for example, place under ‘animaux
astrologiques’ nos 388 (orange translucent carnelian; inscription
tthdpgli on obverse; back blank); 389 (yellow jasper; inscription
ôethmencheilianbô on reverse); and 390 (hexagonal yellow
jasper; inscription ôrthmenchiniambôn on reverse). See, more
recently, R. Casal García, Colección Glíptica del Museo Arqueológico
Nacional, 2 vols, Bilbao, 1990, ad. no. 453 (rose carnelian inscribed
with a scorpio olivaceus) ‘probablemente … zodiacal’, or Henig and
MacGregor (n. 42), ad. no. 13.25 (light brown jasper with scorpion
and the usual magical name, but no star): ‘Perhaps Scorpio as a
cure for diseases of the genitals’.
50 Michel (n. 4), 160–5.
51 Bonner (n. 2), 77–8, following S. Eitrem, ‘Der Skorpion in
Mythologie und Religionsgeschichte’, Symbolae Osloenses 7 (1928),
53–82, allows that these stones were designed ‘to protect against
the painful sting of the scorpion’, but adds that they may have
‘come under the influence of systematic astrology’ according to
which each zodiac sign governed a special part of the human body;
since the region assigned to Scorpio was the genital organs, it is
60 | ‘Gems of Heaven’
possible that these amulets ‘were valued not only as a protection
against scorpions, but also as a remedy for sexual disorders and
disabilities’. See also Barb (n. 39), 305; M. Henig, ‘An intaglio and
sealing from Blackfriars, London’, Antiquaries Journal 60 (1980),
331–2, at 331, with pl. LXa; J. Śliwa, Egyptian Scarabs and Magical
Gems from the Collection of Constantine Schmidt-Ciążyński ,
Warsaw, 1989, ad. no. 128. The astrological image is further
transformed when Tiberius, born under this constellation, makes
it a personal symbol of sorts that appears on the shields of his
Praetorian guards; see A. Hamburger, ‘Gems from Caesarea
Maritima’, ‘Atiqot 8 (1968), 1–38, pls I–VIII, at 23.
52 For the use of scorpion images to ban scorpions in Pharaonic Egypt,
see G. Pinch, Magic in Ancient Egypt, London, 1994, 97 and 142–6,
with pl. 76, and C. Andrews, Amulets of Ancient Egypt, London, 36,
who comments on Fifth Dynasty examples: ‘at so early a date the
probability is that they were worn as apotropaic protection against
this noxious reptile’s sting’. For images of scorpions on papyrus
amulets from Graeco-Roman Egypt, see M.N. Tod, ‘The scorpion in
Graeco-Roman Egypt’, Journal of Egyptian Archaeology 25 (1939),
55–61, and (for Aramaic ones from the Cairo Geniza) see G. Bohak,
‘Some “mass produced” scorpion-amulets from the Cairo Genizah’,
in Z. Rodgers, M. Daly-Denton and A. F. McKinley (eds),
A Wandering Galilean: Essays in Honor of Sean Freyne, Leiden,
forthcoming. Sometimes amulets have a more complicated
‘combat scene’, like that of Heracles and the lion: they depict the
scorpion being trampled and/or eaten by a large bird. For textual
evidence see Kyranides 1.7.17–21 (a flamingo on top; this is ‘a good
phylacterion against all venomous animals’) and 1.24.100–7 (a
swallow on top ‘turns away every venomous animal, reptile and
quadruped’) and Geoponica 1.7.8, which claims that a flamingo
standing on a scorpion inscribed on copper pyrites bans venomous
creatures. For extant examples of similar gems, see, e.g., Sliwa (n.
51), no. 72 (yellow jasper with falcon-headed person treading on
large scorpion) and no. 142 (red jasper with rooster standing on
scorpion). This motif of trampling scorpions also derives from
Egypt; see R.K. Ritner, The Mechanics of Egyptian Magic, Chicago,
1993, 127–8, with fig. 8, and Pinch (ibid.), 143–6.
53 M. Gramatopol, Les pierres gravées du Cabinet numismatique de
l’Academie Roumaine (Collection Latomus 138), Brussels, 1974, no.
392.
54 Barb (n. 39), 305, n. 4, was the first to adduce this important text.
The text uses the word ‘agate’ to describe this stone, which by
modern parlance would mean a chalcedony or jasper that had
curved bands of colour. I know of one near parallel: Śliwa (n. 51),
no. 128, a yellow onyx in three layers with a scorpion carved on one
side (in the onyx the layers are straight, in the agate curved). But as
I have stressed throughout both ancient and modern terminology
for these kinds of stones is inexact and the word ‘agate’ in this text
could refer to any gold or tawny jasper.
55 The recipe (discussed above in n. 43) for an amulet depicting a
swallow standing on a scorpion in Kyranides 1.24.100–7 adds that
in addition to keeping venomous animals at bay ‘If someone is
struck by a scorpion and you seal the wound with this seal-ring,
you will put the injured person out of danger’.
56 Yellow is by far the most frequent colour of these amulets, but as in
the case of the other amulets discussed earlier, there is some
variation in the colour and opacity; see, for example, Delatte and
Derchain (n. 21), no. 388 (orange translucent cornaline); Buora and
Prenc (n. 27), no. 148 (orange carnelian from Aquileia); M.-L.
Vollenweider, Deliciae Leonis: Antike geschnittene Steine und Ringe
aus einer Privatsammlung, Mainz, 1994, no. 501 (orange-red
carnelian); Sena Chiesa (n. 48), nos 151 (red carnelian) and 173 (red
jasper); Casal Garciá (n. 49), nos 453 and 454 (both red carnelian).
57 J. Dalrymple, ‘Snakes and scorpions in late-antique Egypt:
remarks on papyri documenting envenomation’, in J. Frösén,
T. Purola and E. Salmenkivi (eds), Proceedings of the 24th
International Congress of Papyrology, Helsinki, 2004, 205–13.
58 A yellow jasper gem inscribed with a scorpion and nothing else
was found in the mud of the Thames river near Blackfriars; see
Henig (n. 51), 331, who adds: ‘the relative frequency with which
scorpions are encountered amongst gems from the Mediterranean
area where the creatures are found and their absence until now on
intaglios from Britain may be significant. The presence of a
foreigner from southern lands would not be surprising in the
vicinity of the quays of Londinium’. For other – negative – evidence,
I should mention that among the 658 gems found in or near the

region of Xanten – published in G. Platz-Horster, Die antike
Gemmen aus Xanten, Cologne, 1987, and Die antike Gemmen aus
Xanten, II, Cologne, 1994, – there are no examples of yellow
scorpions.
59 C.H. Roehrig, ‘Life along the Nile: three Egyptians of ancient
Thebes’, Metropolitan Museum of Art Bulletin (Summer), (2002),
38, fig. 51.
60 A.A. Barb, ‘Bois du sang, Tantale’, Syria 29 (1952), 271–84, at 279–
80, and idem, (n. 39); A.E. Hanson, ‘Uterine amulets and Greek
uterine medicine’, Medicina nei secoli 7 (1995), 281–99, at 290–1; C.
A. Faraone, ‘Magical and medical approaches to the wandering
womb in the ancient Greek world’, Classical Antiquity 30 (2011),
1–32.
61 Hanson (n. 60), 286–7.
62 The drawing is after O. Temkin, Soranus’ Gynecology, Baltimore,
1955, 9, fig. 1, which is in turn based on the Muscio text of about ad
900 (Brussels MS 3714).
63 Delatte (n. 3); Bonner (n. 2), 87; R.K. Ritner, ‘A uterine amulet in
the Oriental Institute collection’, Journal of Near Eastern Studies 43
(1984), 214–19.
64 Of the 191 examples tallied by Michel (n. 4), 334–9, nearly all are
haematite and 151 have orôriouth on their reverse, alone or at the
start of a series of names.
65 Brashear (n. 18) records two possible derivations: (i) from an
Egyptian word for ‘uterus’; and (ii) from the Hebrew phrase ‘Light
of lights’. See Hanson (n. 60), 292–3, for a full review.
66 M.L. Barry, ‘Notice sur quelques pierres gnostiques’, Annales du
Service des Antiquités de l’Egypte 7 (1906), 242.
67 Barry (n. 66), no. 3; Bonner (n. 2), no. 140; Michel (n. 3), no. 379. C.
A. Faraone, ‘Stopping evil, pain, anger and blood: the ancient
Greek tradition of protective iambic incantations’, Greek, Roman
and Byzantine Studies 49 (2009), 73–102, discusses the iambic
Text, Image and Medium
incantation.
68 Delatte (n. 3), nos 33 (‘contract, womb!’) and 34 (‘contract!’);
Michel (n. 3), no. 351 (‘stop moving!’); H. Philipp, Mira et Magica:
Gemmen im Ägyptischen Museum der Staatlichen Museen ·
Preußicher Kulturbesitz. Berlin-Charlottenburg, Mainz, 1986, no.
184 (‘Stop moving, womb!’).
69 Ritner (n. 63), 219–20, and M.G. Lancellotti, ‘Il serpente ouroboros
nell gemme magiche’, in Mastrocinque (n. 6), 71–85, show that the
protective ouroboros serpent is a very old Egyptian device. The
ouroboros also appears on the gem depicted in Pl. 17, but it is worn
and difficult to make out.
70 Ritner ibid.
71 Other gods: e.g. Michel (n. 3), nos 358 (Chnoubis alone), 359–66
(Chnoubis flanked by other Egyptian gods) and 367–80 (various
Egyptian gods).
72 Bonner (n. 2), 7–9 and 22–6; the quotation is on 22.
73 First mentioned explicitly by Theophrastus, but there may be some
references to it in the Hippocratic corpus; see Hanson (n. 60), 290–
2.
74 See the catalogue in Michel (n. 4), 334–41. There are, to be sure,
some examples of Thoth or other Egyptian gods with their hands
on the key and clearly involved in the magical process, but this is
very rare. These and other more rarely appearing Egyptian gods,
like Seth or Chnoubis, often point to different and novel
adaptations of the series; see Ritner (n. 63).
75 Lancellotti (n. 3), 433 and passim. C.A. Faraone, Ancient Greek Love
Magic, Harvard, 1999, 96–131, discusses how these gemstones,
along with knotted cords and other amuletic types of charm- and
love-magic, comprise the sphere philia-magic that is distinct from
eros-magic. The former mimics curative forms (e.g. amulets)
because it aims to heal a pre-existing relationship.
‘Gems of Heaven’ | 61
The Colours of Magical Gems
Attilio Mastrocinque
Introduction
Lapidary books from Assyro-Babylonian to Roman Imperial
times underline the importance of stones and their properties
in the production of amulets. Greek lapidary books and
magical papyri describe several kinds of magical gems and
prescribe the sort of stone to be used. A comparison between
these texts and the surviving gems shows that the major trends
of gem cutting follow the rules of written recipes,1 although
one has to admit that there are many exceptions. Indeed for the
Imperial period we know of only a very limited selection of
recipes although one cannot be certain that all recipes followed
the same rules as the few preserved ones.
A detailed study of each type of semi-precious stone from
the Assyrian to the Imperial Roman period has not yet been
done. We have only the important essay on haematite by
Alphonse Barb2 and a few other contributions on various
questions concerning these stones.3 Research teams, in which
Assyriologists, Egyptologists, and classicists co-operated in a
complete description of every sort of semi-precious stone in its
cultural environment and from an inter-cultural perspective,
would be a welcome development. Here I would like to offer a
contribution on the basis of my recent studies of a large amount
of magical gems in Italy4 and in the Cabinet des médailles,
Paris.5 My experience forces me to recognise some kind of
relationship between parts or organs of the human body and
the colours or forms of some magical gems.6 The relationship
between herbs or coloured clothes and the colour of certain
diseases has been studied within the context of medical magic
in the Middle Ages.7 The same can also be true of gemstones.
Stones which imitated human behaviour were called
ἀνθρωπόμιμοι, ‘they which imitate man’.8 Near Eastern
traditions conceived of colours mostly in association with
objects, and indeed people said ‘colour of the goat’, ‘colour of
the sky’ and so on. In the Assyrian and Hebrew languages only
the specific names for red-brown and green-yellow existed. (By
‘specific’ I mean ‘red’ or ‘yellow’, whereas ‘creamy’ for example
is a non-specific term.) Stones were usually described
Plates 1a-b Haematite with womb on obverse. Rome, private collection
62 | ‘Gems of Heaven’
according to the colour of analogous objects/substances such
as wine or the skin of specific animals;9 they could also be
described as light or dark, matt or glossy.10 The same is true for
the extant ancient lapidary books. For instance, magicians of
the Imperial Roman age called glossy stones ‘male’ and matt
ones ‘female’.11 From the 7th century bc a neo-Assyrian belief is
recorded, according to which there were three heavens: the
most distant was made of luludanîtu, a reddish conglomerate
with black and white specks, the middle heaven of saggilmud, a
dark blue stone, and the lower one of translucent ‘jasper’ with a
hint of blue, pink, green or purple.12
Blood
In the case of amulets the most important colours of stones
were those of human organs and secretions. A well-known case
is that of haematite, whose name indicates that it was the stone
of blood. Barb stressed that haematite was known also as lapis
adamas, that is adamu, an ancient Semitic word which signifies
‘dark red’, ‘red blood’, kindred with damu, ‘blood’. In fact, lapis
adamas13 was also the name of various other red stones, such as
jasper, heliotrope, carnelian,14 but, above all, haematite, which
is red when pulverized. Haematite is the mineral form of iron
oxide (Fe2O3), which was used as a remedy in the treatment of
haemorrhages;15 a large number of those amulets concerned
with uterine health were also made of this material (Pls 1a–b).
Christians used haematite for gems depicting Jesus and the
haimorrhoissa.16 The similarity between blood and haematite
could be connected not only with the powdered stone, which is
red, but also the natural globular form of some specimens,
which were similar to coagulated blood (Pl. 2).17
A number of myths tell the story of divine blood which fell
from heaven and coagulated on contact with the earth or sea.
Hesiod’s Theogony, for example, reports that when Cronos
emasculated Uranos the blood which fell to earth produced the
Gigantes, Erinyes and Meliae.18 Gnostic myths concerning
Adamas and Sophia also told the story of divine blood falling
from heaven. Pliny the Elder says that the Magi used a
Plate 2 Specimen of natural haematite

Plates 3a-b Coral with Medusa head. Verona, Civic Museum
substance called basilisci sanguen, which was also called
Saturni sanguen.19 It was useful against diseases and
efficacious when praying to the gods. A magical papyrus
asserts that haematite is the ‘snake’s stone’,20 a definition which
is very close to the ‘basiliscus stone’ of Pliny.21
Coral was also conceived to be divine blood which had
condensed in the sea. In the 4th book of Ovid’s Metamorphoses
one myth is told, according to which the blood which flowed
from Medusa’s head after her beheading by Perseus, flowed
into the sea, hardened and became coral.22 Also Pliny describes
a type of coral called Gorgonia, after the monster’s name.23 In
fact a series of coral gems with Medusa’s face have been
preserved (Pl. 3a).24 Red jasper could be used as a substitute for
coral, as is proved by gems with the same subject.25 These gems
often have an image of Hekate on the reverse;26 both Gorgon
and Hekate were considered to be powerful characters in
averting demons,27 illnesses and enemies.
The liver
Another name for haematite was hepatitis,28 or ‘liver stone’.
This type of stone was the property of Ares-Mars, a very
dangerous planetary god. According to the doctrines of eastern
sages, each planet was associated with one particular metal; in
the case of Mars (i.e. Ares) it was iron,29 the metal of arms. In
the British Museum are two haematite gems on which the
figure of Ares is engraved as well as the inscription: Ἄρης
ἔτεμεν τοῦ ἥπατος τὸν πόνο(ν): ‘Ares stopped the liver’s pains’
(Pl. 4).30 The choice of haematite was conditioned by the iron
which is in this stone. Its colour was also important, for
haematite, when polished, has the sheen of iron and the gloss
Plates 6a-b Haematite in the form of a kidney. Paris, Cabinet des médailles
The Colours of Magical Gems
Plate 4 Haematite with Ares and the Plate 5 Haematite gem with Ares
inscription: ‘Ares stopped the liver’s and the inscription: ‘Tantalos, drink
pains’. 32 x 25 x 5mm. London, British blood’. 38 x 25 x 4mm. London,
Museum, PE G 112 British Museum, GR 1928,0520.1
of a liver. However, a brown jasper of similar colour could
suffice for the same purpose.31 A small number of haematite
gems depicting Ares (Pl. 5) bear the well-known inscription
inviting Tantalos to drink blood.32 In this case both the stone
and the god were appealed to in order to stop haemorrhages.33
This Ares is not the Greek god, but the result of cultural
synthesis in which a Syrian god assumed the iconography of
the Homeric god. This iconography was used to represent
Syrian gods like Azizos and Arsou, or the warlike god of the
Nabataeans.34
The kidney
There is a variant of haematite known as limonite, whose
colour tends to brown or yellow. It is produced when haematite
is metamorphosed by absorbing water (Fe2O3+ H2O).
Sometimes small blocs of haematite or limonite (oolithes) are
kidney-shaped.35 They also have the colour of this organ, which
is reddish-brown. Pliny the Elder36 knew a stone called ‘Adad’s
kidney’ and it is highly probable that it was either haematite or
limonite. In the Cabinet des médailles in Paris there is one
haematite intaglio whose form is very similar to that of a
kidney (Pls 6a–b, 7).37 On the obverse are engraved two gods,
namely a smiting god and Apollo of Hierapolis-Bambyce
(Syria),38 and Heracles. The reverse depicts Selene and Helios,
with an engraved line between them which divides the stone
into two zones in a similar manner to divinatory terracotta or
bronze Etruscan livers. The Orphei Lithika kerygmata39
mention the mantic properties of siderite, an iron-rich
carbonate, and therefore one could not exclude a mantic
purpose for this intaglio.
Plate 7 Diagnostic drawing of a kidney
‘Gems of Heaven’ | 63
Mastrocinque
Plates 8a–b Chalcedony with Chnoubis and triple-barred S. 17 x 14 x 2mm.
London, British Museum, PE G 402
Chnoubis gems: Chnoubis and liquid secretions
Many gems show the image of the lion-headed snake
Chnoumis. Chnoubis or Chnoumis is the Greek name of the
Egyptian creator god Chnoum. He was especially venerated at
Elephantina, near the Nile’s first cataract, and at Syene as the
god of the Nile flood,40 which began under his influence in the
sign of Leo.41 His image on gems consists of two elements: a
snake’s body and a lion’s head from which emanate rays, and a
charakter formed of a triple-barred S (Pls 8a–b). His relevance
within many religious milieus and his role in protecting the
womb and procreation are related to the fact that he was the
creator. Chnoubis was also considered as the first decan of Leo.
His iconography perhaps depended on that of the decan
Knem(et). One Chnoubis gem depicts the hieroglyph for
water.42 The power of these magical gems resided in the notion
that the god who regulated the Nile flood, could also regulate
menstruation, stimulate the flow of breast milk, and prevent
haemorrhaging, bleeding ulcers and abnormal digestive juices
in the stomach or intestine.43
Campbell Bonner discusses the question of what materials
Chnoubis gems were made of:
The materials actually used for the type of the lion-headed snake
cover a fairly wide range, and yet there are certain manifest
preferences. Commonest of all is chalcedony, white, gray, blue,
pale yellow, and smoky brown; next, (probably) green jasper,
plasma, chrysolite, and prase. There are also some specimens on
agate and on black jasper and obsidian, and I have seen several on
stones that had been so altered by heat, whether purposely or
accidentally, that the original colour and even the material could
not be readily determined. Yellow jasper is rare, red jasper
probably rarest of all. There is so much irregularity about all
magical amulets that we should hesitate to treat an unusual
material as a ground for suspicion of forgery.44
On the basis of my experience, the most common Chnoubis
gems are of chalcedony or prase, i.e. green or white
translucent, semi-transparent stones.45 Bonner is surely right in
maintaining that a choice of stone is evidenced by the majority
of gems, while the exceptions are the minority. I agree also
with his assumption that an ‘irregular’ stone is not necessarily
a forgery.
Next, I would like to discuss the idea of ancient imitations,
which were probably made outside the main areas of
production. Many workshops produced series of ‘normal’
Chnoubis gems following the directions of one or more
magicians, but it was possible for anybody to make a copy from
an engraved specimen or from a written recipe. The result was
often what we recognise as an ‘irregular’ or ‘exceptional’
magical gem. For example, one yellow jasper depicting
Harpocrates46 is such an exception; it has seven vowels
64 | ‘Gems of Heaven’
Plates 9a–b Serpentine amulet with Chnoubis above a womb, and an ibis
above IAW. Paris, Cabinet des médailles
inscribed in reverse writing, as if the intention was to use the
gem as a seal. Both these features (the use of yellow jasper and
reverse writing) are irregular. We know that the designs on
many gems were taken directly from written recipes, as is
proved by textual misunderstandings or mistakes in copying.
In the case of this gem, it is possible that the gem cutter had at
his disposal only the drawing and the letters to be cut, but
because he was used to cutting normal gems with reversed
inscriptions he applied that technique to this amulet.
Another problem arises when we read recipes such as those
in the Kyranides,47 which credit a single amulet with occult
powers over the stomach, womb and kidneys. For example, the
series of heart-shaped stone pendants with an ibis, or an ibis
and Chnoubis48 were supposed to favour the digestion, as
proved by the accompanying inscription: πέσσε πέσσε πέσσε:
‘Digest, digest, digest!’ The colour of these gems is dark as they
were made from steatite, limonite, serpentine, metamorphic
basalt or brown jasper. Intaglios in this series were also meant
to protect the womb,49 which is represented on them (Pls
9a–b). This fact makes the situation more complicated because
we do not know whether the dark colour was chosen for the
health of the stomach, intestine, or womb, and whether the
stone was chosen for Chnoubis or for the divine ibis.
But the problem is not hopeless. If we set aside the few
Chnoubis gems of very unusual material and colour, such as
red jasper, it is possible to look for a logic in the choice of
stones. It is true that many gods on magical gems had different
values according to the religious context in which they were
venerated.50 It is nonetheless also true that magical amulets of
an appropriately coloured stone were seen as an empirical
means of influencing human organs or secretions of the same
colour. One suspects that the diseases of specific human
secretions or organs were regarded as benefitting from stone
amulets because of their similarities. Remedies with the same
colour could indeed be used within different religious and
cultural milieus. Milk was white to Jews, Egyptians, or Romans
and gastric juices were green to Greeks, Syrians, or Spaniards.
Several texts explain what the purposes of such gems were.
Galen51 and Hephaistion52 write that Chnoubis amulets were
effective against stomach diseases. Galen speaks of green
jasper for this purpose. The lapidary book of Socrates and
Dionysius says:
white onyx, completely translucent, like air... inscribe on it the
coils of a serpent, whose front part is the head of a lion emanating
rays. When worn, this stone completely stops stomach pains, and
whatever you eat you will digest well.53

Plate 10 Chalcedony with Chnoubis sign without the bar. Ravenna,
Archaeological Museum
The physician Marcellus Empiricus writes:
here is a physical remedy to stomach pains: engrave in jasper stone
a radiated snake (it should have seven rays), set the stone in a gold
frame and wear it on the neck54... against pleuritic and hip pains…
engrave on a jasper stone Phrygian aerizusa this script: ·SSS·, and
wear it on the neck… .55
Aëtius of Amida also prescribes the use of green jasper against
stomach diseases.56 Here the problem of the relationship
between the literary sources and magical gems arises,57 as
white onyx is never used for Chnoubis gems;58 maybe Socrates
and Dionysius were thinking of chalcedony. The passage of
Marcellus is contradictory because jasper could not be
aerizusa, i.e. a stone as transparent as the air. But we have
already seen that in Mesopotamian traditions the stone which
Greeks called ‘jasper’ was a translucent one. Two Chnoubis
gems are made of rock crystal,59 which could correspond to
Marcellus’ description and be the aerizusa stone. But Marcellus
suggests drawing the sign of Chnoubis without the bar, as three
S’s between two dots. A gem could correspond to this form as
evidenced by a white chalcedony in Ravenna (Pl. 10),60 which
could be supposed to be the aerizusa.
Many workshops in different periods produced gems in a
more conventional shape which depict Chnoubis and bear
inscriptions which instruct the stomach ‘to digest’. In the
Skoluda collection is a haematite gem which depicts Chnoubis,
the womb and the inscription παῦσον πόνον τοῦ στομάχου
Αβρασαξ: ‘Let the stomach pains cease! Abrasax’.61 In this case
the choice of the stone depended on the properties of haematite
rather than on the power of Chnoubis. A very famous specimen
is in the Cabinet des médailles, Paris, and is inscribed with the
words φύλαξον ὑγειῆ στόμαχον Προκλου (‘Keep the stomach
of Proklos healthy!’) (Pls 11a–b).62 The stone is probably a
black-brown serpentine.63
Plates 11a–b Serpentine with Chnoubis and the inscription: ‘Keep the
stomach of Proklos healthy!’. Paris, Cabinet des médailles
The Colours of Magical Gems
Plates 12a–b Green jasper with anguipede and Chnoubis and inscribed: ‘Chnoubis,
digest, digest!’. 33 x 31 x 9mm. London, British Museum, PE 1986,0501.40
One prase gem bears the inscription: ‘Avert from Julian, son
of Nonna, every (abdomen’s) stress, every bad digestion, every
stomach pain!’.64 One green-yellow jasper in the British
Museum has the inscription [Χνοῦ]βις πέσσε πέ[σσε...]:
‘Chnoubis, digest, digest!’ (Pls 12a–b).65 The same inscription
urging ‘to digest’ is cut on a gray-blue chalcedony,66 whereas
one yellow chalcedony has the inscription εσωσιβωει στομάχου:
‘Rescue-help (protection of the) stomach’.67 One may agree with
Bonner that there is some confusion concerning the material
and the colour of stones intended for Chnoubis amulets.
The gem of Proklos and others with personal names were
produced on demand. On the other hand, anonymous gems
were often mass-produced. There is a large number of small
Chnoubis gems, convex on both sides, which are of light
translucent green stones, such as prase, green agate, moss
agate, green chalcedony, or olivine. One could suspect that
they had been issued by a limited number of workshops in the
2nd and 3rd centuries ad, and in a limited number of provinces,
as is suggested by the uniformity of the products. Their colour
is similar to that of gastric juices, and therefore it is possible to
suppose that they were used to give health to the stomach. In
the case of several stomach diseases, such as pylorus hernia,
the stomach can regurgitate clear juices, which could explain
the choice of a transparent stone such as rock crystal.
Intestinal juices
We have seen that several gems favouring the digestion are
made of dark stones. Intestinal or gastric diseases such as
bleeding ulcers produce black juices, which can be seen in the
faeces, and this may be a possible reason for the choice of black
stones. Obviously we are proposing solutions which are
sometimes hypothetical. The difficulty and complexity of this
matter, however, should not induce us to label the choice of
stones as arbitrary or casual. Lapidary books, for example the
Kyranides or the Orphic Lithika, make it clear that rules in this
choice had to be followed. Unfortunately the surviving amulets
are rarely described in these books. In the case of medical
amulets the colour of human organs and secretions provides us
with traces of a taxonomy.
Milk
Milk is white and Chnoubis protected breast-feeding activities.
The treatise on stones by Socrates and Dionysios prescribes:
‘Another onyx stone, completely black. It is useful to pregnant
women and to those who are breast-feeding. On it one should
engrave a three-headed Chnoubis’.68 Two black obsidian stones
depicting Chnoubis have survived,69 but only one gem with a
three-headed snake is known,70 and it is made of white
‘Gems of Heaven’ | 65
Mastrocinque
chalcedony; the inscription on the reverse mentions Chnoubis,
even if the snake has no lion’s heads. For a milk amulet a white
chalcedony is very appropriate. The archaeological evidence
shows that Chnoubis was often represented on milky
chalcedony. Also many small gems of the same material with
the Chnoubis sign are known. The lapis galactitis was a light
stone which was useful for breast-feeding.71 When pulverized,
it was able to colour water white.72 Therefore it is possible that
Socrates and Dionysios spoke of dark stones for pregnancy,
whereas the colour of breast-feeding amulets was another, i.e.
white. Chnoubis is often represented on uterine amulets, at the
centre of three or more Egyptian gods, who are placed over the
female organ (Pl. 9a). All or almost all of these gods are
concerned with conception, pregnancy, and breast-feeding, as
is the case for Isis, Harpocrates and Osiris. The colour of those
gems is always dark and the stone is always haematite,73 which
could explain Socrates and Dionysius’ recipe which speaks of a
black stone.
Conclusion
The long-lasting production of magical gems, the plurality of
workshops and of religious traditions, the naïve ancient
imitations, not to speak of modern fakes, prevent us from
making simple and unproblematic classifications. Moreover,
these problems can lead us to assume that the choice of stones
was an arbitrary or casual factor. Our ignorance should not
however be the reason for denying the existence of rules and of
a logic with regard to these choices. Only a few stones have
been studied for their alleged properties and their documented
use as amulets. The preceding discussion on haematite and
other stones has shown that there was a supposed similarity or
a sort of kinship between these stones and certain bodily
functions. Haematite shared the nature of blood, aetites
(‘eagle-stone’) the nature of a pregnant body, galactitis the
nature of milk. Practitioners pulverized galactitis, mixed it
with water and obtained a liquid which appeared to be milk;
the same occurred with haematite, which resembled
coagulated blood, but when pulverized assumed the colour of
living blood. Lapidary books and other texts by eastern
astrologists and magicians explored the connections between
stones, plants, stars, and parts of the human body. It would
have been illogical if these learned men had not taken into
account the colour of stones. On the contrary, the tradition of
Near Eastern treatises shows that colour was important.
The series of gems discussed above show that haematite
(and its reddish-brown variety, i.e. limonite) was used to
protect red-brown organs, such as the liver and kidney, and
that natural pieces of this stone actually have the shape and
gloss of these organs. The inscriptions on several gems confirm
that they were amulets for the liver. Moreover the Syrian god
Adad was known for his stone kidney, and a Syrian haematite
intaglio is shaped like a kidney. The colour and properties of
stones were supposed to be syntonic with the world of the gods.
My argument is furthered by presenting an experimental
study on Chnoubis gems. They are relatively numerous and it is
possible to single out a few groups according to their colour.
Ancient authors and inscriptions on magical gems explain that
the Chnoubis gems were used to treat diseases or to prevent
them. Stomach or intestinal diseases, pregnancy and the
womb’s other functions, and breast-feeding were dealt with
66 | ‘Gems of Heaven’
thanks to Chnoubis and his gems. The liquid secretions on
which bodily health depended are red (blood), black (when the
intestine is bleeding), green or transparent (the stomach
juices), and white (milk). Chnoubis gems are rarely red or
yellow, whereas they are often white, green, and sometimes
transparent or black. If we reject the idea of casuality, we must
look for a logic or a taxonomy in the choice of colour. A possible
explanation emerges from the colours, because white stones
were suitable for breast-feeding and the iconography of a three-
headed Chnoubis appears on a white gem. This iconography is
known as useful for breast-feeding. Therefore: white gems =
milk. Greek physicians recommended, in case of stomach
diseases, green or transparent stones with the image of this
god. Therefore: green or transparent gems = the stomach,
whose juices are green or transparent. Other physicians
recommended transparent stones for pleuritic and hip
diseases, and therefore another possibility is: transparent gems
= pleura and hips; but in this case the colour was scarcely
related with the disease. Another numerous group of gems
shows this god over the womb, and these gems are of
haematite. They were aimed at controlling the functions of this
organ, and especially pregnancy and menstrual cycles.
Therefore: haematite’s colour = blood and womb. It is possible,
at the end, to hypothesise the use of black stones by taking into
account the colour of blood from intestinal haemorrhages, even
if the black gem of Proklos urges to protect his stomach. The
relatively dark colour of uterine gems could suggest other uses,
for example to protect functions of the womb. The
interpretation of black-coloured stones is uncertain in
comparison with the green, white and transparent ones, which
are mentioned by ancient physicians.
I conclude by saying that we should not characterise as
casual the choice of stones, even though our scant knowledge
only in a few cases allows us to understand an underlying logic
and to outline a simple taxonomy.
Notes
1 Á.M. Nagy, ‘Gemmae magicae selectae. Sept notes sur
l’interprétation des gemmes magiques’, in A. Mastrocinque (ed.),
Gemme gnostiche e cultura ellenistica, Atti dell’incontro di studio
Verona, 22–23 ottobre 1999, Bologna, 2002, 162–9.
2 A.A. Barb, ‘Lapis Adamas’, in Hommages à Marcel Renard, I,
Brussels, 1969, 67–82; see also, C.N. Bromehead, ‘Aetites or the
Eagle-Stone’, Antiquity 81 (1946), 16–22; A.A. Barb, ‘Birds and
Medical Magic. I. The Eagle-Stone’, Journal of the Warburg and
Courtauld Institutes 13 (1950), 316–18. Now there are also the Belles
Lettres editions of lapidary books; see note 3 below.
3 R. Halleux and J. Schamp, Les Lapidaires grecques. Lapidaire
orphique, Kérygme, Lapidaire d’Orphée, Socrate et Denys, Lapidaire
nautique, Paris, 1985; M. Waegeman, Amulet and Alphabet -
Magical Amulets in the first Book of Cyranides, Amsterdam, 1986;
A. Mastrocinque, ‘Die Zauberkünste der Aphrodite. Magische
Gemmen auf dem Diadem der Liebesgöttin (Kyranides I.10)’, in
Th. Ganschow (ed.), Otium. Festschrift für Volker Michael Strocka,
Remshalden, 2005, 223–31.
4 A. Mastrocinque (ed.), Sylloge gemmarum Gnosticarum, II, Rome,
2008.
5 I am preparing a new publication of the collections in this museum.
6 It is well known that amethysts of a violet hue were useful in
avoiding drunkenness; the anemone flower served the same
purpose; cf. A. Önnerfors, ‘Magische Formeln im Dienste
römischer Medizin’, in Aufstieg und Niedergang der römischen Welt
2.37.1 (1993), 186–7.
7 W. Bonser, ‘The Significance of Colour in Ancient and Mediaeval
Magic: With Some Modern Comparisons’, Man 25 (1925), 194–8.
8 Pseudo-Plutarch, De fluviis, 11.4 = 1157 A.

9 B. Landsberger, ‘Über Farben im Sumerisch-Akkadischen’, Journal
of Cuneiform Studies 21 (1967), 139–73, esp. 151. On the 11 basic
colours employed in human speech, see B. Belin and P. Kay, Basic
Colour Terms: their Universality and Evolution, Berkeley, 1969.
More recently see, D.B. Kapp, ‘Basic Colour Terms in South
Dravidian Tribal Languages’, Indo-Iranian Journal 47 (1994), 193–
201.
10 Landsberger (n. 9), 152.
11 Halleux and Schamp (n. 3), 151, Orphei Lithika kerygmata 8; cf. R.
Halleux, ‘Fécondité des mines et sexualité des pierres dans
l’antiquité gréco-romaine’, Revue Belge de Philologie et d’Histoire 48
(1970), 16–25.
12 W. Horowitz, Mesopotamian Cosmic Geography (Mesopotamian
Civilisations 8), Winona Lake, 1998, 9–11, 13–14; G.B. Lanfranchi,
‘La volta celeste nelle speculazioni cosmografiche di età neo-
assira’, in Atti del Convegno internazionale di Archeologia e
Astronomia: L’uomo antico e il cosmo (Roma, 15–16 maggio 2000),
Rome, Accademia Nazionale dei Lincei, 2001, 149–62, esp. 155.
13 Pliny, Nat. Hist., XXXVII.57–58; 61.
14 On the grounds that the odem stone was found on the breastplate
of the Jewish high priest: Barb (n. 2), 71.
15 Barb (n. 2), 72; for haematite as a medical astringent and desiccant:
Theophrastus, De lapidibus, 37; Galen, De comp. med. gen. (12.720
Kühn); Paulinus, Ep. med. VII.3.11; Aetius II.13; Stephanus, Collyr.
13 Studemund; Ps. Dioscor. De lap. II.1, 21 Ruelle.
16 K. Weitzmann (ed.), Age of Spirituality. Late Antique and Early
Christian Art, Third to Seventh Century (exh. cat., The
Metropolitan Museum of Art, 19 November 1977–12 February
1978), New York, 1979, no. 393.
17 Theophrastus, De lapidibus, 37.
18 Hesiod, Theogony, 180–7. On the miracles of Uranos’ blood and of
that of other gods or heroes, see A.A. Barb, ‘St. Zacharias the
Prophet and Martyr’, Journal of the Warburg and Courtauld
Institutes 11 (1948), 35–67.
19 Pliny, Nat. Hist., XXIX.66.
20 Papyri Graecae Magicae XII, 412 (hereafter PGM).
21 P. Gaillard-Seux, ‘Le “sang de basilic” chez Pline l’Ancien (N.H.,
XXIX, 66): résine de genévrier ou hématite?’, L’Antiquité Classique
68 (1999), 227–38. The same papyrus in another passage calls
‘Cronos blood’ the juniper resin: PGM XII, 434.
22 Also, Halleux and Schamp (n. 3), 109–14, Orphei Lithika 20, and
160–2, Orphei lithika kerygmata 20.
23 Pliny, Nat. Hist., XXXVII.164.
24 Mastrocinque (n. 4), nos Pe 20; Vr 25–6; H. Philipp, Mira et Magica,
Mainz, 1986, no. 38; Coralli segreti. Immagini e miti dal mare tra
Oriente e Occidente (Mostra, Potenza 22 giugno–30 ottobre 2006),
Lavello (PZ), 2006, 56; B. Tasser, ‘Eine magische Korallen-
Gemme’, in Herkos. Studi in onore di Franco Sartori, Padua, 2003,
265–71.
25 C. Bonner, Studies in Magical Amulets, Ann Arbor, 1950, 263–4;
A. Mastrocinque (ed.), Sylloge gemmarum Gnosticarum I, Rome,
2003, no. 317.
26 See Nagy (n. 1).
27 For Hekate, see Mastrocinque (n. 25), no. 305.
28 Pliny, Nat. Hist., XXXVII.186.
29 M.P.E. Berthelot, Les origines de l’alchimie, Paris, 1885, pl. 1 and
106; Sch. Pind., Isthm.V.1–2.
30 Bonner (n. 25), 66; S. Michel, Die magischen Gemmen im Britischen
Museum, P. and H. Zazoff (eds), London, 2001, nos 385–6; cf.
A.A. Barb, ‘Bois du sang, Tantale’, Syria 29 (1952), 279–83.
31 A. Delatte and Ph. Derchain, Les intailles magiques gréco-
égyptiennes, Paris, 1964, no. 461.
32 H. Seyrig, ‘Invidiae medici. 1. La faim de l’ibis et la soif de Tantale’,
Berytus 1 (1934), 4 and fig. 3; Bonner (n. 25), 87–9, 276, D144; Barb
(n. 30); A. Mastrocinque, ‘Studi sulle gemme gnostiche. VIII. “Bevi
sangue, Tantalo”’, Zeitschrift für Papyrologie und Epigraphik 130
(2000), 137–8; Michel (n. 30), nos 382–4; S. Michel, Die Magischen
Gemmen, Berlin, 2004, 294; A. J. Festugière, ‘Pierres magiques de
la collection Kofler (Lucerne)’, Mélanges Université de St. Joseph
Beyrouth 37 (1961), 287–93 = A.J. Festugière, Etudes d’histoire et de
philologie, Paris, 1975, 151; cf. idem, ‘Amulettes magiques’, Classical
Philology 46 (1951), 86–9.
33 To cure not only haemorrhages, but also to regulate menstruation:
C. Faraone, ‘Does Tantalus drink the Blood, or not?: an enigmatic
series of inscribed hematite gemstones’, in U. Dill and C. Walde
(eds), Antike Mythen. Medien, Transformationen und
The Colours of Magical Gems
Konstruktionen (in honour of Fritz Graf), Berlin/New York, 2009,
203–28.
34 H.Seyrig, ‘Antiquités syriennes. 89. Les dieux armés et les Arabes
en Syrie’ and ‘Appendice III. Le culte d’Arès en Syrie’, Syria 47
(1970), 77–112, esp. 110–12; H.J.W. Drijvers, Cults and Beliefs at
Edessa ( EPRO 82), Leiden, 1980, 146–74; C. Augé, in LIMC II.1, s.v.
Ares (in peripheria orientali) (1984), 493–8.
35 Barb (n. 2), 72.
36 Pliny, Nat. Hist., XXXVII.186: ‘Adadu nep<h>ros <sive> renes,
eiusdem oculus, digitus; deus et hic colitur a Syris’.
37 Henry Seyrig saw it in Aleppo and bought it before giving it to the
Cabinet des médailles: H. Seyrig, ‘Antiquités syriennes. 40. Sur
une idole hiérapolitaine’, Syria 26 (1949), 17–28; Drijvers (n. 34), 32
and pl. XXXIV.3.
38 Macrobius, Saturnalia, I.17.66–7.
39 Halleux and Schamp (n. 3), 18, 156–7.
40 D. Bonneau, La crue du Nil, Paris, 1964, 232–3.
41 D. Wortmann, ‘Kosmogonie und Nilflut’, Bonner Jahrbucher 166
(1966), 85–7; R. Merkelbach, ‘Astrologie, Mechanik, Alchimie und
Magie im griechsch-römischen Ägypten’, Begegnung von
Heidentum und Christentum im spätantiken Ägypten, Riggisberger
Berichte 1, Riggisberg, 1993, 59. Here is not the place to discuss in
detail the origins of Chnoubis-Chnoumis and his relationship with
the decan Knem(et).
42 Delatte and Derchain (n. 31), no. 352.
43 Mastrocinque (n. 25), § 20.
44 Bonner (n. 25), 60.
45 I have done a rough statistical analysis of Chnoubis gems, in order
to class them into several groups, each of which has a restricted
range of colours and transparencies. This statistical evidence is
based on data assembled by Simone Michel in her book, Die
Magischen Gemmen (n. 32). They are as follows: white and
transparent: 70 examples of white, yellowish or greenish
chalcedony, rock crystal, smoky or yellowish quartz. White and
opaque: four examples of light steatite, yellow or white jasper.
Green and transparent: 49 examples of prase, moss agate,
chrysoprase, olivine and beryl. Green and opaque: 46 of green
jasper, green-yellow jasper, plasma, heliotrope, nephrite, jade,
green onyx. Black or dark brown, opaque or scarcely translucent:
10 examples of ‘black jasper’, obsidian, black glass, basalt, or an
unidentified black stone. It is not to forget that gems labelled as
black jasper almost always are stones other than jasper; see
A.A. Barb, review of Delatte and Derchain (n. 31), Gnomon 41.3
(1969), 298–307.
46 Delatte and Derchain (n. 31), no. 135. The same is true of an onyx
with the anguipede rooster: Mastrocinque (n. 4), fig. 47.
47 Kyranides I.22, 102 (on chrysites stone): D. Kaimikis (ed.), Die
Kyraniden, Meisenheim am Glan, 1976.
48 Seyrig (n. 32), 2–3, fig. 2; Bonner (n. 25), D77–82.
49 Delatte and Derchain (n. 31), 188–90.
50 Harpocrates, for instance, was venerated not only by Egyptians,
but also by Greeks, Romans and Jews: see, A. Mastrocinque, From
Jewish Magic to Gnosticism, Tübingen, 2005, 77–9. On Harpocrates
(i.e. Horus the child) and the (non-orthodox) Jews: Talmud,
Avodah Zarah, III 3 (IX) (1934), 566–8, Ed. L. Goldschmidt); cf.
M. Hadas-Lebel, Le paganisme à travers les sources rabbiniques des
IIe et IIIe siècles, in: ANRW II/19.2, 1979, 405; N. Belayche, Iudaea-
Palaestina. The Pagan Cults in Roman Palestine (Second to Fourth
Century), Tübingen, 2001, 162–4 and passim. On Jews who
venerated or revered Serapis in various forms: Talmud, as quoted
above; G. Mussies, ‘The Interpretatio Judaica of Sarapis’, in
M.J. Vermaseren (ed.), Studies in Hellenistic Religions (EPRO 78),
Leiden, 1979, 189–214; for Christians: Historia Augusta, Vita
Saturnini 8.
51 Galenus, XII, 207, (ed. K.G. Kühn, Leipzig, 1821–1833).
52 Hephaistion, IV 1, (ed. D. Pingree, Leipzig, 1973–1974, II, 143).
53 Theophrastus, De lapidibus 35; Halleux and Schamp (n. 3), 170.
54 Marcellus Empiricus, De medicamentis, XX 98 (ed. E. Liechtenhan,
Basel, 1917, 354).
55 Ibid., XXIV 7 (Liechtenhan ibid., 412). In PGM IV, 1264 reference is
made to a sign of this kind with a single S, useful for driving away
demons.
56 Aëtius of Amida, Libri medicinales II.18 (ed. A. Olivieri, Leipzig/
Berlin, 1939–1950).
57 It is impossible to deal here with the relationships or otherwise
between the ancient and modern names of semi-precious stones.
‘Gems of Heaven’ | 67
Mastrocinque
For example, jasper is opaque in modern descriptions, but ancient
lapidaries speak sometimes of it as a semi-transparent stone.
58 A green one is described by A. De Ridder, Collection De Clercq:
Catalogue, T. VII, 2 parte. Les pierres gravées, Paris, 1911, no. 770.
59 Michel (n. 30), nos 325–6; cf. also Delatte and Derchain (n. 31), 86
(Chnoubis without rays on the head, and, on the other side, a fish
and a bird). See the paper by Faraone this volume, Pl. 2.
60 Mastrocinque (n. 4), no. Ra 23.
61 S. Michel, Bunte Steine - Dunkle Bilder: “Magische Gemmen”,
Freiburg, 2001, 124–5, pl. 24.145; eadem 2004 (n. 32), 259–60:
no.11.3.e.
62 Delatte and Derchain (n. 31), no. 80 = Mastrocinque (n. 4), no. 168.
63 It is described as a black jasper by Michel 2004 (n. 32), 259, no.
11.3.e.
64 C. Bonner, ‘A Miscellany of engraved Stones’, Hesperia 23 (1954),
149, no. 36; Michel 2004 (n. 32), 259, no. 11.3.e.
65 Michel (n. 30), no. 338.
66 Bonner (n. 25), 267, D 83.
67 Delatte and Derchain (n. 31), no. 89.
68 Socrates and Dionysius, De lapidibus 36 = Halleux and Schamp
(n. 3), 171.
69 Bonner (n. 25), 267, D 85; Mastrocinque (n. 4), no. Pe 11.
68 | ‘Gems of Heaven’
70 E. Zwierlein-Diehl, Magische Amulette und andere Gemmen des
Instituts für Altertumskunde der Universität zu Köln, Papyrologica
Coloniensia XX, Opladen, 1992, no. 18. Cf. E. Zwierlein-Diehl,
Antike Gemmen und ihr Nachleben, Berlin/New York, 2007, 213,
461, pls 178, 802.
71 Pliny, Nat. Hist., XXXVII.162: ‘Galactitis ex uno colore lactis est…
sunt qui smaragdum albis venis circumligatum galactiten vocent’.
See also Isidore, Origines, XVI.10.4; 4.20; Dioscorides, V.132;
Aëtius of Amida (n. 56), II.17 (I, 162 Olivieri); Orphei Lithica 2 (=
Halleux and Schamp [n. 3], 92–4); Orphei Lithica kerygmata 2 (=
Halleux and Schamp [n. 3], 147–8); Psellus 8, 202 (F. de Mély and
C.E. Ruelle, Les lapidaires de l’Antiquité et du moyen âge, II, Paris,
1898). In modern Europe milk stones are white: G. Bellucci,
Amuleti italiani antichi e contemporanei, Perugia, 1912; Halleux
and Schamp (n. 3), 304, n. 5.
72 Halleux and Schamp (n. 3), 302, n. 3.
73 A. Delatte, ‘Etudes sur la magie grecque, III–IV. (La clef de la
matrice)’, Musée Belge 18 (1914), 5–88, esp. 76–9; Bonner (n. 25), 85;
A.A. Barb, ‘Diva matrix’, Journal of the Warburg and Courtauld
Institutes 16 (1953), 193–238; J.-J. Aubert, ‘Threatened Wombs:
Aspects of Ancient Uterine Magic’, Greek, Roman and Byzantine
Studies 30 (1989), 421–49; Faraone (n. 33).
Magic and Medicine
Gems and the Power of Seals
Véronique Dasen
A number of recent studies have explored the fluidity of the
boundaries between medical, religious, and magical therapies
in Classical antiquity,1 but the implications of this interplay
have yet to be fully investigated. In this paper, I would like to
pursue reflections on these interactions. I will more specifically
focus on the notion of sphragis (seal), common to both
practices, and explore the double meaning of the word which
casts light on an important aspect of the cultural context of
magical gems and could elucidate one of their operating
modes.
Material evidence of the relationship between ancient
magic and medicine is manifold. On the one hand, medical
instruments may show divine or magical devices ensuring the
success of the practitioner. Besides Asclepios, the figure of
Heracles is common. His presence is partly explained by his
fame for his courage and endurance, partly by his competence
as alexikakos, ‘evil’s averter’, partly by the genealogy of
Hippocrates. Some believed that Hippocrates was descended
from Asclepios through his father, and from Heracles through
his mother. An apocryphal letter to Artaxerxes compares
Hippocrates, who defeats ‘wild’ and ‘bestial’ diseases, with
Heracles, the champion of dangerous animals. Divinised,
Hippocrates allegedly received in Greece the same honours as
Heracles and Asclepios.2 Roman period coins from Cos depict
on the obverse a seated Hippocrates, inscribed with his name,
and on the reverse the bust of Heracles holding a club.3 It is thus
no surprise to find allusions to Heracles on medical
instruments, especially on items used for painful operations
requiring great skill; some handles of surgical knives from
Pompei depict his bust, the knotty handles of embryo hooks
and needles for cataract couching imitate the hero’s club,
whereas retractors end in the shape of a lion’s head, possibly of
the Nemean lion.4 Heracles thus helped ‘taming’ pain as he
mastered wild animals, also promoting the patient’s resistance
and chances of survival. Collyrium stamps for eye-salves too
can bear divine or magical figures, such as the stars and moon
also found on magical gems (Pls 1–2).5
On the other hand, magical gems often refer to medical
practices. They share a common imagery of the body, displayed
on gems. Uterine gems are thus carved with a cupping device,6
a visual metaphor for the womb used in medical texts, such as
Plates 1a-b Steatite or green serpentine (46 x 20 x 12mm). Avignon, Palais du
Roure
the Hippocratic treatise on Ancient Medicine:
Again, cupping instruments, which are broad and tapering, are so
constructed on purpose to draw and attract blood from the flesh.
[…] Of the parts within the human frame, the bladder, the head,
and the womb are of this structure. These obviously attract
powerfully, and are always full of a fluid from without.7
The image of a cupping vessel is also the conventional emblem
of the medical profession during the Graeco-Roman period.
The device thus possessed a supplementary value; it added
medical authority to the efficacy of the magical procedure.
Other literary medical metaphors can be detected on gems,
such as the image of the octopus, representing the womb in
medical texts and on uterine gems.8 These interactions are no
coincidence: they reflect a wide therapeutic system which
could combine magical and medical remedies without
antagonism, and in a complementary way.
Sphragis
The double meaning of the word sphragis throws an interesting
light on the nature and function of healing stones, pointing to
other possible connections between medical and magical
therapies. Sphragis usually designates a seal ring or stamp.9
Physicians also had stamps: oculists used to impress solid sticks
of eye ointments with a stamp, usually made of greenish-black
steatite, carved with a text, cut in reverse, on the flat face of
each edge. The content of the inscription provides the name of
the person who probably invented the salve, the name of an
affliction, and the name of the salve for its treatment,
sometimes adding how to use it.10
The word sphragis also has another meaning for
practitioners: it denotes the result of stamping, namely not just
the impression of the stamp, but the remedy itself. A sphragis is
thus a stamped pill, called in Greek trochischos, in Latin
pastillus.11 In the reign of Tiberius, Celsus describes the famous
sphragis or pill of Polyidus, perhaps named after the legendary
seer and healer Polyidus:
But the pastil of Polyidus called the ‘seal’, sphragis autem
nominatur, is by far the most celebrated. It contains split alum
4.66g, blacking 8g, myrrh 20g, lign aloes the same, pomegranate
heads and ox-bile, 24g each; these are rubbed together and taken
up in dry wine.12
Plate 2 Red jasper (15 x 13 x 4mm). London, British Museum, PE 1849,1127.16
‘Gems of Heaven’ | 69
Dasen
Similarly, the physician Galen uses sphragis as a synonym for
collyrium (eye-salve): ‘For inflammed eyelids, apply a collyrium
mixed with water, that some call a sphragis’.13
An oculist stamp from Reims in France confirms that the
word sphragis could designate a remedy: it names the
impressed dried salve stick not collyrium, as expected, but
sfragis in Latin transliteration, demonstrating that the Greek
term was well understood in 2nd–3rd century ad Roman
Gaul:14
d galli (s)fragis ad aspritudin(em).
d galli (s)fragis ad impet(um) lippit(udinis)
Sphragis of Decimus Gallus Sestus for trachoma
Sphragis of Decimus Gallus Sestus at the onset of inflammation
Terra Lemnia
The analogy between stamped pills and stone gems extends far
beyond the common use of the word sphragis. Like gems, pills
could bear pictures, some of them being very similar to those
found on medical magical gems. The most famous, and the
most ancient, sphragis of classical antiquity was sealed clay,
made of earth collected on the island of Lemnos in north-
eastern Greece. Lemnian earth was highly reputed as an
antidote with wide-ranging healing properties, from eye-
diseases to stomach pains and the bites of venomous animals.
The pill was characterised by its reddish colour – and by a
stamped image. Pliny defines the earth as a red ochre, rubrica
Lemnia:
In medicine it is a substance ranked very high. Used as a liniment
round the eyes it relieves defluxions and pains, and checks the
discharge from eye-tumours; it is given in vinegar as a draught in
cases of vomiting or spitting blood. It is also taken as a draught for
troubles of the spleen and kidneys and for excessive menstruation;
and likewise as a remedy for poisons and snake bites and the sting
of sea serpents; hence it is in common use for all antidotes.15
Many ancient authors discuss the healing qualities of Lemnian
earth that could also reduce inflammations, heal up recent and
malignant wounds and soothe chronic pains.16 Its styptic
properties are observed by Cassius Felix (5th century ad) who
recommends Lemnian seals against blood spitting.17 Theodorus
Priscianus (5th century ad) also prescribes it against
haemorrhage as does Mustio (6th century ad) against
gynaecological bleeding.18
Thanks to the recent analysis by two geologists, Hallas and
Photos-Jones,19 we know today that the typical red colour of
Lemnian clay is due to the presence of haematite, a powerful
red pigment (c. 5%). They also found that Lemnian pills could
work as a medicine because of its other components:
montmorillonite (c. 40%), a clay with a strong absorbing
power, very efficient for the removal of toxins, also used
externally, and kaolin (35%), another healing clay efficient
against soft tissue inflammation, and an absorbent when taken
in case, for example, of gastro-enteritis. The earth also
contains alum (20%) with well-known haemostatic and anti-
bacterial properties.
The most intriguing fact about the Lemnian clay sphragis is
that it looked like a gem because it was stamped with an image,
that of a goat. Dioscorides underscores the role of the goat, and
reports that the presence of its blood explained the colour of
the earth:
70 | ‘Gems of Heaven’
What they call Lemnian earth is brought up from a certain
cavernous underground passage and mixed with goat’s blood; the
people there, after moulding it and stamping it with the figure of a
goat, call it sphragis. It is an uncommonly effective antidote for
deadly poisons when drunk with wine, and, when taken ahead of
time, it forces one to vomit the poisons. It is suitable for the strokes
of venomous animals and for their bites. It is mixed with
antidotes.20
We have a precious eye-witness in the person of the physician
Galen himself who wrote a detailed account of his second
journey to Lemnos.21 Lemnian earth was one of the 37
ingredients of his famous mithridatium. He wanted to see how
the product was exploited and manufactured before buying it
for his own practice :
I also sailed to Lemnos and for no other reason than to get the
Lemnian earth or ‘seal’ (sphragis) whichever it is called. This has
been thoroughly described in the ninth book of my treatise On the
Properties of the Simple Drugs.22
Galen first describes the ritual performed by the priestess of
Artemis, and confirms that she stamped an image on the clay:
The priestess collects [the earth], to the accompaniment of some
local ceremony, no animal being sacrificed, but wheat and barley
being given back to the land in exchange. She then takes it to the
city, mixes it with water so as to make moist mud, shakes this
violently and then allows it to stand […]. She takes small portions
and imprints upon them the seal of Artemis [the goat]; then again
she dries these in the shade till they are absolutely free from
moisture […]. This then becomes what all physicians know as the
Lemnian Seal.23
Galen was intrigued by the description of Dioscorides:
I had once read in the works of Dioscorides and others that the
Lemnian earth is mixed with goat’s blood, and that it is out of the
mud resulting from this mixture that the so-called Lemnian seals
are moulded and stamped. Hence I conceived a great desire to see
for myself the process of mixture […] in order to see in what
proportion blood was mixed with the earth.24
On the spot, the enigma was soon solved: ‘All who heard this
question of mine laughed’.25 No goat’s blood was added, the red
colour was natural. As we know thanks to Hallas and Photos-
Jones, it is due to the presence of haematite. A book providing a
respected medical authority was brought to Galen:
I got a book from one of them, written by a former native, in which
all the uses of the Lemnian earth were set forth. Therefore I had no
hesitation myself in testing the medicine, and I took away twenty
thousand seals.26
Galen then goes on describing the astringent and dessicative
action of Lemnian earth on animal bites, ulcers, persistent
pains and swellings, and explains how to employ the seals for
external and internal use. They had to be dissolved in a liquid,
such as vinegar, wine, or oxymel, until it has a mud-like
consistency, ‘like these pastilles (trochisci) which are made in
various ways’.27 Mixed with vinegar it was applied to a wound.
As an antidote against poisoning, it had to be drunk, added to a
special preparation. The long-lasting fame of Lemnian clay,
used as a kind of panacea, extended beyond antiquity. In post-
medieval and modern times, it was no longer collected by the
priestess of Artemis, but blessed by the church.28
No clay sphragis from Lemnos is preserved, but we find a
reflection of it on a gem from the Seyrig collection in the
Cabinet des médailles in Paris (Pls 3a–b).29 A she-goat is
carved, not on a reddish clay, but on a haematite, a stone which
produced a red colour too. On the reverse, we find the

Plates 3a–b Haematite, 13 x 9mm. Paris, Cabinet des médailles (Seyrig
collection)
expression pauson, ‘stop, put an end to’, which could refer to
the bleeding stopped by the power of haematite, or to the relief
of any pain. The formula pauson ponon occurs on other medical
gems, such as a haematite gem from the Skoluda collection
addressing Chnoubis ‘pauson ponon tou stomachou’ (Pl. 4).30
The choice of the stone carved with the goat is not a
coincidence: haematite, or ‘bloodstone’, was credited with
qualities very similar to those of Lemnian earth. It was highly
reputed as a blood-stauncher; it could also cure eye diseases
and venomous bites, says the Orphic Lapidary.31 Dioscorides
has a similar description, arguing that:
It has properties that are astringent, that warm somewhat, that
thin, and that wipe off scars and roughness in the eyes with honey.
With a woman’s milk it is good for opthalmia, for rents, and for
bloodshot eyes.32
The manner of using it provides another parallel between
haematite stone and Lemnian clay. Like Lemnian pills, it was
advised to drink the stone broken and mixed with a liquid, such
as water, or applied with other ingredients, such as honey or
human milk.33 This procedure explains why a large number of
haematite gems are found broken: they were taken as a
medicine, as were other stones with medical properties, but in
lesser quantities.34 In a medical context, brittleness was even
regarded as a quality for haematites. Dioscorides thus asserts
that:
Haematite is of excellent quality when it breaks easily as if of its
own accord and when it is hard, uniformly strong, and free of any
dirt or veins.35
In sum, the picture on Lemnian seals has a revealing parallel
on a magical medical gem. The stone in the Seyrig collection
could be identified as a kind of Lemnian seal, not impressed,
but carved with a she-goat, not in red clay, but in a stone with
similar qualities.36 Did gem carvers intend to imitate the
famous clay pill? They may have followed the more general
custom of stamping precious medical products.
Plate 4 Haematite, 46.2 x 24.9 x 5.8mm. Plate 5 Pot fragment, from Aquincum (size of the printed
Skoluda Collection M085 gem: 19 x14mm). Present location unknown
Magic and Medicine
Medical sphragides
How widespread were medical sphragides, apart from the
Lemnian one, and do other magical gems look like them? A
number of remedies with pictures can be traced, some
presenting images also found on magical gems.
Galen reports a remedy from a lost treatise of Asclepiades
the Younger (1st century ad): ‘The yellow remedy of Antigonos,
called little lion because it was printed with the image of a
lion’.37 In the same treatise, Asclepiades also mentions a crow
seal, korakinè sphragis, a remedy good for mouth or throat
troubles;38 the name may refer to its black colour or to the
image of a crow. Another example occurs in a 1st-century ad
Egyptian papyrus where Servilius explains to Nemesion, a
wealthy man from Philadelphia, that he bought for him a
‘stone’ (litharion) of silphium, printed with the image of
Harpocrates;39 a very common iconographic type on magical
gems.40 In the same period, the Pliny the Elder tells us that:
‘Now indeed men also are beginning to wear on their fingers
Harpocrates and figures of Egyptian deities’.41
Remedies prepared in a magical context could also be
stamped, like normal drugs, with an image, but this time
explicitly magical. One of the Greek Magical Papyri offers a
description of the preparation of a collyrium made of animal
and plant material (field mouse, dappled goat, dog-faced
baboon, ibis, river crab, moon beetle, wormwood, and a clove
of garlic), duly stamped, like regular remedies, but with a ring
bearing the image of Hecate and a magical name:
Blend with vinegar. Make pills, kolluria, and stamp them with a
completely iron ring, completely tempered, with a Hecate and the
name Barzou Pherba.42
Apart from solid sticks of salve, containers of precious
medicine were also impressed with an image certifying its
authenticity, such as the famous lykion pots, miniature jars
around 2–3cm high, containing a much valued liquid extracted
from a shrub from the buckthorn family, originally from Lycia
in Asia Minor. The most ancient jars seem to be as early as the
3rd or 2nd century bc and are stamped with the word ‘Lykion’,
occasionally with the name of the druggist or owner,
sometimes also with the head of Asclepios with or without a
radiating diadem.43 The label proved that the druggist was
selling the genuine product, an alleged wonder drug, effective
as an astringent, good against ophthalmic inflammation,
ulcerations, and bleeding.
The image of Asclepios and Hygieia impressed on a pot
found in Aquincum (Pl. 5) could indicate that the vessel was
Plates 6a–b Brown agate, 22.5 x 18 x 3.5mm. London,
British Museum, PE G21
‘Gems of Heaven’ | 71
Dasen
Plate 7 Carnelian, 19 x 16mm. London, British Museum, PE 1859,0301.118
also a container for a medicine.44 Unfortunately, it is so
fragmentary that no conclusion can be drawn, but it is
interesting to note that the image was made with a gem,
perhaps magical, as the type exists, as on a dark brown agate
from the British Museum (Pl. 6a) showing Chnoubis on the
reverse (Pl. 6b).45
Two gems demonstrate the intertwining between medical
and magical sphragides. A deep orange carnelian gem, carved
with a retrograde inscription, was thus used as a stamp to mark
a collyrium for the eyes (Pl. 7).46 The inscription is short, but
typical of collyrium stamps: herophili/ opobalsamvm. The
name Herophilus may designate the druggist who invented the
salve. It is also the name of the famous Alexandrian physician
who worked on the anatomy of the eye and carried out the first
dissection of the eye.47 The druggist may have attributed the
salve to him in order to increase the fame of its product, or a
physician himself took the name of his famous predecessor.
The second term, opobalsamum (opobalsaminum), is a well-
attested drug from the balsam-tree efficacious against eye
diseases.48
Eyesight is central in the scene, carved with the image of
Athena seated, looking at a tragic mask, as if it were an active
persona. As M. Pardon-Labonnelie demonstrated,49 the image
contains several references to the power of eyesight. First the
eyes of Athena were reputed for their special colour, glaukos,
greenish-blue, but also, according to Plutarch and Pausanias,
she saved Lycurgus from losing a wounded eye. Lycurgus in
return introduced in Sparta the cult of Athena Ophtalmitis or
Oplitetis.50
A round jasper from Wroxeter with a name and a
prescription, but no image, provides another example of a gem-
like (or pill-like?) stamp for dried salve sticks.51
Plates 8a-b Haematite (25 x 15 x 4mm). London, British Museum, PE G 87
72 | ‘Gems of Heaven’
Sphragis Theou
The word sphragis occurs not only on regular medical stamped
pills or collyria, but is carved on magical gems. We find it on
the well-known 4th–5th century ad series of so-called
‘Solomon’ gems. The type depicts on one side a horseman,
often labelled ‘Solomon’, spearing a prostrate female figure (Pl.
8a). The reverse usually bears the inscription sphragis theou,
‘Seal of God’ (Pl. 8b).52 The motif of the rider may derive from
Horus stabbing a crocodile personifying evil, or the hunting
emperor struck on coins, though Solomon is not in military
costume.53 The device is nearly always carved on haematite, a
choice so far unclear.
The expression sphragis theou is traditionally interpreted as
referring to the magic seal-ring which Solomon received from
Iahweh to repel the vampire-like demons assaulting him
during the construction of the Temple of Jerusalem. The gems
are usually explained as depicting how Solomon masters a
female demon harmful to women and children, present in all
Mediterranean folklores. Different names are proposed for the
woman, such as Gello, Gylou, or Abyzou.54 The role of Solomon,
however, was not limited to women and children’s protection;
as Spier points out, he controlled all evils.55 Thus, the reverse of
a haematite in the British Museum is carved with the
inscription stomachou designating his power over pains of the
belly (Pl. 9b),56 which fits well with the haematite’s potency for
or against internal bleeding, like Lemnian earth.
The double meaning of the word sphragis introduces a new
reading of the ‘Solomon’ series which could explain the
preference for haematites: sphragis theou could also mean ‘the
medicine of god’. ‘Solomon’ haematites are often found broken,
most likely because they were used as a drug, as we saw above.
One may guess that, like pills, the broken part of the gem was
pulverized and drunk mixed with a liquid.57
It may be noted that the iconography of the horseman
subduing the female demon appears when the figure of
Heracles mastering the lion disappears. Solomon seems to have
taken over the capacity of the hero. Like Heracles, who
controlled the roaming of the womb (compared with a wild
animal), variants depict Solomon with the hystera formula.58
Solomon had power over all diseases inflicted by demons,59
including the fear of poisoning, mastered by haematites, like
the red Lemnian earth.60
Plates 9a-b Haematite (20 x 12 x 2mm). London, British Museum, PE G 439

Conclusion
In conclusion, I suggest that magical medical gems could be
conceived as sphragides, that is, as stone remedies. The practice
of carving this category of gems with images can be
understood in the light of the custom of stamping costly
medical drugs. For both medical and magical sphragides, the
image certified the authenticity and quality of the medicine. It
also increased the value of the gems, as we should not forget
that medical gems without any device did also exist. Carved
medical gems were luxury products as were precious stamped
products. A chronological coincidence is worth noting:
references to stamped remedies begin in the Hellenistic period
and intensify with trade and exportation in the early Empire,
when magical gems with a specific iconography develop.61 The
emergence of both genres may be interrelated. Stone and clay
pills may also have had a similar destiny: manufactured in one
place, sold or used by itinerant practitioners in another.
Notes
1 See e.g. C. Nissen, Entre Asclépios et Hippocrate. Etude des cultes
guérisseurs et des médecins en Carie (Kernos suppl. 22), Liège, 2009;
V. Nutton, ‘From medical certainty to medical amulets: three
aspects of ancient therapeutics’, in W.F. Bynum and V. Nutton
(eds), Essays in the History of Therapeutics (Clio Medica 22),
Amsterdam, 1991, 13–22; V. Nutton, Ancient Medicine, London,
2004. On gems, see e.g. M.G. Lancellotti, ‘Médecine et religion
dans les gemmes magiques’, Revue de l’histoire des religions 218
(2001), 427–56; A. Mastrocinque, ‘Medicina e magia. Su alcune
tipologie di gemme propiziatorie’, in A. Marcone (ed.), Medicina e
società nel mondo antico. Atti del convegno di Udine (4–5 ottobre
2005), Grassina, 2006, 91–100; S. Michel, Die magischen Gemmen.
Zu Bildern und Zauberformeln auf geschnittenen Steinen der Antike
und Neuzeit, Berlin, 2004, 146–202; Á.M. Nagy, ‘Daktylios
pharmakites. Magical healing gems and rings in the Graeco-
Roman world’, in  I. Csepregi and Ch. Burnett (eds), Ritual Healing
in Antiquity and in the Middle Ages, London, 2011 (in press).
2 Hippocrates, Letter II (9.314-5 L. = E. Littré, Oeuvres complètes
d’Hippocrate, Paris, 10 vols, 1839–1861); see also Pliny, Nat. Hist.
VII.123 (trans. H. Rackham), London, 1952; J. Jouanna, Hippocrate,
Paris, 1992, 32–3.
3 E.g. A. Verbanck-Piérard (ed.), Au temps d’Hippocrate. Médecine et
société en Grèce antique, Mariemont, 1998, 221, I. 22 (fig.).
4 E.g. L.J. Bliquez, ‘The Hercules motif on Greco-Roman surgical
tools’, in A. Krug (ed.), From Epidaurus to Salerno, Symposium held
in Ravello, April 1990, (PACT 34), Rixensart, 1992, 35–50; idem,
Roman Surgical Instruments and Other Minor Objects in the
National Archaeological Museum of Naples. With a Catalogue of the
Surgical Instruments in the «Antiquarium» at Pompei by Ralph
Jackson, Mainz, 1994, 83–4, 99–106, nos 40–1, fig. 20.
5 Collyrium stamp: J. Voinot, Les cachets à collyres dans le monde
romain, Montagnac, 1999, no. 240 (stars and moon); gem: S.
Michel, Die Magischen Gemmen im Britischen Museum, London,
2001, no. 92 (stars and moon); see also, A. Mastrocinque (ed.),
Sylloge gemmarum gnosticarum, vol. II, Rome, 2008, Ts 25 and 26
(stars and moon).
6 C. Bonner, Studies in Magical Amulets, Chiefly Graeco-Egyptian,
Ann Arbor, 1950, 79–94; V. Dasen and S. Ducaté-Paarmann,
‘Hysteria and metaphors of the uterus’, in S. Schroer (ed.), Images
and Gender. Contributions to the Hermeneutics of Reading Ancient
Art (OBO 220), Fribourg and Göttingen, 2006, 239–61.
7 Hippocrates, Ancient Medicine 22 (1.629 L.; trans. W.H.S. Jones,
Loeb Classical Library, Cambridge, MA, 1868).
8 V. Dasen, ‘Métamorphoses de l’utérus, d’Hippocrate à Ambroise
Paré’, Gesnerus 59 (2002), 167–86; eadem, ‘Représenter l’invisible:
la vie utérine sur les gemmes magiques’, in V. Dasen (ed.),
L’embryon humain à travers l’histoire. Images, savoirs et rites,
Gollion, 2007, 41–64.
9 P. Chantraine, Dictionnaire étymologique de la langue grecque.
Histoire des mots, nouvelle édition avec un supplément sous la dir. de
A. Blanc, Ch. de Lamberterie, J.-L. Perpillou, Paris, 2009, 1041.
10 M.J. Roulière-Lambert, A.-S. de Cohën and L. Bailly (eds), L’œil
Magic and Medicine
dans l’antiquité romaine. Exposition Lons-le-Saunier, Musée
d’archéologie, 31 janvier-4 avril 1994, Lons-le-Saunier, 1994;
R. Jackson, ‘Eye medicine in the Roman Empire’, in Aufstieg und
Niedergang der römischen Welt, II, 37.3 (1996), Berlin-New York,
2228–51, at 2238–43; Voinot (n. 5), 1–49; M. Pardon-Labonnelie,
‘La préparation des collyres oculistiques dans le monde romain’, in
F. Collard and E. Samama (eds), Pharmacopoles et apothicaires: les
“pharmaciens” de l’Antiquité au Grand Siècle, Paris, 2006, 41–58.
11 D. Gourevitch, ‘Collyres romains inscrits’, Histoire des sciences
médicales 32 (1998), 365–72; M.-H. Marganne, ‘Les médicaments
estampillés dans le corpus galénique’, in A. Debru (ed.), Galen on
Pharmacology: Philosophy, History, and Medicine. Proceedings of
the Vth International Galen Colloquium, Lille, 16–18 March 1995,
Leiden, 1997, 153–74; eadem, ‘Les médicaments estampillés dans la
littérature médicale latine’, in P. Defosse (ed.), Hommages à Carl
Deroux, II, Prose et linguistique, médecine, Brussels, 2002, 536–48.
12 Celsus De medicina 5.20.2 (trans. W.G. Spencer, Loeb Classical
Library, Cambridge, MA, 1935). For further literary references to
this sphragis, see Marganne 2002 (n. 11), 537–8.
13 Galenus De compositione medicamentorum secundum locos 4 (12.751
K. = C.G. Kühn, ed. and Latin trans., Galeni opera omnia, 20 vols,
Leipzig, 1821–1833).
14 CIL XIII 76; Voinot (n. 5), no. 158 ; Marganne 2002 (n. 11), 546.
15 Pliny, Nat. Hist. XXXV.33–4.
16 H. Hasluck, ‘Terra Lemnia’, Annual of the British School at Athens 16
(1909–10), 220–31; L. Taborelli, ‘A proposito della genesi del bollo
sui contenitori vitrei. Note sul commercio delle sostanza
medicinali e aromatiche tra l’età ellenistica e quella imperiale’,
Athenaeum. Studi periodici di letteratura e storia dell’Antichità 63
(1985), 198–217, at 202–5.
17 Cassius Felix, De medicina 39.7 (Lemnias sfragitidos) (ed. V. Rose,
Leipzig, 1879).
18 Theodorus Priscianus, Euporistes 3.7.29 (Lemnia sfragitide) (ed.
V. Rose, Leipzig, 1894); Mustio, Sorani Gynaeciorum uetus
translatio Latina, 30.83 (sfragitida) (ed. V. Rose, Leipzig, 1882).
19 A.J. Hall and E. Photos-Jones, ‘Accessing past beliefs and practices:
the case of Lemnian earth’, Archaeometry 50/6 (2008), 1034–49.
20 Disocurides, De materia medica 5.97 (= Pedanius Dioscorides of
Anazarbus, De materia medica, trans. and ed. L.Y. Beck,
Hildesheim, Zürich, New York, 2005). On the associations of goats
with Artemis, girls and medicine, see P. Brulé, ‘Héraclès à l’épreuve
de la chèvre’, in idem, La Grèce d’à côté. Réel et imaginaire en miroir
en Grèce antique, Rennes, 2007, 255–81.
21 The date is debated, between ad 162 and 167.
22 On Antidotes 1.2 (14.8 K.= Kühn [n. 13]); A.J. Brock (trans. and ed.),
Greek Medicine. Being Extracts Illustrative of Medical Writers from
Hippocrates to Galen, London, 1929, 199.
23 On Simple Drugs 9.2 (12.169–70 K. = Kühn [n. 13]); Brock (n. 22).
192.
24 Ibid., 9.2 (2.171 K.); Brock (n. 22), 192–3.
25 Ibid., 9.2 (2.173–4 K.); Brock (n. 22), 194.
26 Ibid., 9.2 (12.174 K.); Brock (n. 22), 194.
27 Ibid., 9.2 (12.176 K.); Brock (n. 22), 195.
28 Hasluck (n. 16).
29 C. Bonner, ‘Amulets chiefly in the British Museum’, Hesperia 20
(1951), 301–45, at 342, no. 73 (not illustrated). I am very grateful to
Attilio Mastrocinque for providing me with the image.
30 S. Michel, Bunte Steine - dunkle Bilder: ‘Magische Gemmen’,
Munich, 2001, no. 145.
31 R. Halleux and J. Schamp, Les lapidaires grecs: lapidaire orphique,
kérygmes lapidaires d’Orphée, Socrate et Denys, lapidaire nautique,
Damigéron-Evax, Paris, 1985: lapidaire orphique 21.
32 De materia medica (n. 20), 5.126.1.
33 Halleux and Schamp (n. 31): lapidaire orphique 21.645–76;
kérygmes lapidaire d’Orphée 22; Damigéron-Evax IX.
34 Some are also burnt, see Pliny, Nat. Hist. XXXVII.139 (antagates);
Michel (n. 1), 151.
35 De materia medica (n. 20), 5.126.1.
36 I leave aside the question of the choice of the goat as the seal of
Artemis, and the healing properties of the animal, but it may be
noted that the she-goat was much appreciated in medicine and
magico-medical treatments: see Pliny, Nat. Hist. XXVIII.130 on the
properties of her milk.
37 R.J. Durling, A Dictionary of Medical Terms in Galen, Leiden, 1993,
219; Marganne 1997 (n. 11), 165. On magical gems, see Michel (n. 5),
no. 253 (red jasper with a lion in profile); no. 280 (yellow jasper
‘Gems of Heaven’ | 73
Dasen
with a lion, inscribed iaw; on the reverse a woman, inscribed
pausia perhaps derived from pauson?).
38 Asclepiades ap. Galen, De compositione medicamentorum per
genera 5.11 (13.826, 4-7 K. = Kühn [n. 13]); Marganne 1997 (n. 11),
166.
39 H. Cuvigny, Papyrus Graux II (P. Graux 9 à 29), Geneva, 1995, no.
10, 22–8, esp. lines 8–9; Marganne 1997 (n. 11), 153.
40 Michel (n. 5), no. 112 (dark green jasper with the child seated on a
lotus flower, a hand to his mouth, the head crowned with the sun
disc or the pschent).
41 Pliny, Nat. Hist. XXXIII.41.
42 PGM IV 2691–2 = K. Preisendanz, Papyri graecae magicae. Die
griechischen Zauberpapyri, 3 vols, Leipzig and Berlin, 1928, 1931,
1941 (Eng. trans. by N. O’Neil in H.D. Betz (ed.), The Greek Magical
Papyri in Translation, Including the Demotic Spells, Chicago and
London, 1992 (2nd edn), 88, n. 331).
43 See, E. Sjövqist, ‘Morgantina: Hellenistic medicine bottles’,
American Journal of Archaeology 64 (1960), 78–83, at 80, pl. 19, fig.
8.
44 I. Wellner, ‘Aeskulapius és Hygieiát ábrázoló gemma
Lenyomatával díszített edény Aquincumból (un vase orné de
l’empreinte d’une gemme représentant Esculape et Hygie trouvé à
Aquincum)’, Archaeologiai Értesító 92 (1965), 42–4.
45 Michel (n. 5), no. 319; see also, A. Delatte and Ph. Derchain, Les
intailles magiques gréco-égyptiennes, Paris, 1964, 179, no. 235.
46 First published by C.W. King, Antique Gems and Rings, II, London,
1872, 20, and discussed by M. Pardon-Labonnelie, ‘Les
thérapeutiques oculistiques romaines, entre survivances et
métamorphoses. L’exemple de la thérapeutique du vert’, in
H. Duchêne (ed.), Survivances et métamorphoses, Dijon, 2005, 111–
32, at 130–1, fig. on 124. See also, R. Jackson, Catalogue of Greek
and Roman Medical Collections in the British Museum, in
preparation. I thank Ralph Jackson for providing me with the
results of the stone analysis.
47 See H. von Staden, Herophilus. The Art of Medicine in Early
Alexandria, Edition, Translation and Essays, Cambridge and New
York, 1989.
48 On the drug, see Voinot (n. 5), 47–8, no. 87–8; Jackson (n. 10), 2240.
For a similar inscription on a conventional stamp, see E.
Esperandieu, ‘Recueil des cachets d’oculistes romains’, Revue
archéologique 24 (1894), 58, no. 7: Herophili opob(alsamum);
74 | ‘Gems of Heaven’
Pardon-Labonnelie (n. 10), 45.
49 Pardon-Labonnelie (n. 46).
50 Pausanias, Description of Greece (trans. W.H.S. Jones and H.A.
Ormerod), Cambridge, MA, 1926, 3.18.2; Plutarch, Lycurgus 11. See
also Pausanias, ibid., 2.2.4.2 (Athena oxyderkes).
51 Voinot (n. 5), no. 43. Two circular or cylindrical examples were also
found in Enns and Ipswich: Voinot (n. 5), nos 216 and 247. I would
like to thank Ralph Jackson for these references.
52 Michel (n. 5), no. 436. On the series, see Bonner (n. 6), nos 294–328;
J. Spier, ‘Medieval Byzantine magical amulets and their tradition’,
Journal of the Warburg and Courtauld Institutes 56 (1993), 25–62.
53 Bonner (n. 6), 210.
54 P. Perdrizet, Negotium perambulans in tenebris. Etudes de
démonologie gréco-orientale, Strasbourg, 1922; I. Sorlin, ‘Striges et
Geloudes. Histoire d’une croyance et d’une tradition’, Travaux et
mémoires du Centre de recherche d’histoire et civilisation de Byzance
11 (1991), 411–36; Spier (n. 52), 33–9; S.I. Johnston, ‘Defining the
dreadful. Remarks on the Greek child-killing demon’, in M. Meyer
and P. Mirecki (eds), Ancient Magic and Ritual Power, Leiden-New
York-Cologne, 1995, 361–87.
55 Spier (n. 52), 44.
56 Michel (n. 5), no. 447.
57 E. Zwierlein-Diehl, Magische Amulette und andere Gemmen des
Instituts für Altertumskunde der Universität zu Köln, Opladen,
1992, 50. See PGM III (n. 42), 188: ‘Grind up a magnet’; on ingesting
magical powers: PGM I (n. 42), 231–248: ‘Wash the papyrus and
drink the water’.
58 On Heracles and Omphale on magical gems, see V. Dasen, ‘Le
secret d’Omphale’, Revue archéologique 46 (2008), 265–81. See for
example the inscriptions of the hystera formula on a silver pendant
in Spier (n. 52), 30, nos 15–24, 33, pls 2a–b, 3a.
59 See the bronze pendant with Solomon on one side and the Evil Eye
attacked by animals on the other side: Bonner (n. 6), nos 298–303;
Spier (n. 52), 62.
60 I note that inscriptions relating to the belly or stomach occur on all
types of haematite gems. The image of the reaper, for example,
may be inscribed with pepte (instead of schiôn, ‘for the hips’) or
stomachou: Michel (n. 5), no. 427. In the Orphic Lapidary 21.675–
679, haematite also secures success and victory.
61 Taborelli (n. 16), 216–17.
Magical Gems and Classical Archaeology
Árpád M. Nagy
Introduction
Today, research on magical gems is characterised by a strange
duality: it is both flourishing and stagnant. Much has certainly
changed in recent years. Following the publication in 1950 of
Campbell Bonner’s aere perennius work, Studies in Magical
Amulets, decades passed before any significant leaps forward
were made. Around the turn of the millenium, however,
magical gems are more topical than ever. To mention only a
few examples from this year (2009): the Getty Colloquium in
February,1 the present conference, and the establishment of an
international research group connected with the exhibition of
magical gems scheduled to open in Budapest in December 2011.
Notwithstanding the renewed large-scale exploration of
the sources, this revival of interest is essentially confined to the
discipline of comparative and historical religion, where it has
provided many new interpretations.2 Classical archaeology,
however, remains largely untouched by it. Richard Gordon tells
a story that shows just how untouched. A few years ago, one of
the most distinguished archaeological institutes immediately
refused to invite a scholar on hearing that he was planning to
deliver a lecture on magical gems. ‘Such things do not form
part of Classical archaeology’ – ran their summary reasoning.3
It is worth stopping for a moment and (instead of easy
tabloid moralising) consider the moral of the story, since this
attitude points to a fundamental problem. Even though
magical gems (the surviving ones number 4000–5000 pieces)
constitute a large source-group within the material remains of
Classical antiquity, they have not been fully integrated into the
scholarly record. I believe that the most important task for us
today is to accomplish that integration. Magical gems should be
considered a simple genre of material culture, like Chalcidian
vases, Samnite bronzes, or Attic sarcophagi.
In my opinion there are three main reasons for Classical
scholarship’s instinctive rejection of magical gems. They are
considered to be ‘ugly’, ‘magical’, and – since they have no
workshops, chronology or context – ‘unfathomable’ for the
archaeologist. These reasons stand in the path of research like
three increasingly massive barricades. My paper aims to tackle
them one by one.
1. ‘Magical gems are ugly’
This problem is easily dealt with. The ‘ugliness’ of magical
gems is by now only a diminishing theme in a centuries-old
trial symbolically opened by the founding father of Classical
archaeology, Johann Joachim Winckelmann, when he
relegated magical gems beyond the frontier of ancient art.
They are, he wrote ‘…nicht würdig, in Absicht der Kunst, in
Betrachtung gezogen zu werden’.4 It is in the spirit of
Winckelmann’s anathema that his spiritual heirs, who consider
the study of ancient art the primary task of Classical
archaeology, aim to place magical gems, sometimes quite
literally, outside the scope of their profession. As a synecdoche,
one can mention the well-known fact that in the Berlin
Museum magical gems are kept in the Egyptian Department,
while the British Museum once stored them in its Medieval
collection.5 The aesthetic quality of magical gems, however,
has never been subjected to independent analysis. Their
alleged ‘ugliness’ serves mainly as an emotional support for the
second reason. Of course they are ‘ugly’ because they are
‘magical’ i.e. alien from the admired Greek spirit of rationality.
2. ‘Magical gems are magical’
This charge against magical gems is much more severe, and is
decisive even today. I believe, however, that what we have here
is a misunderstanding. It is worth briefly reviewing the
question.6
As known, magical gems are defined by three constituent
features: magical names (voces magicae, logoi), signs
(charakteres), and unique iconographic schemes (e.g. anguipes,
Chnoubis). These are complemented by two structural features
of equal importance. First, the inscription is not engraved in
mirror-writing, and is thus legible on the gem itself, not on the
impression. Secondly, the reverse and the edges of the gem are
also often engraved, even though most magical gems were
ringstones, with only the obverse visible. So much for
morphology. It has never been clarified, however, what the
definition ‘magical gem’ actually defines. The term ‘magical
gem’ suggests that the category includes all precious stones
used for a magical purpose – whatever is understood by the
word magic. But this is not the case.
Definitio fit per genus proximum et differentiam specificam.
If we use this scholastic method to define the position of
magical gems, the differentia specifica will obviously be the
above-mentioned constituent features, while the genus
proximum must be the whole group of talismans, that is:
jewellery used for magical purposes.7 It is worth taking our
sources concerning talismans as a starting point, first
synchronically, with examples from the period of the Roman
Empire, and then diachronically, examining evidence from the
Classical and Hellenistic periods. These sources attest the
existence of other groups of objects than magical gems.
In the Roman period there are at least three further
iconographic categories to be distinguished. The Lithika, or
literary works on precious stones, show that the majority of
representations recommended for gems and rings belong to the
usual iconographical schemes of the Graeco-Roman pantheon,
and contain neither magical signs nor names. Thirty-four out
of 45 representations in the Lithika are like that.8 The next
group is represented by the recipes listed in the Kyranides.9
These are not magical gems either, since their decoration
typically consists of combinations of bird and fish or snake,
arranged around the letters of the Greek alphabet. Finally,
‘Gems of Heaven’ | 75
Nagy
Plates 1a–b Nicolo gem. St Petersburg, The State Hermitage Museum, Inv. no. Ж.1517
there is a separate group of gems which have a completely
general iconography, but whose inscriptions still explicitly
identify them as amulets. The inscriptions usually contain a
wish of a general nature to make the amulet’s wearer lucky and
attractive.10 An example is the nicolo gem of the Hermitage,
one of the most famous talismans (Pls 1a-b). It aimed to cure
gout by taking the traditional image of Perseus and imagining
him chasing Podagra, that is, by invoking a historiola and
transforming it into an image.11 The nicolo gem from the
Carthage treasure was presumably a talisman of the same
type: it bears an image of Isis Pelagia holding out the sails of a
ship; next to her an inscription in mirror-writing: NAVIGA FELIX
(Pl. 2).12 Based on the image and the text, the piece functioned
as an amulet, ensuring safe sea-travel. These gems, too, lack all
the constituent features identified above.
It is by no means surprising that the boundaries between
different talisman types are not strictly drawn. Maryse
Waegeman, for example, was able to connect recipes in the
Kyranides with related Zaubergemmen.13 I would like to add
two more pieces to the list she compiled. One of them is a gem
(‘ferruginous sandstone with malachite veins’) in the British
Museum that bears the image of a peacock on its obverse, with
a branch and an engraving of uncertain meaning at its feet.
The reverse shows the incised letters aiw (Pls 3a-b).14 Although
Simone Michel, who published the gem, correctly referred to
the possibility of a Christian interpretation and a connection
with the phoenix, a connection with the tau recipe in the
Kyranides seems more relevant: ‘In the stone (sc. taites) a
peacock is engraved walking on a sting-ray and under the stone
the cry of the peacock which is aiw ’. The angry green veins
shining forth from the brown stone rhyme well with the
colourful feathers of the peacock. The branch on the obverse
Plates 3a–b Sandstone gem, 20 x 18 x 2mm. London, British Museum, PE G
506
76 | ‘Gems of Heaven’
Plate 2 Nicolo gem, max. D. 64mm. London, British
Museum, PE AF 329
may be the floral component typical of the Kyranides, while the
‘undefinierbarer Gegenstand’ below the bird’s feet may refer to
the sea creature. The other is an unpublished jasper gem in the
Hermitage. Its obverse is decorated with an eagle grasping a
snake – without any hint as to its being a talisman. The reverse,
however, bear voces magicae and charakteres (Pls 4a-b). One of
the recipes in the Kyranides recommends just this combination
of iaspis gem and an eagle grasping a snake.15
To sum up: there were other types of talismans besides
magical gems in the Roman period. It must also be understood
that magical gems as a genre are not defined by their function.
Magical gems were mounted primarily in rings. The use of
magic rings in Greek culture is mentioned in the written
sources as early as the 5th century bc. They were represented
on the Athenian comic stage in plays by Eupolis, Aristophanes,
Ameipsias, Plato comicus, and Antiphanes. According to the
Epidaurian Iamata even the god Asklepios made use of magic
rings. Hesychius and the Aristophanes-scholia even preserve
the ancient designation for them: daktylios pharmakites, which
means that in their own culture magic rings – as opposed to
Zaubergemmen – constituted an independent category. The use
of talismans and amulets is thus far from being alien to Greek
culture. The novelty of magical gems lies not in their function,
but in their form.
Whereas the perspective of classicizing aesthetics relegated them
to the periphery, nowadays we are beginning to see them in terms
of a new, still developing paradigm as something central. Magical
gems are today understood as a thoroughly Hellenized genre of
magika. They constituted a medium in which the magical
traditions of different cultures could unite to create a
transnational expertise in the Mediterranean.16
Other talismanic objects are very difficult or impossible to
identify on an archaeological basis. After all, as we have seen,
Plates 4a–b Jasper gem. St Petersburg, The State Hermitage Museum, Inv. no.
Ж.6673

Plates 5a–b Chalcedony gem, 28 x 20 x 5mm. London, British
Museum, PE G 65
the iconographical schemes of Graeco-Roman art were also
appropriate for amulets. In the case of magical gems, however,
the magical function is recognisable through the constituent
features mentioned above. This is precisely the characteristic
that offers a chance for further analysis. And this takes us to
the third problem.
3. ‘Magical gems are unfathomable’
Three factors are usually mentioned here: the lack of context,
workshop and date. It will suffice to touch upon the first one
briefly. It is known that of several thousand magical gems, only
a few were recovered from a proper archaeological context.
Intensive excavation in the Mediterranean has brought to light
only a few examples.17 Most of them come from tomb-deposits,
which merely mark the last phase of their ancient life-cycle,18
giving no information about the way that they were used in
life. Only a few dozen have a provenance, which again sheds no
light on their diffusion in antiquity, since magical gems, like all
precious stones, were continuously collected until the modern
period. This situation, however, has its hitherto-unnoticed
silver lining. The corpus of surviving magical gems – precisely
because of their continuous collection, and not excluding
perhaps the possibility of large unique finds19 – is by and large
complete. This opens scope for statistical analysis: a promising
field that will hopefully be explored by future researchers.
I suggest that we deal with the other two problems, the
question of dating and workshops, simultaneously. As for
dating, what we read in the 1822 work of Christian Gottlieb
Heyne still holds true. Production of Zaubergemmen probably
began in the late Republican period, and ended with the
decline of ancient glyptics in Late Antiquity.20 This can be
complemented by the fact that the appearance of charakteres
and magical names on other genres of magika from the 1st
century bc provides a terminus post quem for magical gems –
naturally without necessarily fixing the beginning of their
production.21 Following the suggestion of Max Pieper in 1934,22
workshops are generally localised in Alexandria, although this
theory remains largely unsubstantiated. In fact, the situation is
much worse today: even the recognition of related gems causes
difficulty. We do not know where and when magical gems were
produced, and which pieces belong together. Concerning these
two areas, even the smallest certain answer is a step forward.
In what follows, I propose two possible methods that may lead
towards a solution to the problem of workshops and dating.
Their novelty lies mainly in that they constitute a return to
methods suggested by Campbell Bonner (1876–1954),23 but
whose elaboration was left to his followers.
As a starting point, we have to ask what is the significance
Magical Gems and Classical Archaeology
Plate 6 Blackish green jasper gem, 13 x 11 x 3mm. The Hague, Royal Coin
Cabinet, Inv. no. 1049
of the constituent features of magical gems, and what made the
invention of this genre of gems – so radically different from
other types of talismans – necessary. In other words, what is
the difference between Zaubergemmen and other amulet-
gems?
In my opinion the main difference lies in the fact that
magical gems reflect magical ritual in a way radically different
from other types of amulet-gems. The latter elude
identification precisely because they are iconographically
indistinguishable from the other Graeco-Roman gems. It is
only the inscription that marks the Perseus-gem in St
Petersburg or the Isis-gem in London as an amulet, and not the
iconography of the hero or the goddess in itself. The elements
of the related rite or incantation cannot be deduced from them
because these gems continue the tradition of talismans general
until the 1st century bc: a tradition whose essential
characteristic was its orality, and which was replaced by a
literary tradition whose initial diffusion is reflected in the
prevalence of voces magicae and charakteres.24 The magical
gems are direct visual and inscribed representations of magical
praxeis. The elements engraved on the gems: images, magical
names and signs which occur together in no other context, are
connected to each other only in and through the praxis in
which they are established and whose ritual requires their
presence. This is true even in cases where we can see that the
gem in question is an improvisation or bricolage (the tropaion
gems discussed below are an example); but also when we are
dealing (as in the case of the phoenix gems, also discussed
below) with a magical rite rooted in centuries of theological
tradition. One can, following Véronique Dasen, conclude that
the magical gem is the sphragis (‘seal’) of the praxis.25 In
consequence, one optimal basis for identifying magical gems is
the reconstruction of a praxis which can be realised through
formal or structural analysis of the gem.
Today, magical gems are identified either by type or
function.26 The former, where a characteristic iconographic
feature such as the tropaion or the phoenix becomes the basis
of classification is more frequent. Groups thus generated
contain all examples which bear the selected motif. This is the
categorising principle in Simone Michel’s 2004 corpus,27 for the
collection of which all scholars must be grateful. However, it is
perhaps worth focusing on structure instead of elements. This
will bring us closer to reconstructing the praxis that produced
the gems.
For example, magical gems decorated with the ‘trophy’
motif were divided into six groups by Michel. The third group
was characterised as follows: ‘Tropaion als Zentralmotiv, mit
Figur am Boden, Löwe’.28 Three pieces were listed in this group.
‘Gems of Heaven’ | 77
Nagy
Plates 7a–b Dark green jasper gem, 26 x 22.4 x 4.4mm. Paris, Bibliothèque
Nationale, Cabinet des médailles, Inv. no. 2175
One, in London, is a chalcedony stone of blueish colour (Pls
5a–b).29 On the obverse is the trophy, a lion, and a recumbent
figure; on the reverse is an anguipede and a gorgoneion.30 On
both sides there are stars and magical names, as on the
following gems. The Hague gem is a blackish-green jasper
stone and half the size of the previous one. Pl. 6).31 Only one
side is engraved, with a ‘trophy on thunderbolt and dead
soldier, the inscription is an atypical vox magica: ΑΡΛΜΧΛΡΝ.
The Paris gem was slightly reworked after its production, since
its polished edge cuts through the inscription, its present size is
thus slightly smaller than the original (Pls 7a–b).32 It is a green
jasper. On one side we have an anguipede and the gorgoneion;
on the other the tropaion and a lion. The list may be
complemented with an unpublished dark green jasper gem in
the collection of the Hermitage (Pls 8a–b).33 It is about the
same size as the London gem, while the Paris example must
also have been this size. On its obverse it bears an anguipede,
below it the Trigrammaton inscribed in a tabula ansata; on the
reverse, the trophy and the lion.
The four pieces actually fall into two categories: the Hague
gem is only half the size of the others, only one side bears
decoration, the magical name is completely different, and so is
the style of engraving. The other three pieces, however, are
worth a closer inspection. Considering the motifs represented,
the three gems are variations on a theme. The composition
engraved on the obverse of the London gem appears on the
reverse of the other two, while the corpse beneath the lion is
lacking. The obverse of the two latter gems is a variation on the
reverse of the London piece; the Paris example also preserves
the gorgoneion; the Hermitage gem shows the Trigrammaton.
The voces inscribed on these gems do not belong among well-
known logoi. Despite their differences they are obviously not
independent from each other (Pl. 9).
The three gems are very closely connected in style (Pl. 10).
To start with the anguipede’s sides, the scheme of the rooster’s
Plates 8a–b Dark green jasper gem. St Petersburg, The State Hermitage
Museum, Inv. no. Ж.6742
78 | ‘Gems of Heaven’
Plate 9 Disposition of voces on the London, Paris and St Petersburg gems
head is identical. The engraving of the whip is also identical:
the handle rises up diagonally from the hand shaped like a
crab’s pincers, then hangs towards the forearm in a broken line.
The engraving of the shield is shaped like an inverse C in all
three cases. Instead of the porpax we can see long straight
lines. The snake’s legs were engraved by the master in similar
wavy lines. Both gorgoneia lean a bit to the right. They belong
to the same iconographic variant: the face is framed by locks of
hair, with double incisions at the sides to mark the snakes.
Above the temples on both sides are spread wings represented
in the same manner. On all three gems the trophy is a long
pole, strengthened on both sides by diagonal supports. The
representation of the helmets follows the same model. All the
shields are represented frontally, with a boss in the middle.
Other characteristic motifs of tropaion representations (such as
a spear or greaves) are missing.
The lion’s nose is in each case marked by a single and
separate incision cut diagonal and straight; the bristled mane is
indicated by 3-3 incisions. The hind legs of the animal form a
V-shape, the forelegs are parallel. The baseline is marked on
two gems: in both cases it is shown on one side of the trophy,
below the lion’s hind legs. The arrangement of the voces and
the stars is also similar.
Lastly, a few words about the technique of engraving. The
palaeographical connection between the three gems is best
attested by the incision of the epsilon. The anguipede’s
backward-leaning torso and high-raised arm also cuts out an
area of similar shape from the background.
Plate 10 The London, Paris and St Petersburg gems

There are two conclusions we can draw from the analysis of
these pieces. The first one is a suggestion: they may have been
created by the same hand. Based on the characteristic
iconographic feature, we can call him the ‘Tropaion master’.34
The second is a recognition which continues an old tradition of
Classical archaeology, namely that the other optimal way for
classifying magical gems is attribution.
The analysis of the three pieces also sheds light on the
daktyliographos – magos relationship. It is known that magical
gems were generally produced as follows: the daktyliographos
engraved the gem on the basis of a sketch given to him by the
magos, who had designed the praxis. These three gems
however point to a daktyliographos more creative than that,
who easily produced variations on the same basic type and who
produced three unique pieces from iconographic features used
like ‘Lego’ blocks. He could thus meet the contradictory basic
requirements set for talismans, still typical of the magic-
market. The vendor could sincerely reassure his customer: the
talisman you bought follows ancient knowledge, but is the
result of the newest improvements, and you are the only person
to have it! In general, it is clear that there are hardly any
absolutely identical pieces among magical gems (for the best
example, see below). The magos – daktyliographos relationship
is thus not at all that of ‘designer’ and ‘maker’. It is worth taking
an old suggestion of Hanna Philipp seriously: some makers of
gems were, it seems, expert designers as well.35
Analysis of iconographic syntax also offers new possibilities
for dating magical gems. Three phoenix gems (preserved in the
Taubman Medical Library, University of Michigan, Ann Arbor,
the Museum of Fine Arts, Budapest, and the Skoluda collection
in Hamburg: Pl. 11) belong to a frequently discussed type.36
Their composition is essentially identical, which makes them
exceptional in the extant corpus of magical gems. The type was
deciphered by Campbell Bonner: according to a story
mentioned by Herodotus (II 73) the phoenix buries his father’s
body in an egg kneaded from myrrh, then takes it to Heliopolis,
Plate 11 Haematite gems with phoenix: Ann Arbor (Michigan), Taubman
Medical Library, ex-Bonner 29; Budapest, Museum of Fine Arts, Classical
Collection, Inv. no. 55.154; Hamburg, Skoluda Collection, Inv. no. M095
Magical Gems and Classical Archaeology
where he places it in the sanctuary of the sun god.37 These gems
are engraved in their central zones with middle Egyptian
representations of the sun god: a crocodile, a phoenix on a
column which rests on an ovoid object, a scarab with
outstretched wings. The pairs of animals on the two sides
(bird, scorpion and snake) represent Isis and Nephthys.38
Further analysis attempted to reconstruct the Egyptian
background of the story, pointing out how thoroughly the
image engraved on the gem was thought through (for example:
the Egyptian name of Heliopolis is Iunu – City of Columns).39
The magos – the gem – commands the client’s stomach in the
name of the deity thus invoked to do what it has to: pepte –
digest! However, we can even go one step further. Herodotos
talks about an egg. The Michigan and the Hamburg gems also
show an ovoid incision complemented by two diagonal strokes
on each side. I assume that the two strokes refer to the akhet-
hieroglyph meaning ‘horizon’.40 The gem thus both evokes
Herodotos’ story, and complements the egg-motif with the
akhet-sign which generally represents the rising sun, thus
joining the two related motifs.
The third piece, the Budapest gem, is different: it shows a
regular sphere. Its maker used a motif well known in Roman
period iconography: a divinity holds a phoenix standing on a
globe, which was a symbol of the empire continuously reviving
itself. The daktyliographos then placed this scheme in a new
context, where the initiated viewer interpreted the globe as an
egg. The stylistic relationship is close between these pieces.41
The three gems – exactly because of the stylistic cross-relations
between them – are to be placed at the same chronological
moment, and are not copies of the same model from different
periods. If this is the case, then this otherwise identical
iconographic type had two contemporary variants: one with a
sphere, the other with an egg. The daktyliographos had taken
the globus – phoenix scheme known from imperial coinage,
and deconstructed it (to use a modern idiom). He gave it a
completely new meaning, and created two variants of it: one
more Greek, the other more Egyptian.42
The globus – phoenix scheme appeared in imperial coinage
in the age of Hadrian, first in a series minted in ad 121/122.43
This is thus the terminus post quem date for the phoenix gems.
It is worth noting again that representing the sun god as a
phoenix may have won a strange topicality in the age of
Antoninus Pius, when the return of the bird was expected
(ad 139) (Pl. 12).44
Plate 12 Detail of sestertius of Antoninus Pius; after AD 141. Hungarian
National Museum, Coins Collection, Bitnitz 358
‘Gems of Heaven’ | 79
Nagy
I believe that these archaeological methods offer
possibilities as yet unexploited for studying magical gems. As
mentioned before, both were developed by Campbell Bonner,
and perhaps this fact may also support their validity. A magical
papyrus would have recommended them as follows: ‘tested
methods for studying magical gems – invented by the greatest
symmagos’.
Acknowledgements
The English translation is by Kata Endreffy (Budapest), for whose
careful work I am much indebted; thanks also to Peter Agócs
(Cambridge) for reading a draft of the English text. I would also like to
express my deep gratitude to the Hungarian National Research
Foundation (OTKA) (K 81332), the Deutsches Archäologisches Institut
in Athens, the Ecole Française d’Athènes and especially the American
School of Classical Studies at Athens for their generous support.
Notes
1 ‘Colloquium on Magical Gemstones’, Malibu, Getty Villa, 13
February 2009.
2 The most significant recent publications are: S. Michel, Die
Magischen Gemmen im Britischen Museum, London, 2001; A.
Mastrocinque (ed.), Gemme gnostiche e cultura ellenistica,
Bologna, 2002; S. Michel, Die Magischen Gemmen, Berlin, 2004;
A. Mastrocinque (ed.), Sylloge gemmarum gnosticarum I, Rome,
2003, and II, Rome, 2007; E. Zwierlein-Diehl, Antike Gemmen und
ihr Nachleben, Berlin–New York, 2007, 210–31.
3 R. Gordon, ‘The power of stones: Graeco-Egyptian magical
amulets’, Journal of Roman Archaeology 21 (2008), 714.
4 J.J. Winckelmann, Geschichte der Kunst des Altertums, Vienna,
1776 (2nd edn), 95; new edn: A.H. Borbein et al. (eds), Mainz am
Rhein, 2002, 91.
5 For the story of the Berlin gems, see: H. Philipp, Mira et Magica,
Mainz am Rhein, 1986, 2–3. For the London gems: R. Gordon,
‘Magical Amulets in the British Museum’, Journal of Roman
Archaeology 15 (2002), 666–7.
6 The following are a summary of a study currently only accessible
online, and expected to appear in the Micrologus Library:
‘Daktylios pharmakites. Magical healing gems and rings in the
Graeco-Roman world’, in Ch. Burnett and I. Csepregi (eds), Ritual
Healing in Antiquity and the Middle Ages, Florence, 2011
(forthcoming; see http://www2.szepmuveszeti.hu/pegasos/
lexikon.php?id=91http://www2.szepmuveszeti.hu/pegasos/
lexikon.php?id=91).
7 The words talisman and amulet are used here as synonyms.
8 The currently used edition of the Lithika: R. Halleux and J. Schamp
(eds), Les Lapidaires grecs, Paris, 1985. List of the talismans
suggested in the Lithika: Á.M. Nagy, ‘Gemmae magicae selectae.
Sept commentaires sur l’interprétation des gemmes magiques’, in
Mastrocinque 2002 (n. 2), 170–6. See also S. Perea Yébenes, ‘Magic
at Sea: Amulets for Navigation’, in R.L. Gordon and F. Marco Simón
(eds), Magical Practice in the Latin West, Leiden/ Boston, 2010,
457–86, esp. 457–60 and 473–81.
9 Cf. M. Waegeman, Amulet and Alphabet. Magical Amulets in the
First Book of Cyranides, Amsterdam, 1987.
10 For the best summary for the ‘ep agathó, dos charin tó phorounti’
type gems, see: A. van den Hoek, D. Feissel and J.J. Herrmann,
‘Lucky Wearers: A Ring in Boston and a Greek Epigraphic Tradition
of Late Roman and Byzantine Times’, Journal of the Museum of Fine
Arts, Boston 6 (1994), 41–62.
11 Á.M. Nagy, ‘Egy kismestermű: a szentpétervári Perseus-gemma’,
in Gy. Rugási and B. Somlyó (eds), Grádicsok éneke. Tanulmányok
Tatár György 60. születésnapjára, Budapest, 2007, 68–77 (in
Hungarian).
12 C. Metzger in F. Baratte et al. (eds), Le trésor de Carthage :
contribution à l’étude de l’orfèvrerie de l’Antiquité tardive, Paris,
2002, 85–6, pl. II, fig.h.t. 5. For the Isis Pelagia-schema on a gem,
see: Philipp (n. 5), 63, no. 73.
13 Waegeman (n. 9), especially for the letters epsilon, eta, kappa and
nu, respectively 40–6, 56–64, 78–87 and 102–9.
14 Michel 2001 (n. 2), 291–2, no. 471. The peacock-snake motif without
magical names appears on gems as early as the Classical period:
the earliest example is a jasper scaraboid in the British Museum.
80 | ‘Gems of Heaven’
Most recently: H.C.L. Wiegandt, Die griechischen Siegel der
klassischen Zeit, Frankfurt am Main, 2009, Katalog 121, Epc2.
15 Inv. no. Ж.6673. Cf. Kyranides I 9, 12–16: ‘Engrave in the jasper a
kite rending a snake to pieces…, it stops every stomach complaint
and it permits one to eat a lot with a good digestion. Moreover it
has other powers: wear it on your chest and you will see’: from
Waegeman (n. 9), 71, s.v. iota.
16 Nagy (n. 6).
17 The first list was compiled by Hanna Philipp: Philipp (n. 5), n. 18 on
8–10; This was significantly expanded by Roy Kotansky, ‘The
Chnoubis Gem from Tel Dor’, Israel Exploration Journal 47 (1997),
ns 9–18 on 258–60. The latest supplement is: Michel 2004 (n. 2), 2,
n. 7.
18 As a starting point for the characteristically long ancient history of
gems, see: Zwierlein-Diehl (n. 2), 249–50.
19 That this possibility should always be considered was justified by
Roman period finds recovered in Zeugma, Anatolia a few years
ago: more than a 100,000 (!) gem impressions were found,
including seal impressions made with magical gems. The first
news of the finds is in: M. Önal, Clay Seal Impressions of Zeugma,
Ankara, 2007 (non vidi).
20 C.G. Heyne, Akademische Vorlesungen über die Archäologie der
Kunst des Alterthums, insbesondere der Griechen und Römer,
Brunswick, 1822, 520.
21 As a summary: W. Brashear, ‘The Greek Magical Papyri: an
Introduction and Survey; Annotated Bibliography (1928–1994)’, in
W. Haase (ed.), Aufstieg und Niedergang der römischen Welt II, 18:5
(1995), 3430.
22 ‘Die Abraxasgemmen’, Mitteilungen des Deutschen Archäologischen
Instituts, Abteilung Kairo 5 (1934), 119–43.
23 He was the first to discuss the question of pairs in the
supplementum of his book: ‘Amulets chiefly in the British Museum’,
Hesperia 20 (1951), 303–14 (‘Ancient Replicas and Modern
Imitations’). It is worth evoking the memory of the great American
scholar in another aspect as well. Today, in the time of industrial
academic publishing, Bonner could hardly have had a successful
scholarly career. He wrote a single monograph on magical gems,
and that too, only at the end of his fertile career. His work remains
the best synthesis of the subject even after half a century. For a
biography of Bonner, see: W.W. Briggs, Biographical Dictionary of
North American Classicists, Westport (CT) and London, 1994, 54.
24 A good starting point is C.A. Faraone, ‘Handbooks and
Anthologies: The Collection of Greek and Egyptian Incantations in
Late Hellenistic Egypt’, Archiv für Religionsgeschichte 2 (2000),
195–214. On the question of orality and literacy from a wider view-
point: G. Bohak, Ancient Jewish Magic, Cambridge, 2008, 143–4.
25 See the paper by Dasen in the present volume. I am grateful to her
for kindly allowing me to read it while still in manuscript.
26 Thus most recently C. Sfameni, ‘Magic in Late Antiquity: the
Evidence of Magical Gems’, in D.M. Gwynn and S. Bangert (eds),
Religious Diversity in Late Antiquity, Leiden–Boston, 2010, 445.
27 Michel 2004 (n. 2).
28 Ibid. 333, 53. Group 1.c.
29 Michel 2001 (n. 2), 179–80, no. 287.
30 By obverse I am referring to the larger, by reverse the smaller side
of the gem.
31 M. Maaskant-Kleibrink, Catalogue of the Engraved Gems in the
Royal Coin Cabinet, The Hague, The Hague, 1978, 355, no. 1119.
32 A. Delatte and Ph. Derchain, Les intailles magiques gréco-
égyptiennes, Paris, 1964, 210, no. 286. The pictures were kindly
sent to me by A. Mastrocinque.
33 Inv. no. Ж.6742. I am grateful for the kind assistance of E.
Arsentyeva in granting permission to publish the piece.
34 At the same time it also draws attention to a possible way of
making progress: by creating a common database of high
resolution images, gems now dispersed around the world would be
easily and freely accessible for study. This would make the mutual
analysis of style and technique possible, which is a sine qua non
prerequisite for this method.
35 Philipp (n. 5), 12.
36 The Ann Arbor gem: C. Bonner, Studies on Magical Amulets, chiefly
Graeco-Egyptian, Ann Arbor, 1950, 270, no. 103; the Budapest gem:
Nagy (n. 8), 162–9; the Hamburg gem: S. Michel, Bunte Steine –
Dunkle Bilder: >Magische Gemmen<, Munich, 2001, 83–4, no. 89.
I am indebted to the generous assistance of B. Shipman, who made
the study of the Ann Arbor gem possible.

37 Bonner (n. 36), 60.
38 Nagy (n. 8), 162–9.
39 Á.M. Nagy, ‘Le phénix et l’oiseau-benu sur les gemmes magiques’,
in S. Fabrizio-Costa (ed.), Phénix: mythe(s) et signe(s), Bern et al.,
2001, 57–84.
40 R.H. Wilkinson, Reading Egyptian Art, London, 1994, 135. I am
grateful for the kind help of Péter Gaboda in clearing up this
question.
41 Instead of detailed analysis, it will suffice here to draw attention to
a few things: the relationship of the crocodile, the column and the
birds, the similar composition of the reverse, the shape of the
Magical Gems and Classical Archaeology
letters. All these, however, are still insufficient to attribute the
gems.
42 This phenomenon is not unknown on magical gems: Horus may,
for example, have curly hair, like that of Eros, and he may also be
bald, wearing a single side-lock of youth – we have a Greek and an
Egyptian version. See Michel 2004 (n. 2), 70–1.
43 The largest collection of coins with a phoenix: R. van den Broek,
The Myth of the Phoenix according to Classical and Early Christian
Traditions, Leiden, 1972, pls 6–8. See also R. Volkommer, LIMC VIII
(1997), 987–90, s.v. Phoinix III.
44 See Nagy (n. 39), 71–2 and n. 49.
‘Gems of Heaven’ | 81
Studies on Magical Amulets in the British Museum
Simone Michel-von Dungern
Numbering about 650 items, the collection of magical amulets
in the British Museum is still the largest in the world. Eight
years ago my two-volume catalogue of this collection was
published and consequently, my doctoral dissertation appeared
in print in 2004, based on nearly 3,000 magical gems in
accessible collections in Europe, the USA, and in private
possession.1 After classifying this source material into groups
by comparing colours, materials, shapes, iconography, and
inscriptions, it is possible to re-examine the discussions or
interpretations of the basic motifs, individual themes, and
several details of the matter afresh to gain further
understanding. This paper will focus on the astrological
aspects of magical gems.
In the Papyri Graecae Magicae (hereafter PGM), astrology is
called ‘Mathematikos’ and predominantly covers the field of
individual astrology, which arose during the 3rd century bc.2
Astrology, and primarily decanmelothesia (see p. 83), also
played an important role in medical use in magical occult
doctrines. Since there is a close relationship between magical
gems and ancient texts on magic, such as the Greek magical
papyri, one can expect congruent aspects to come up on
amulets. Astrological aspects, how­ever, seem to appear seldom
in images and inscriptions of magical amulets. Of course the
zodiac is sometimes found on magical gems especially in com­
bination with the sun god, Greek Helios3 or Egyptian
Harpokrates (Pl. 1),4 where the sun god is seen as the ruler of
the cosmos, as represented by the signs of the zodiac. More
often the sun god is surrounded by different animals (Pl. 8)5
that might stand for the older Egyptian Dodekaoros system
which will be discussed below.
Plate 1 Brown jasper. Hamburg,
Skoluda Collection
Plates 2a–b Green jasper, 14 x 16 2mm.
London, British Museum, PE G 431
82 | ‘Gems of Heaven’
Some examples of gems associated with astrological signs
and horoscopes seem to be adapted to magical use later, for
example, by adding an inscription. A dark green jasper ring
stone in the British Museum’s collection shows a ram, jumping
to the left, and looking back in profile (Pls 2a–b).6 Above the
animal, a crescent moon is opened to the left with an eight-
pointed star inside. Underneath the belly of the animal is the
symbol of a circle surrounded by a three-sided rectangle. In the
field around the animal, one can read the letters of the magi­cal
name ‘(a)brasax’ but without the first letter, a capital alpha.
The symbol underneath the ram could be this missing letter as
well as the astrological symbol for ‘house’. Aries is the leader of
the zodiac in addition to being the constellation of the
beginning of spring (equinox). The zodiacal Aries is, in general,
usually shown like this: jumping in profile with its head turned
back to Taurus.7 This green jasper, however, is the only
example known to me of a magical gem showing this zodiacal
sign.
A yellow jasper also in the British Museum’s collection
shows a crab seen from above, orientated vertically to the top
of the stone (Pls 3a and 3c).8 On illustrations of the
constellation Cancer, the head and the claws point toward Leo,
and occasionally to Gemini. In circular compositions it is
directed towards the outer periphery.9 The animal inscribed on
the gem is flat and plump with four parallel pairs of legs angled
to the left and right. Two eyestalks are indicated on the head,
as well as the frontal extremities with claws. The body is
patterned with small bead-like indentations to create the visual
effect of a crab shell. In the field are a crescent moon opened to
the top and some letters. On the reverse, a Greek inscription in
Plates 3a–d Yellow jasper, 24 x 18 x 6mm. London, British Museum, PE G 123

Plates 4a–d Yellow jasper, 12 x 16 x 4mm. London, British Museum, PE G 415
three rows states: ‘Romula gave birth to Sosipatria’ (Pls 3b and
3d). Obviously Sosipatria was born under the sun sign of
Cancer. A peculiar aspect of this gemstone is its shape: the
obverse is flat, but the reverse is convex with a rounded bevel.
As a rule, gems with this crab motif are flat on both sides.10 The
animal is always shown from this perspective and direction.
The preferred material for this motif is always yellow jasper,
and in the majority of cases the magical word bapxai is incised
on the reverse as seen on a further gem in the British Museum’s
collection (Pls 4b and 4d).11 ‘Barcha(i)’ sounds a little bit like
the magical name ‘pchorbazanachay’ given for the constellation
Cancer by the PGM.12 Some academics assume it is an
association with the Persian word ‘Barza’ for ‘bright light’.13 In
the PGM ‘Barza’ is mentioned in alliance with Venus or the
goddesses of the moon, which are occasionally shown on
gemstones together with the crab.14 Cancer is the astrological
house of the moon and the PGM recommends the making of
phylacteries when the moon is in Cancer.15 These stones clearly
show that certain magical words of unknown meaning were
regarded as powerful for certain purposes and when used
under certain conditions. As a rule, color, material, design and
inscription all follow a consistent pattern. The frequent
combination with a specific stone material, color and magical
inscription implies a deeper magical content than just a
zodiacal sign or horoscope.
There is another group like this: yellow jaspers with the
picture of a scorpion and variations of the inscription
ôpthmenchiniambô(n) as seen on another gem in the British
Museum’s collection (Pls 5a–b).16 The scorpion always has
eight legs and a semi-circular tail ending with a stinger. On
amulets it is also depicted as seen from above, vertically
oriented to the top of the stone. The constellation of Scorpio, in
contrast, is normally directed to the right and faces Libra in the
West. It can also be found facing Sagittarius and sometimes
even the periphery. The constellation consists of the central
star ‘Antares’ on the body, three stars on the forehead, and
several bright stars creating the tail with the stinger. If a star or
a scale is above the claws of the scorpion, the meaning of the
picture may be astrological.17 The scale originally was
combined with the constellation Scorpio and was only later
Plates 5a–b Yellow jasper, 15 x 11 x 2mm. London, British Museum, PE G 180
Studies on Magical Amulets in the British Museum
separated to become its own sign.18 The scorpion may be the
horoscope sign of the wearer, but its frequent combination with
a specific stone material, colour and magical inscription again
implies a deeper magical content than just a horoscope. It is
probable that these amulets closely followed a prescription set
down in a common, much-used magical book. The Arabic
Picatrix, for example, or the lapidary of Socrates and Dionysios,
recommend yellow stones for protection against the painful
sting of scorpions.19 It therefore suggests that these stones may
have been intended for this purpose. On the principle of
‘similia similibus’ yellow jaspers incised with the picture of a
scorpion could have been a talisman against scorpion stings.20
In addition, there could also be a connection with the field of
medical astrology, as Campell Bonner has previously suggested
in his book, Studies in Magical Amulets, published in 1950.21
The field of medical astrology, especially decanmelothesia,
was based on the belief that the body’s ‘microcosm’
corresponded to the ‘macrocosm’ of the planets and stars that
gave order to the seemingly random course of life and health.22
The so-called Holy Book of Hermes to Asclepius assigns a
specific part of the human body to each constellation of the
zodiac,23 which governs and influences for good or evil, as well
as illustrated by diagrams called ‘Men of Signs’ in medical
books of the early 15th and 16th centuries. Each sign of the
zodiac represents an area of the body, starting with Aries
ruling the head, Taurus ruling the throat area, and so on. The
signs continue down the body in their natural order ending
with Pisces ruling the feet. To avoid a specific illness, the book
recommended cutting the shape and character of the
corresponding decan in the stone related to it, which was to be
worn as an amulet along with its herbs and plants. Since the
region assigned to Scorpio included the genital organs, it is
possible that the type previously considered was valued not
only as a protection against scorpions but also as a remedy for
sexual disorders and disabilities. An interesting fact is that
gems with images of other signs like Pisces, Gemini, Virgo, and
so forth, have never been found.
On diagrams, Leo is shown near the stomach, which
corresponds to occult doctrines where the first decan of Leo
rules diseases afflicting the area of the heart.24 The decan’s
name is Chnoumos and is described as having the face of a lion
with solar rays and its whole body as that of a serpent, coiled
and standing erect. It is recommended engraving this decan on
an agate, to set it in whatever one chooses with a piece of the
‘lion-foot’ plant underneath, and to wear it.25 In a second text,
preserved only in Latin, this description at the appropriate
juncture is offered:
The first decan of Leo has the face of Saturn. Its name is Zaloias. It
rules the stomach. It is a large serpent in the form of a lion; it (the
serpent) has solar rays around its head.26
‘Gems of Heaven’ | 83
von Dungern
Plates 6a-d Green jasper, burned, 15 x 12 x 3mm. London, British Museum,
PE G 397
Indeed, the Chnoubis-snake is a common motif on magical
gems: 33 in the British Museum’s collection27 belong to this type
(Pls 6a and 6c).28 The image is always accompanied by a
symbol of three S’s crossed through with a horizontal line (Pls
6b and 6d), which may be the astrological symbol of the decan
god which can originally be traced back to the Egyptian decan
Kenmet.29 In most cases, the stones used are greenish quartz
and milky chalcedony, convex on both sides and worn as rings
or pendants. A passage by the well-known Greek doctor
Claudius Galen from his book On the Composition and
Specificity of Simple Remedies (10.19; XII) reads as follows:
Some authorities ascribe special properties to particular varieties
of stone – special properties of the sort that the green jasper, for
example, actually does possess. Worn (about the neck) it heals the
stomach and the mouth of the womb. Some people also set the
stone in a ring and inscribe on it the radiate serpent, precisely as
king Nechepso prescribes in his fourteenth book.30
However, only six out of about 600 in­scrip­tions relating to the
stomach and the digestive system have been found on gems
with this Chnoubis motif. Therefore, again, the meaning of this
figure surely signifies something more than just the meaning of
a medical astrological decan god. New results are to be
expected from Joachim Friedrich Quack’s forthcoming book,
Beiträge zu den ägyptischen Dekanen und ihrer Rezeption in der
griechisch-römischen Welt.
A further question is whether gems with images of other
healing decans exist. Following the Latin list of decans of
Hermes Trismegistos, the second decan of Aries is named
‘Sabaôth’ and has the face of a sparrow hawk. In his right hand
he carries a jar of water and in his left, a sceptre with a sparrow
hawk sitting on the top. On his head he wears a lotus with stars
around it and he stands on a turtle. This description is
comparable with a group of yellow jaspers showing a similar
figure, but excluding the latter details.31 With the exception of
the Chnoubis snake, it seems that there are no other figures
shown on gems corresponding to the ancient writings from the
decans.32
It is likely that older Egyptian astrological systems
84 | ‘Gems of Heaven’
Plate 7 Diagram of the Dodekaoros after Teukros and the ‘Daressy Zodiac’
influenced ancient magic and magical amulets more than the
zodiac signs known today. Before we continue, we have to turn
back to an astrological doctrine called ‘Dodekaoros’ known to
us, for example, from the writings of the astrologer Teukros
from the 1st century bc, and also from a marble plaque sighted
and squeezed in Cairo about 1901 by the French Egyptologist
Georges Daressy (1864–1938).33 The object is designed with a
Roman-period Greek zodiac, on which the Greek zodiacal signs
are associated with animals. Two concentric bands enclose a
central area and 12 radial lines divide the bands into 12
individual sectors. The outer band has a depiction of the signs
of the Greek zodiac, and the inner band depicts the 12 animals
that are associated with the zodiacal signs according to the
Dodekaoros system.34 Following the description given by
Teukros, the pairs pictorially depicted (Pl. 7) are:
Aries (ram, with belt): cat (sitting).
Taurus (bull): dog (or jackal).
Gemini (twins, man and woman): serpent.
Cancer (crab): scarab or crab.
Leo (lion): donkey/ass.
Virgo (virgin): lion walking.
Libra (balance, borne by a man): goat (gazelle).
Scorpio (scorpion): bull/ox.
Sagittarius (archer, centaur): falcon.
Capricorn (goatfish): baboon/ape.
Aquarius (waterman): ibis
Pisces (fish): crocodile.
The PGM contain reminiscences of the Dodekaoros system, too,
but transferred the twelve animals to the hours of the day, and
considered them to be different shapes of the sun god changing
from hour to hour.35 Helios has the form of a cat in the first
hour, the form of a dog in the second hour, the form of a serpent
in the third hour, and so on. The order of the animals in every
hour of the day is in accordance with the Dodekaoros system
mentioned above (Pl. 7). On the con­trary, only some of the
animals mentioned above are actually depicted on magical
gems: goats, serpents, crocodiles, falcons, scarabs, lions, ibises,
and sometimes donkeys are shown (Pl. 8).36 In the way of
ancient Egyptian plural forms there are always three animals
to represent the entire species. Though the sequence of these

Plate 8 Haematite. Hamburg, E.
Sossidi Collection
animals is different from the Dodekaoros system, it always
remains the same on magical gems even if the placement is
shifted: sometimes the crocodiles are below, sometimes on the
upper left, but are always followed by the snakes, then the
goats, and so on. The circle of animals on magical gems seem
to have an astrological and topo­graphical significance: for
instance, scarabs symbolise the morning and the East, falcons
the noon, the air and the South, goats the evening, earth and
the North, snakes and crocodiles the night, water, the
underworld and the West.37 Nevertheless, as is foreshadowed
by the following examples of some associations and possible
astrological combinations, there are surely more motifs
containing astrological aspects on magical gems than those
recognised to date.38
A dark green jasper in the British Museum shows Eros
riding on a lion (Pls 9a and 9c).39 This image is common on
Roman ring stones but seldom used for magical gems. This
particular image of Eros is usually identified with Harpokrates,
the young sun god, who is also shown riding a lion on magical
gems with a large bundle of thunderbolts underneath.40 The
gem was probably adapted for magical use by adding an
inscription (Pls 9b and 9d). Slightly misspelled, it reads:
pyroeis iaô moui srô (planet) Mars, Iaô, lion, Srô (ram).41 Two
stars in the field characterise the astrological aspects of the
image on the gem, which are further supported by the names
of planets and decans inscribed on the reverse. According to
the papyri, Srô is the decan sitting on top of the head of
Cosmos.42 The meaning of ‘buck’ or ‘ram’ in pre-Greek times is
the name of the second decan of Capricorn, and in the Greek
horoscope the name of the first decan of Aries. Srô and also
moui occur in the similar inscription serphoutmouisrô
(‘Lotus, lion, ram’), which is one of the names of the Egyptian
Plates 9a–d Dark green jasper, 18 x 24 x 4mm. London, British Museum, PE G
255
Studies on Magical Amulets in the British Museum
Plate 10 Red-brown jasper.
Hamburg, E. Sossidi Collection
sun god, reflecting his shape in the morning (lotus), at noon
(lion), and in the evening (ram).43 On some amulets, the
inscription serphoutmouisrô occurs together with pictures of
the lotus blossom, the lion and the ram in combination with the
motif of a human figure whose head is replaced by seven snake-
like lines on his shoulders: Akephalos, ‘the headless’, a further
manifestation of the sun god.44 A brown-red jasper in a private
collection also shows the headless god perched on a lion
instead of the sun god (Pl. 10).45 On other gems with similar
motifs, the ‘headless’ is replaced by a baboon (Pl. 11a),46 usually
combined with a picture of Kronos-Saturn on the reverse, and
also with the commonly-used inscription normally found with
this god: the Sisisrô-Logos (Pl. 11b).47
The exchangeability of ‘the headless’ with the baboon as
well as the combination with Kronos-Saturn and Sisisrô-Logos
has astrological parallels. Both hermetic lists of decans place
Akephalos as first decan of Capricorn.48 According to the
Dodekaoros system, however, the baboon correlates with
Capricorn (Pl. 7). Furthermore, Capricorn/­A ke­pha­los/­baboon
is the day house of Saturn. The words Srô, Isrô, and Sisrôi
included in the Sisisrô-Logos are traded in the decan lists as
names for the second and third decan of Capricorn, or the
subsequent first decan of Aquarius.49 An example from the
Getty Museum shows a bearded male figure pouring water
from a jug in his right hand onto a bundle of lightning bolts in
his left (Pl. 12a).50 The figure is standing on a lotus blossom and
Plates 11a–b Greenish-brown-yellow jasper. Malibu, J.P. Getty Museum
Plates 12a–b Greyish stone, unidentified. Malibu, J.P. Getty Museum
‘Gems of Heaven’ | 85
von Dungern
Plate 13 Chalcedony. Ann Arbor,
Kelsey Museum
is flanked by a lion and a ram. On the reverse, slightly
misspelled, is the inscription: serphoutmouisrô (‘Lotus, lion,
ram’) (Pl. 12b). Following the Holy Book of Hermes to Asclepios
the attributes of the figure, the lightning bolts and water jug,
belong to the third decan of Gemini. This decan, according to
the book, should be engraved on heliotrope (bloodstone) and is
associated with the sun; his astrological house is Aquarius and
he also has a connection to Saturn, whose astrological night
house is Aquarius.51 According to the Latin list of the decans of
Hermes Trismegistos, Aquarius is responsible for the feet, and
his decans for podagra (gout).52 Some degrees of Aquarius – for
example as late as the 13th-century Astrolabium Planum by
Pietro d’Abano ad, the 2nd, 3rd, and 19th degree – are
associated with Perseus carrying Medusa’s head.53 Since the
inscription combined with this motif on a sardonyx from the
Hermitage reads ‘Flee po­­dagra, Perseus pursues you!’,54 it is
again brought into question if, regarding gems, iatromagical
contents and the after-effects of decanmelothesia are still
frequently misjudged.
I’d like to conclude with a final example of the astrological
aspects of magical gems and their modern after-life. Campbell
Bonner published a smoky-gray chalcedony, both faces convex
without bevel, from the Kelsey Museum, Ann Arbor (Pl. 13).55
The work is extremely crude but the design is identifiable:
Ophiouchos, oriented to the right with the man in front view,
agrees with pictures of this constellation. This so-called 13th
constellation is found between Scorpio and Sagittarius;
therefore the bird sitting on the figure’s head could possibly be
a falcon, the Dodekaoros­equivalent of Sagittarius (Pl. 7).56
Other examples with this design also show Ophiouchos
standing on a scorpion, signifying his position over Scorpio.57
In the field are eight-rayed ring signs, characters, and an
imperfect star; under the picture among indistinct letters the
seven vowels are arranged as a palindrome. The gem proves
the ‘Serpent-Holder’ to have been used as an amuletic device
and, as Bonner professed, basing himself on Manilius (5,
389–93) and Firmicus Maternus (8, 15, 1), was believed to give
protection against poisonous snakes.58 A haematite in the
British Museum set in a modern ring also shows this
Plates 14a–b Haematite, 20 x 17mm (bezel). London, British Museum PE G 388
86 | ‘Gems of Heaven’
constellation (Pls 14a–b).59 Campbell Bonner had already
recognised that the lapidary has taken the trouble to set, as
nearly as possible, the right number of stars in the proper
places on the man’s body following a traditional arrangement
closely resembling the description of the constellation by
Eratosthenes, with slight discrepancies relating to the arms.60
Since ancient descriptions also make clear that the snake’s neck
was held by the man’s left hand, Bonner also assumed that the
gem was meant to be viewed as an impression, but also
mentioned that amulets like this, with a few exceptions, were
meant to be looked at directly.
There is another reason for this right-left reversed image.
The model for this engraving was Hyginus’ Poeticon
astronomicon, one of the primary ancient literary sources on
the constellations. During the Renaissance the work was
usually attributed to the Roman historian C. Julius Hyginus,
but since the text describes 47 of 48 constellations in a similar
order to that of the catalogue in Ptolemy’s Almagest, it may
have been done by another Hyginus and date from the 2nd
century ad or later. The first formal edition was published in
1482 in Venice, Italy, by Erhard Ratdolt (Clarissimi uiri Hyginii
Poeticon astronomicon opus utilissimum), who also
commissioned a series of woodcuts depicting the constellations
to accompany Hyginus’ text.61 These figures provided the
prototypes for the constellations at the beginning of the
printing of star maps in the 16th century. The beautiful web
version of this book by the Linda Hall Library, Kansas City, also
adds hand-coloured woodcuts of an edition from 1485 (Pl. 15).62
In both editions, from 148263 and 1485, the ‘Serpent-Holder’
turns in a three-quarter position to the left holding the snake’s
neck in his right hand. In the later Venetian edition from 1488
by Tommaso de Blavi, however, the figure was shown correctly
right-left reversed, as described in ancient writings.64
Whenever the gem was engraved, the lapidary seems to have
followed the woodcuts of the older editions without any
detailed knowledge of astrology. Designs and inscriptions on
gems created in the modern period have continued in the
tradition of ancient amulets which were nurtured by
alchemists and magicians during the Medieval period and by
others throughout the centuries to our own day. Apart from
that, the field of the astrological aspects of magical gems is still
a rewarding desideratum for future research.
Notes
1 S. Michel, Die magischen Gemmen im Briti­schen Museum, 2 vols (P.
and H. Zazoff eds), London, 2001; S. Michel, Die Magi­schen
Gemmen. Zauberformeln und magi­sche Bilder auf ge­schnittenen
Stei­nen der Antike und Neuzeit (Habilitationsschrift; Studien aus
dem Warburg-Haus Band 7), Berlin, 2004; Reviewed: R. Gordon,
Plate 15 Hand-coloured woodcut, Kansas City, Linda Hall Library

Journal of Roman Archaeology 15 (2002), 666–70; R.W. Daniel,
Zeitschrift für Papyrologie und Epigraphik 142 (2003), 139–42;
J.F. Quack, Gnomon 76 (2004), 257–62; C.A. Faraone, American
Journal of Archaeology 111.1, Jan. (2007), Online reviews, books
(PDF); B. Overbeck, Sehe­punkte 6 (2006), no. 1 [15.01.2006, PDF];
E. Zwierlein-Diehl, ‘Les intailles magiques’, Pallas. Revue d´études
antiques 75 (2007), 249ff.
2 W.M. Brashear, ‘The Greek Magical Papyri: an Introduction and
Survey; Annotated Bib­li­ogra­phy (1928–1994)’, in Aufstieg und
Niedergang der römi­schen Welt 18,5 (1995), 3380– 684, at 3456. On
astrology and magic: J.F. Quack, ‘Zum ersten Astrologischen
Lapidar im Steinbuch des Damigeron und Evax’, Philologus 145/2
(2001), 337ff.; Michel 2004 (n. 1), 161, n. 810.
3 C. Bonner, Studies in Magical Amu­lets, chiefly Graeco-Egyptian,
Ann Arbor, 1950, 320, pl. 21, D 391; Michel 2001 (n. 1), nos 245–7,
150ff., pl. 35.
4 S. Michel, Bunte Steine – Dunkle Bilder: ‘Magische Gemmen’.
Ausstel­lungskatalog zur gleichnamigen Wander­ausstellung,
Munich, 2001, 31, pl. 3,18; Michel 2004 (n. 1), 271 (19.1.g).
5 Michel 2004 (n. 1), 274 (19.4.b), pl. 30,1 (for further examples, see:
75f., 273ff. (19.4); Michel 2001 (n. 1), nos 125–30, 133, 82ff., pls 18–19.
6 Michel 2001 (n. 1), no. 341, 215, pl. 50.
7 For the constellation in general: H.G. Gundel, ‘Zodiakos: Tier­k reis­
bilder im Al­­tertum. Kosmische Be­züge und Jenseits­vor­stel­lungen
im anti­ken All­tagsleben’, Kul­t ur­ge­schichte der Antiken Welt 54
(1992), 16ff., 27, 29, 68ff., 154.
8 Michel 2001 (n.1), no. 94, 62, pl. 13.
9 For the constellation in general: H.G. Gundel, Weltbild und Astrolo­
gie in den griechischen Zauber­papyri, Munich, 1968, 52ff.; idem (n.
7), 27, 70, 154.
10 Michel 2001 (n. 1), nos 93–6, 61ff., pls 13–14; Michel 2004 (n. 1), 303
(33.1).
11 Michel 2001 (n. 1), no. 95, 62, pl. 14.
12 H.D. Betz (ed.), The Greek Magical Papyri in Translation. Including
the Demotic Spells, Chicago, 1986; K. Preisendanz (ed.), Papyri
Graecae Magicae. Die griechischen Zauber­papyri, I and II, Leipzig,
1928 (hereafter PGM); 2nd edn. rev. by A. Henrichs, Stuttgart,
1973; III Index (by E. Diehl and S. Eitrem (1941), PGM VII 802ff.
13 Preisendanz (n. 12), PGM IV 2920, VII 802ff.; Betz (n. 12), 93, n. 368.
14 Michel 2004 (n. 1), 155ff., n. 808 and n. 812.
15 Betz/Preisendanz (n. 12), PGM VII 284ff., III 276ff.
16 Michel 2001 (n. 1), nos 342–4, 216f., pls 50–1; Michel 2004 (n. 1), 163f.,
331f. (51.1.a), col. pl. IV,3, pl. 66,1 (Getty Inv. no. 85.AN.370.32).
17 Michel 2004 (n. 1), 332 (51.1.b, 51.1.c).
18 Ibid., 163, n. 834–5; Gundel (n. 7), 20, 27, 31, 54, 72.
19 H. Ritter and M. Plessner (trans.), Picatrix: Das Ziel des Weisen von
Pseudo-Magriti, London, 1962, 32, 56; F. de Mely, Les lapidaires de
l’antiquité et du moyen âge, Paris, 1898, tome II 1: Les lapidaires
grecs 1, 177.
20 A. Delatte and Ph. Derchain, Les in­tailles ma­g iques gréco-égyptien­
nes, Biblio­thèque Na­tionale, Paris, 1964, 268ff.; A.A. Barb, Gnomon
41 (1969), 305, n. 4 and n. 5.
21 Bonner (n. 3), 77ff., 200; S. Eitrem, ‘Der Skorpion in Mythologie
und Religionsgeschichte’, Symbolae osloenses 7 (1928), 53ff., 470ff.
22 Michel 2004 (n. 1), 159f.; J.F. Quack, ‘Dekane und Gliederver­
gottung. Altägyptische Traditionen im Apokry­phon Johannis’,
Jahrbuch für Antike und Christentum 38 (1995), 98, 104ff., n. 49.
23 W. Gundel, Dekane und Dekanstern­bilder, Darmstadt, 1969 (2nd
edn), 374; Michel 2004 (n. 1), 160f.
24 Gundel (n. 23), 279, n. 1, 374ff., 388.
25 Gundel (n. 23), 376; H.M. Jackson, ‘The Lion Becomes Man. The
Gnostic Leontomorphic Creator and the Pla­tonic Tradition’, Society
of Biblical Litera­t ure Dis­serta­t ion Series 81 (1985), 82.
26 Gundel (n. 23), 381.
27 Michel 2001 (n. 1), nos 304–38, pls 45–50; Michel 2004 (n. 1), 166f.,
255ff. (11), pls 67–9; S. Michel, ‘Der Nycheya Bolbach-Logos. Zu
einer neuen magischen Formel und ihrer Bedeu­tung’, in A.
Mastrocinque (ed.), Gemme gnostiche e cultura ellenistica. Atti
dell‘Incontro di studio: Verona, 22–23 ottobre 1999, Bologna, 2002,
119–34; R. Kotansky and J. Spier, ‘The “Horned Hunter” on a Lost
Gnostic Gem’, Harvard Theological Review 88, Nr. 3 (1995), 315–37;
Jackson (n. 25), 74ff.; A. Mastrocinque, ‘Un’altra imagine
transculturale: Chnoubis’, in S. Estienne, D. Jaillard, N.
Lubtchansky and Cl. Pouzadoux (eds), Image et reli­g ion dans
l’Antiquité Gréco-romaine. Actes du Colloque de Rome 11–13
décembre 2003 (Collection du Centre Jean Bérard, 28), Naples,
2008, 391–7; see also the paper by Véronique Dasen in this volume.
Studies on Magical Amulets in the British Museum
28 Michel 2001 (n. 1), no. 305, 195, pl. 45.
29 Michel 2004 (n. 1), 168, n. 859; O. Neugebauer and R.A. Parker,
Egyptian Astronomical Texts, vol. III: Decans, Pla­nets, Constellations
and Zodiacs, Providence and London, 1969, 105ff., 157, 160.
30 Bonner (n. 3), 54f.; Jackson (n. 25), 77f.
31 Gundel (n. 23), 379; Michel 2004 (n. 1), 170, 266f. (16.2.a), pl. 27,3.4.
32 New results are to be expected from J.F. Quack’s study: Beiträge zu
den ägyptischen Dekanen und ihrer Rezep­t ion in der griechisch-
römischen Welt (forthcoming).
33 F. Boll, Sphaera, Leipzig, 1903, 6ff., 31ff.; the present location of the
‘Daressy Zodiac’ remains unknown; Daressy’s squeeze is now kept
in the French Archaeological Institute in Cairo: G.D. Thompson,
History of Constella­t ions, (G) Greek Constellations, 14: Sphaera
Barbarica and Sphaera Graecanica: http://members.optusnet.
com.au/gto­siris/page11-14.html#top.
34 H. Brugsch, ‘Die Kapitel der Verwandlungen’, Zeitschrift für
ägypti­sche Sprache und Alter­t umskunde 5 (1867), 21–6; Betz/
Preisendanz (n. 12), PGM III 500ff., IV 1596ff.; Gundel (n. 9), 3ff.;
Neugebauer and Parker (n. 29), no. 80, 103.
35 Betz/Preisendanz (n. 12), PGM IV, 1646ff.
36 For further examples see n. 5 above.
37 Michel 2004 (n. 1), 64ff., 74ff.
38 Ibid., 171ff.
39 Michel 2001 (n. 1), no. 257, 158f., pl. 37.
40 Red-brown jasper: Bonner (n. 3), 288, pl. 10, D 211.
41 Bonner (n. 3), 150, 200, n.78.
42 Betz/Preisendanz (n. 12), PGM IV 1011.
43 Michel 2004 (n. 1), 172, n. 885; R. Merkelbach and M. Trotti,
Abrasax. Ausge­wählte Papyri religiö­sen und magi­schen In­halts I
(Ab­handlungen der Rheinisch-West­fä­li­schen Aka­de­mie der Wis­
senschaften. Son­der­reihe Papyro­lo­gica Co­loniensia XVII/1),
Opladen, 1990, 79 on XII, 79 (Lit.); Betz (n. 12), PGM 35, n. 132;
Bonner (n. 3), 164f., 200; Preisendanz (n. 12), PGM VII 498, PGM
XXXVI 351.
44 Delatte and Derchain (n. 20), 45, n. 2, no. 43, 50f.; Bonner (n. 3),
200, n. 75; Michel 2004 (n. 1), 172f., n. 887, 302 (32).
45 Michel 2004 (n. 1), 173, n. 888, 322 (42.4.a), pl. 59,2.
46 Getty Inv. no. 83.AN.437.52: Michel 2004 (n. 1), 322 (42.4), col. pl.
III,8, pl. 59,1.
47 Michel 2004 (n. 1), 173, n. 889–90, 322 (42.4), for Sisisrô-Logos:
idem, 173, n. 890, 183f., n. 951, 304 (35.1.b), 322 (42.4), 329f. (48.1),
486f., 521. For Kronos see also, R. Kotansky, ‘Kronos and a New
Magical In­scrip­tion Formula on a Gem in the J.P. Get­t y Museum’,
The Ancient World 3 (1980), 29–32.
48 Gundel (n. 23), 166, 232, n. 2, 378, 382.
49 See n. 47 above; Gundel (n. 23), 77f.
50 Getty Inv. no. 83.AN.437.45: Michel 2004 (n. 1), 172, 345 (57.2.b), pl.
65,1.
51 Gundel (n. 23), 363, 375f.
52 Ibid., 286, 382.
53 Ibid., 147.
54 Sard­onyx (nicolo): O. Neverov, Antique Intaglios in the Hermit­age
Collection, Leningrad, 1976, no. 143, 80; see Nagy’s paper this
volume, pl. 1; Michel 2004 (n. 1), 174f., n. 899–901, 269 (18.2);
A. Önnerfors (ed.), Antike Zau­ber­spruche, Stuttgart, 1991, no. 30,
60f.; R. Kotansky, ‘Incantations and Pray­ers for Salvation on
Inscribed Greek Amulets’, in C.A. Faraone and D. Ob­bink (eds),
An­cient Greek Magic and Re­li­­gion, Oxford, 1991, 118f., n. 86; J. Spier,
‘Medieval Byzantine Magical Amu­lets and their Tradition’, Journal
of the Warburg and Courtauld Institutes 56 (1993), 45, 61, n. 163.
55 Bonner (n. 3), 234f., 313, pl. 19, D 353.
56 Gundel (n. 9), 4f.
57 Delatte and Derchain (n. 20), nos 383–4, 268; Bonner (n. 3), 233f.,
312, pl. 19, D 352.
58 Bonner (n. 3), 232ff.
59 Ibid., 232f., 312, pl. 19, D 351; Michel 2001 (n. 1), no. 346, 218f., col.
pl. V, pl. 51; Michel 2004 (n. 1), 164f., 328 (46.5), pl. 65,4.
60 Bonner (n. 3).
61 http://www.lindahall.org/services/digital/staratlases.
shtml#top: Hyginus, Poeticon Astronomicon, 1482, 1485 and 1512.
62 Browse the 1485 edition – D 6 recto. Serpentarius woodcut (http://
contentdm. lindahall.org/ cdm4/docu­ment.php?CISOROOT=/
star_atlas&CISOPTR=1659&CISOSHOW=1535).
63 Browse the 1482 edition – D 7 recto. Serpentarius woodcut (http://
contentdm.lindahall.org/ cdm4/docu­ment.php?CISOROOT=/
star_atlas&CISOPTR=1534&CISOSHOW=1410).
64 http://www.atlascoelestis.com/ara1488.htm, (Felice Stoppa).
‘Gems of Heaven’ | 87
‘Grylloi’1
Kenneth Lapatin

Grylloi (singular gryllos; or grylli, sing. gryllus) is the

conventional term used to denote hybrid figures that often
appear on ancient gems. There is no precise definition of a
glyptic gryllos, but such images are nonetheless
comprehensible as a class. Still, the difficulty of defining and Plate 1 P.-J. Mariette, Traité des
pierres gravées, II, Receuil des pierres
delimiting the category of grylloi may be one of the reasons gravées du Cabinet du Roy, Paris,
that these gems are often overlooked, despite being quite 1750, pl. 72: intaglio with conjoined
numerous, chronologically long-lived, and geographically heads of a man, lion, and ram
surmounted by a comic mask with
wide-spread. Grylloi are typologically varied and inherently elephant tusks and a dove, an eagle,
intriguing; they deserve and richly reward further attention – and, in the field, a pedum, caduceus,
and club. Compare with Pl. 2
collectively as well as individually.
I will discuss below the origins and history of the word
gryllos and its erroneous application to gems. A better – or
more neutral – way to define this imagery is as ‘hybrids’, or, if
one prefers to employ a more elevated vocabulary, ‘fantastic
mixtures’, ‘animal aggregates’, ‘polycephalous monsters’, and
even ‘tiermorphic symplegmata’; they have also been called
‘mask-animal gems’, ‘auspicious’ and ‘ludic’ motifs, and, more
Plate 2 Red jasper intaglio with
simply, ‘composites’.2 Of course, there are many composites and conjoined heads of a bearded man
hybrids in ancient imagery: centaurs and sirens, minotaurs and and lion surmounted by a comic mask
with a raven and eagle, and a pedum,
griffins, pegasoi and chimaeras. In fact, in many early modern
caduceus, and club in the field. 14 x
sources, such as Pierre-Jean Mariette’s 1750 treatise on gem- 11mm. London, British Museum, GR
carving, what we call a gryllos is called a ‘chimere’ (Pl. 1). But 1814, 0704.1455
while a gryllos might be called a chimaera, a chimaera3 isn’t a
gryllos, nor are the other kinds of mythological hybrids
mentioned above, nor even less common hybrids, like Abraxas
(on which see other contributions to this volume), which have
also occasionally been linked to the grylloi.
The simplest of these hybrids – iconographically – are
combinations of two or more heads or, perhaps better, faces:
either human or humanoid with one another, disposed
circularly (Pl. 3) or superimposed; animal with animal (Pl. 4); Plate 3 Red jasper intaglio of four
conjoined bearded heads. Atlanta,
or, most commonly, human or humanoid with animal (Pls Michael C. Carlos Museum at Emory
5–6). These images are often playful in spirit, as diverse University, 2008.031.076
elements seem to take on an alternate life, and many have a
Boschian or Arcimboldo-esque quality. (On the associations of
these and other early modern artists with ancient grylloi see
further below.) Of humanoids on glyptic grylloi, silens and
satyrs (both bearded and beardless) are quite common, as are Plate 4 Amethyst intaglio with
masks, and of animals, birds (eagles, cocks, peacocks, ibises, conjoined heads of bridled horse,
goat, and boar, surrounded by a snake
etc.) appear most frequently, but horses, boars, goats, snakes,
with tail in its mouth. 14 x 17mm.
lizards, dolphins, dogs, and even elephants, are not London, British Museum, GR 1824,
uncommon. Images, moreover, are often blurred or hidden, for 0301.93
example, the third face at the bottom of a red jasper now in
Atlanta, whose thin beard is the tusk of the elephant (Pl. 7).
Not only are combinations of multiple faces prevalent, but also
admixtures of additional elements. Heads of horses often
sprout atop a satyr’s head, while a ram’s or boar’s
conventionally extends backwards, and the whole is perched Plate 5 Jasper (?) intaglio with
conjoined heads of eagle and silen.
on bird legs, so that these hybrid creatures frequently are fully Atlanta, Michael C. Carlos Museum at
formed – even if extremely multifarious. In fact, tails of various Emory University, 2008.031.066

88 | ‘Gems of Heaven’
 ‘Grylloi’

Plate 6 Red Jasper with conjoined
heads of eagle and silen. Atlanta,
Michael C. Carlos Museum at Emory
University, 2008.031.067
Plate 7 Red jasper with conjoined heads
of an elephant holding a caduceus in its
trunk, a silen, a youth, and a bearded
male. Atlanta, Michael C. Carlos Museum
at Emory University, 2008.031.112
Plate 8 Green jasper with conjoined
heads of a bearded man, ram, and
horse on bird legs; ears of wheat in
ram’s mouth; star and crescent moon
in field. Atlanta, Michael C. Carlos
Museum at Emory University,
2008.031.074
Plate 9 Black jasper with
conjoined heads of a bearded
man, ram, and horse on bird legs;
purse in ram’s mouth; caduceus in
field. 10 x 9mm.London, British
Museum, GR 1814,0704.1457

kinds are sometimes also included, and these can be combined
with yet other animal parts (Pls 8–11). In short, glyptic grylloi
can join elements of as few as two creatures – often only faces –
or more than a dozen components, especially if we count
further attributes held in the mouths or trunks of the many
animals, such as sheaves of wheat, wreaths, cornucopia, a
thyrsus or caduceus, and so on.
The origins of glyptic grylloi seem to lie in the eastern
sphere, and scholars have identified a series of what we might
call proto-grylloi in Achemenid, Carthaginian, Sardinian, and
late Archaic east Greek glyptic art (Pl. 12).4 Other grylloi, which
stand on bird legs and are topped with the heads of horses,
recall earlier images of hippalectyra (an outlandish hybrid
mentioned in a fragment of Aeschylus’ lost Myrmidons (61) and
in three plays of Aristophanes: Peace 1175; Birds 800; and Frogs
932). Although such iconography is common in late Hellenistic
and Roman glyptic, it is important to recognise that it also
exists in other media, such as Greek vases5 and especially
Hellenistic and Roman wall painting and other decorative
arts,6 to which I shall return. For now, I want to highlight that
the iconography is rich and diverse, though there are certainly
some patterns. Additionally, I must stress that my work on this
material remains in its preliminary stages. What I present here
are not conclusions so much as possible directions for further
study – strategies for approaching this class of material, as well
as individual objects.7 I must also emphasise that I do not by
any means pretend to have collected the entire corpus of
glyptic grylloi, but that my observations emerge from perusal
of the catalogues of various collections and other literature.
Thus, without claiming to have compiled a statistically valid
corpus, it seems that some types of stones were used more
frequently than others, and we should keep in mind the
potential value of correlating this information with what has
been presented in other papers in this collection, especially
those concerning the so-called magical gems.8 Red jasper
(26%), sard (20%), and carnelian (18%) appear to be the stones
most frequently used for glyptic grylloi, although these
numbers will necessarily change as more data is analysed.
Sardonyx (10%) and nicolo (8%) appear less often; and
amethyst (4%), chalcedony (2%), agate (2%), lapis lazuli (2%),
and various other stones (8%) are not so frequently employed.
My study to date also suggests something about the frequency
of individual iconographic elements, though the following
numbers, too, must not be taken as gospel: birds (58%) appear
most frequently, followed by male heads (42%), heads of silens
and rams (each 34%), horses (28%), boars (14%), goats (10%),
helmets (8%), dolphins, lizards, and serpents (each 6%),
elephants and dogs (each 4%).
Now, what are we to make of specific images on these
gems? As Erika Zwierlein-Diehl9 and others10 have observed,

Plate 10 Amethyst with Plate 11 Heliotrope intaglio with conjoined heads of a silen and Plate 12 Clay seal impressions from Ur, with heads of men
conjoined heads of a silen, ram surmounted by horse ridden by Eros all above an eagle conjoined with goats, rams, lion, eagles and ducks. London,
unicorn, and long-beaked bird tearing at a dead rabbit; another rabbit, chased by a dog, in British Museum
on bird legs; ear of wheat in mouth of horse and ear of wheat in mouth of ram; with mouse,
silen’s mouth; mouse in lizard, moon and stars in the field. Derek J. Content collection
unicorn’s mouth; spear, staff,
star, and three balls in field. 15
x 11mm. London, British
Museum, GR1814,0704.1460

‘Gems of Heaven’ | 89
Lapatin
Plate 13 Roman marble relief Plate 14 Intaglio depicting various
depicting various creatures attacking creatures attacking the Evil Eye in
the Evil Eye, in J. Russell, Duke of Gori (n. 23), pl. 95, no. 8
Bedford, Outline Engravings and
Descriptions of the Woburn Abbey
Marbles, London, 1822, pl. 14
interpretation can sometimes seem quite simple: on a mid-1st-
century bc carnelian in Vienna,11 for example, an eagle with a
crown in its beak might symbolise Victory; a silen mask joyous
festivals; ram’s head and ear of wheat plentiful meat and bread;
a caduceus success. Some have explained the silens as a form of
protective oscillum, and cocks as solar symbols. Thus one
possible interpretative path for the glyptic grylloi is to read
them rather like those gems carved with unconnected luck
symbols, such as a slightly later carnelian, also in Vienna.12 But
can the images present on glyptic grylloi all be explained so
simplistically? What about others, such as elephant heads and
mice; bells and phalloi; or different combinations?
Some scholars13 have adduced a remark of Plutarch about
the active power of vision in regard to the Evil Eye:
What I have said shows why the so-called amulets (τὸ τῶ̄
λεγομένων προβασκανίων γένος) are thought to be a protection
against malice. The strange look of them (διὰ τὴν ἀτοπίαν τη̑ς
ὄ̓ ψεως) attracts the gaze, so that it exerts less pressure upon its
victim (Quaestiones conviviales, 5.7.681F).14
Certainly some of the elements on glyptic grylloi may function
this way, and birds, crabs, scorpions, lizards, snakes, lions, and
dogs, for example, appear in marble reliefs intended to protect
against the Evil Eye and similar imagery also features on gems
(Pls 13–14). Most glyptic grylloi, however, seem to be neither
modes of distraction, nor on the attack, though they are, to be
sure, a conglomeration of elements – and frequently strange
ones at that. Perhaps a more useful approach is to move beyond
reductive one-to-one equivalents and consider more fully the
implications of the hybridity that lies at the very heart of the
grylloi. But before doing so, I want to return to nomenclature
and consider some approaches to these objects
historiographically.
Pliny the Elder, in the 35th book of his encyclopaedic
Natural History (XXXV.37 [114]), mentions a painting by the
4th-century bc Alexandrian painter Antiphilos, a rival of
Apelles: ‘idem [Antiphilus] iocosus nomine Gryllum deridiculi
habitus pinxit, unde id genus picturae grylli vocantur’ (‘poking
fun, he painted a person named Gryllos in a ludicrous costume,
whence pictures of that kind are called grylloi’). To my
knowledge, this passage is first explicitly linked to gems by the
extremely well-read J.J. Winckelmann, who uses the term
‘Grylle’ as an alternate to ‘Chimére’ in his unillustrated
90 | ‘Gems of Heaven’
catalogue of the collection of Baron von Stosch.15 Most of the
gems we would call grylloi are listed in his seventh class,
‘Animals’, under the sub-heading ‘Animaux Chimériques’,16 but
five are placed earlier, in his second class: ‘Sacred Mythology’:
667. Jaspe rouge. Un Amour monté sur une espéce de Chimére, ou
de Grylle, composé des parties de differents animaux & d’un grand
masque. Selon… Pline, Antiphile fut l’inventeur de ces sortes de
monstres que l’imagination avoir inventés dans la peinture des
anciens: Antiphilus. …iocoso nomine Gryllum ridiculi habitus
pinxit, unde hoc genus picturae Grylli vocantur.
Numbers 668–71 depict the same subject ‘plus un Trident’ and
‘avec un Liéve sur la barb du masque’.
Winckelmann does not comment further on the passage
from Pliny, but the name Gryllos is well-attested in other
ancient texts. It belonged, for example, to Xenophon’s father
and one of his sons, and also appears in several inscriptions
(Xenophon, Hellenika 7.5.15–17; Diogenes Laertius 2.48, 52–55;
Pausanias 1.3.4, 5.6.5, 8.9.5, 8.11.6).17 Its connotations, however,
are far from noble: it means ‘grunter’ and hence ‘pig’. Some
scholars, therefore, have suggested that the specific Gryllos in
Antiphilos’ painting was given a pig’s head, apparently taking
Pliny’s habitus to mean ‘condition’ rather than ‘costume’, and
one is reminded of the famous Roman parody painted on the
wall of a villa near Stabiae of a dog-headed Aeneas escaping
Troy.18 Other ancient literary sources, such as Phrynichos of
Bithynia, Philodemos, Plutarch, and John Chrysostomos also
mention grylloi. While Plutarch’s essay, popularly titled
Γρύλλος, features Odysseus in dialogue with one of his men
whom Circe turned into a pig, Phrynichos (Sophistic
Preparations 58B) distinguishes between the meanings of the
root with one lambda versus two: γρυ̑λος is ‘pig’, and γρυλίζω
is ‘to grunt’, while γρύλλος is the performer of an Egyptian
dance, the γρύλλισμος. Still, the form with two lambdas seems
to have been applied not only to dancers (perhaps wearing pig
masks?), apparently in fashion in Alexandria, but also images
of misshapen figures, and might have denoted a kind of
grotesque, like dwarf or pygmy dancers, such as those depicted
in bronze statuettes recovered from the ancient shipwreck off
the coast of Mahdia in modern Tunisia, or other misshapen
figures. Thus, perhaps, Philodemos (Rhetorika 2.297S) defines
γρυλλογραφει ̑ν in contrast to καλοὺς δημιουργει ̑ν πίνακας i.e.,
‘to draw caricatures’. In any event, there seems to be
considerable confusion in the Greek about what the term
gryllos means: a pig, a grunter, a kind of dance, a performer of
that dance, a caricature. Indeed in Oxyrhynchus Papyrus 2331
the term, with two lambdas, seems to refer to accompanying
illustrations and thus to signify ‘cartoon’.19
To complicate matters further, in Latin gryllus or grillus
means cricket or grasshopper; likewise grillo in modern Italian.
Thus ancient images of these similar-looking insects, in various
media, have been associated with the term. A wall painting
found in Herculaneum in October 1745 (Pl. 15) that depicts an
inversion of the natural order – a parrot pulling a grasshopper
in a chariot (rather than eating it) – was explicitly associated
with Pliny’s reference to Antiphilos’ painting in the first volume
of Le Antichità di Ercolano Esposte.20 Moreover, it has been
suggested that some 200 years earlier, in the 1520s, when
Michelangelo drew two crickets on a sheet with various
caricatures, he was referring to the paintings of Antiphilos.21
Certainly by the mid-16th century the term was current in

Plate 15 ‘Grillo su biga trainata da pappagallo’, wall painting from
Herculaneum, engraved in Delle Antichità di Ercolano I, Naples, 1755 [1757],
257, pl. XLVII
artistic circles: Don Felipe de Guevara called a work by
Hieronymous Bosch a ‘grillo’, referring explicitly to Antiphilos;
Antony van Dyck labeled Adriaen Brouwer a ‘grillorum pictor’
in his Iconographia of 1632–44; the term was also used to
characterise the works of other artists, such as Pieter Bruegel
and Arcimboldo; and some of the jokes attached to the
Pasquino in 17th-century Rome were also called ‘grilli’.22
In the second volume of his massive compendium of gems
in Florence, A.F. Gori23 uses the term Grylli to denote images of
insects, i.e., ‘Locustae, & Grylli citharoedi, sistulae cantores…’
(e.g. Pl. 16). Such images are also included in Tassie’s late
18th-century compendium24 (Pl. 17), where they are called
‘Sauterelles & Cigales’ (‘Grasshoppers & Locusts’), neither grylli
nor grylloi. What we consider grylloi, Tassie and Raspe call
‘Chimeras, symplegmata, or antics’.25 Long before Walt Disney’s
Plate 17 Plaster impressions of 12 gems depicting grasshoppers and crickets, some engaged in human activities, included in James Tassie and Erich Raspe, A
Descriptive Catalogue of a General Collection of Ancient and Modern Engraved Gems, Cameos and Intaglios, Taken from the Most Celebrated Cabinets in Europe…,
London 1791, 143–4, 705–6, nos 1906–11, 13341–6
Plate 18 Sard with a grasshopper in a Plate 19 Sard with a grasshopper
chariot drawn by two butterflies. 7 x wearing a lion-skin and playing on a lyre.
11mm. London, British Museum, 13 x 11mm. London, British Museum,
GR 1814,0704.1446 GR 1814,0704.1447
‘Grylloi’
Plate 16 Sard intaglio of a gryllo carrying
two baskets past a sundial mounted on a
pillar in Gori (n. 23): pl. 96, no. 4
Jiminy Cricket, who was modeled on the grillo parlante, a
minor character in Carlo Collodi’s 1883 Le avventure di
Pinocchio, which presumably drew on the long Italian tradition
mentioned above, insects on gems were given human
attributes. In addition to those collected by Gori and by Tassie
we might cite three Roman examples in the British Museum
(Pls 18–20).26 Composite grylloi, in our sense, that feature
entomological components, are less common. One might cite a
siren with a grasshopper’s body and griffin-headed tail on a
gold ring from Bliznitsa now in the Hermitage;27 a locust with
the head of Pan carrying a fish and two hares accompanied by
a scorpion and serpent on a late Hellenistic banded agate in the
Fitzwilliam Museum;28 and a grasshopper with the head of a
ram carrying two stalks of wheat as well as a cornucopia
whence emerge an ibex and a bee on a yellowish chalcedony in
the Metropolitan Museum.29 There are also some examples on
which insect wings frame the face of humanoid figures (Pl. 21).30
Plate 20 Sard with a grasshopper with Plate 21 Sard with a face
an Amazon’s shield and an axe. 15 x framed by insect wings. 9 x
13mm. London, British Museum, 8mm. London, British Museum,
GR 1814,0704.1448 GR 1923,0401.324
‘Gems of Heaven’ | 91
Lapatin

Kaufmann has suggested that Arcimboldo was familiar

with Rudolph II’s collection of ancient gems,31 and Bosch’s
composite creatures are in many ways similar to hybrid glyptic
grylloi, but I have found no evidence of the term gryllos applied
to ancient gems depicting hybrids, rather than caricatures,
before the 18th century. When Abraham van Goorle first
published his collection of engraved gems and finger rings in
Antwerp in 1601 two of his intaglios (nos 46 and 186, Pl. 22),
which we would call grylloi, were not categorised. Later
Plate 22 Left: agate with conjoined heads of two bearded males and an
publishers of his collection, in Latin (1707) and French (1778), elephant holding a trident in its trunk; Right: sard with conjoined heads of a
however, do describe and interpret these gems. In Jacob silen and horse with a sheaf of wheat in its mouth on bird legs; both in Abraham
van Goorle, Dactyliotheca, seu, Annulorum sigillarium quorum apud priscos
Gronovius’ Latin edition, number 46 is called ‘Fascinum ex tam Graecos quam Romanos usus e ferro, aere, argento & auro promptvarivm:
duobus humanis vultibus & elephantino, cuius proboscide accesserunt variarum gemmarvm quibus antiquitas in sigillando uti solita
tenetur tridens, forsan contra pericula maris: modo singula bene scalptvrae, Antwerp, 1601, nos 46 and 186

sint expressa…’, while in the later French edition it is merely

‘Figures Maritimes’.32 Number 186 is described in 1707 as
‘Fascinum ex senili & equino capite, caudaque & pedibus
gallinacei…’ with further references to Jean Chifflet’s
monograph on Socrates (on which see below), while in 1778 it is
characterised as ‘Têtes de vieillard & de cheval, pieds & queue
de coq’.33 Other early authors call grylloi ‘Chimeras’, as did Plate 23 Carnelian with conjoined
heads of a youth with a boar, in
Mariette and Winckelmann, but Leonardo Agostini, writing c. Agostini (n. 34), pl. 13
1667, separated one out from the crowd, cleverly identifying it
as Meleager on account of conjoined human and boar’s heads
(Pl. 23).34 This identification is repeated by several authors,
including Paolo Alessandro Maffei,35 who does not follow his
predecessor indiscriminately, however. He calls one of the
grylloi Agostini labled a ‘Chimera’ an ‘Abraxas’ (cf. Pls 24–25).36
Likewise Johann Jakob Baer37 questions the identification of Plate 24 Nicolo with conjoined heads
this figure as a chimaera, as it does not conform to ancient of silen, cock and ram on bird’s legs
trampling a dolphin with palm branch
descriptions of that beast as having a lion’s body, a goat coming
sprouting from its tail; rabbit in ram’s
from its torso, and a serpent tail; he proposes instead that it be mouth; cornucopia behind cock in
called an amulet or Abraxas: Agostini (n. 34), pl. 243

secundum fabulam antiquam, ex capite & pectore leonine, ventre

caprino & cauda draconis compositum portentum referre debeat:
Sed in nostrate nec leonine nec serpentini quicquam apparet.
Many late 17th- and 18th-century authors, like Gronovius,
invoke the name of Socrates when describing glyptic grylloi,
following Jean Chifflet, whose learned treatise, Socrates, sive,
De gemmis eius imagine coelatis iudicium (1662, Pl. 26) was
much read and praised. Chifflet cites ancient authors whose
descriptions of the Athenian philosopher associate him with Plate 25 The same intaglio as in Pl.
Silenus and thus identifies the bearded heads on gems as 24, reversed and relabeled ‘Abraxas’
in Maffei (n. 35), pl. 20
Socrates. He explains other heads conjoined to the
philosopher’s as those of his followers or enemies. Hence some
grylloi were interpreted as representing Socrates and the
young Alcibiades (Pl. 27), or Socrates and his wife Xanthippe (a
silen with a comic mask, Pl. 28), or Socrates and Athena above
the visages of his accusers Anytus, Meletus, and Lycon (Pl. 29).
Chifflet admits a greater difficulty interpreting gems with an
elephant head, whether holding a caduceus or palm in its trunk
(Pl. 30), but he sees the beast’s ugliness as analogous to the
philosopher’s and cites the elder Pliny’s comments regarding
the virtues and probity of the animal.38 Chifflet also explains
the presence in other gems of avian and other animal elements
(Pl. 31), although apparently ludicrous (‘Ludicrum, ut apparet
emblema’), by quoting Socrates’ famous last words, recorded in
Plato’s Phaedo (118a), that he owes a cock to Asclepius. Eagles,
meanwhile, signify victory, and ram’s and horse’s heads, too, Plate 26 Plate 6 and title page of J. Chifflet, Socrates, sive, De Gemmis euis
are ascribed special meanings with the authority of ancient imagine coelatis iudicium, Antwerp, 1662

92 | ‘Gems of Heaven’
 ‘Grylloi’

Plate 27 Chifflet (p. 26), pl. 1, no. 4: Plate 28 Chifflet (p. 26), pl. 2, no. 8: Plate 29 Chifflet (p. 26), pl. 2, no. 6: ‘Socrates and Athena’:
‘Socrates and Alcibiades’ ‘Socrates and Xanthippe’ ‘Ex imo Palladis dependent Melitvs, Anytvs, & Lycon’

literary sources. Chifflet even cites Pliny’s passage on

Antiphilos, though not to identify the gems as grylloi, but
rather as evidence for the existence of caricatures in antiquity.
Other learned commentators also invoke ancient literary
texts to explain the imagery of the grylloi. In his early
18th-century Gemmae selectae antiquae (Pl. 32), Jacob de
Wilde39 makes reference not only to images of Socrates, but also
to Abraxas, and the opening lines of Horace’s Ars Poetica:
GEMMA CXXXVI. Caput Socratis junctum capitibus hominis &
equi habens pedes galli gallinacei; similem fere habes Macarii
Abraxan pag. 36. & Horat. de arte Poëtica v.1.
A.F. Gori40 continues to use the term ‘monstrum
Chimaericum’, translating it into Italian as ‘mostro, o Chimera’
and, like Wilde, quotes the description of a composite beast at
the opening of Horace’s Ars Poetica, although he, too, notes that
the iconography does not match that of the gem he describes
(Pl. 33). In an earlier work, however, Gori only references the
beast fought by Bellerophon.41
Gori was a prolific author. It was he who called gems
engraved with grasshoppers grylli, and he is also the first
author I have encountered to use the term grylloi to describe
Plate 30 Chifflet (p. 26), pl. 4 our hybrid gems, although he cites neither Pliny nor any other
authority beyond, it seems, common usage (‘ut aiunt’). At the
end of his ‘third class’ of gems in Florentine collections he
treats in a single paragraph 27 grylloi that are illustrated on
three and a half large plates drawn by Domenico Compaglia
and engraved by Baldassare Gabbuggiani (Pls 34–35):42
Reliquae vero Gemmae aenigmaticae in eadem Tabula ex-
pressae, & in sequentibus quae huic Classi finem imponunt, quae,
ut aiunt, Gryllos, & Griphos, seu imagines monstrosas humano-
rum capitum vel Socratis, vel feminarum brutis animalibus vel
aeris, vel terrestribus, vel marinis, sive aquatilibus implexas
praeferunt, alio pertinent; atque, ut pereruditi viri senten-
tia est, ab aliquo virorum, fortassis superstitiosorum, Phratrio,
sive Secta originem habuere, quae portentosis hisce symplegmati-
Plate 31 Chifflet (p. 26), pl. 6, nos 21–22
bus ostendere voluit varios hominum affectus, virtutes, & vitia
quae proxime bestiis accedunt. Neque vero id erit suasu diffi-
cile, quum omens sciant Aesopum suis confictis Fabulis Philo-
sophiae moralis dogmata tradidisse. Hos vero portentorum au-

Plate 32 Carnelian with conjoined Plate 33 Cameo of a ‘Mostra, o

heads of silen, bearded male, and horse Chimera’ in Gori (n. 23), pl. 68,
on bird legs, described as ‘Abraxas’ in compared in the text to a hybrid
J. de Wilde, Gemmae selectae antiquae image described in the opening of
e museo Jocobi de Wilde, Amsterdam, Horace’s Ars Poetica
1703 [1707], no. 136

‘Gems of Heaven’ | 93
Lapatin
Plate 34 Diverse ‘Grylli’ in Gori (n. 42), pl. 50
ctores Aegyptios imitatos fuisse facile credam, quibus hic bel-
luarum, & hominum complexus familiaris fuit ad exprimendas
virtutes, & vitia; quod prae aliis docet eximius locus apud
Porphyrium, quem laudat summus vir Spanhemius in Disserta-
tione V. de Prestantia, & Vsu Numismatum antiquorum; qui
etiam duos nummos Nicomediensium adfert,in quibus draco-
nes cum humano capite sculpti sunt. Adeundus, si quis plura cu-
piat, Ioannes Chifletius in Socrate, qui recondita haec aenigma-
ticarum Gemmarum emblemata illustrat.
Gori cites Chifflet, among others, and presents many possible
lines of interpretation of these gems, from superstitious
amulets to illustrations of animal fables, like Aesop’s. But as
Karl August Böttiger noted in 1804, the term gryllos, when
applied to hybrid figures that appear on gems and elsewhere, is
erroneous and anachronistic. The situation is similar to what
classical archaeologists today conventionally call the ‘Daidalic’
style of 7th-century bc Greek art, which actually has little to do
with the legendary Bronze Age craftsman Daidalos. As Jeffrey
Hurwit43 observed, ‘the style is called Daidalic only because it
has to be called something, and the name…happens to be
convenient’. The anecdote recorded by Pliny, Natural History
XXXV.37 [114], appears to be another example of a literary text
falsely adduced as a key to an image – or group of images. And
not only does the ancient term gryllos not apply to our material,
it fails to provide a useful window to interpretation, for it
overlooks the hybridity at the core of our corpus.
Two other passages in ancient literature are also often
invoked in discussions of glyptic grylloi, and although they do
not specifically use that term, either in Greek or in Latin, they
are, at least, somewhat relevant because they do, at least, deal
94 | ‘Gems of Heaven’
Plate 35 Diverse ‘Grylli’ in Gori (n. 42), pl. 51
with the depiction of hybrid creatures. Vitruvius, writing for
the Emperor Augustus, famously complains about ‘unnatural
things’ (monstra) that had begun to appear in contemporary
wall painting:
Sed haec, quae ex veris rebus exepla sumebantur, nunc iniquis
moribus improbantur. Nam penguntur textoriis monstra potius
quam ex rebus finitis imagines certae: pro columnis enim
statuuntur calami, pro fastigiis appagineculi strianti cum crispis
foliis et volutis, item candelabra aedicularum sustinentia figurgas,
supra fastigia eorum surgentes ex radicibus cum volutis teneri
flores habentes in se sine ratione sedentia sigilla, non minus
coliculi dimidiata habentes sigilla alia humanis, alia bestiarum
capitibus.
But these which were imitations based on reality are now
disdained by the improper taste of the present. On the stucco are
monsters rather than definite representations taken from definite
things. Instead of columns there rise up stalks; instead of gables,
striped panels with curled leaves and volutes. Candelabra uphold
pictured shrines and above the summits of these, clusters of thin
stalks rise from their roots in tendrils with little figures seated
upon them at random. Again, slender stalks with heads of men and
of animals attached to half the body (De architectura 7.5.3).44
These lines have occasioned much commentary, but what
Vitruvius objects to is not illusionism per se, but architecturally
unsustainable acts ostensibly performed by thin reeds
supporting heavier elements, which, as with hybrids, represent
things that ‘neither are, nor are able to be, nor have ever been’
(‘haec autem nec sunt nec fieri possunt nec fuerunt’, 7.5.4).
They are against the laws of nature, of which architecture,
throughout Vitruvius’ treatise, is presented as a logical
extension. Hybrid creatures and architectural fallacies, for

Plate 36 Pavanozetto table leg depicting a
winged panther. H. 26.50 in. London, British
Museum, GR 1805,0703.454 (Sculpture 2529)
Vitruvius, serve as negative exempla of compositional
incoherence and empty artifice, irrationality and instability.45
Similarly, Horace’s Ars Poetica, cited in the context of
glyptic grylloi by de Wilde and Gori, opens with the following
lines:
Humano capiti cervicem pictor equinam
iungere si velit, et varias inducere plumas
undique collatis membris, ut turpiter atrum
desinat in piscem mulier formosa superne,
spectatum admissi risum teneatis, amici?
If a painter chose to join a human head to the neck of a horse, and to
spread feathers of many a hue over limbs picked up now here now
there, so that what at the top is a lovely woman ends below in a
black and ugly fish, could you my friends, if favoured with a private
view, refrain from laughing?(Ars Poetica I, 1–5).46
Bernard Frischer warns that this sentiment should not be
naively taken to present the poet’s own attitude, but may,
rather, represent an important part of the construction of the
persona of a narrator who is out of touch with current trends, ‘a
tiresome and old-fashioned pedant’.47 The poem continues:
credite, Pisones, isti tabulae fore
librum persimilem, cuius, uelut aegri somnia, uanae
fingentur species, ut nec pes nec caput uni
reddatur formae. ‘pictoribus atque poetis
quidlibet audendi semper fuit aequa potestas.’
scimus, et hanc ueniam petimusque damusque uicissim;
sed non ut placidis coeant immitia, non ut
serpentes auibus geminentur, tigribus agni.
Believe me, dear Pisos, quite like such pictures would be a book,
whose idle fancies shall be shaped like a sick man’s dreams, so that
neither head nor foot can be assigned to a single shape. ‘Painters
and poets,’ you say, ‘have always had an equal right in hazarding
anything.’ We know it: this license we poets claim and in our turn
we grant the like; but not so far that savage should mate with tame,
or serpents couple with birds, lambs with tigers (Ars Poetica I,
6–13).48
Horace’s concern here is with texts, not paintings, but such
couplings, animal combinations ‘whose idle fancies’ are
‘shaped like a sick man’s dreams’ find parallels not only in
Augustan painting, but also in glyptic works of the sort we have
been exploring. What Horace – or his narrative persona –
seems to object to is their absence of unity of subject, as well as,
in literature, of diction. As both Aristotle (Poetics 1458b.11–13)
‘Grylloi’
Plate 37 Marble table supports (trapezophoroi) from the House of Gaius
Cornelius Rufus at Pompeii depicting winged, horned lions each on a single
foot, Pompeii, Uffico Scavi 43371
and Quintillian (8.3.59–60) observe, this easily leads to
riddling speech and ends in incomprehensibility.49 Lack of
verisimilitude, as well as lack of consistency through the
proper integration of parts to whole, too, is objectionable.
Aristotle (Poetics 1460b 10–11) wrote that poetry and painting
should represent ‘things as they were or are; things as they are
said or seem to be; or things as they should be’. The fantastic,
being impossible in nature (α͗ δύνατα in Greek), makes such
imagery improper in the kind of poetry Horace advocates.50 But
such surreal images were clearly quite popular in Augustan
Rome, appearing, for example, in the Augustan villa across the
Tiber beneath the Villa Farnesina, which has been associated
with Agrippa or some other imperial personage, as well as in
the House of Augustus on the Palatine, not to mention on the
walls and even table legs (Pls 36–37) found in luxury villas and
homes of the well-to-do around the Bay of Naples.51 Whether
Horace himself is posturing or does actually object to such
imagery matters little outside the scope of literary theory, but
ancient literary theory does help to illuminate some of the
issues raised by the grylloi. Horace was in dialogue with
Aristotle, who in the Poetics wrote of the unexpected arousal of
wonder through τὸ θαυμαστόν. This is ‘generated in a
particularly effective manner by the marvellous’ which, in
turn, is characterised as ἄλογον (irrational), ἄτοπον
(strange), α͗ δύνατον (impossible), and ψευ̑δος (false).52 Ancient
literary theorists largely reject such excesses, but they most
certainly appealed to the carvers of glyptic grylloi and other
visual artists and to their customers. In fact, it could be that the
very distance of such imagery from truth made these objects
appealing.53
Grylloi and other hybrids, whether in painting or in glyptic,
bend the laws of nature, diverge from the conventions of
mimesis, engender – or even celebrate – the marvellous and
with it possibilities of metamorphosis, provocation, and
de-stabilisation.54 Such marvels may stimulate pleasure, as
Aristotle noted, but more is going on here. Their ‘gross
infringements of the natural order’ creates a paradox, and
paradox, in ancient thinking, could lead to a new
understanding of the world.55 These hybrids and the gems on
which they appear also seem to possess unworldly power and
‘Gems of Heaven’ | 95
Lapatin
could not only protect, but also, perhaps, impart a special
charisma to those who deployed them.56 Indeed, although both
Vitruvius and Horace’s narrator object to such images, they
recognise their potency, the ability of these apparitions to
beguile.57 For Vitruvius, they are monstra – not ‘monsters’ in the
modern sense, but something beyond nature. And for Horace,
they are like a ‘sick man’s dream’ (‘uelut aegri somnia’). We see
the powerful, mystical – if that is the right word – associations
of such hybrids perhaps nowhere better than in the so-called
‘Aula dell’Orante’ on the Caelian Hill beneath the Church of SS
Giovanni and Paolo, where they appear in 3rd-century ad wall-
paintings alongside the image of a praying man.58 Here they are
explicitly linked to some power beyond that of our world.
Hanging from cords, associated with Christian imagery –
adding to, or even overloading that imagery – they are highly
resistant to conventional reductive interpretations – and this in
the very same period of the greatest popularity of glyptic
grylloi.
Taking a step back in time, it is interesting to note that the
potentiality inherent in hybrids – as being unfixed – is also
present in one of the earliest literary references to them, where
they are viewed not so much as the opposite of nature, as the
origins of species. The Presocratic Sicilian philosopher
Empedocles of Acragas explored theories of ‘mechanism vs.
teleology’ in the formation of animals and suggested that in
early times animal parts, such as heads and limbs, were
generated separately and then eventually combined into more
integrated organisms. He held that such independent body
parts were alive and sentient on their own, but when combined
into organisms only those that formed stable and successful
animals survived. In fact, Empedocles reputedly said that some
of the animals ‘were like dream images, with the parts growing
together’59 – perhaps influencing the formulation of Horace.
In any case, Empedocles’ fragment 57 (RP 173a) mentions
‘many heads sprung up without necks and arms wandering
bare and bereft of shoulders’. Fragment 57 (RP 173b) recounts
that ‘many creatures with faces and breasts looking in different
directions were born; some, offspring of oxen with faces of
men, while others, again, arose as offspring of men with the
heads of oxen, and creatures in whom the nature of women and
men was mingled’. Empedocles was renowned for his
penetrating knowledge of nature and gained a reputation for
marvelous powers, including curing diseases, averting
epidemics, and raising the dead (e.g., Diogenes Laertius 8.58–
61, 67–70). He was later considered a magician who could
control storms, defeat evil, and overcome old age, and his ideas
must have had some appeal, for centuries after his death the
Roman poet Lucretius took pains to dispute them in De rerum
natura, denying the existence – ever – of hybrids:
sed neque Centauri fuerunt, nec tempore in ullo / esse queunt
duplici natura et corpore bino / ex alienigenis membris compacta;
But neither were there centaurs, nor at any time can there be
animals of twofold nature and double body, put together of limbs
of alien birth, (5.878–80);60
aut rabidis canibus succinctas semimarinis / corporibus Scyllas et
cetera de genere horum, / inter se quorum discordia membra
videmus;
or Scyllas either, with bodies half of sea-monsters, girt about with
ravening dogs, or any other beasts of their kind, whose limbs we
see cannot agree one with another, (5.892–4);61
96 | ‘Gems of Heaven’
nam quod multa fuere in terris semina rerum / tempore quo
primum tellus animalia fudit, / nil tamen est signi mixtas potuisse
creari / inter se pecudes compactaque membra animantum;
For because there were in the earth many seeds of things at the
time when first the land brought forth animals, yet that is no proof
that beasts of mingled breed could have been born, or limbs of
living creatures put together in one, (5.916–19).62
I do not propose here to promote a new, Empedoclean
solution to the puzzle of the so-called grylloi, or any other fixed
interpretation, but rather to suggest that Empedocles’ use of
hybrids demonstrates how such imagery can provide a means
of working out diverse issues. Indeed, the various contexts of
the texts that mention hybrids and the wide diffusion of the
wall-paintings and gems that depict them suggest that
hybridisation was both pregnant with meaning and – this is key
– intended to evoke a heuristic response in viewers. For Cicero
(de natura Deorum 1.105) and other critics, hybrids are vain
thoughts, the products of empty imagination (‘motus inanis’),
but such imagery begs to be puzzled out and makes work for
the viewer through its inherent complexity and deliberate
absence of immediate comprehensibility. Thus, I would like to
suggest that the instability of glyptic grylloi was a key part of
their power and contributed to their multi-valence.
Experimentation seems to have been welcome. More, in short,
was more.
Of course, in order to understand the grylloi better we must
also consider still additional associations of their imagery: the
zodiacal and astrological implications of their elements; the
role of animals in moralising tales and popular fables, proverbs
and medicine;63 not to mention possible punning that might
have evoked the names of their owners. To be sure, it is not
difficult to recognise standard compositional elements to
which a patron might request the addition of personally
significant items. And then there are other, more general
considerations: the frequently small size of grylloi chock-full of
so many elements; the implications of reproducibility through
impression, for excavated sealings demonstrate that the grylloi
were not just amulets,64 but closely bound up with personal
identity; the possible associations of stones with particular
images, as mentioned above – or with particular practices; the
presence of inscriptions;65 and the relevance of findspots: are
elephants, for example, more prevalent in eastern gems? Or
less? And what of elephants on those found in Britain, or
elsewhere in the West66 versus those found in the East?67
Considerable work remains to be done.
This seminar has demonstrated that we are only just
beginning to recognise the diverse powers of ancient
gemstones and their imagery. Certainly this is true of no class
of object more than the so-called grylloi.
Notes
1 I am grateful to Chris Entwistle for the invitation to participate in
the London conference; to Marden Nichols for information,
encouragement, and insight at early stages of my investigations;
and to my fellow seminar participants for their suggestions. I have
attempted to incorporate the latter into this version of my paper,
but it nonetheless remains much as presented in May 2009. Special
thanks are due to Jasper Gaunt, Derek Content, Richard Gordon
and P. Corby Finney for generously providing images of gems in
their care and further references (although I have attempted to
furnish here neither a complete bibliography of the subject nor a
comprehensive corpus of examples).

2 E.g. K.A. Böttiger, Journal des Luxus und der Mode 7–8 (1804)
(reprinted in K.A. Böttiger, Kleine Schriften archäologischen und
antiquarischen Inhalts II, Dresden, 1838, 460–1); M. Maaskant-
Kleibrink, Classification of Ancient Engraved Gems, Leiden, 1975,
242; M. Henig and M. Whiting, Engraved Gems from Gadara in
Jordan: the Sa’d Collection of Intaglios and Cameos (Oxford
University Committee for Archaeology Monograph no. 6), Oxford,
1987, 31.
3 See for example, M. Iozzo et al., The Chimaera of Arezzo, Florence,
2009.
4 J. Boardman, ‘Disguise and exchange in Eastern imagery’, in
T. Potts, M. Roaf and D. Stein (eds), Culture through Objects:
Ancient Near Eastern Studies in Honour of P.R.S. Moorey,
Cambridge, 2003, 123–31; A. Roes, ‘New light on the Grylli’, Journal
of Hellenic Studies 55 (1935), 232–5; A. Blanchet, ‘Recherches sur les
“Grylles”. A propos d’une pierre gravée, trouvée en Alsace’, Revue
des Études anciennes 23 (1921), 43–51, at 50; A. Furtwängler, Die
antiken Gemmen: Geschichte der Steinschneidekunst im klassischen
Altertum, Leipzig and Berlin, 1900, 113–14, 352–3.
5 See G. Hafner, ‘Neue Mischwesen des 4. Jahrhunderts’, Wiener
Jahreshefte 32 (1940), 25–34.
6 G. Sauron, ‘Le monstres, au coeur des conflits esthétiques à Rome
au Ier siècle avant J.-C.’, Revue de l’Art 90 (1990), 35–45; S. Yerkes,
‘Vitruvius’ monstra’, Journal of Roman Archaeology 13 (2000), 234–
51; V. Platt, ‘Where the wild things are: locating the marvellous in
Augustan wall painting’, in P. Hardie (ed.), Paradox and the
Marvellous in Augustan Literature and Culture, Oxford, 2009,
41–74.
7 I have not been able to consult S. Verberk, Grylloi en Kombinaties op
Romeinse Ringstenen (Doctoraalscriptie Klassieke Archeologie
Universiteit van Amsterdam) [unpublished doctoral thesis
submitted August 27, 1993].
8 I.e. Gordon, Faraone, Mastrocinque, Dasen, Nagy, Michel.
9 E. Zwierlein-Diehl, Antike Gemmen und ihr Nachleben, Berlin,
2007, 142.
10 E.g., Böttiger (n. 2); Blanchet (n. 4), 44.
11 E. Zwierlein-Diehl, Die antiken Gemmen des Kunsthistorischen
Museums in Wien, III, Munich, 1991, no. 2112 = Zwierlein-Diehl (n.
9), 142, fig. 590.
12 Ibid., fig. 591.
13 E.g., G.M.A. Richter, Metropolitan Museum of Art. Catalogue of
Engraved Gems: Greek, Etruscan, and Roman, Rome, 1956, 114.
14 Plutarch, Quaestiones conviviales, trans. P.A. Clement and H.B.
Hoffleit, Loeb Classical Library, Cambridge, MA, 1969.
15 J.J. Winckelmann, Description des pierres gravées du feu Baron de
Stosch, Florence, 1760, (anastatic reprint Baden-Baden 1970), 130,
no. 667.
16 Ibid., 559–62, nos 207–51.
17 P.M. Fraser and E. Matthews et al. (eds), A Lexicon of Greek Personal
Names, Oxford, 1987–, I.111, II.97, IIIb.94.
18 P. Zanker, The Power of Images in the Age of Augustus, Ann Arbor,
1988, 209, pl. 162.
19 H.G. Liddell and R. Scott, A Greek English Lexicon9, Oxford, 1996,
391; P. Chantraine, Dictionnaire étymologique de la langue grecque,
Paris, 1999, 238; Blanchet (n. 4); E. Lobel and C. Roberts, ‘Verses on
the Labours of Heracles’, Oxyrhynchus Papyri 22 (1954), no. 2331
and pl. XI; W. Binsfeld, Grylloi; ein Beitrag zur Geschichte der
antiken Karikatur, diss. Cologne, 1956; idem, ‘Γρύλλοι’, RE Suppl.
IX (1962), 76–8; D. Page, ‘P.Oxy. 2331 and Others’, Classical Review
7 (1957), 189–92; P. Maas, ‘The GRULLOS Papyrus’, Greece and
Rome 5 (1958), 171–3; G. Becatti, ‘Grylloi’, Enciclopeda dell’Arte
Antica 3 (1960), 1065–6; H. Bartels, ‘Grylloi’, in E. Kunze (ed.),
Olympia Bericht VIII, Berlin, 1967, 250–62. F. Perpillou-Thomas,
‘P.Sorb. inv. 2381: γρύλλος καλαμαύλης χορός’, Zeitschrift für
Papyrologie und Epigraphik 78 (1989), 153–5; S. Pfisterer-Haas, ‘Die
bronzene Zwergentänzer’, in G. Hellenkemper Salies et al. (eds),
Das Wrack: Der antike Schiffsfund von Mahdia, Bonn, 1994, 483–
504; J. Hammerstaedt, ‘Gryllos. Die antike Bedeutung eines
modernen archäologischen Begriffs’, Zeitschrift für Papyrologie
und Epigraphik 129 (2000), 29–46; L. Herchenroeder, ‘Τί γὰρ
τοῦτο πρὸς τὸν λόγον; Plutarch’s Gryllus and the so-called
Grylloi’, American Journal of Philology 129 (2008), 347–79, esp.
350–8.
20 O.A. Bayardi (ed.), Le Antichità di Ercolano Esposte, I, Le pitture
antiche di Ercolano e contorni, incise con qualche spiegazione,
Naples, 1757, 245–7.
‘Grylloi’
21 W.E. Wallace, ‘Instruction and originality in Michelangelo’s
drawings’, in A. Ladis and C. Wood (eds), The Craft of Art:
Originality and Industry in the Italian Renaissance and Baroque
Workshop, Athens, GA, 1995, 113–33, at 132, figs 2–3, n. 11.
22 E.H. Gombrich, Norm and Form: Studies in the Art of the
Renaissance, London, 1966, 113; S. Miedma, ‘Grillen van
Rembrandt’, Proef 3 (1974), 74–5; J. Bruyn, ‘Problemen bij Grillen’,
Proef 3 (1974), 82–4; S. Miedma, ‘De Grillen’, Proef 3 (1974), 84–6;
R. Ruurs, ‘Adrianus Brouwer: Gryllorum Pictor’, Proef 3 (1974),
87–8; S. Alpers, ‘Realism as a comic mode: low-life painting seen
through Brodero’s eyes’, Simiolus 8 (1975/6), 115–44, at 119, n. 15;
I. Lavin, ‘Bernini and the art of social satire’, in I. Lavin (ed.),
Drawings by Gianlorenzo Bernini from the Museum der Bildenden
Künste Leipzig, Princeton, 1981, 27–54, at 45–6, 54, nn. 60–4;
H. Bredekamp, ‘Grillenfänge von Michelangelo bis Goethe’,
Marburger Jahrbuch für Kunstwissenschaft 22 (1989), 169–80, at
169–71; T.D. Kaufmann, Arcimboldo: Visual Jokes, Natural History,
and Still-life Painting, Chicago, 2009, esp. 39, 104, 171; for grylloi in
the Middle Ages see, J. Baltrusaitis, Le moyen age fantastique:
antiquités et exotismes dans l’art gothique, Paris, 1955.
23 A.F. Gori, Museum Florentinum exhibens insiniora vetustatis
monumenta quae Florentia sunt Ioanni Gastoni Etruriae Magno
Duci dedicatum, 2. Gemmae antiquae ex thesauro Mediceo et
privatorum dactyliothecis Florentiae: exhibentes tabulis C. imagines
vivorum illustrium et deorum cum observationibus Antonii Francisci
Gorii publici historiarum professoris, Florence, 1732, 147–8.
24 J. Tassie and E. Raspe, A Descriptive Catalogue of a General
Collection of Ancient and Modern Engraved Gems, Cameos and
Intaglios, Taken from the Most Celebrated Cabinets in Europe; and
Cast in Coloured Pastes, White Enamel, and Sulphur, London, 1791
[also on-line at http://www.beazley.ox.ac.uk/gems/tassie].
25 Ibid., 709.
26 H.B. Walters, Catalogue of Engraved Gems and Cameos, Greek,
Etruscan and Roman in the British Museum, London, 1926, nos
2549–51, 2562.
27 Furtwängler (n. 4), 3.146, pl. 104; J. Boardman, Greek Gems and
Finger Rings, London, 1970, pl. 721.
28 B80: M. Henig et al., Classical Gems: Ancient and Modern Intaglios
and Cameos in the Fitzwilliam Museum, Cambridge, Cambridge,
1994, 84, no. 146.
29 Inv. no. 81.6.160: Richter (n. 13), 115, no. 549; 1920, no. 276.
30 See also M.L. Vollenweider, Catalogue raisonné des sceaux,
cylindres, intailles et camées, II. Les portraits, les masques de
théâtre, les symboles politiques, Mainz, 1979, 340, no. 378, pl. 107.7,
where the insect wings form the figure’s beard.
31 Kaufmann (n. 22), 39.
32 A. van Goorle, Dactyliotheca seu Gemmarum Annulorumque ex
Abrahami Gorlaei cum explicationibus Jacobi Gronovii, Leyden,
1707, 9; 1778: 2.
33 Ibid., 27; 1778: 2, 7.
34 L. Agostini, Le gemme antiche figurate, Rome, 1686, 7 = L. Agostini,
Gemmae et sculpturae antiquae, Franequerae, 1694, 34, no. 37.
35 P.A. Maffei, Gemme antiche figurate, vol. 4, Rome, 1707–9, 29–30.
36 Ibid., vol. 2, 40–5, pl. 20.
37 J.J. Baier, Gemmarum affabre sculptarum thesaurus, Nuremberg,
1720, 13–14, pl. 4, no. 5.
38 This gem, once in the collection of Elisabeth-Charlotte of the
Palatinate, Duchess of Orléans, was incorrectly described in the
1727 sale catalogue, Description sommaire des pierres gravées et des
médailles d’or antiques du cabinet de feue madame, Paris, 7, as ‘Têtes
à trois faces opposées, du front d’une desquelles sort en maniere
d’ornement de coëffure, un Serpent qui port un Caducée...’. It is
now in the Hermitage, see: J. Kagan and O. Neverov, Le destin d’une
collection. 500 pierres gravées du Cabinet du Duc d’Orléans,
St Petersburg, 2001, 106, no. 164/67.
39 J. de Wilde, Gemmae selectae antiquae e museo Jocobi de Wilde,
Amsterdam, 1703 [1707], 125.
40 A.F. Gori, Le gemme antiche di Anton-Maria Zanetti di Girolamo,
Venice, 1750, 134, pl. 48.
41 A.F. Gori (n. 23), 80, Index Rerum s.v. Chimaera.
42 A.F. Gori, Museum Florentinum exhibens insiniora vetustatis
monumenta quae Florentia sunt Ioanni Gastoni Etruriae Magno
Duci dedicatum, 1. Gemmae antiquae ex thesauro Mediceo et
privatorum dactyliothecis Florentiae: exhibentes tabulis C. imagines
vivorum illustrium et deorum cum observationibus Antonii Francisci
Gorii publici historiarum professoris, Florence, 1731, 104, pls 48–51.
‘Gems of Heaven’ | 97
Lapatin

43 J. M. Hurwit, The Art and Culture of Early Greece, 1100–480 bc, Addendum
Cornell, 1985, 188–90. Another early appearance of a glyptic gryllos is an agate
44 Vitruvius, De Architectura, trans. F. Granger, Loeb Classical
intaglio published by M. Antoine Le Pois, councillor and
Library, Cambridge, MA, 1970.
45 See for example: Sauron (n. 6); Yerkes (n. 6); M. Citroni, ‘Horace’s physician to the Duke of Lorraine in 1579. It features three
Ars Poetica and the marvellous’, in Hardie (ed.) (n. 6), 31–40, at 38; heads (of a bearded humanoid, a ram, and a horse with branch
Platt (n. 6); M.F. Nichols, Vitruvius and the Rhetoric of Display in mouth) atop a cock’s legs, and is explained as representing a
(Ph.D. thesis, University of Cambridge, 2009).
46 Horace, Ars Poetica, trans. H. Rushton Fairclough, Loeb Classical journey to the three parts of the world, i.e., Africa, Asia, and
Library no. 194, Cambridge, MA, 1999 (repr.). Europe, represented, respectively, by the horse, ram, and
47 B. Frischer, Shifting Paradigms: New Approaches to Horace’s Ars human head. For the horse, we are told, appears on some
Poetica (American Philological Association, American Classical
ancient African coins, and Europe is the region ‘plus polie,
Studies, no. 27), Atlanta, 1991, 68ff.; cf. Citroni (n. 45), 40, n. 42.
48 See n. 46. ornee, civile & humaine que les deux autres’. Of course, ‘Autres
49 Frischer (n. 47), 70–1. ay-ie ouy interpreter autrement & bien differemment ceste
50 Citroni (n. 45), 19–20; Platt (n. 6). figure’.
51 Sauron (n. 6); M.R. Sanzi Di Mino et al., La Villa della Farnesina in
Palazzo Massimo alle Terme, Rome, 1998; I. Iacopi, La Casa di
Augusto: Le Pitture, Rome, 2007; Platt (n. 6).
52 Citroni (n. 45), 25–6; Platt (n. 6).
53 Although Horace’s very precise description of a hybrid creature in
the proem of the Ars Poetica sounds like a gryllos, I have
encountered no such creature depicted on an engraved gem or in
any other medium. The absence of such imagery in ancient art may
be attributed to the conventional, if varied nature of the glyptic
grylloi. The fact that no post-antique gem engraver appears to have
realised Horace’s creature is perhaps due to the proscriptive power
of the poet’s emphatic negative assessment of the image he so
vividly describes.
54 Platt (n. 6), 44.
55 P. Hardie, ‘Introduction’, in Hardie (ed.) (n. 6), 1–18, at 9, 14.
56 Platt (n. 6), 44–5.
57 For other links between Horace and Vitruvius see M.F. Nichols,
‘Social status and the authorial personae of Horace and Vitruvius’,
in L.B.T. Houghton and M. Wyke (eds), Perceptions of Horace: a
Roman poet and his readers, Cambridge, 2010, 109–22.
58 A. Englen et al., Case romane e Antiquarium sotto la Basilica dei SS.
Giovanni e Paolo al Celio, Rome, 2004, 11; see also: P.C. Finney, The
Invisible God. The Earliest Christians on Art, Oxford, 1994, 222, on
such imagery in the Callistus catacomb. M. Antoine Le Pois, Discours sur les medailles et grauevres antiques,
59 B. Inwood, The Poem of Empedocles: A Text and Translation with an principalement romaines, Paris,1579, pl. d, no. 6
Introduction, Toronto, 2001, 188.
60 Lucretius, De rerum natura, ed. and trans. C. Bailey, Oxford, 1947.
61 Ibid.
62 Ibid.
63 See, for example, Herchenroeder (n. 19); I have not had the
opportunity to consult two recent collections: I. Boehm and P.
Luccioni (eds), Le médecin initié par l’animal: animaux et médecine
dans l’Antiquité grecque et latine. Actes du colloque international
tenu à la Maison de l’Orient de la Méditerranée-Jean Pouilloux les 26
et 27 octobre 2006. Collection de la maison de l’orient et de la
Méditerranée 39. Série littéraire et philosophique 1, Lyon, 2008;
M. Fansa (ed.), Tierisch moralisch. Die Welt der Fabel in Orient und
Okzident. Begleitschrift zur Sonderausstellung des Landesmuseums
Natur und Mensch Oldenburg vom 22. Februar bis zum 01. Juni 2009.
(Schriftenreihe des Landesmuseums Natur und Mensch, Heft 63),
Wiesbaden, 2009.
64 Boardman (n. 27), 234–5, 322; Boardman (n. 4); A. Invernizzi et al.,
Seleucia al Tigri. Le impronte di sigillo dagli Archivi, Alexandria,
2004; cf. Maaskant-Kleibrink (n. 2), 242.
65 Blanchet (n. 4), 48–50.
66 E.g. M. Henig, A Corpus of Roman Engraved Gemstones from British
Sites, Oxford, 1974, 54–6, nos 373–85; G. Dembski, Die antiken
Gemmen und Kameen aus Carnuntum, Vienna, 2005, 1084, 1089,
1096–7, 1104.
67 G. Sena Chiesa, Gemme del Museo Nazionale di Aquileia, Aquileia,
1966, 342–3, nos 1005, 1009–10; Henig and Whiting (n. 2), 31–2, nos
309–10, 313.

98 | ‘Gems of Heaven’
Engraved Gems from Sites with a Military Presence in
Roman Palestine
The cases of Legio and Aelia Capitolina
Orit Peleg-Barkat and Yotam Tepper
Introduction
The sixth and tenth Roman legions stationed in Roman
Palestine left behind many artefacts including some unique
intaglios and cameos. Recent surveys and excavations near the
historical site of Megiddo resulted in the identification of the
sites of Legio (i.e. the camp of the Legio VI Ferrata), the nearby
Jewish village of Kefar ‘Othnay and the Romano-Byzantine city
of Maximianopolis.1 Several of the intaglios and cameos
unearthed in these sites are highly intriguing. Past excavations
in Aelia Capitolina, namely the city of Jerusalem in the 2nd to
3rd centuries ad, where the camp of the Legio X Fretensis was
located, uncovered another interesting group of intaglios, only
recently published.2 In this paper we would like to address the
significance of the devices engraved on these gemstones, as
well as their contribution to our understanding of the sites
where they were found.
The two sites – Legio and Aelia Capitolina – were inhabited
during the 2nd to 4th centuries ad by a rather heterogeneous
population of both civilians and military personnel, pagans
and Christians, and in the case of Legio – also Jews and
Samaritans. These groups had common, as well as
distinguishing, attributes and their co-existence highlights
mutual cultural influences. Since gem cutting is an artistic
medium with each gem having an individual character
reflecting the owner’s personal preferences and individual
taste, the study of the intaglios from Legio and Aelia Capitolina
provides a glimpse into the personal world of some of the
inhabitants of these ancient sites. In discussing gems from
Roman Britain, Henig has already pointed out the close
connection between the type of site – military, civic, religious,
industrial, funerary, or hoard – and the subjects engraved upon
the gems found in it.3 We could, therefore, expect the devices
appearing on the intaglios and cameos found in Legio and
Aelia Capitolina to reflect the character of the population of
these two sites.
Our discussion focuses on three groups of finds; the first
group includes a silver ring set with an intaglio bearing the
image of Abrasax, as well as several types of copper-alloy rings,
one set with a small cabochon. These rings were recently found
in an excavation led by Yotam Tepper on behalf of the Israel
Antiquities Authority between 2004 and 2008, on the Megiddo
police-station hill, which has been identified as the location of
Kefar ‘Othnay.4 The second group includes two cameos, seven
intaglios (one made out of copper-alloy), and one copper-alloy
ring with an engraved design, and comes from the collection of
the nearby Kibbutz Giv’at Oz. The entire collection is comprised
of random finds made by Aryeh Drori, a Kibbutz member, in
the fields surrounding the Kibbutz, all within the boundaries of
the ancient districts of Legio, including the military camp,
village and Romano-Byzantine city.5 The third group includes a
dozen intaglios discovered during excavations led by Benjamin
Mazar on behalf of the Hebrew University between 1968 and
1978 south of the Temple Mount in Jerusalem.6
Kefar ‘Othnay
The settlement at Kefar ‘Othnay is mentioned in the Mishna,
compiled c. ad 200, as the southern Halkhic boundary of
Jewish Galilee (Mishna, Gittin, 7:7). It is also mentioned in the
Geography of Ptolemy of Alexandria (V, 16.4), and appears on
the Peutinger Map (section 9). Six ritual baths and several
fragments of stone vessels revealed at the site, which were
characteristically used by Jews in the early Roman period, are
further evidence that Jews constituted the main component of
the population at the site.7 From the rabbinic rulings regarding
Samaritans and their activities in Kefar ‘Othnay, it appears that
their presence in the village was significant during the 2nd to
4th centuries ad (Mishna, Gittin, 1:5). Several finds, such as a
beautiful bronze Lar statuette,8 indicate that pagans were also
resident at the site. The excavations have exposed a house that
apparently served as the residence for those centurions who
did not live in the nearby legionary camp, probably because
they had families. The identification of the residents of this
building as Roman soldiers or officers is based on several
artefacts, such as military equipment and bread stamps
bearing Latin inscriptions.9 The most significant evidence,
however, is a Greek inscription on a mosaic floor exposed in the
western wing of the building, mentioning the centurion
Gaianus, also called Porphyrus, who paid for the paving of the
hall. According to another inscription, the hall served as a
Christian prayer hall.10
During the excavations at Kefar ‘Othnay a broken silver
ring set with a haematite intaglio was found at the northern
part of the site, beside a white mosaic floor, in a room
presumed to be an industrial installation dating to the 3rd
century ad (Pl. 1). The device is of Abrasax (or Abraxas), a solar
Plate 1 Silver ring set with a haematite intaglio bearing the image of Abrasax
found at Kefar ‘Othnay. 16 x 11mm
‘‘Gems of Heaven’ | 99
Peleg-Barkat and Tepper
Plate 2 Copper-alloy rings found at Kefar ‘Othnay. D. 16 – 22mm
deity that was common on gems and amulets during the 3rd
and 4th centuries ad.11 He appears with the head of a cockerel,
the body of a cuirassed Roman officer, and serpent-like legs –
perhaps in turn symbols of the upper air, the earth and the
underworld. In his right hand he grasps what appears to be a
club (though the flail is his more customary attribute),12 and in
his left is a round shield, depicted in a somewhat obscure
fashion.13
The identification of this figure with Abrasax is based on an
abundance of amulets from museum collections and
archaeological sites, mainly in Egypt and Syria, bearing his
image alongside an engraving of his name.14 The name was
related in the past to the Gnostic doctrine of Basilides, an early
2nd century ad Christian religious teacher in Alexandria.15 Yet
objections have been raised concerning this interpretation and
Abrasax gems are now regarded as pagan amulets and
instruments of magic, following the deciphering of Egyptian
magical papyri, in which many of the unintelligible names of
the Abrasax gems reappear, alongside directions for making
and using gems with similar figures and formulas for magical
purposes.16 Still, the frequent appearance of Abrasax in Jewish
magical texts demonstrates its assimilation into the Jewish
magical tradition. In these texts Abrasax is conceived of as the
name of a very powerful angel or celestial power that was not
only incorporated into the Jewish magical tradition, but in
some cases even seems to have been entirely Judaised in the
process.17 The fact that the figure of Abrasax survived both on
kabbalistic amulets and in Christian circles during medieval
times seems to indicate that its use was common amongst Jews,
Christians and pagans alike.18
Beside the Abrasax silver ring, about 20 copper-alloy rings
were found in different areas of the excavation (Pl. 2). The
rings differ in section and diameter and although some of them
might have had a variety of functions, at least 13 circular rings
with diameters of 16–22mm, seem to have been originally used
as finger-rings. These rings could either have been used as
100| ‘Gems of Heaven’
Plate 3 Copper-alloy ring with cabochon Plate 4 Key-ring found at Kefar
found at Kefar ‘Othnay. D. 17mm ‘Othnay. L. 20mm, D. 17mm
betrothal rings by pagan residents19 or as plain rings for Jewish
women. According to Jewish sources the custom of betrothal or
wedding rings amongst Jews only began later, in the 6th or 7th
centuries ad, and while men usually wore seal rings, women
normally wore plain bronze or iron rings.20 Four of the rings
that have a clear archaeological context dated to the Roman
period were found in courtyards, where domestic activities
took place that might have required the removal of rings.
One of the rings is set with a small pinkish-orange stone
held in place with a small pin and is decorated with vertical
notches on the bezel (Pl. 3). This type of ring has parallels
mostly in the Early Roman period, and probably dates from the
1st century ad.21 Another interesting ring found at the site is of
a key-ring type (Pl. 4). The key-ring was rather an elegant and
expedient solution to the lack of pockets in Roman clothes; key-
rings served a dual purpose – being both decorative and
functional. The key was likely for the lock of a small box or
casket.22 One such key-ring was found with fragments of the
lock and casket itself at Elsenham, Essex.23 Another key-ring,
similar to the one from Kefar ‘Othnay, was found at the Cave of
the Letters in the Judaean Desert, dated before ad 135, together
with another key-ring of a different design.24 Jewish sources
indicate that this type of ring was a feminine accessory. For
example, in Tosefta Shabbat 4:11 it is said that a woman should
not go out of her house on Shabbat with a key on her finger.
Kibbutz Giv’at Oz collection
The assortment of intaglios and cameos on display today at
Kibbutz Giv’at Oz is part of an archaeological collection that
was initially the private collection of Aryeh Drori, one of the
founders of the Kibbutz.25 Later Drori handed over the
collection to the Kibbutz and a small museum was erected for
its exhibition. The items that comprise the collection were
found between 1945 and the early 1970s in the fields of the
Kibbutz, near the police-station hill (Kefar ‘Othnay), in the
environs of the modern Megiddo Junction and adjacent areas
(Legio).
Two cameos are included in this assemblage. One is of a
Medusa head engraved on a white agate and the second is of a
bust of a lady. The Medusa head is slightly turned to left (Pl. 5).
The hair is schematically rendered, and so are the wings and
snakes that are supposedly tied under the neck. A similar
rendering of the hair and facial features appears on an onyx
cameo dated to the 2nd century ad in the Sa’d collection of
engraved gems from Gadara in Jordan.26 The Medusa head is
the single most popular type on Roman cameos and a very
 Engraved Gems from Sites with a Military Presence in Roman Palestine
Plate 5 White agate cameo with a Medusa Plate 6 White agate cameo with a feminine
head, Kibbutz Giv’at Oz collection. 19 x17mm bust, Kibbutz Giv’at Oz collection. 14 x 10mm
Plate 8 Carnelian engraved with the image of Artemis and
impression, Kibbutz Giv’at Oz collection. 10 x 8mm
large number have survived in all parts of the Roman Empire,
primarily dated to the late 2nd and 3rd centuries ad. They were
variously set in pendants, rings and earrings. The round shape
of the Medusa head was convenient for this medium and the
image of her head was believed to ward off the evil eye.27 The
second cameo is of finer workmanship. It depicts a Roman lady
with her head in profile to right, while the shoulders are only
partly turned to the right (Pl. 6). She is wearing a stephane, a
beaded necklace, and what seems to be a pala, or some other
outer garment over a tunic. Her hairstyle points to a date in the
Hadrianic period.28
The intaglios in the collection of Kibbutz Giv’at Oz are of a
variety of subjects and raw materials. A red carnelian engraved
with the image of Venus Victrix is the best in quality (Pl. 7).29 It
is broken at the bottom and chipped in a few places, suggesting
that it might have been taken out of its bezel by force by
someone who was more interested in the ring, perhaps made of
gold, than in its setting. Venus is seen from the back, nude but
for a small cloth around her hips. She is leaning with her left
elbow on a small column at the foot of which lies a round
shield. In her outstretched right hand is a tasseled helmet,
while her left hand clutches a sword. Behind her left arm there
is a spear, while before her stands Cupid with an arrow in his
hand – a rather uncommon feature on gemstones.30 The gem is
engraved in the Imperial Classicising style and probably should
be dated to the 1st or early 2nd centuries ad.
Another 2nd century ad carnelian intaglio depicts Artemis,
wearing a short chiton, running to the right on a ground line
Plate 7 Red carnelian engraved with the image of
Venus Victrix, Kibbutz Giv’at Oz collection. 11 x 11mm
Plate 9 Iron ring set with a carnelian depicting a schematized Tyche-Fortuna (?) and its
impression, Kibbutz Giv’at Oz collection. 11 x 7.5mm
(Pl. 8). She is raising the bow with her right outstretched arm
and with her left hand she is reaching for an arrow in the
quiver on her back. At her feet is a hound. Such depictions are
very common on gemstones and close parallels have been
found in Caesarea, Gadara and elsewhere.31 Together with the
armed Venus Victrix the hunting Artemis would have been a
suitable device for the rings of Roman soldiers.
A carnelian of a lesser quality, and probably of a later date
in the 3rd century ad, depicts a goddess dressed in a long
chiton, perhaps a schematised Tyche-Fortuna (Pl. 9). The stone
is still set in its iron bezel. A nicolo intaglio depicts a bust of
Zeus facing right (Pl. 10). Zeus appears bearded and crowned
with a laurel wreath that ends in projecting leaves above the
forehead.
An intaglio that was most certainly owned by a woman
rather than a male soldier is a carnelian engraved with the
name Ziala in positive (Pl. 11). Similar names, and especially
Zoila (Ζωιλα), were very common in the eastern Mediterranean
during the Hellenistic and Roman periods, up until the 4th
century ad. The name Zoila appears on an inscription from the
Jewish cemetery at Beth She’arim and in other sites from
Roman Palestine.32 It seems likely that the gem was originally
set in a ring of a Jewish woman from Kefar ‘Othnay or
Maximianopolis.
Another intaglio that might have originated from a finger-
ring of one of the Jewish residents of the area is a translucent
white chalcedony with a convex profile engraved with a simple
chalice or cup (skyphos), from which issues a vertical stem (Pl.
‘‘Gems of Heaven’ | 101
Peleg-Barkat and Tepper
Plate 10 Nicolo intaglio depicting a bust of Zeus and its impression, Kibbutz
Giv’at Oz collection. 11 x 9.2mm
12). The execution is very simple and includes no details or
textures, but the general layout is reminiscent of a yellowish
light brown glass intaglio from Masada depicting a palm tree
issuing out of a similar cup.33 A similar cup is also depicted on a
combination intaglio found at the foot of the Western Wall in
Jerusalem, which is the only gem found in that excavation that
may be attributed with some certainty to a Jew at the time of
the Second Temple period. It depicts two confronted
cornucopiae flanking a rounded elongated stand upon which
sits a cup/skyphos with fruit, apparently three pomegranates.34
The three motifs – the cornucopiae, the stand resembling a
trunk of a palm tree and the skyphos with the pomegranates –
are all familiar motifs in Jewish art during the Second Temple
period: the cornucopiae are the most conspicuous symbols on
Hasmonean coins and the pattern of a pair of cornucopiae
facing each other, connected at their bases by a pomegranate,
is an original Hasmonean creation.35 The palm tree trunk is
also a motif known from Jewish numismatics and art. A
parallel appears for example on two clay bullae in the Israel
Museum sealed with the seal of Alexander Jannaeus.36 Even
though these symbols often appear in combination with pagan
symbols on early Roman gemstones, it seems that the
intentional choice of symbols, which fit in well with the world
of Jewish symbols of the time, suggests that the gem was
ordered by a Jew.
Also in the collection are a copper-alloy setting for a ring
and a copper-alloy ring with engraved decoration. The first is
decorated with the image of Harpokrates crowned by an orb
and sitting atop a lotus bloom in a papyrus barque (Pl. 13). He
holds his left hand to his mouth and a flail in his right hand.
Plate 12 Translucent white chalcedony with a convex profile engraved with a
simple chalice or cup (skyphos)and its impression, Kibbutz Giv’at Oz collection.
11.5 x 11mm
102| ‘Gems of Heaven’
Plate 11 Carnelian engraved with the name Ziala in positive, Kibbutz Giv’at Oz
collection. 7.9 x 9.9mm
Two fish or perhaps some other sea creatures are depicted
below the barque. The name Harpokrates is the Hellenised
version of the Egyptian phrase meaning ‘Horus the child’, son
of Isis and Osiris. He was represented as a naked child, with a
side-lock of youth and his finger to his mouth. According to
Egyptian mythology, Isis revived her murdered husband Osiris
to conceive a child. She fled to the Delta to give birth, hiding
from her brother Seth, who was intent on seizing the throne of
Egypt. When her son, Harpokrates, was born he was attacked
by snakes, crocodiles and scorpions sent by his uncle. He was
protected by the gods, and given power over dangerous
creatures. The power Harpokrates had over dangerous animals
meant that he was regarded as a protective deity and this might
be the reason for his frequent appearance on magical amulets.37
The other copper alloy-ring seems to depict a combination of
two masks back to back, probably of a silen and a satyr. Such
combinations are common on intaglios, perhaps due to the
apotropaic qualities of Dionysian masks.38
Aelia Capitolina
The ‘Temple Mount Excavations’ conducted by Benjamin Mazar
between 1968 and 1978, on behalf of the Institute of
Archaeology at the Hebrew University, took place at the foot of
the southern and western retaining walls of the Temple Mount.
During the excavations a dozen Roman intaglios were found in
various areas. The area south and south-west of the Temple
Mount was not, until recent years, regarded by scholars as
being of much importance with regard to the urban plan of
Jerusalem during the 2nd to 4th centuries ad.39 It was even
claimed that this area was outside the boundaries of Aelia
Plate 13 Copper-alloy setting for a ring with the image of Harpokrates and its
impression, Kibbutz Giv’at Oz collection. 13 x 9.5mm
 Engraved Gems from Sites with a Military Presence in Roman Palestine
Capitolina.40 However, since the processing of the finds for
publication began, it has become clear that this area was an
important and integral part of the city at that time.41 Several
complexes exposed during the excavations can be assigned to
this period, including a large bathhouse and a bakery with
ovens built of tiles bearing the Legio X stamp. Considerable
quantities of fragmentary Latin and Greek inscriptions, marble
sculptures, bronze figurines, and tiles and bricks bearing the
Legio X stamp were unearthed throughout the excavated areas,
testifying to the character of its ancient inhabitants. Due to the
military character of some of the finds and architectural
features, several scholars promoted the idea that the camp of
the Legio X Fretensis was moved to the vicinity of the Temple
Mount from its former location at the western part of the city
after the establishment of Colonia Aelia Capitolina in ad 131.42
The corpus of Roman gemstones from the Temple Mount
excavations was published in 2003 in the Palestine Exploration
Quarterly and therefore will be discussed only briefly here in
order to give a broader context to the assemblage from Legio.
As in Legio, it is clear that the gemstones do not constitute a
uniform assemblage originating from a single workshop; rather
they represent an accumulation of individual finds dating from
the entire Roman period.
Some of the designs appearing on the gemstones can be
associated with the Roman army.43 This fact corresponds with
other artefacts with a military character found at the site.
Mars, the god of war, had a cult following among the soldiers;
the Tyche of Aelia Capitolina was linked by the city’s coinage
with the legionary standards; the eagle was the symbol of the
legion and the figure of a horseman can be associated with the
army’s auxiliary forces. Nonetheless, no gemstone bears an
obvious military symbol, such as the legionary standard or an
inscription citing the name of the legion. For this reason, we
can only say that the connection of most of the designs with
symbols of the Roman army suggests that they belonged to
soldiers.
Throughout the Empire dozens of nearly identical, mass-
produced intaglios bearing the images of Mercury and Mars
have been found, similar to those from the Temple Mount
Excavations.44 All the other gemstones from the site, however,
are unique versions of popular subjects found elsewhere. One
such example is the addition of the ear of corn near the image
of Sol; a horned animal held by an eagle and the arrangement
of masks in a Dionysian combination, are likewise unique.45 No
parallels have been found in the corpus of Roman glyptics for
two of the gemstones, whose depictions exist only in other
media. These are the portrayals of Aelia Capitolina, which
appears on the city’s coinage,46 and of a snake wrapped around
an altar, which has parallels in wall paintings.47 If our
identification is correct then a further gemstone portraying
Telemachos riding on a quadriga constitutes the sole depiction
currently known of this subject.48 The fact that some of these
gemstones are unique suggests that they were cut to order. The
production of specially ordered gemstones compelled the artist
to invest a greater amount of time in their engraving, so these
gemstones were certainly expensive. One can conclude from
the unique characteristics of the designs and the fine quality of
most of them that they belonged to persons of comparatively
high economic standing.
Conclusions
The Abrasax intaglio found during the excavations on the
Megiddo Police Station Hill seems to have belonged to a Jewish
resident of or a passer-by at Kefar ‘Othnay, while the various
copper-alloy rings were in the possession of either Jewish
women or of men and women from other ethnic and religious
groups residing at the site. The intaglios and cameos in the
collection of Kibbutz Giv’at Oz probably originate from the
finger-rings of soldiers stationed at the Roman military camp
and also from civilians – male and female, Jewish and pagan –
who inhabited the nearby vicus adjacent to the Roman
legionary camp, the city of Maximianopolis and Kefar ‘Othnay.
In the case of the intaglios from Aelia Capitolina – regardless of
whether the camp of the Legio X was situated at the area where
the intaglios were found – they seem to point, together with
other finds, to the presence of Roman soldiers and veterans
either stationed nearby or just passing through on their way to
the bathhouse or bakery.
Gemstones have been found throughout the entire Roman
Empire and portray all aspects of life. Despite their minute size,
they are rich illustrations of life in the Roman world, and in
particular provide evidence of individual, unofficial cults.
Although it is impossible, of course, to write the social history
of Aelia Capitolina or Legio based on these small finds, there
can be no doubt that, gemstones, alongside sculpture and
mosaics, reflect the religious diversity of the population of both
sites at that time.
Notes
1 Y. Tepper, Survey of the Legio Area near Megiddo: Historical and
Geographical Research, MA thesis, Tel Aviv University (Hebrew),
2003; Y. Tepper, ‘Legio, Kefar ‘Otnay’, HA−ESI 118 (2006), http://
www.hadashot-esi.org. il/report_detail_eng.asp?id=363&mag_
id=111. Y. Tepper and L. Di Segni, A Christian Prayer Hall of the
Third Century ce at Kefar ῾Othnay (Legio) – Excavations at the
Megiddo Prison 2005, Jerusalem, 2006. Salvage excavations on the
Megiddo Police-Station Hill were carried out between 2004 and
2008 on behalf of the Israel Antiquities Authority. We would like to
thank the Israel Antiquities Authority (IAA) for its permission to
publish the Abrasax intaglio, as well as other finger-rings found at
the site. All the photographs in this article were taken by Yotam
Tepper.
2 O. Peleg, ‘Roman Intaglio Gemstones from Aelia Capitolina’,
Palestine Exploration Quarterly 135:1 (2003), 52–67. Another
interesting assemblage of intaglios, also recently published
(M. Hershkovitz and S. Amorai-Stark, ‘The Gems from Masada’,
Masada VIII: The Yigael Yadin Excavations 1963–1965 Final Reports,
Jerusalem, 2007, 217–32) was found in Masada. Here also the
intaglios came from rings that were worn both by the Jewish rebels
and Roman soldiers.
3 M. Henig, A Corpus of Roman Engraved Gemstones from British
Sites (BAR Series 8), Oxford, 2007, (3rd edn), 55–70.
4 Tepper 2003 (n. 1); Tepper and Di Segni (n. 1).
5 Tepper 2003 (n. 1), 31–2, 83–5, fig. 20.
6 Peleg (n. 2); O. Peleg-Barkat, ‘The Roman Intaglios’, in E. Mazar
(ed.), The Temple Mount Excavations in Jerusalem, 1968–1978
Directed by Benjamin Mazar, Final Reports IV: The Tenth Legion in
Aelia Capitolina (Qedem series, no. 52), Jerusalem, 2011, 255–304.
7 Tepper and Di Segni (n. 1), 13, 17, 27.
8 Ibid., 22.
9 G.D. Stiebel, ‘Roman Military Artefacts’, in Tepper and Di Segni (n.
1), 29–31.
10 Tepper and Di Segni (n. 1), 31–42.
11 On the god Abrasax, see A.A. Barb, ‘Abrasax-Studien’, in
Hommages á Waldemar Deonna (Coll. Latomus 28), Brussels,
1957,67–86; on the Abrasax gems, see M. Le Glay, ‘Abrasax’, LIMC I
(1981), H.C. Ackerman and J.-R. Gisler (eds), 2–7, pls 6–14.
12 Other examples of Abrasax holding a club do exist, for example on
‘‘Gems of Heaven’ | 103
Peleg-Barkat and Tepper
a gem from Chichester: M. Henig, ‘Archbishop Hubert Walter’s
Gems’, Journal of the British Archaeological Association 136 (1983),
56–61, at 59.
13 Many of the Abrasax amulets were engraved with images or
inscriptions on their reverses: see for example, E.R. Goodenough,
Jewish Symbols in the Greco-Roman Period II: The Archaeological
Evidence from the Diaspora, New York, 1953, figs 1078–83, 1085,
1087–1109; M. Maaskant-Kleibrink, Catalogue of the Engraved
Gems in the Royal Coin Cabinet The Hague – The Greek, Etruscan
and Roman Collections, The Hague, 1978, figs 1099–1104). An
attempt conducted by Mimi Lavie from the metal conservation lab
at the Hebrew University to examine the reverse of the discussed
gem using X-ray bore inconclusive results.
14 Reading the letters as numbers, the name Abrasax totals 365, as in
the number of days in a year. The name may derive from Hebrew,
possibly Abra (Arba?) Sabaoth = Four (=yhwh, the tetra-
grammaton?) Sabaoth (‘of hosts’): M.W. Meyer and R. Smith,
Ancient Christian Magic – Coptic Texts of Ritual Power, Princeton
NJ, 1999, 389.
15 C.W. King, The Gnostics and Their Remains, Ancient and Mediaeval,
London, 1887 (2nd edn), 245–6. While it is true that the Basilidian
Gnostic system included a cosmic deity called Abrasax, related to
365 heavens and equated with ‘the God of the Jews’ (Irenaeus,
Against Heresies 1.24.7), it is hardly the case that Basilides invented
this name: B.A. Pearson, Gnosticism and Christianity in Roman and
Coptic Egypt, New York and London, 2004, 257.
16 C. Bonner, Studies in Magical Amulets, Chiefly Graeco-Egyptian,
Ann Arbor, 1950, 135; Pearson (n. 15), 257–8.
17 Á.M. Nagy, ‘Figuring out the Anguipede (‘Snake-Legged God’) and
His Relation to Judaism’, Journal of Roman Archaeology 15 (2002),
159–72; G. Bohak, Ancient Jewish Magic – A History, Cambridge,
2008, 249–50.
18 Goodenough (n. 13), 246.
19 An iron ring without a stone was a common betrothal gift in
Roman times. Married women might also wear rings, but ‘wedding
rings’ were not considered essential (L. Cleland, G. Davies, and
L. Llewellyn-Jones, Greek and Roman Dress, from A to Z, London-
New York, 2007, 162).
20 Bavli, Shabbat, 62a; L.Y. Rahmani, ‘On Some Byzantine Brass
Rings in the State Collections’, ‘Atiqot 17 (1985), 168–81;
T. Grossmark, Jewellery and Jewellery-Making in the Land of Israel
in the Time of the Mishnah and Talmud, PhD dissertation, Tel Aviv
University (Hebrew), 1994, 50–7.
21 F. Henkel, Die römischen Fingerringe der Rheinlande und der
benachbarten Gebiete, Berlin, 1913, 19–22, pl. VII: 127, 134.
22 K. Lester and B.V. Oerke, Accessories of Dress: an Illustrated
Encyclopedia, Mineola NY, 2004, 332.
23 E. Swift, Roman Dress Accessories, Princes Risborough, 2003, 30,
fig. 29.
24 Y. Yadin, Judean Desert Studies – The Finds from the Bar-Kokhba
Period in the ‘Cave of Letteres’, Jerusalem (Hebrew), 1963, 95–6, fig.
33:44–5, pl. 26: 44–5.
25 We would like to thank Aryeh Drori for his permission to examine
and publish the intaglios and cameos in the collection. See also
Tepper 2003 (n. 1), 83–5.
104 | ‘‘Gems of Heaven’
26 M. Henig and M. Whiting, Engraved Gems from Gadara in Jordan –
The Sa’d Collection of Intaglios and Cameos, Oxford, 1987, 38, fig.
408.
27 J. Spier, Ancient Gems and Finger Rings – Catalogue of the
Collections of the J. Paul Getty Museum, Malibu, CA, 1992, 161.
28 A.T. Croom, Roman Clothing and Fashion, Stroud, 2000, 100–1, fig.
47.9.
29 This is a very common motif on gems and especially on green
plasma stones. The type is connected with the Julian family and
appears on Early Imperial coins (S. Amorai-Stark, ‘Gems, Cameos
and Seals’, in R. Gersht (ed.), The Sdot-Yam Museum Book of the
Antiquities of Caesarea Maritima – In Memory of Aharon Wegman,
Tel-Aviv, 1999, 87–113 (Hebrew with English summary, 12*–13*), at
91, fig. 11; Maaskant-Kleibrink (n. 13), 221.
30 Maaskant-Kleibrink (n. 13), 221, fig. 531.
31 Amorai-Stark (n. 29), 91, fig. 12; Henig and Whiting (n. 26), 12, figs
70–5.
32 Y. Porath, E. Yannai and A. Kasher, ‘Archaeological Remains at
Jatt’, ‘Atiqot 37 (1999), 1–78 (Hebrew with English summary, 167*–
9*), at 55–6; I. Roll and E. Ayalon, Apollonia and Southern Sharon –
Model of a Coastal City and its Hinterland, Tel Aviv (Hebrew), 1989,
48–9; M. Schwabe and B. Lifshitz, Beth She’arim II: The Greek
Inscriptions, Jerusalem (Hebrew), 1967, 43, 93, nos 123, 214.
33 Hershkovitz and Amorai-Stark (n. 2), 222–3, fig. 13. See eadem, this
volume, Pl. 8.
34 Peleg (n. 2), 63–4, figs 1:11, 2:11.
35 Y. Meshorer, A Treasury of Jewish Coins from the Persian Period to
Bar-Kochba, Jerusalem, 1997, 37–8 (Hebrew).
36 N. Avigad, ‘Two Bullae of Jonathan, King and High Priest’, in
H. Geva (ed.), Ancient Jerusalem Revealed, 1994, Jerusalem, 257–9.
37 G. Pinch, Handbook of Egyptian Mythology, Santa Barbara, 2002,
146–7.
38 Peleg (n. 2), 62, figs 1: 10, 2: 10.
39 Y. Tsafrir, ‘The Topography and Archaeology of Aelia Capitolina’,
in Y. Tsafrir and S. Sfrai (eds), The History of Jerusalem, The Roman
and Byzantine Periods (70–638 ce), Jerusalem, 1999, 115–66
(Hebrew), at 160–1.
40 D. Bar, ‘The Southern Boundary of Aelia Capitolina and the
Location of the Tenth Legion’s Camp’, Palestine Exploration
Quarterly 130 (1998), 8–19, at 14–16, fig. 1; Z.Y. Eliav, ‘A “Mount
without a Temple”: The Temple Mount from 70 ce to the Mid-Fifth
Century – Reality and Idea’, PhD dissertation, Hebrew University
of Jerusalem, 1998, 107–9, ill. B (Hebrew).
41 E. Mazar, ‘The Camp of the Tenth Roman Legion at the Foot of the
South-West Corner of the Temple Mount Enclosure Wall in
Jerusalem’, in A. Faust and E. Baruch (eds), New Studies on
Jerusalem, Proceedings of the Fifth Conference, Bar-Ilan University,
Ramat-Gan, 1999, 52–67 (Hebrew).
42 Mazar (n. 6), passim.
43 Peleg (n. 2), 54–8.
44 Ibid., 58.
45 Ibid., figs 1:3,8,10; 2:3,8,10.
46 Ibid., figs 1:2; 2:2.
47 Ibid., figs 1:7; 2:7.
48 Ibid., 59–60, figs 1:6, 2:6.
Selected Antique Gems from Israel
Excavated Glyptics from Roman-Byzantine Tombs1

Shua Amorai-Stark and Malka Hershkovitz

Introduction The first intaglio, set in an iron ring, is of transparent red

This paper is primarily concerned with gems from excavated, glass and depicts a portrait of Harpocrates in profile to left, his
stratified tombs, with a few parallels from other stratified right hand raised towards his mouth (?), with a small
locations. The chosen gems were in use by the local population cornucopia at his back (Pl. 3). The ring comes from the tomb’s
in Late Hellenistic to Early Byzantine Palestine. Several early phase, and most likely dates from the 1st century bc.2
hundred glyptic items have been excavated in Israel, many of Harpocrates is a fairly common figure on Roman gems, but not
which are still awaiting publication. Less than 30 have been in a Jewish burial context. He is depicted in diverse postures
found in tombs, and we discuss here only those pieces that have and scenes on gems, magical gem-amulets and metal rings
been published.
Certain excavated sites dating from our period yielded
gems in substantial numbers, others only a few to single gems.
The best example is Roman-Byzantine Caesarea Maritima,
although none of the gems from this large harbour city come
from unequivocal burial contexts. The tombs represent
different burial types, customs and traditions: burial caves,
rock-cut tombs, single burials, mausoleums, cemeteries; burial
in stone ossuaries, clay, wood and lead coffins.
As opposed to gems found during excavations at Masada,
Gamla and other places where the majority of recovered
glyptics belonged to Roman soldiers, the gems discussed here
were in use by the local population from the 1st century bc to
the 4th century ad. Some gems were found without their
mounts, but a considerable number were unearthed in their
settings, principally iron and copper-alloy rings. Only
Plate 1 Stone ossuary from Tomb Two, Mount Scopus, Jerusalem. Israel
established and affluent people could afford burial in ossuaries Antiquities Authority (hereafter IAA)
or coffins placed within caves or rock-cut tombs. Therefore, at
least in Roman Palestine, pagans and Jews buried with iron
rings set with intaglios also belonged to the middle to upper
economic strata of the population.

Gems found in Jewish tombs

Only four gems were excavated in Jewish tombs (1st century bc
to 4th century ad). They provide a small but meaningful group
of the published gems and are of particular interest because the
majority of Jews in the Second Temple Period and during the
Roman to Byzantine periods were not buried with intaglios,
thereby adhering to the second commandment (Exodus 20:
4–5; Deuteronomy 5: 8–9).

Gems from two tombs on Mount Scopus, Jerusalem

The Mount Scopus cemetery was part of the vast Second
Temple Period Jewish burial grounds surrounding Jerusalem.
Perhaps the most significant finds from there are two intaglios
set in iron rings which come from two rock-cut cave, family
tombs with kokhim and pits containing secondary burials of
the bones in communal bone depositories and ossuaries made
from Jerusalem stone. Many of the ossuaries in both tombs
were decorated, some were inscribed, others plain (Pls 1–2).
Both tombs represent typical Jerusalem Second Temple Period
Jewish burial customs. They date from the 1st century bc up to
the destruction of Jewish Jerusalem by the Romans in ad 70. Plate 2 Stone ossuary from Tomb Two, Mount Scopus, Jerusalem. IAA

‘Gems of Heaven’ | 105

Amorai-Stark and Hershkovitz
Plate 3 Red glass intaglio (and impression) with a bust of Harpocrates from
Tomb One, Mount Scopus, Jerusalem
from Israeli and nearby sites, some dating from the late 1st
century ad and most from the 2nd to 3rd century ad.3 However,
the sub-type of a Harpocrates bust is not a frequent one. The
only known parallel, from a neighbouring site, is a 1st-century
ad nicolo from Gadara (Pl. 4).4 The second intaglio from Mount
Scopus, a carnelian, depicts another solar deity: the bust of
young Apollo in profile to left, with short hair held by a thin
diadem (a bay laurel?), and the slight remains of a diagonal
branch rising from his shoulder (Pl. 5).5 This subject in diverse
engraving styles is a very common motif on gems dating in the
main from the Hellenistic period up to the 2nd century ad.
However, these Apollo bust gems usually do not originate in
unequivocal Jewish contexts. For example, the two specimens
found at Masada probably belonged to Roman soldiers (Pls
Plate 4 Silver ring with a nicolo Plate 5 Iron ring with carnelian
intaglio with a bust of Harpocrates intaglio with a bust of Apollo from
from Gadara. Sa’d Collection Tomb 2, Mount Scopus, Jerusalem. IAA
Plates 6a-b Intaglios with a bust of Apollo from Masada. Institute of
Archaeology, Hebrew University, Jerusalem
106 | ‘Gems of Heaven’
Plate 7 Intaglio with an amphora from Plate 8 Intaglio with a chalice and
Masada. Institute of Archaeology, palm tree from Masada. Institute of
Hebrew University, Jerusalem Archaeology, Jerusalem
6a-b).6 It is unclear what meaning the affluent Jews buried
opposite the Temple Mount invested in these two pagan motifs
which had different aspects of solar significance. Perhaps for
their Jewish owners these devices alluded to belief in the
victory of light and possibly, therefore, also over death? It is
important to note that in both of these Mount Scopus burials
members of Jewish families chose for their burials iron rings
set with gems illustrating figurative ‘portraits’ which can easily
be recognised as deities of solar symbolism.
The finds from Gamla and Masada show that in their
settlements Jews of the Second Temple Period adapted symbols
known from the repertory of Jewish art of this period, or motifs
from the Graeco-Roman repertoire which could more easily be
invested with non-pagan meaning (Pls 6–10).7
Plate 9 Intaglio with three ears of corn from Gamla. IAA
Plate 10 Intaglio (with impression) with a grape vine from Gamla. IAA

Plate 11 Haematite intaglio with a reaper similar to examples from Tiberias,
Shiqmona and Beit She’an. London, British Museum, PE G46
Qiryat Shmuel, a quarter of Tiberias, on Lake Galilee
The third excavated intaglio was discovered in a Jewish
mausoleum situated on the edge of a large Jewish cemetery at
Tiberias. This mausoleum was constructed during the late 1st
to early 2nd century ad. This magical haematite was
discovered in the debris that had accumulated during
earthworks on the site. It depicts a well-known theme: on the
obverse is a reaper bending forwards about to cut four stalks of
corn with his sickle; on the reverse is inscribed in Greek: cxiwn
(‘for the hips’) (Pl. 11). The inscription, known from other
similar gems, indicates that this amulet was used as remedy for
pains such as sciatica or lumbago. The engraving style of this
amulet, taken with other finds from the site, comparison with
coins and other reaper amulets, date it to the early to mid-2nd
century ad.8
The majority of reaper amulets seem to come from the
eastern Mediterranean region (Syria, Palestine, Asia Minor or
Egypt), but only a few of these have secure provenances. So far,
seven reaper amulets from sites in Israel have been published:
another smaller piece comes from Roman Tiberias; one from
Roman-Byzantine Caesarea Maritima (Pl. 12); three from
Byzantine Shiqmona; and finally, one from Byzantine Beit
She’an.9 The relative popularity of this basically rustic motif
shows that many people suffered from backaches and that, of
all the agricultural tasks common in this region, reaping was
considered the prime source for such aches. Analysis of these
seven reaper gems and their contexts shows that locally reaper
gem-amulets were considered to possess a powerful magical
medical potency by people of all beliefs, and were locally in
common use from at least the 2nd century ad up to the 6th
century ad. It appears that humans have always suffered from
backache and not much has changed.
Plate 12 Drawing of a haematite intaglio with reaper from Caesarea Maritima.
IAA
Selected Antique Gems from Israel
Plate 13 Iron ring with intaglio with a Plate 14 Copper-alloy ring with lion
lion and a star from Iksel. IAA from Iksel. IAA
Iksel, Beit Shearim area
The fourth intaglio set in a broken iron ring comes from a
Jewish cave burial at the post-Second Temple period Jewish
cemetery of Beit Shearim, one of the largest Jewish centres of
this period. This cemetery functioned as the most important
burial place in 2nd to 4th-century ad Israel both for the local
Jewish population, its leaders and for affluent Jews of the
diaspora. The intaglio depicts a standing lion to left with a star
above (Pl. 13). A copper-alloy ring with an engraved motif
representing a different lion on its bezel (a lion to left holding
down a rigid human figure resembling prey or a felled enemy)
was also found in this burial (Pl. 14).10
Although the material and engraving styles of these two
lions differ, both rings are of similar basic Roman types,
common mainly from the late 2nd to the 3rd and 4th centuries
ad; both represent engraving styles typical of the Late Roman
to Early Byzantine period in the region; and both themes are
typical of magical amulets. These two amuletic rings date from
the 3rd–4th century ad.
Lions in different postures and scenes with additional
symbols are one of the most common animal subjects depicted
on Early Roman to Byzantine period gems. Finds from other
sites show that during Late Roman to Early Byzantine times the
lion subjects represented by the Beit Shearim gem and copper-
alloy ring became popular among the general population.
Lions now commonly appear on stone amulets, metal rings and
glass pendant amulets (Pls 15–16).11 Their diverse symbolism
primarily alludes to the lion’s solar significance, to his
astrological and magical associations. However, no example of
a gem engraved with a lion device has as yet been discovered in
a 4th to 7th century ad pagan or Christian burial site in Israel. It
is interesting that of the four gems discovered in Jewish burials
Plate 15 Glass amulet with lion and Plate 16 Copper-alloy ring with lion
star from Hirbet Kenes. IAA from Sajur
‘Gems of Heaven’ | 107
Amorai-Stark and Hershkovitz
Plate 17 Iron ring with agate intaglio
with Athena-Minerva holding Nike
from Hagoshrim. IAA
dating from the 1st century bc to the Late Roman period three
bear devices with solar significance (the two pre-ad 70
Jerusalem specimens and the Beit Shearim gem), two are
magical amulets (the Tiberais and Beit Shearim gems), and all
depict motifs used by the general local population.
Gems from pagan tombs
Due to their fairly large number we only discuss here examples
from burial sites in the Galilee and the Jerusalem area, during
the Hellenistic to Late Roman periods.
Examples from burial sites in Galilee
Kibbutz Hagoshrim (upper northern Galilee)
The site is situated below the Golan Heights, on the bank of the
river Dan, a tributary of the river Jordan. Three gems were
unearthed in one burial cave which had acted as a burial
ground for pagans from Late Hellenistic to Early Roman times.
Nails and iron carrying-loops? show the deceased were usually
buried in wooden coffins with their jewellery, dating the burial
activity in the cave up to the second half of the 1st century ad.12
Four rings were found of which three are broken iron rings set
with gems and one a copper-alloy example without a gem. The
gems are two agates and a glass intaglio (all of type F2) and
they depict three typical early Roman subjects. The agates
portray an Athena-Minerva of the ‘Parthenos’ type, standing to
right and holding a small Nike (Pl. 17), and a hound chasing a
hare (Pl. 18). The engraving style of these two agates dates
them to the 1st century ad, but not necessarily to the same
workshop. Athena in diverse sub-types is one of the most
popular subjects depicted on Late Hellenistic and Roman
glyptics in our region. The scene of the dog leaping after a hare
probably reflects actual hunting practices. The lush natural
Plate 18 Iron ring with agate intaglio Plate 19 Cameo with leaping hare
with a hound chasing a hare from from Caesarea. IAA
Hagoshrim. IAA
108 | ‘Gems of Heaven’
Plate 20 White glass intaglio
with a sphinx from Hagoshrim.
IAA
flora of the Hagoshrim area still to this day provides splendid
opportunities for hunting. Yet, since the hare was in addition
considered a symbol of plenty, the owner of this ring might
have chosen the device also as a symbol of his position, as a
wish for plenty in life as well as in his afterlife. Another 1st to
2nd century ad intaglio with the device of a hound chasing a
hare and a 3rd-century ad cameo depicting a leaping hare with
a tree behind were found at Caesarea (Pl. 19).13
The subject of the white glass intaglio from Hagoshrim is a
seated sphinx to left, with its right leg above an unidentified
object (head?) (Pl. 20). It too probably dates from the 1st
century ad or slightly earlier. A carnelian representing another
seated sphinx was excavated at the nearby Jewish settlement of
Gamla in a private house context (up to ad 67) (Pls 21a–b).14
Qadesh (upper eastern Galilee in the Qadesh valley)
Tel Qadesh is the largest city in the mountainous region of
northern Israel: it dates from the Early Bronze Age to the Late
Roman period. In the 1st century bc to the 1st century ad it was
populated since the Hellenistic period mainly by people of
Phoenician origin. Burials belonging to the city’s population
were found in the vicinity of the 2nd-century ad temple of Bal-
Shamin, which is located below the city.
The finds in this cave include the remains of a copper-alloy
jewellery box, silver bracelets, beads and six copper-alloy rings,
suggesting that the deceased was a female (Pl. 22). Of these six
rings, two are set with gemstones, now broken, which were
probably engraved. The third ring had an intact intaglio
depicting a star and crescent moon (Pl. 23). The interesting
fact about this intaglio is its material: a piece of shell, an
uncommon gem material in this period. It comes from a 1st to
2nd century ad level in the Roman burial cave.15
Plates 21 a–b Carnelian intaglio (and impression) with a sphinx from Gamla.
IAA
 Selected Antique Gems from Israel

Plate 22 Contents from a copper-alloy jewellery box and tomb from Qadesh. IAA

Plate 23 Copper-alloy ring with intaglio with star and crescent from Qadesh. IAA

‘Gems of Heaven’ | 109

Amorai-Stark and Hershkovitz
Plate 24 a–b Red jasper intaglio with olive tree and Greek inscription from
Nahariyyah. IAA
The crescent moon with varying numbers of stars was a
fairly frequent device on 1st to 2nd century ad Roman intaglios.
It is found, for example, on three carnelians from Gadara.16 In
Qadesh the star and crescent moon presumably relate to the
worship of Baal-Shamin (‫שמין‬-‫ = בעל‬Lord of the Sky) and
perhaps also to aspirations relating to life after death. In this
context it is interesting to note that the Qadesh temple of Bal-
Shamin might have been connected also with the cult of the
dead.
Givat Usishkin, Nahariyyah (on the northern coast)
The site comprises the remains of a large Roman-Byzantine
settlement with many excavated burials which include burial
caves with alcoves and arcosolia. The single intaglio found
there is a red jasper (Type C6 or C1) depicting an olive tree with
fruits and leaves. Flanking the trunk is a positive Greek
inscription: ‘Mercy of the Lord’ (Pls 24a–b). The other finds in
the tomb included glass vessels, clay lamps and especially
coins, indicating that the burials were of pagans.17 The gem’s
epitaph is obviously directed to a male divinity. This unique
magical amulet most likely dates from the first half of the 3rd
century ad. The identity of the ethnic or religious back-ground
of its owner is unclear.18 Although Roman period gems
depicting a single tree are known, this is the only securely
excavated example discovered in Israel. Its device and
inscription perhaps play on the Greek words for mercy – eleos,
and for olive – elaion, a connection which in itself cannot
identify the owner’s beliefs. Rahmani in his discussion of
Jewish and gnostic legends concluded that this remarkable
magical-amulet is perhaps of ‘gnostic’ origin, with a Jewish
background assuring its owner, in life or death, of the mercy of
the Christian Lord.19 Yet if this gem is addressed to the
Christian Lord it would be the only glyptic find with a Christian
background from an excavated grave to have been discovered
in Israel. A small number of gems of secure 4th to 7th century
ad dates were found in Israel in different excavations, but all of
them originate in settlements or Christian living structures.
Furthermore, although the Christian usage of some of these
excavated gems is assured, none of them bear clear-cut
Christian symbols or religious motifs. The olive tree is
connected with several pagan gods. Due to the close
geographic and administrative connections between the site
and the city of Tyre and because the olive tree was the holy tree
of Tyre and of its titular god Melqart-Heracles, the gem might
well be an amulet directed to this divinity.
110 | ‘Gems of Heaven’
Plate 25 Copper-alloy ring with glass Plate 26 Copper-alloy coin with
intaglio with a camel from camel from Bostra. IAA
Nahariyyah. IAA
Asherat (south-east of Nahariyyah, in the mountainous area of
western Galilee)
At this site were discovered three Roman burial caves with clay
coffins. A single glass gem set in a 1st to 2nd century ad type of
copper-alloy ring was found in cave burial no. 1.20 The jewellery
finds from this cave included gold earrings, copper-alloy
bracelets, and four rings: two broken iron rings, and two
copper-alloy ones. The device on the glass gem (type F2) set in
one copper-alloy ring is identified by us as a camel facing left
(Pl. 25). The camel is an uncommon Roman intaglio motif and
to the best of our knowledge this is the only intaglio depicting a
camel from this region. The camel does appear, however, on
Roman and Nabataean coins. The animal’s coarse engraving
style, ring type, the jewellery found with it and coin
comparanda (Pl. 26) date this gem to the later 1st to 2nd
century ad.
Berit Achim (inland to the north of Acre)
A single broken 1st to early 2nd century ad gem was found in
one of the burials in a Hellenistic-Roman burial field on this
site.21 It was discovered with other jewellery such as a pair of
gold earrings, a mirror, glass bottles and stone weaving whorls.
This burial appears to have been the resting place of a female.
The gem, made of black glass (type F4) depicts the Disokouroi,
standing frontally and facing each other, holding swords and
broken spears, on a ground line. This version of the standing
Dioskouroi, a common Roman gem motif, occurs also in
Gadara.22
Givat Yassaf (near Kibbutz Lochami H’Getaot, south of Acre)
The site is the remains of an ancient town dating from the Early
Bronze Age to the Byzantine period and was probably an
agricultural centre. A single Roman red carnelian intaglio, set
in a broken copper-alloy ring of 2nd to 3rd century ad type, was
discovered in a clay coffin placed in an alcove (niche no. 7) of a
rock-cut burial room cut within a natural cave. The cave is
located to the east of the city. Other burials in this room were
also within clay coffins. Collectively the small finds in the
various alcoves of this burial chamber comprise copper-alloy
bracelets, a simple loop ring, glass plates and bottles, clay
lamps and coins.23 The gem, which depicts Nike-Victory about
to crown a trophy (Tropaion), was found among the bones of
the deceased together with late 2nd to early 3rd century ad
Roman provincial coins (ad 198−222, of Caracalla and Julia
Domna). The gem dates from the same period. Similar diverse
representations appear on 2nd to 3rd century ad coins from

Tyre. The image of Nike-Victory with a trophy is also found on
gems from Gadara.24 Glyptic depictions of Nike-Victory in
different postures and scenes and as a component of
pantheistic goddesses are among the most common glyptic
finds from different sites in Israel. Due to the gem’s motif and
absence of jewellery in the alcove, the deceased was possibly a
male, perhaps from a military background.
Intaglios from burials in the Jerusalem area
Site of the Jerusalem International Convention Centre
(Binyanei Ha’uma), Jerusalem
The site began as a Jewish village on the outskirts of ancient
Jerusalem, became a Roman military site of the Legio X
quarries, and later was a monastic complex. A single gem was
found in a mixed locus leading to an entrance of a burial cave
which dates from the Herodian to the Byzantine periods.25 The
intaglio, an opaque black jasper (type F1) depicts a standing
Tyche-Fortuna Pantea with cornucopia and rudder, Demeter’s
corn-ears in her right hand, and probably a coarse version of
Isis’ crown and knot (Pl. 27). Although the engraving is fairly
worn and somewhat coarse, on stylistic grounds and
comparisons with other locally found gems, the preferred date
for this jasper is the 1st to 2nd century ad.
Tyche-Fortuna, primarily combined with Demeter as on
this gem, but also as a pantheistic Tyche-Fortuna combined
with three to four other goddesses engraved in diverse styles, is
encountered frequently amongst gems discovered in Israel.
Some of these intaglios originate from other Jerusalem
locations (Pls 28 and 29).26 The ownership and burial context
of this small gem (7 x 5mm) is uncertain. In view of the
popularity of this motif amongst Roman soldiers, and the use
Plate 27 Black jasper intaglio with pantheistic goddess from Binyane Ha’uma,
Jerusalem. IAA
Plate 28 Intaglio with pantheistic Plate 29 Intaglio with pantheistic
goddess from Hirbet Hadass, goddess from ‘Dominus Flevit’,
Jerusalem. IAA Jerusalem. Jerusalem, SBF Museum
Selected Antique Gems from Israel
Plates 30 a-b Gold ring with nicolo intaglio with a shrimp from Manahat,
Jerusalem. IAA
of the site by the Legio X from the 1st century ad to the early 3rd
century ad, its owner could well have been a Roman soldier.
Manahat (near Holyland, on the western outskirts of
Jerusalem)
A single intaglio set in a typical Roman gold ring (Pl. 30a) was
discovered on the left little finger of a young male (25 years
old) buried with an older male (35 years old) in a lead coffin
decorated with erotes at the base of a Roman rock-cut shaft
tomb. The lead coffin, the burial of two males in one coffin
within a short period, the gold-embroidered outer garment
which clad the elder male, the jewellery and other finds in the
coffin date the grave up to ad 240, and identify the men as
Romans who probably settled on the outskirts of Jerusalem,
and were possibly connected with the caravan cities to the
north of Israel, perhaps Palmyra or Dura.27
The gem, which is a nicolo (type F2), depicts a decapod
crustacean, probably a shrimp rather than a prawn (Pl. 30b).
The shrimp (Palaemon squilla) is a familiar Roman gem motif
found on other intaglios from Israel and nearby regions such as
Caesarea and Gadara.28
Nahal Raqafot (western outskirts of Jerusalem)
At this burial site were found three intaglios set respectively in
two gold rings and a gold brooch. Two rock-cut tombs were
unearthed, each with a single wood cast lead coffin
strengthened with wooden uprights. One of the coffins
contained the skeletal remains of two males, presumably
Roman soldiers stationed in Jerusalem, buried with rich gold
jewellery and gold-ornamented leather fabric.29 The two gold
rings were discovered near the left hand and hip of the same
male placed in this coffin. The first is a 2nd-century ad type
solid gold ring with a nicolo depicting a cuirassed elephant
walking to left with raised trunk representing a parade
elephant (Pl. 31). Coin and gem depictions of elephants from
the public shows at Rome are known, but to date this is the only
glyptic example of the motif found at an Israeli site. The
Plate 31 Gold ring with nicolo intaglio with elephant from Nahal Raqafot,
Jerusalem. IAA
‘Gems of Heaven’ | 111
Amorai-Stark and Hershkovitz

Plate 32 Glass intaglio with goatherd milking a goat from Nahal Raqafot,
Jerusalem. IAA

Plate 33 Intaglio with goatherd milking a goat from Bab el Hawa. IAA Plate 34 Gold brooch set with an onyx cameo with Athena from Nahal Raqafot,
Jerusalem. IAA

nearest coin analogy from the Roman mint to this gem’s the world usually come from more northern regions. The
depiction of parade elephants in detail and style is a coin of polychrome brooch discovered in Dura-Europos is such an
Commodus dating to ad 183/4. This ring and gem date from the example (Pl. 35).33 So far the Nahal Raqafot brooch is a unique
later 2nd century ad and their place of origin is most likely example from an Israeli site of this rich polychrome brooch
Rome. type. Yet a contemporary gold foil diadem roughly set with gem
The second ring is a hollow gold ring with a wide elliptical stones and their imitations was excavated in a lead coffin
loop filled with a paste and with a slightly raised centre line of burial at the northern site of Kefar Giladi (Pl. 36).34 The Kefar
an earlier 2nd century ad type. Its gem is a loose, slightly Giladi diadem might be of local workmanship; the Nahal
broken worn blue glass. The device is a goatherd sitting to left Raqafot brooch is more likely an import, plausibly from Syria
under a tree and milking a goat standing to left (Pl. 32). It dates or the Black Sea area.
from the early to mid-2nd century ad. The image of a single
goatherd was a popular subject on Roman glyptics and is Conclusion
encountered in diverse engraving styles and details; it is also a Most of the gems originate from pagan tombs; a few were
favourite motif on local gems (to date a total of five gems have found in Jewish tombs; no gem was found in an unequivocally
been published, with three examples coming from Caesarea). Christian grave. The four gems discovered in indisputable
The Bab el Hawa goatherd stone, found at a 4th to 6th century Jewish tombs attest to the fact that from the 1st century bc to
ad, predominantly agricultural, settlement in Galilee (Pl. 33)30 the 3rd to 4th century ad few Jews chose to use rings set with
is secure evidence that this rustic motif which probably reflects figurative and animal intaglios in their private surroundings,
actual agricultural practices continued in the region well into and to be buried with them. The use of intaglios and cameos in
Byzantine times. the Late Hellenistic-Roman period, popular throughout the
The polychrome brooch was discovered together with a Empire, is reflected locally by the glyptic finds from pagan
pair of gold earrings and the remains of a gold diadem lying on
the closed lead coffin discussed above which contained the two
rings. (Compare the gold earring found at the burial with a
similar earring discovered at another early 3rd century ad
tomb at Shmuel Ha-Navi, Jerusalem.31) These jewels were most
likely torn from the garments of mourners at the time of the
burial – a custom connected with interment in lead coffins in
early 3rd century ad Jerusalem. An onyx cameo depicting the
helmeted bust of Athena to right and a round shield with a
gorgoneion in the centre is set in the brooch’s rich gold frame;
the frame is decorated with filigree wire, emeralds and garnets
(Pl. 34). Part of Athena’s nose and upper lip were chipped off in
antiquity. The style of this Athena is fairly coarse. The subject
of a helmeted Athena bust cut in differing details and
workmanship was a favourite 2nd to 3rd century ad cameo
motif. For example, a better executed cameo with the same
subject without its mount was found in the Jerusalem area.32
Such 2nd to 3rd century ad onyx cameos were usually set in
gold brooches. Yet polychrome brooches of all sub-types set Plate 35 Gold brooch set with a magical intaglio from Dura-Europos.
with cameos or large intaglios are rare finds and in our part of Damascus, National Museum of Damascus

112 | ‘Gems of Heaven’

 Selected Antique Gems from Israel

Plate 36 Gold diadem from Kefar Giladi. IAA

tombs; these items date mainly up to the Late Roman period. Periods at Hagoshrim’, ‘Atiqot 38 (1999), 223–4 (in Hebrew).
The general comparative decline of the use of gems during the 13 S. Amorai-Stark, ‘Gems, cameos and seals’, in R. Gersht (ed.), The
Sdot-Yam Museum Book of the Antiquities of Caesarea Museum, Tel-
Byzantine period is also reflected in the glyptic finds from
Aviv, 1999, 87–114, at 97–8, fig. 23 (in Hebrew with English
Israeli sites: only a few excavated gems come from Byzantine abstract).
Israel, all originating from settlements and compounds. 14 Amorai-Stark and Hershkovitz (n. 6).
15 G. Edelstein, ‘Two Roman-Period Burial Caves near Tel Qedesh’,
‘Atiqot 43 (2002), 99–105 (in Hebrew).
Notes 16 Henig and Whiting (n. 3), nos 322–4.
1 We wish to express our gratitude to Uzi Dahari, the IAA Deputy for
17 L.Y. Rahmani, ‘A Magic Amulet from Nahariyyh’, Harvard
permission to use the material; to Adi Ziv, Curator of the IAA Bet-
Theological Review 74/4 (1981), 387–90.
Shemesh Late Periods Store Rooms for her help and advice in
18 Y. Meshorer, Testimony, Jerusalem, 2000, 17, no. 2; J. Spier, Late
locating the gems; and to Clara Amit, an IAA photographer for
Antique and Early Christian Gems, Wiesbaden, 2007, 52, n. 57.
taking new photographs of some of the gems for this article.
19 Rahmani (n. 17).
2 L.Y. Rahmani, ‘Jewish Rock-Cut Tombs on Mount Scopus’, ‘Atiqot
20 H. Smithline, ‘Three Burial Caves from the Roman Period in
14 (1980), 49–54, mistakenly identified as Fortuna.
Asherat’, ‘Atiqot 33 (1997), 47–61 (in Hebrew).
3 Caesarea: A. Hamburger, ‘Gems from Caesarea Maritima’, ‘Atiqot 8
21 G. Edelstein, ‘A Selection of the Hellenistic-Roman Cemetery at
(1968), 1–38, nos 116–17; S. Amorai-Stark, Engraved Gems and Seals
Berit Achim,’ Atiqot 43 (2002), 76–98.
from Two Collections in Jerusalem: The Studium Biblicum
22 Henig and Whiting (n. 3), nos 236−8; cf. also Hamburger (n. 3), no.
Franciscanum Museum Gem Collection (SBF) and the Pontifical
54.
Institute Museum Gem Collection (PBI), Jerusalem, 1993, SBF, no.
23 A. Abu Uqsa, ‘A Burial Cave from the Roman Period East of Givat
129; Gadara: M. Henig and M. Whiting, Engraved Gems from
Yasaf’, ‘Atiqot 33 (1997), 41–2 (in Hebrew).
Gadara in Jordan. The Sa’d Collection of Intaglios and Cameos,
24 Henig and Whiting (n. 3), nos 141−2 (1st–2nd century ad).
Oxford, 1987, no. 217; Carmel Area: S. Amorai-Stark and M.
25 B. Arubas and H. Goldfus, ‘A Gemstone’, in B. Arubas and H.
Hershkovitz, ‘A Roman Ring depicting Hermes Psychopompos
Goldfus (eds), Excavations on the Site of the Jerusalem International
from the Carmel Area’, forthcoming.
Convention Centre (Binyanei Ha’Una), Journal of Roman
4 Henig and Whiting (n.3), no. 218.
Archaeology, Supplementary Series 60, 2005, 15–26.
5 V. Sussman, ‘A Jewish Burial Cave on Mount Scopus’, in H. Geva
26 H. Khalaily and M. Avisar, ‘Khirbit Adasa: A Farmstead of the
(ed.), Ancient Jerusalem Revealed, Jerusalem, 2000, 226–30.
Umayyad and Mamluk Periods in Northern Jersualem’, ‘Atiqot 58
6 M. Hershkovitz and S. Amorai-Stark, ‘The Gems from Masada’, in
(2007), 91–122, fig. 18, from Hirbet Adessah, northern Jerusalem
J. Aviram, G. Foerster, E. Netzer and G.D. Stiebel (eds), Masada
(in Hebrew); Amorai-Stark (n. 3), SBF, no. 96, from Dominus
VIII: The Yigael Yadin Excavations 1963–1965, Final Reports,
Flevit, Mount of Olives.
Jerusalem, 2007, 217–28, nos 4–5, 1st century bc–1st century ad
27 L.Y. Rahmani and J. Gath, ‘A Roman Tomb at Manahat, Jerusalem’,
7 Gamla: S. Amorai-Stark and M. Hershkovitz, ‘The Gems and
Israel Exploration Journal 27 (1977), 209–14.
Jewellery from Gamla’, in D. Syon (ed.), Gamla, Final Report (IAA),
28 Hamburger (n. 3), no. 157; Henig and Whiting (n. 3), no. 385.
Jerusalem, forthcoming; Masada: eadem (n. 6), nos 12−13. See
29 L.Y. Rahmani, ‘Roman Tombs in Nahal Raqafot – Jerusalem’,
also, M. Hershkovitz, ‘Gemstones’, in H. Geva (ed.), Jewish Quarter
‘Atiqot 11 (1976), 81–90; Y. Israeli and M. Tadmor, Treasures of the
Excavations in the Old City of Jerusalem Conducted by Nahman
Holy Land, Ancient Art from the Israeli Museum (exh. cat., The
Avigad, 1969–1982. II: The Finds from Areas A, W and X-2. Final
Metropolitan Museum of Art), Jerusalem and New York, 1988,
Report, Jerusalem, 2003, 296–301, at 299–300.
236–8.
8 F. Vitto, ‘A Jewish Mausoleum of the Roman Period at Qiryat
30 M. Hartal, ‘Bab el-Hawa’, Hadashot Arkheologiyot 97 (1991), 6–8
Shemu’el, Tiberias’, ‘Atiqot 58 (2008), 7–29.
(in Hebrew).
9 Cf. ibid., 22.
31 L.Y. Rahmani, ‘Roman Tombs in Shmuel Ha-Navi Street,
10 A. Mokary, ‘Iksel’, Hadashot Arkheologiyot 115 (2003), 27 (in
Jerusalem’, Israel Exploration Journal 10 (1960), 140–8, pl. 21:D.
Hebrew).
32 M. Henig, The Content Family Collection of Ancient Cameos, Oxford,
11 For an example of a glass pendant amulet from a tomb at Hirbet
1990, no. 134.
Kenes, Galilee, see: L. Porat, ‘Quarry and Burial Caves at H. Kenes
33 M. Fortin, Syria, Land of Civilizations (exh. cat., Quebec), Quebec,
(Karmiel)’, ‘Atiqot 33 (1997), 81–9, pl. 3:6 (in Hebrew); also,
1999, no. 339.
E. Brown, C. Dauphin and G. Hadas, ‘A Rock-Cut Tomb at Sajur’,
34 Israeli and Tadmor (n. 29), no. 123.
‘Atiqot 25 (1994), 112.
12 R. Ovadia, ‘A Burial Cave of the Hellenistic and Early Roman

‘Gems of Heaven’ | 113

Christian Gems from Portugal in Context
Graça Cravinho and Shua Amorai-Stark
Introduction
The gem corpus from present-day Portugal, discussed by Graça
Cravinho in her forthcoming doctoral thesis, totals more than
920 glyptic items; rings and other related objects bring the
number of finds to over 1,000 pieces. The majority of these
finds are Roman gems (440 items) and rings. The total of
engraved gems dating from the Late Antique period is around
110 pieces (i.e. c. 12.2% of the corpus); intaglios and rings
dating from the same period include specimens depicting
motifs of Roman and barbarian origin engraved in differing
styles. The majority of these are not discussed in this paper.
This paper focuses on selected examples from within this
group of 3rd to 7th century ad finds. They come from various
Portuguese locations and regions, all with diverse historical,
political and geographic backgrounds. In order to stress the
context, we present gems and related material from seven
excavated sites (moving from north to south in modern
Portugal) as well as a few incontestably Christian intaglios
from illegal excavations. The majority of these selected
examples may have been used by Christians, but this is far from
certain. Allocating Christian ownership and meaning to most
examples is problematic due to two major problems typical of
glyptic and finger-rings dating from Late Antiquity, but
especially to the early centuries within the above time frame in
which various other finds are also encountered which come
from different regions of the Late Roman and Early Byzantine
Empires.
Many common Roman motifs, such as fish, dolphins, hares,
birds, palm-branches and simple geometric motifs such as ‘V’,
‘X’ or crosses, were engraved on the Portuguese specimens and
continued to be employed in both the 3rd and 4th centuries ad.
Christians adopted most of the above-mentioned motifs and
gave them a Christian meaning. However, in general, Christian
symbolism before the Constantinian period was often
deliberately ambiguous, and this ambiguity sometimes
persisted well into the 4th century ad, or these motifs
continued to be then used by pagans. For instance, because the
fish and the dolphin had many different levels of meaning for
pagans, Christians and Jews alike, ranging from generic
everyday meanings to diverse layers of religious-theological
symbolic significance, we cannot assign any of the non-
inscribed single fish or dolphin gems from Portugal to
members of one religion or to an explicit group of peoples.
Since these motifs fall within the larger range of ambiguous
themes chosen by Christians we prefer to call them ‘ambiguous’
or ‘neutral’ motifs, that is motifs engraved on gems which
might have been used by persons of various beliefs in these
periods. When, as is the case of the majority of finds from
Portugal discussed in this paper, the finds depict these
‘ambiguous’ or ‘neutral’ motifs without additional evidence for
unquestionable Christian motifs, symbols, monograms, scripts
114 | ‘Gems of Heaven’
and so forth, their Christian ownership is debatable.
The other major problem encountered is also inherent to
many finds. The materials, forms and engraving style of many
of the gems decorated with these ‘ambiguous’ motifs frequently
cannot be assigned with any certainty to a limited period, since
these features were applied to gems over several centuries and
such gems continued to be used after the Constantinian era.
When such gems are found without their mounts, in mixed
pagan and Christian contexts, ascribing them to a given
century is problematic. This chronological problem also
pertains to certain ring-types in which some of the gems from
Portugal were set as, like the intaglios, they too were in
circulation for some centuries. Consequently they too cannot
be closely dated without explicit stratigraphic evidence or an
archaeological association with intrinsically dateable artefacts.
Below, we summarise the historical and political
background of each of the seven sites, emphasising the
evidence for their Christian background, the context in which
the glyptic items were found, and whenever there is evidence
of related, small uncontestable Christian items, we present it. It
should be stressed that whenever the illustrated items lack a
reliable archaeological context which can help with their
chronology, as well as with assigning a Christian use to the
items, we prefer to indicate the widest date range for the item
in question, leaving the matter of the patron’s religious
affiliation open.
Historical background
In general, the eastern parts of the Iberian peninsula were
under Roman rule from the 3rd century bc onwards. The
indigenous peoples living in the regions of present-day
Portugal (mainly Lusitanians1 and Gallaecians) were subdued
only by the late 1st century bc (Pl. 1).2 Their territories were
then integrated by Augustus into the imperial provinces of
Lusitania and Tarraconensis. After the peninsula’s division into
several conventi iuridici under Vespasian, those territories
became, under Diocletian, the Roman provinces of Lusitania
and Gallaecia (Callaecia). By the end of the 3rd century ad,
south Gallaecia and Lusitania (except the eastern territories,
now Spain) almost formed the modern boundaries of Portugal
(Pl. 2).
In the 3rd century ad the majority of the population in
Portugal were pagans, but there is evidence for the presence of
Christians in the country from at least the 2nd century ad. By
the middle of the 3rd century ad, churches with their
accompanying bureaucracy of deacons, presbyters and bishops,
were established in Lusitania. Yet, as in other regions, only
with Constantine and the Council of Nicaea (ad 325) did the
persecution of Christians in Portugal cease. From the early 5th
and during the 6th century ad, the Iberian peninsula was
attacked by various barbarian peoples (Pl. 3), the most
Plate 1 Map of pre-Roman Iberia
Cravinho

important for our discussion being the Sueves and Visigoths.

The Sueves became Christians during their occupation of the
peninsula, first as Arians and then as Catholics; the Visigoths
arrived in the peninsula as Christians. Both had local Christian
kingdoms. The south of Portugal continued under Byzantine
rule until the 7th century ad (Pl. 4). From the beginning of the
8th century (ad 711) Portugal was gradually conquered from
the south by North African Muslims; by the mid-8th century ad
all of Portugal was under Muslim rule. Their influence was
strongest in the areas to the south of Coimbra, because the
northern areas had been intentionally devastated and
depopulated by the Visigoths in order to prevent the total
occupation of the peninsula by the Muslims.

Provenanced finds: the single fish/dolphin device: Braga

The first excavated site from which we present gems is Bracara
Augusta, capital of Roman Gallaecia (all the territories to the
north of the river Douro, the ‘Gold River’) and later the political
and intellectual capital of the Christian kingdom of the Sueves.
Economically and socially active during the 4th century ad, in
the course of which the city became a diocese,3 by the end of
this century, its responsibilities as a political and adminis-
Plate 2 Map of Portugal showing cities under Roman rule until the early 5th
trative centre increased. Braga’s prosperity continued during
century AD its occupation by both the Sueves (from ad 411) and the
Visigoths (from ad 585 to 715).4
Many Early Christian remains were and are still found in
Braga, including a Christian basilica situated under the
present-day cathedral, where two glass vessels decorated with
the chi-rho, and a sarcophagus, dated to the end of the 4th or
the beginning of the 5th century ad, were found.5 By this time
Plate 3 Map of Iberian peninsula Braga was the most important cultural centre of the whole
showing early 5th century AD
barbarian invasions Christian peninsula as evidenced, among other things, by the
many documented eminent Church personalities, authors and
intellectuals coming from the city;6 by the fact that it was the
centre of religious disputes between followers of Orthodoxy,
Priscillianism, Arianism and other groups;7 and by the fact that
monasticism was already important there. However, no gem
with an obvious Christian motif or inscription has as yet been
excavated in Braga.
Bearing in mind the highly stylised, non-naturalistic
manner of depicting dolphins and fish in Roman art in general,
which is also reflected in the schematic representations of the
marine images discussed below, it is not always possible to
ascertain whether these gems depict a fish or a dolphin.
A nicolo paste with a fish/dolphin, dating from the late 2nd
to 4th centuries ad (Pl. 5), was found during an archaeological
survey at Carvalheiras insula, in a location where a group of
Roman houses from the Flavian period were unearthed, the
most important of which became a small thermae in the first
half of the 2nd century ad. This area, which is fairly close to the
Early Christian basilica, was abandoned at the end of the 4th or
the beginning of the 5th century ad.8 A further six gems with
the single fish/dolphin motif, likewise without inscriptions or
definite Christian elements, were found at other Portuguese
sites with continuous Roman to Early Christian occupation
(Pls 6–11).
Plate 6 shows a carnelian found at Ammaia, an important
Roman city of about 22ha (Pl. 12) located on a gentle and fertile
slope irrigated by the river Sever (a tributary of the river Tejo),
Plate 4 Map showing barbarian and Byzantine areas of influence by the ‘ribeira dos Alvarrões’ (a small river), several springs

116 | ‘Gems of Heaven’


Plate 5 Nicolo paste with dolphin/fish from Plate 6 Carnelian with dolphin/fish from Ammaia,
Braga, late 2nd–4th century AD. Braga, late 2nd–3rd century AD. Lisbon, Museu Nacional
Museu D. Diogo de Sousa de Arqueologia (hereafter MNA)
Plate 8a Carnelian with dolphin/fish from Alentejo, Plate 8b Cast of Plate 8a
3rd century AD. Formerly in the Barreto Collection
Plate 10 Nicolo with fish/shark from Alentejo, 2nd–3rd century AD. Formerly
in the Barreto Collection
(‘Olhos de Água’, ‘Olheirão’) and wells.9 Ammaia was
connected to other civitates in Lusitania and throughout the
Iberian peninsula by an excellent road system.10 Its territory
was rich in gold, silver, lead, haematite and quartz. The local
exploitation of quartz in Roman times, referred to by Pliny
(Nat. Hist., XXXVII, 24; 127), has been demonstrated by the
discovery of some quartz quarries within the city perimeter
(for example, in Pitaranha and Naves, Porto da Espada) and by
the many quartz fragments uncovered recently during current
excavations.11 In previous studies we have already shown that
during Imperial times there were glyptic workshops in
Ammaia12 and it is feasible that perhaps they continued into the
4th century ad. The discovery of a number of Early Christian
lamps (one decorated with the chi-rho, dated to the 4th–
mid-5th century ad), as well as of tegulae and imbrices with
Christian symbols testify to the existence of Christians in
Ammaia from at least the 4th century ad onwards. Also
important is a gravestone (now lost) dedicated to Optatus
[Famulus Dei], dated to 25 June, ad 513.13 The existence of a
Christian ‘temple’/church in the city or its environs is fairly
likely, perhaps situated in the vicinity of the church of São
Salvador de Aramenha.14
Plate 7 shows another nicolo paste from Conimbriga,15 a
Christian Gems from Portugal in Context
Plate 7 Nicolo paste with fish from Conimbriga,
2nd–early 3rd century AD. Conimbriga, Museu
Monográfico de Conimbriga (hereafter MMC)
Plate 9 Carnelian with fish from Alentejo, 3rd–4th
century AD. Formerly in the Barreto Collection
Plate 11 Carnelian with fish from Alentejo, 3rd–4th century AD. Formerly in
the Barreto Collection
Roman city 40km from the sea. Pls 8–11 illustrate four further
intaglios – three carnelians16 and one nicolo17 – from illegal
excavations which until recently belonged to the now dispersed
Barreto Collection; they are said to have come from different
sites in the interior of the Alentejo region.18 Of these, the
carnelian shown in Pl. 8 was found near Borba or Estremoz
(Alentejo), probably in Veiros.19 Veiros is located to the north-
east of Estremoz, to the south of Civitas Ammaiensis and
connected to Merida by an important road network. Its
economy probably depended on the agricultural products
coming from the rich villae of the surrounding area (for
instance, Silveirona and Santa Vitória do Ameixial), the raising
of horses (mainly attested in the Torre de Palma villa)20 and
probably on the marble trade which is attested in all the
surrounding areas (the so-called ‘marble triangle’: Estremoz-
Borba-Vila Viçosa).21
These seven fish gems date from the 2nd to the 4th
centuries ad. Thus, the single fish/dolphin device is one of the
most common ‘ambiguous’ gem motifs found on intaglios from
Roman or Early Christian Portugal. Taking into account the
importance of fishing in the ancient economy this is perhaps
not surprising. It is interesting to note that these seven gems
are made either from a blue glass imitating nicolo (3), which is
‘Gems of Heaven’ | 117
Cravinho
Plate 12 Plan of Roman Ammaia
an appropriate background colour for the depiction of fish, or
from red-orange stones (4). Spier’s four single fish gems dating
from the Early Christian period are likewise all engraved on
red-orange carnelian or sard.22
An interesting carnelian set in a modern ring, from an
unknown Portuguese site, represents a more likely Christian
motif: a dove perched on a column flanked by fish (Pls 13a–b).
The design closely resembles two carnelians published by
Spier.23 Yet, because the central element of the design is neither
a cross or a simple column, and as it appears to be positioned
on a globe, the identity of its owner as a Christian is uncertain.
The hare device and the chi-rho monogram: Lancobriga
(modern Fiães)
Two obviously Christian rings (Pls 14–15), one set with a gem,
were excavated during the 1970s in Lancobriga. The site is
situated in northern Lusitania below the river Douro, which
acted as the border between Gallaecia and Lusitania. It was a
Bronze and Iron Age castrum, then, in the Late Roman-Suevian
period, an important hamlet (identified by the Portuguese
excavators as the ubi of Lancobriga cited in the Antonini
Plates 13a–b Carnelian (and impression) with a dove on a column, two fish
flanking an altar, a globe, crescent and star; unknown provenance, 3rd–4th
century AD. Private collection
118 | ‘Gems of Heaven’
Itinerarium).24 Large quantities of 4th-century ad coins and
glass were uncovered,25 and both rings were found together
with Hispanic sigillata and North African ceramics dating from
this period.
The first is a gold signet ring set with a nicolo depicting a
hare with, above, the reversed inscription aviti,26 the genitive
of Avitus, a common Roman name in Portuguese Gallaecia and
Lusitania.27 A single chi-rho is engraved on one of the ring’s
shoulders (Pl. 14); the rest of the hoop is inscribed with similar
letters to the bezel inscription, but due to wear is not entirely
legible. The ring represents one of the many sub-types of a
common, mainly 3rd-century ad ring-type in both bronze and
gold. Other sub-types of this ring, primarily in gold and usually
with more elaborate/decorative hoops and bezels than the
Lancobriga example, became fashionable during the 4th–6th
centuries ad. Some of these, which are likewise set with
nicolos, depict what might be considered as ‘ambiguous’
images, others indisputable Christian motifs.28
The hare, which is commonly found on both Roman
intaglios and cameos, is usually depicted on Late Roman
glyptics with a vegetal element, branch, fruit or tree.29 The
calm hare is depicted ‘seated’ on a short ground-line. The motif,
the style of engraving and lack of detail, and the letter
formation of the word aviti, all suggest that this gem is to be
dated to the later 2nd–3rd century ad (given the ring form the
latter date is more likely). This ring was perhaps a gift to Avitus
wishing/assuring his prosperity and good luck (the hare motif)
in his earthly actions and life, and perhaps also in the afterlife.
The combination in one jewel of the owner’s name as a sealing
device, the pagan hare motif for prosperity and the Christian
symbol and inscription on the hoop, shows that common motifs
of pagan origin continued to be used as beneficial, valid, semi-
magical motifs by Christians, especially before Christianity
became the formal religion of the Empire as well as in the ad
330s and 340s. Thus the hare, just like other Roman animal
glyptic motifs, appears to have first become a general ‘neutral’
motif used by patrons of different religious beliefs, and then (or
concurrently) a motif accepted by Early Christians, at least in
northern Portugal.
Because this ring was found with ceramics datable to the
period ad 350–400,30 provided that the ring is contemporaneous
with the pottery, then this is the latest possible date for it. The
gem, however, as discussed above, is probably earlier (late 2nd
or more likely 3rd century ad), so the ring could have been
made in this period. An alternative possibility is that the earlier
nicolo was reset in a ring of later date. The phenomenon of
Plate 14 Gold ring set with nicolo from Lancobriga, 3rd–4th century AD.
Private collection

Plate 15 Amber ring with chi-
rho from Lancobriga, 5th
century AD. Private collection
re-setting intaglios in new rings became more frequent in the
Early Byzantine period when the emphasis shifted from
intaglios to rings.31 Because it is not known whether this
intaglio is contemporaneous with the ring, dating both to the
3rd century ad, or whether this is a 3rd-century ad gem which
was set in the ring at a later date, perhaps during the 4th
century ad, we prefer to give it a 3rd–4th century ad date.
The other ring found at Lancobriga is an amber ring whose
circular bezel is engraved with a chi-rho (Pl. 15).32 Amber rings
occur in the Roman ring repertoire and are predominantly of
Baltic origin.33 Yet, to the best of our knowledge, to date this is
the only known amber ring with the chi-rho device. Owing to
the historical events in the region of Lancobriga during the
early 5th century ad, when the country was under Suevian
rule, the owner of this engraved ring was most likely a Sueve.
Because amber existed in Lusitania in the 13th century,
according to Muslim texts,34 and is still present to this day for
Christian Gems from Portugal in Context
example on the Cascais, Sesimbra and Sintra coasts,35 the
amber for the ring might derive from a local or nearby source.
The bird with branch device: Conimbriga
Conimbriga is a large Roman city36 (Pls 16–17) situated in the
north-western part of Lusitania.37 Already an important trading
centre in pre-Roman times,38 Conimbriga became from the
Augustan period onwards a major road junction as it was
crossed by the important via Olisipo-Bracara Augusta (that
went as far north as Astorga). This via connected it with the
other major cities of Lusitania and with the cities of other
provinces (such as Clunia), allowing movements of people from
one place to another,39 and facilitating the import of various
products such as marble and limestone from quarries in
Lusitania, glass and pottery from other provinces and jewellery
from the Byzantine Empire.40 The discovery of three intaglios
depicting Mercury may testify to the importance of
Conimbriga’s commercial activities,41 and perhaps also to the
existence of a cult of this god supported by the merchants of
the city.42
The city was attacked and destroyed by the Sueves in the
second half of the 5th century ad.43 The Sueves populated and
ruled it as Christians until it was devastated by the Visigoths at
the end of the 6th century ad. Although during its occupation
Plate 16 Plan of Conimbriga: the
Augustan wall
Plate 17 Plan of Conimbriga: the Late
Roman wall
I – Houses (Casa do Mosaico das
Suásticas and Casa dos Esqueletos);
II – Baths (Termas extra-muros); III
– Paleo-Christian Basilica; IV –
House of Cantaber; V – Forum; VI –
Insulae; VII – Baths (Termas de
Trajano); VIII – Insula and Baths
(aqueduct area); IX – House (Casa
dos Repuxos); X – Amphitheatre
‘Gems of Heaven’ | 119
Cravinho
Plate 18 Bronze ring with cross from Plate 19 Bronze ring with palm-
Conimbriga, 5th–6th century AD. branch from Conimbriga, 5th–6th
Conimbriga, MMC century AD. Conimbriga, MMC
by the Sueves, Conimbriga was more important than
Aeminium (Conimbriga was a diocese44 while Aeminium was a
parish),45 comparatively few architectural remains (perhaps
originating from the area of the Early Christian basilica) testify
to the city’s importance during its ‘Suevo-Visigothic’ period.46
By the 8th century ad the site was gradually abandoned and by
the 10th century it had disappeared.47 Conimbriga then became
a necropolis, a ‘stone quarry’ and an agricultural area for the
local population who left behind some ancient Portuguese
coins, signs of their sporadic occupation of the site or their
transit through it.
Excavations attest to the existence of copper alloy, pottery,
wood, brick, and stone workshops and textile artisans in the
city or its vicinity.48 Pieces of raw glass (dating from the second
half of the 1st–2nd century ad) and several fragments of
refractory clay combined with glass waste demonstrate the
existence of glass workshops.49 Therefore, can we also infer the
production of glass pastes in the city? Furthermore, a crucible
for the refining of gold was uncovered in the central peristilum
of the so-called ‘Casa de Cantaber’.50
There is ample evidence for the presence of Christians on
the site as witnessed, for example, by a 4th-century ad lamp
decorated with a chi-rho,51 the abandonment and partial
destruction of the amphitheatre at the end of the 3rd or the
beginning of the 4th century ad, perhaps imposed by the
ecclesiastical authorities, and the transformation of the pagan
temple in the forum into a Christian space with its own
necropolis.52 Furthermore, many 4th–6th-century ad ceramics
(Pls 24–25), lamps, rings (Pls 18–21) and liturgical items (Pls
22–23) decorated with Christian motifs such as the chi-rho, the
cross, the palm branch, or inscribed Christian names like
Emmanuel, were found.
However, both devices on the relevant gems from
Conimbriga belong to the group with ‘ambiguous’ motifs. They
include the above-mentioned fish intaglio (Pl. 7), and a nicolo
Plate 20 Bronze ring with cross within Plate 21 Bronze ring with palm-
lozenge from Conimbriga, 5th–6th branch from Conimbriga, 4th–6th
century AD. Conimbriga, MMC century AD. Conimbriga, MMC
120 | ‘Gems of Heaven’
Plate 22 Bronze stamp with chi-rho from Conimbriga, 4th–6th century AD.
Conimbriga, MMC
set in a gold ring (Pl. 26) engraved with the figure of a bird/
parrot before a small branch: a well-known pagan motif
adapted to Christian use. Because on the Conimbriga example
another branch – a palm branch, a further motif with pagan,
Jewish and Christian symbolic significance – is engraved on
the ring’s upper shoulder (Pl. 26), it is not clear whether this
ring and its gem with their seemingly ‘neutral’ motifs belonged
to a Christian. For example some of the copper-alloy rings
found in Conimbriga bear similar or identical devices to those
ornamenting the shoulders of the gold ring under discussion,
that is the palm branch, which appears on the bezels of simple
metal rings of a common Roman ring-type found at the city
(Pls 19 and 21) which might well date from the 2nd to 3rd
century ad.53 Even the basic cross motif engraved on a ring
found at Conimbriga is of a common type which circulated
during the 2nd to 6th centuries ad (Pl. 18). If it comes from the
earlier centuries within this range, it might represent the
simple geometric cross encountered at times on Roman copper-
alloy rings of this type and on other simple ring-types.54
However, the gold ring-type (Pl. 26) and its excavation context
allows for the possibility that its owner was indeed a Christian.
This late 3rd-4th century ring was excavated in 1995 in the
course of a research programme in the garden of the peristilum
under a mosaic in the largest private house discovered in
Conimbriga, the ‘Casa de Cantaber’, which is situated near the
inner city wall and is adjacent to the 4th–6th century ad
Christian basilica. According to the excavator, Virgílio Correia
(who is the Director of the city’s museum) the gold ring with
parrot intaglio (Pl. 26) seems to have been intentionally hidden
within the underlayer of the mosaic. The intentional
concealment of the ring makes sense only if it was made before
the house was first abandoned, that is in the second half of the
5th century ad, just before the Sueves invaded the city.55
Occasionally rings of this type or similarly shaped gold rings
with plant-decorated shoulders are set with gems engraved
with Christian motifs.56
Plate 23 Bronze handle with inscription from Conimbriga, 6th century AD.
Conimbriga, MMC

Plate 24 Pottery fragment stamped with cross,
5th–6th century AD. Conimbriga, MMC
Plate 25 Painted ceramics, 4th–5th century AD. Conimbriga, MMC
Two other gems with different versions of this basic bird-
and-branch motif set in common Roman–Late Roman and Late
Roman–Early Byzantine type gold rings, likewise without a
clear-cut Christian symbol, monogram, etc., were collected
from other local Portuguese sites in the course of archaeo-
logical excavations (Pls 27–28). Whether these rings were
owned by pagans or Christians is unclear. Typologically the
earliest ring (Pls 27a–b) is a basic Roman ring-type, whose sub-
type, as represented by our ring, was widespread during the
1st–2nd century ad, a date which would therefore favour a
pagan patron. We include it here as evidence for the existence
of the common Roman bird-and-branch motif on gold rings
coming from Portugal dating from before the 3rd century ad.
Such gems and rings presumably influenced the continued
popularity of this motif on late 3rd–4th century gems set in
gold rings (Pl. 26) and, at least in Portugal, also on 4th–5th
century gold rings such as Pl. 28. Rings of the same basic type
as this second, more elaborate ring were in use above all during
the 4th–5th century ad. Thus the probability that this specimen
(Pl. 28), like the Conimbriga ring (Pl. 26), belonged to a
Christian is fairly high.
Plate 26 Gold ring set with nicolo Plate 27a Gold ring with nicolo of
from Conimbriga, late 3rd–4th unknown provenance, 3rd–4th
century AD. Conimbriga, MMC century AD. Private collection
Christian Gems from Portugal in Context
Christian symbols on rings not set with gems
The Portuguese Gem Corpus includes only one 5th century ad
intaglio engraved with an unequivocal Christian symbol: the
cross (see below Pl. 36a). Examples of gems displaying most
other Christian motifs (monograms, symbols, narrative scenes
or incontestably Christian inscriptions) engraved on 3rd–5th
century ad gems and current in other regions are lacking.57
However, a meaningful group of rings depicting on their bezels
certain Christian motifs, primarily the chi-rho monogram, have
been found on Portuguese soil (see also the ring from
Lancobriga, Pl. 15 above). We have chosen to include in our
discussion examples of such rings discovered at sites in which
3rd–6th centuries ad intaglios were found because of the well-
known connections between gems and metal rings during the
Late Roman and Early Byzantine periods, which was
presumably also true within the Christian Sueve and Visigothic
kingdoms.
This may be incidental. However, it might also point to a
trend within the Portuguese Christian communities towards a
preference for all-metal rings over metal rings set with gems
representing definite Christian symbols, especially from the
later 4th century ad. In Portugal this decline in the usage of
rings set with intaglios seems to be also connected with the
invasions from the North and the consequent disruption or
even breaking up of trade routes and commerce with the
Orient. The occurrence of this phenomenon in Portugal is of
special interest since the gold and copper mines and perhaps
also the quarrying of quartz, continued to be active in the
region throughout these periods (as well as in post-Byzantine
times). Therefore, the decline of gem-set rings in Portugal
during this period also appears to be a question of choice and
fashion. Whatever the reasons for this phenomenon in
Portugal, it also contributes additional information with
regards to the overall decline in usage of intaglios within the
Early Byzantine Empire.
Rings from Ribas castrum and the Early Christian necropolis of
Silveirona
Ribas castrum
The first ring is a cast copper-alloy ring engraved with the
chi-rho monogram (Pls 29a–b) found at Gallaecian Ribas
castrum (Argeriz, Valpaços), a Late Bronze Age and Iron Age
fortification situated near Aquae Flaviae to the north-east of
Braga. A few items found on the site (a Roman fibula and
Plate 27b Profile of Plate 27a Plate 28 Gold ring with nicolo of
unknown provenance, 4th–5th century
AD. Lisbon, MNA
‘Gems of Heaven’ | 121
Cravinho
Plate 29a Bronze ring with chi-rho from Ribas castrum, late 4th–5th century
AD. Vila Real, Museu de Arqueologia e Numismática de Vila Real
Pl. 29b Detail of Pl. 29a
sigillata ware) suggest the site was also, at some stage,
occupied during the Roman period.58
The sign within a circular border on the ring bezel
identifies it as a Christian jewel. The majority of known metal
rings with this Christogram come from the western part of the
Empire and date primarily to the 4th century ad.59 Rings of the
Ribas castrum type and of similar forms are known from the
late 4th century onwards.60 Depictions of the chi-rho monogram
as well as other Christian motifs within a circular border on
round bezels are well known. Such rings date mainly from the
4th century ad and most often originate in the Balkans and
North Africa.61 On a sub-group of these bronze rings the
circular border is not a plain band but a decorative ‘linear-dot’
border similar to that of the Ribas castrum bezel.62 A rare
aspect of the Ribas castrum ring is the fact that its chi is turned
into a frontal face by a linear dot pattern (Pl. 29b). This stylistic
peculiarity perhaps testifies to barbarian artistic notions
combined with Christian beliefs. This combination seems to
point to a local workshop and maybe to transitional times.
Although the Ribas castrum ring has in principle an upper date
limit of the 7th century, both its context within the site and its
typological features, suggest that it comes from the late
4th–5th century ad and is most likely of Christian Sueve origin.
Necropolis of Silveirona
The other ring is a gold (or gilded copper-alloy) ring of a typical
Late Roman-Early Byzantine type (Pl. 30). It was a stray find
recovered from the Christian necropolis dated to the 6th
century ad located at the Herdade (= farm) of Silveirona, in
the small village of Santo Estêvão, near Estremoz, Alentejo.
Silveirona was connected by several roads to important towns
in Lusitania, such as Ebora (present day Évora) and Ammaia
(the road between them crossed the site). This necropolis is
situated close to an earlier Roman necropolis dating from the
2nd–4th/5th centuries ad – the distance between these two
cemeteries is only 300m.63
Plate 30 Gold or gilded copper- alloy Plate 31 Bronze earring from a
ring with a cross-like motif from cemetery at Silveirona, 6th century
Silveirona, 6th century AD (?). Lisbon, AD. Lisbon, MNA
MNA
122 | ‘Gems of Heaven’
The two necropoleis, discovered and excavated in 1934 by
Manuel Heleno,64 are both situated near a large Roman villa
(‘Villa da Coelha’).65 Many finds were uncovered in the Roman
necropolis in which 86 tombs, mainly inhumations, were
unearthed. They include, for example, local or regional
ceramics, sigillata, glossy North African sigillata and glass
vessels dating from the 2nd–5th centuries ad. Among the latter
dated finds is a bowl engraved with graffiti representing a
palm branch and a cross.66 Some of the gravestones (for
example those of Laberia, 1st–2nd century ad; Lucius Valerius
Maxumis, 2nd century; and Quintus Fabius Tryphon Argyrius,
3rd century) were re-used in the Christian necropolis.
More than 80 tombs were found in this 6th-century ad
cemetery, 35 of which were studied by Manuel Heleno. The
excavated materials in this later necropolis are mainly personal
objects such as earrings (Pl. 31)67 without any specific Christian
symbolism. The gravestones have precise dates, for example:
ad 517 (Sabinus), ad 531 (Veranianus), ad 543 (Savinianus) and
ad 544 (Talassa). Some architectonic elements may suggest the
existence of a mausoleum or a funerary basilica or a local cult
to a martyr, now forgotten.68
The Silveirona ring was found in the necropolis area as an
isolated find (Pl. 30).69 Its central geometric device is a cross-
like motif which, however, may well be yet another example of
a simple ‘ambiguous’ geometric motif inherited from the
Roman decorative repertoire and which survived into the Early
Byzantine period. Like the Ribas castrum ring, the motif is
created with rows of stamped dots flanked by the typical Late
Antique ring-and-dot design. This dotted circle motif is
frequently employed as a decorative element on metal objects,
especially on those originating in the eastern part of the
Roman and Byzantine Empires. Such ‘cross’ or X-like motifs
continued to be applied to small metal items up to the 7th
century ad, when the occurrence of such motifs declines. This
all-metal ring with its raised shoulders is another ring sub-type
with truncated shoulders close to the above ring (Pl. 29). Its
low round flat bezel which ‘sits’ on a rectangular semi-bezel,
and its flat-sectioned hoop which widens towards this semi-
bezel and directly continues to it as one unit, suggest that the
ring falls within a 3rd–4th century ad date range. It may belong
to the latter period, but as it comes from a Christian cemetery,
it might be an heirloom.
Occasionally other pre-Constantinian ring sub-types close
to this basic ring-type set with gems carry a Christian
inscription or motif.70 Yet the possibility that this is a later ring
dating more closely to the Christian cemetery in which it was
found or that it is even a 6th century ring should be
considered.71 Rings of this basically Late Roman type
ornamented with geometric motifs continued into the Early
Christian era. Its discovery in a 6th-century context allows for
the possibility that this is also perhaps such a case. Rings with
truncated elevated shoulders continued to be made during the
5th–6th century ad and variants with similar shoulders are
also found in the early Islamic ring repertoire, chiefly among
rings originating in the Near East. The only thing which
identifies these rings as Islamic is the fact that their inscriptions
are written in Kufic script.72
Another copper-alloy ring, which has a plain cross on its
truncated shoulders, came from the recent excavations in the
thermae at Ammaia (Pls 32a–b). This fragment seems to be

Plate 32a Copper alloy ring from Ammaia, Plate 32b Detail of Plate 32a
5th century AD
another 5th-century ring which plausibly belonged to a
Visigothic Christian.
Stamped glass pendants
The two following items (Pls 33–34) are yellow glass pendants
stamped on the front with ‘neutral’ animal motifs. Such glass
pendants enjoyed a wide circulation in some parts of the
eastern Mediterranean and appear occasionally also in the
western parts of the former Roman Empire.73 To date these are
the only excavated examples of this glass pendant type found
in Portugal. The Portuguese examples, which represent well-
known motifs, common within the general group of such glass
pendants, come from two different sites in Lusitania,
Igaeditania and Alcochete.
Igaeditania (Idanha-a-Velha)
This very important Roman city (perhaps founded before
Augustus and capital of the Civitas Igaeditanorum) is located in
the interior part of central Lusitania in a region rich in tin and
gold;74 its roads connected it to all parts of the peninsula.75 It
became at an early stage an episcopal city. A church basilica
was built there in the 4th century ad or perhaps even earlier.76
Ferreira de Almeida77 states that the ground plan of the
basilica’s ‘sanctuary’ has similarities with other 6th-century ad
cult buildings in North Africa, the Adriatic region and Syria,
and that the fine quality of the construction of this ‘sanctuary’
in opus quadratum reflects a contemporary Syrian technique.
The glass pendant depicts a roaring lion standing to right
(Pl. 33). It was excavated in the ancient cathedral.78 Its shape,
motif and context date it to the mid- to late 4th century ad.79
The lion motif is yet another ‘ambiguous’ device encountered
frequently in Late Antiquity on small items such as gems, rings
and glass pendants, coming mainly from the East, and used by
people of diverse religious beliefs including Roman and
Sasanian pagans, Christians and Jews. The find context of this
pendant suggests it may perhaps be an item invested with
Plate 33 Glass pendant with lion from Plate 34 Glass pendant with scorpion
Igaeditania, 4th century AD. Idanha- from Alcochete, 4th century AD.
a-Nova, Centro Cultural Raiano Location unknown
Christian Gems from Portugal in Context
Plate 35 Gold ring set with a garnet from Alcochete, 6th–7th century AD.
Location unknown
Christian meaning, but for its owner whether Christian or of
some other religious affiliation it probably also had an
astrological significance. The existence of a 4th-century church
basilica in the city, presenting characteristics which suggest a
Syrian influence in its building, is the most likely indication
that a Syrian element was present in the city.80 This might
explain the discovery of a typical eastern Mediterranean glass
pendant on the site.
Alcochete
This Roman-Byzantine port and important trading centre is
located on the south-western shore of the River Tejo opposite
Lisbon. Finds from the site include many fragments of Late
Roman glass and several Visigothic items, for example, an
architectonic element, a gold coin issued in Merida by Egica
(ad 698–700) and a fine Roman-Byzantine type gold ring set
with a plain garnet cabochon (Pl. 35).81
The pendant, which is stamped with a scorpion (Pl. 34),
was discovered in Tomb no. 26 in the sizeable necropolis of
Porto dos Cacos located above the port. This necropolis is dated
by the coins, sigillata, amphorae and lamps excavated there to
the 3rd up to the beginning of the 5th century ad.82 The
scorpion, an ancient amuletic motif in the East, is less common
on glass pendants than other animal subjects such as lions,
bulls, lambs, birds and frogs.83 The shape of this glass pendant
favours the view that it is a late 4th–early 5th century object.
However, since the other finds in the tomb suggest a 4th
century date (perhaps even early 4th century?) it is safe to date
this glass pendant also to this period. The motif doubtless had
an astrological meaning and perhaps also a medical/magical
one. Although the identity of its owner as a Christian is far
from certain, it is not inconceivable that he/she might have
been a pagan (perhaps a follower of Mithras Tauroctonos), a
Jew or a Christian of eastern origin. Both pendants, which
carry further devices of an ambiguous nature, were probably
imported from the East, perhaps Israel, Lebanon or Syria.
Plate 36a Gold ring set with a garnet, Plate 36b Profile of Plate 36a
late 5th century AD. Private collection
‘Gems of Heaven’ | 123
Cravinho
Plates 37a-c Chalcedony with standing figure holding a cross and inscribed, 6th–7th century AD.
Private collection
Christian glyptics from illegal excavations
The final four gems (Pls 36–39) come from recent illegal digs
in Portugal and are published here for the first time. Since at
least three are asssuredly of Christian origin and the amber
bead is a fairly rare find, we choose to end our discussion with
them. Typologically, the gold ring set with a garnet (Pls 36a–b)
depicting a cross belongs to Spier’s garnet group dated to the
later 5th century ad.84 Furthermore, this ring appears to fit in
particularly well with a sub-group of this: the Visigothic Group.
All the specimens of this sub-group published by Spier are said
to come from Spain and are believed by him to be ‘imitations
cut in Visigothic workshops in Spain’ of similar garnets made in
Byzantine workshops of the late 5th century ad.85 Our ring
might likewise be a product of such Visigothic workshops
(perhaps located in modern Portugal).
However, the possibility exists that our specimen is the
product of a Byzantine workshop. This is because the garnet
intaglio of our jewel differs from all the garnets identified by
Spier as belonging to these proposed Visigothic-Spanish
workshops by the fact that it is still set in its original ring,
whereas the rings in which Spier’s specimens were set, are no
longer available for inspection and comparison; and by the fact
that it was discovered in Portugal, whereas Spier’s garnets
engraved with the same type of cross are said to come from
Spain. Since our gold ring is of a basic Late Roman ring type,
which became common during the Early Byzantine period,
when it is encountered set with garnets engraved with a cross
throughout the Byzantine world,86 the plausible eastern
Byzantine origin of our ring might also be supported by the
historical background: the end of Roman rule in most regions
of present-day Portugal and the later re-conquering by the
Byzantines of southern Portugal which ended only in the 7th
century ad.
The other three examples, which present different figural
motifs, are evidence of the growing importance of the cross as
a motif in 6th–7th century ad glyptic representations.
Although each of the figural devices harks back to late Classical
models, their simplified (Pl. 37), and/or exaggerated (Pl. 38),
and/or decorative and schematic (Pls 39a–b) styles are by now
far removed from Late Antique prototypes. They are most
likely products of Christian Visigothic origin.
Roman-Byzantine beads are rarely, if at all, engraved with
figural motifs.87 The bead’s material, amber, and the style of its
engraving suggest it is either of Visigothic production (perhaps
from local amber, see above Pl. 15), and as such the bust
probably depicts a Visigothic leader, or is an import into
Portugal from the Merovingian kingdoms to the north.88
In conclusion we can say that the finds discussed in this
124 | ‘Gems of Heaven’
Plate 38 Agate with an angel from Alentejo, 6th–7th
century AD. Formerly Barreto Collection
paper show a preference in the 2nd–4th centuries ad for the
depiction of ‘ambiguous’ motifs, such as the bird with a branch
and the single fish/dolphin motifs. Very few of these are
undisputedly Christian; the pagan or Christian ownership of
the rest is unclear. A few of these items with an ‘ambiguous’
motif may show eastern influence; other finds attest to the
popularity of the chi-rho on 4th–5th century gems and rings
among Roman Christians, but particularly among Christian
Sueves. A few excavated rings and gems, mainly originating
from sites in northern Portugal, show that in the early 5th–6th
centuries some artisans and patrons preferred personal jewels
ornamented with Christian symbolism executed in a
‘barbarian’ style. Provenanced, and mainly unprovenanced
finds, attest to the use of the Roman geometric cross motif
which was perhaps invested with a Christian meaning, (and
Christians may have seen it as a representation of the Christian
cross from the late 4th century onwards), but particularly to its
growing application on small personal items dating to the
5th–7th centuries ad, either as the sole motif or as an attribute
of figural motifs.
Notes
1 For an example of the Roman view about Lusitania and the
Lusitanians see Strabo, Geography, Book III, Chapter 3, 3.
2 Although the war between the Lusitanians and Romans began in
194 bc, the effective occupation of their territory came about only
in 16 bc. Even so, a clandestine rebellion occurred as late as 1 bc.
3 At the end of the 3rd to early 4th century ad (J. Alarcão, Portugal
Romano, Lisbon, 1987 [4th edn], 188), a local presbyter (Luxurius)
is mentioned in the Concilium of Illiberris, present day Elvira,
Spain (C. Ferreira de Almeida, ‘Notas sobre a Alta Idade Média no
Noroeste de Portugal’, Revista da Faculdade de Letras – Série de
História, I série, III (1972), 113–36, at 124.
4 M. Martins, Bracara Augusta: cidade romana, Unidade de
Arqueologia da Universidade do Minho, Braga, 2000, 8–9.
5 Alarcão (n. 3), 188–9; but see also, C.A. Ferreira de Almeida, ‘Arte
paleocristã da época das Invasões’, História da Arte em Portugal II,
Lisbon, 1986, 9–35, at 13.
6 Ferreira de Almeida (n. 5), 12–13; J. Maciel, ‘A arte da Antiguidade
Tardia (Séculos III–VIII, ano de 711)’, História da Arte Portuguesa 1
(2007), 101–47, at 121–3; J. Marques, ‘Peregrinos e peregrinações
Plates 39a–b Amber bead with a bearded male in profile on one side, stylised
standing figures at least one of which topped with a human head on the other;
from Alentejo, 6th–7th century AD. Private collection

medievais do ocidente peninsular nos caminhos da Terra Santa’, in
Estudos em homenagem a João Francisco Marques, Porto, 2001, 103–
21, at 105–9.
7 Alarcão (n. 3), 189–90; J. Massana and P. Capdevila, Religion and
Policy in the Coexistence of Romans and Barbarians in Hispania
(409–589), Barcelona, 2000–2, 196–216.
8 M. Martins, Bracara Augusta. A casa romana das Carvalheiras
(Roteiros Arqueológicos, 2), Braga, 2000, 7, 32.
9 It was located in the Museu Nacional de Arqueologia (Lisbon)
among the other archaeological artefacts from Ammaia. In 1970
(the date of its first publication), it was neither registered nor had
any inventory number; see: M.A. Graça and J.S. Machado, ‘Uma
colecção de pedras gravadas – elementos para um Catálogo Geral’,
Actas do II Congresso Nacional de Arqueologia, II, Coimbra, 1970,
383, no. 14; R. Parreira and C.V. Pinto, Tesouros da Arqueologia
Portuguesa (Catálogo), Museu Nacional de Arqueologia, Lisbon,
1980, 17, no. 159; G. Cravinho, Glíptica romana em Portugal
(Dissertação de Mestrado em História de Arte. Faculdade de
Ciências Sociais e Humanas da Universidade Nova de Lisboa),
Lisbon, 1999, no. 53.
10 S. Pereira, A Cidade Romana de Ammaia: Escavações Arqueológicas
2000– 2006, Lisbon-Marvão, 2009, 34–5.
11 S. Pereira, ‘Da Pré-História à Romanização na freguesia de S. S. de
Aramenha’, Ibn Maruan 13 (2005), 25–61, at 53.
12 G. Cravinho and S. Amorai-Stark, ‘A Jewish Intaglio from Roman
Ammaia, Lusitania’, Liber Annuus 56 (2006), 521–46, at 533–43.
13 S. Pereira, A Civitas Ammaiensis: Escavações Arqueológicas 2000–
2005 (Dissertação de Mestrado em Arqueologia. Faculdade de
Letras da Universidade de Coimbra), Coimbra, 2009, 117.
14 Pereira (n. 11), 61.
15 Cravinho (n. 9), no. 23; eadem, ‘Peças glípticas de Conimbriga’,
Conimbriga 40 (2001), 141–98, at 184–5, no. 19.
16 One of them is published: Cravinho (n. 9), no. 97.
17 G. Cravinho and R. Casal Garcia, ‘Anillos romanos de la colección
Barreto (Lisboa)’, Gallaecia 21 (2002), 223–43, at 233, no. 23.
18 The evangelisation of southern Portugal (an almost flat and
‘cosmopolitan’ region) during the Late Roman period was earlier
and easier than that of the north. Even so, the appearance of
isolated Christian objects in the south seems to indicate that
during this period (before the Edict of Milan) Christian religious
ceremonies often had to be celebrated in secret without arousing
the displeasure or even wrath of the pagan majority; these
ceremonies took place within private houses (an Early Christian
custom found in other regions of the Roman Empire, for example
in Dura Europos), and/or within small Christian communities.
19 Near Veiros there is also the site of Castelo Velho de Veiros – a
Celtic oppidum (5th-2nd century bc) situated at the confluence of
two small rivers (the ‘ribeiras’ of Alcariça and Ana Loura). It was
occupied in the Republican period and up to the 2nd-1st century
bc, when it was abandoned (no sigillata or other objects of Imperial
date were found: J. Arnaud, ‘Castelo Velho de Veiros (Estremoz) –
notícia da sua identificação’, Revista de Guimarães 78 (1–2) (1968),
61–84; idem, ‘O “Castelo Velho” de Veiros (Estremoz). Campanha
preliminar de escavações de 1969’, Actas das I Jornadas
Arqueológicas II, Lisbon, 1970, 309–28; it was re-occupied only in
Late Roman times, especially during the 3rd-4th century ad
(Arnaud 1968, ibid., 63, 74; M. Maia, Romanização do Território
hoje Português a Sul do Tejo [Dissertação de Doutoramento em Pré-
História e Arqueologia. Faculdade de Letras da Universidade de
Lisboa], Lisbon, 1987, 11–12).
20 The breeding and trade of horses (which were even sent to Syria)
still has echoes in the area, more precisely in the nearby Alter do
Chão, where an equestrian school and a local museum about
horses (Museu do Cavalo) exists.
21 This region’s marble was exploited from Roman times (Alarcão [n.
3], 140; H. Guerreiro, Exploração Subterrânea de Mármores –
Aspectos Geotécnicos (Dissertação de Mestrado em Georrecursos –
área de Geotecnia. Instituto Superior Técnico (Universidade
Técnica de Lisboa), Lisbon, 2000, 7, 177) and quarrying continues
to this day (A.M. Galopim de Carvalho, ‘Alguns aspectos da
Geologia do Alentejo’, Actas do II Encontro Regional de Educação,
Évora, 2004, 2; L. Lopes, ‘Recursos Naturais – Rochas
Ornamentais: o Mármore’, Alentejo Litoral, 2007, 2 (http://www.
alentejolitoral.pt).
22 J. Spier, Late Antique and Early Christian Gems, Wiesbaden, 2007,
41, nos 190–4.
Christian Gems from Portugal in Context
23 Ibid., nos 297–8.
24 A. Mendes Corrêa, ‘Nótulas Arqueológicas – Estação luso-romana
em Fiães’, Revista de Estudos Históricos II (1925), 89–97; C.A.
Ferreira de Almeida and E. Santos, ‘O Castro de Fiães’, Revista da
Faculdade de Letras – Série de História, I série, II (1971), 147–68;
idem, ‘O Castro de Fiães (II)’, Revista da Faculdade de Letras – Série
de História, I série, III (1972), 207–14.
25 Ferreira de Almeida and Santos 1971 (n. 24), 162.
26 C.A. Ferreira de Almeida, ‘Cerâmica Romana de Fiães’, Revista da
Faculdade de Letras – Série de História, I série, III (1972), 129–30;
idem (n. 6), 14, 21.
27 Hispania Epigraphica online: Inscription nos 6729; 11900; 18950;
18962; 20024; 20080; 20385; 20424; 20515; 20671; 21235; 21257;
21321; 21334; 21353; 21354; 21358; 21398; 21412; 21418; 21420; 21457;
21934; 22010; 22185; 22690; 22691; 22966; 23443; 23766; 24918.
28 Cf. Spier (n. 22), nos 15 (a nicolo with the bust of a Constantinian
prince, from Bulgaria, early-mid 4th century ad), 240 (gold ring
set with a nicolo with two fish and an anchor, 3rd or 4th century
ad) and 542 (a garnet with a cruciform monogram, 6th century ad
or slightly later). See also Pl. 36.
29 Cravinho (n. 9), no. 235: carnelian in a silver ring engraved with a
hare with grape cluster in front, from Portugal, the ex-Barreto
Collection; S. Amorai-Stark, ‘Gems, cameos and seals’, in R. Gersht
(ed.), The Sdot-Yam Museum Book of the Antiquities of Caesarea
Museum, Tel Aviv, 1999, 87–114, at 97, fig. 23, with palm or tree,
from Caesarea Maritima, Israel, with lists.
30 ‘Sigillata clara D, estampada, do estilo A de Hayes’: Ferreira de
Almeida (n. 26), 130, n. 81.
31 Cf. O. Peleg-Barkat, ‘A Gold Ring with a Carnelian Intaglio and a
Glass Pendant from Tiberias’, forthcoming.
32 Identified by the excavator as amber and dated to the early 5th
century ad: Ferreira de Almeida (n. 5), 10, 21. He gives no
information as to the precise find-spot and its context.
33 E. Gagetti, ‘Anelli di Età Romana in Ambra e in Pietre Dure’, Arte e
materia – Studi su Oggeti di Ornamento di Età Romana. Quaderni di
Acme 49 (2001), 191–485, at 235.
34 A. Rei, ‘O Gharb al-Andalus em dois geógrafos árabes do século
VII/XIII: Yâqût al-Hamâwî e Ibn Sa‘îd al-Maghribî’, Medievalista
online, 1, Instituto de Estudos Medievais. Universidade Nova de
Lisboa, Lisbon, 2005, 4; 5; 8; 12; 16; 21.
35 http://www.altmann-zuzarte.com/PT/ambar_pt.htm
36 Its early Roman wall (Augustan?), whose perimeter is more than
2 km, surrounds an area of 23 hectares.
37 Its occupation during the Chalcolithic period and, mainly, from
the 9th–8th century bc is uncontested. Subdued by the army of
Decimus Junius Brutus (in 138 bc), the oppidum of Conimbriga
(which means ‘a fortified city on a stony mount’) was redesigned in
the 1st century bc under Augustus and given new structures. The
most important was an aqueduct, since there was no source of
water. Under the Flavians (between ad 70–80), the city acquired
the status of municipium, taking the name of Flavia Conimbriga,
and was totally renovated. At the end of the 3rd century ad a new
wall was built which reduced the city’s perimeter on its eastern
side.
38 V. Correia, ‘Conimbriga. A camada pré-romana da cidade (Notas
de uma exploração de dez dias em Condeixa-a-Velha)’, O
Archeólogo Português 21 (1916), 252–64.
39 As evidenced, for example, by the funerary inscriptions found in
Conimbriga: A.M. Alarcão and S. Ponte, Colecções do Museu
Monográfico de Conimbriga (Catálogo), Coimbra, 1984, nos 534–5;
R. Étienne et al., Fouilles de Conimbriga, II (Épigraphie et
Sculpture), Paris, 1976, nos 15, 31, 33–4; E.W. Haley, Migration and
Economy in Roman Imperial Spain, Barcelona, 1991, 110, n. 362.
40 Alarcão and Ponte (n. 39), nos 403.25–403.26.
41 Cravinho 2001 (n. 15), 166–71, nos 11–12; 188–90, no. 23.
42 Ibid., 144. However, to date, no inscription dedicated to Hermes-
Mercury has been uncovered at Conimbriga.
43 The description of the traitorous Sueves’ attacks, the destruction
of houses, of the aqueduct and of part of the wall, the conquest of
the city and its suburbs and the capture of some of its population
(especially the noble and wealthy Cantaber, his wife and children),
is reported by Hydatius (Chronicon § 229; 241), the bishop of Aquae
Flaviae (J. Alarcão, ‘Notas de arqueologia, epigrafia e toponímia –
III’, Revista Portuguesa de Arqueologia 8 (2005), 293–311, at 306).
44 In the Actae of the 1st and 2nd Concilium of Bracara Augusta (in ad
561 and 572, respectively), the bishop of Conimbriga (Lucentius) is
‘Gems of Heaven’ | 125
Cravinho
mentioned among the participants: J. Alarcão, Conimbriga, O
Chão Escutado, Lisbon, 1999, 29. Furthermore, several Christian
gravestones, dated to ad 522, 538 and 541, testify to the Christianity
of the city in those times: Alarcão and Ponte (n. 39), nos 546–8
respectively.
45 In ad 585, when the Visigoths conquered the Sueve Kingdom
(J. Alarcão, ‘A História, depressa contada, do povoamento da
região de Coimbra desde tempos proto-históricos aos fins do
Século XII’, Trabalhos de Arqueologia 38 (2004), 12–36, at 17), the
local bishop, his Curia and many inhabitants of the city moved to
flourishing Aeminium (present-day Coimbra), which gradually
took Conimbriga’s name.
46 Alarcão and Ponte (n. 39), 117; Alarcão (n. 44), 17.
47 Alarcão (n. 44), 29.
48 Alarcão and Ponte (n. 39), 21–32, and for quarries in its vicinity: 47.
49 Ibid., nos 172–3.
50 V.H. Correia, ‘Conimbriga, casa atribuída a Cantaber: trabalhos
arqueológicos 1995–1998’, Conimbriga 40 (2001), 85–140, at 98,
2.2.11. C 11.
51 Alarcão and Ponte (n. 39), no. 551.
52 A. De Man, ‘Sobre a Cristianização de um Forum’, Al-Madan 13
(2005), 2nd series, Adenda VI, 5. The cult of the 3rd-century ad
martyr saint Christina in Condeixa-a-Nova (a small town near
Conimbriga) is said to be an influence of the Early Christian times
of the city, according to oral information given to us in 2005 by the
priest of the local church.
53 For example, seven identical or very similar slender copper-alloy
rings with the same single palm branch motif were found at
Roman Caesarea Maritima: cf S. Amorai-Stark and M.
Hershkovitz, Intaglios, Plain Rings and Finger Rings from Caesarea
in the Handler Collection, Tel Aviv, forthcoming, Ch. III.
54 Ibid., cf. four examples from Caesarea Maritima.
55 Correia (n. 50), 113, 2.2.29 C 28: ‘o seu achado é fruto de ocultação
intencional no mosaico. Tal ocultação apenas será explicável se
tiver ocorrido com a casa em fase de abandono ou, pelo menos, em
fase de não voltar a sofrer reparações’. The house of Cantaber has
seen many seasons of excavation. Some of the earliest ones took
place in the 1930s, before the systematic excavation of the city
started in 1962. Pls 18 and 20 (inv. A 720) came from the old pre-
1962 excavations. Pl. 19 and the liturgical objects (Pls 22–23) were
found at the city’s forum during the Portuguese-French
excavations: the signet ring (inv. 66.7) was found in 1966 and the
handle of the patera (inv. 64.78) in 1964. See ibid.
56 Spier (n. 22), no. 324, a mottled green jasper with the Good
Shepherd, two sheep, a fish, a tree and an inscription, said to be
from Tartus, Syria.
57 For their occurrence within the overall corpus of Early Christian
gems see: Spier (n. 22), chapters 3–8.
58 A. Freitas, ‘A Cerca de Ribas: Ribas – Freguesia de Argeriz –
Concelho de Valpaços’, Revista de Guimarães 99 (1989), 319–67.
59 Spier (n. 22), 183.
60 Ibid., R63.
61 Ibid., 183, nos R27–32 (with the Christogram), R46, R88–9.
62 Ibid., R30, but see also R32, R88 in which the dots are within a
double plain band.
63 The change of the religious context probably explains the creation
of a new funerary space, in a close but distinct place from the
earlier pagan one: C. Fabião, M. Dias and M. Cunha, Sit Tibi Terra
Levis – Rituais Funerários Romanos e Paleocristãos em Portugal
(Catálogo de Exposição), Museu Nacional de Arqueologia, Lisboa,
Lisbon, 2008, 28.
64 M. Heleno, ‘Um quarto de século de Investigação Arqueológica’,
O Arqueólogo Português, n.s., 3 (1956), 221–37, at 232.
65 These excavations are still unpublished and no final report was
issued. However, the excavation field notes, preserved in several
note books, are kept in the Museu Nacional de Arqueologia
(Lisbon). For remaining questions regarding the chronology of the
Early Christian necropolis in relation to the earlier pagan one see
also: M. Dias, ‘Fragmentos de inscrições paleocristãs, inéditas, da
colecção epigráfica do Museu Nacional de Arqueologia e
Etnologia’, O Arqueólogo Português, Série IV, 5 (1987), 225–32, at
229–30, 232.
66 Fabião, Dias and Cunha (n. 63), 30–1.
67 E.g., in one of the tombs a young woman was buried wearing a ring
126 | ‘Gems of Heaven’
and earrings. The discovery in the 6th-century ad necropolis of a
North African lamp and Imperial period coins, which could still
have been in use in the 6th century ad, reinforce and underline the
fact that this later necropolis is in fact a continuation of the older
one (Fabião, Dias and Cunha [n. 63], 40 and 50–1).
68 Ibid., 48.
69 Ibid., 50. According to Melanie Cunha (‘As Necrópoles de
Silveirona [Santo Estêvão, Estremoz]. Do mundo funerário
romano à Antiguidade Tardia’, O Arqueólogo Português
(Suplemento 4), Lisbon, 2008, no. 150), this must be ring no. 91
mentioned in the second note book of Manuel Heleno’s excavation
field notes, found together with a coin in the second or third ‘lot’
(of the necropolis) on 31 May 1934.
70 E.R.I. Bingol, Museum of Anatolian Civilizations: Ancient Jewellery,
Ankara, 1999, no. 181, gold, glass cabochon, from the excavation at
Kebanm Agin, Turkey, 3rd century ad; Spier (n. 22), no. 94, silver,
carnelian, IHC/OY, 3rd century ad; no. 293, bronze [?], carnelian,
two fish flanking a ‘cross’, from a 3rd-century ad grave at Novo-
Otradnoe on the sea of Azov.
71 Cunha (n. 69), 179, no. 150, dates this ring to the 5th–7th centuries
ad.
72 Cf. M. Wenzel, The Nasser D. Khalili Collections of Islamic Art, Vol.
XVI, Oxford, 1993, nos 54, 57 and 60.
73 D. Barag, ‘Stamped Pendants (355–407)’, in M. Spaer, Ancient Glass
in the Israel Museum. Beads and Other Small Objects, Jerusalem,
2001, 173–6; idem, ‘Late Antique and Byzantine Glass’, in R.B.
Bianchi (ed.), Reflections on Ancient Glass from the Borowski
Collection, Bible Lands Museum, Jerusalem, Mainz am Rhein, 2002,
305–28, at 309– 21.
74 Gold was mined as early as Roman times: J. Alarcão, O Domínio
Romano em Portugal, Lisbon, 1988, 129.
75 A.F. Pereira, ‘Vestígios do passado em Idanha-a-Velha: IV – Ruínas
de ruínas ou estudos igeditanos’, O Archeologo Português, Série I,
30 (1938), 186–204, at 192–5; F. Almeida, Egitânia: História e
Arqueologia, Lisbon, 1956; idem, ‘Antiguidades da Egitânia –
alguns achados dignos de nota’, Arqueologia e História, 8ª série, 11
(1965), 95–101; idem, Ruínas de Idanha-a-Velha, Civitas
Igaeditanorum, Egitânia – Guia para o visitante, Lisbon, 1977; A.
Marques de Sá, Civitas Igaeditanorum: Os Deuses e os Homens,
Idanha-a-Nova, 2007, 25–8.
76 F. Almeida, Arte Visigótica em Portugal (O Arqueólogo Português,
n.s., 4), Lisbon, 1962; Ferreira de Almeida (n. 5), 43–7.
77 Ferreira de Almeida (n. 5), 45–6.
78 Almeida 1965 (n. 75), 98–9, pl. III, no. 1.
79 Barag 2002 (n. 73), 317–8, nos LA-27–9.
80 This conclusion is supported by an inscription found in Idanha-a-
Velha reading: arreno crescentis f(ilio) libiensi mav(r)illa
celeris lib(erta) marito f(aciendum) c(uravit): Marques de Sá
(n. 75), no. 118.
81 This ring, which comes from Tomb no. 8, was found together with a
coin of Severus Alexander (ad 222–35), a dish of glossy sigillata D
(ad 290–300/575) and a lamp of Type 30b Dressel-Lamboglia
dating from the end of the 3rd-early 4th century ad (A. Sabrosa,
‘Necrópole romana do Porto dos Cacos (Alcochete)’, in G. Filipe
and J. Raposo (eds), Ocupação Romana dos Estuários do Tejo e do
Sado, Lisbon, 1996, 283–300, at 298.
82 Sabrosa (n. 81), 287.
83 Barag 2002 (n. 73), 307.
84 Spier (n. 22), ch. 8, especially nos 514 (belongs to his Byzantine
sub-group) and 547 (belongs to his Visigothic sub-group, similar
cross devices).
85 Ibid., 93–4, nos 547–50.
86 For example, ibid., nos 15 (from Bulgaria), 474 (from Asia Minor)
and 542 (unknown provenance).
87 The largest Roman group of beads depicting faces and other
figural motifs are the mainly early Roman mosaic glass beads:
L.S. Dublin, The History of Beads from 30,000 bc to the Present,
London, 1995, 60–1.
88 For an example of a 6th-7th century ad Byzantine amulet gem
depicting a thin elongated figure surrounded by other motifs, cf.
A. Krug, ‘Ein frühbyzantinische Amulettgemme’, in K.M.
Cialowicz and J.A. Ostrowski (eds), Les Civilisations du Bassin
Méditerranéen. Homages à Joachim Sliwa, Instytut Archeologii UJ,
Cracow, 2000, 395–402, Ill. 1.
Intaglios and Cameos from Gaul in the
3rd and 4th Centuries AD
Hélène Guiraud
At the conference I began by presenting 17 unpublished pieces:
too small a number, but useful as a point of departure in
addressing various aspects of glyptics in Gaul (within the
boundaries of France) during the Late Roman Empire, taking
into account the published intaglios and cameos. The results of
this study are presented here.
Of the 1560 objects listed to date, most of them published,1
about 250 can be dated to the 3rd to 4th centuries ad. The term
‘about’ expresses the difficulties involved in the precise dating
of an intaglio or cameo, however tight the context of the find.
In dating these pieces I have used the normal archaeological
criteria. Firstly, the stratigraphic context which allows a piece
to be dated by association with glass, pottery and sometimes
coins: closed contexts such as tombs and hoards are best (Pls
4–7),2 and give at least a t.a.q. The next level of reliability
relates to the dates of abandonment or destruction of a site (Pl.
1).3 Sometimes only the geographic location is known. Even in
the best of contexts any dating must be treated with caution: an
intaglio or cameo is a jewel which may be a gift or family
heirloom with a very long life. The settings are another useful
chronological indicator (Pls 2–3).4 In the period with which we
are dealing, the choice of material, limited compared with
earlier periods, contributes little to the dating problem. Style
and motif too give only an approximate indication of date, even
if it is sometimes possible to identify stylistic groups of
intaglios. However, certain pieces may be placed more
precisely: for moulded glass pastes/nicolo, for example, other
intaglios maybe identified which are better dated and which
are based on the same original.5 Sometimes a motif or detail,
especially women’s hairstyles,6 places an intaglio
chronologically. The various criteria, even the less precise
ones, when combined together7 enable one to classify a certain
number of pieces (the 250 described below) within the period
of the Late Roman Empire. A few intaglios, where the range
suggested by two or three criteria is too great, may indicate
that the piece had a long life or was re-used.
Amongst the 1,560 pieces inventoried for Gaul, those dating
from the 3rd to 4th centuries ad account for only 16% of the
total. During this period most pieces (80%) date to the 3rd
century ad and just 20% of intaglios datable to the 4th century
ad show signs of the decline of glyptics seen in the Empire as a
whole. This uneven distribution is also due to the nature of the
find-spots: in fact, hoards (34% of our objects) were buried in
the 3rd century ad8 but, on the other hand pieces recovered
from cemeteries, rural and urban areas represent 15%, 14% and
13% respectively; 22% are from known sites but with imprecise
locations. Regional distribution varies according to the nature
of the find-spot (Map): hoards were generally found in the
Rhône area and in a second zone around the Oise, and the rest
dispersed around the country. Cemeteries are in the north-east
and these finds may be from very early excavations, or from the
19th century or recent excavations.9 Overall, the areas of
circulation – Rhône, Saône and towards the Rhine – which
were already privileged in other eras, are the richest, while the
Narbonne region, where the object count was high from the 1st
century bc, loses its importance.10
Amongst the materials used, carnelian, as during the
preceding centuries, is the most valued of the stones (13%), but
to a lesser degree (30% from all periods). The other stones
show the same decline: nicolo (12% as against 17%), jasper (6%
against 9%), and especially the rarer stones like amethyst,
plasma and garnet, while onyx and sardonyx still make up 8%.
While gemstones were less numerous (47% against 70%), the
proportion of glass pastes was even greater than that of gems,
and amongst the pastes glass/nicolo stands out (47% against
53% of the total). This decline in the use of gemstones and the
preference for glass/nicolo pastes is difficult to explain
satisfactorily. One explanation might be that a good number of
the supply routes of gems from India were closed. There were
certainly mines in the West, perhaps in the Alps, but were these
enough? The trend was not caused by any economic decline
amongst consumers: the 3rd century ad, despite moments of
crisis, and the 4th century ad were buoyant periods, and in the
rich hoards glass/nicolo pastes sit side by side with carnelians
and nicolos, and in necklaces and in earrings unengraved glass
pastes are deposited together with emeralds and pearls.
As in other periods, divinities figure largest amongst the
motifs identified (54%), especially in the 3rd century ad;
amongst these Victory (16 examples) occurs more often then
hoards
cemeteries/tombs
sanctuaries
towns
rural settlements
various
Map Distribution of intaglios and cameos by type of site
‘Gems of Heaven’ | 127
Guiraud
Plate 1 Cameo from Parville (Eure). Rouen, Musée départmental des
Antiquités
Mercury and the satyrs, but in my classification of 1988 of 1,015
pieces, Eros (love) was most numerous (he now occupies sixth
place), ahead of the satyrs and Mercury; the popularity of
satyrs, related to Bacchus, and of Mercury, may perhaps be
explained by their protective role, but Victory by reason of
provenance is hard to relate to the military world and perhaps
represents an aspect of Roman power. Men and animals each
represent 15% of the total amongst motifs; warriors and
hunters on the one hand, lions and eagles on the other. All
together the repertoire diminishes, just reproducing the same
motifs, but there is still a wide variety of animals appearing on
one or two examples: the squirrel, peacock etc. The
predominance of the lion and eagle may be explained, as in
previous periods, by their rich symbolism. Heroes are less
numerous (8%); oddly, Theseus ranks first (Pl. 6), when one
might have expected Hercules (Pl. 2), so frequently figured on
sarcophagi. Various subjects (vessels, cornucopia, etc.) do not
appear often, and the image of two crossed hands occupies first
place (six examples): here too the family and political
symbolism of the dextrarum iunctio explains this choice and
probably also the ease of engraving the motif. Few magical
gems come from the listed sites;11 perhaps the lion holding a
bovine head in his jaws or under a foot had some prophylactic
power if engraved on yellow jasper.12 Early Christianity seems
to have left no trace except for the chi-rho engraved on a green
jasper from a barbarian grave of the second half of the 4th
century ad.13 Explaining the choice of motifs is always a chancy
business, even if explanations can be suggested for some
subjects. In this matter, the nature of the object, family gift or
personal possession, new purchase or re-use, confuses the
issue – and even more, the fact that one scarcely ever knows
who owned it.
Plates 4–5 Cameos from the Eauze (Gers) treasure. Eauze, Musée
archéologique
128 |’Gems of Heaven’
Plates 2–3 Two rings from l’Ill, Strasbourg
Two categories of wealthy owner may be identified,
without necessarily distinguishing them from the faceless
crowds who bought ‘only’ glass/nicolo pastes. First there are
the treasure-owning occupiers of villas or rich domus in the
towns (Pl. 1); their wealth is not necessarily indicative of choice
of expensive materials or stylistic quality, as in treasures such
as those of Eauze or St-Boil, glass/nicolo paste would rub
shoulders with gems, and in the well-known south-western
villas like Chiragnan or Seviac,14 few intaglios or cameos were
found: on account of the light weight and small bulk of these
jewels, the owners must have taken them away with them
whatever the circumstances. Even if these people were known
by name, Libo in Eauze, or by occupation, in Eauze or
Beaurains, the choice of motif is only explicable in general
terms.
The second category of rich owner is perhaps identifiable in
the rich cemeteries of north-eastern Gaul. The burials are those
of barbarian chiefs, Germanii integrated into the army and of
their families who laid the proof of their wealth and status
along with the deceased – and also, for us, the signs of their
appropriation of Roman cultural forms. The presence of a
glass/nicolo paste decorated with a lion in a grave at
Homblières (Aisne) may seem meaningless, but an intaglio-set
ring is a Roman custom and the deceased lady was
accompanied by figures of Jupiter, Mars and Hercules
embellishing a bronze casket.15 Some motifs have little
meaning, aesthetically in particular, but their presence
signifies the embracing of romanisation – like the
undistinguished glass paste showing two actors in front of a
herm on a silver ring of the 4th century ad placed in a female
grave in a Germanic cemetery.16
Plates 6–7 Glass/nicolo pastes from the Eauze (Gers) treasure. Eauze, Musée
archéologique, no. 86580–1

The craftsmen who made these intaglios and cameos, in
stone or glass, are little known, as they were in preceding
periods. To date, glyptic workshops have not been found in
Gaul, just traces of mounting at the workshop of a jeweller or
bronze-smith.17 If the centre of Aquileia was still producing
engraved gems at the beginning of the 3rd century ad, then the
Rhine and the rich and busy centres of Cologne and Trier
should be looked at. The distribution of moulded replicas
seems to connect the glass/nicolo pastes of France to these
regions of Germania. If on occasion some gems resemble others
distributed along the Rhine, some amongst the Eauze treasure
for example, the same affiliation cannot, however, be claimed
for the intaglios.18
To conclude, can one establish a source for the production
of glyptics in 3rd- to 4th-century Gaul ad? Not for material or
motifs. As far as findspots are concerned, Gaul stands out from
neighbouring regions in the large number of 3rd century ad
hoards; it is in these, but not exclusively, which are found the
earliest intaglios and cameos, amassed and certainly re-used.
In addition, the number of cameos, especially from hoards or
from rich villas, is significant: representing half of the cameos
known from Gaul, most date from this period and they amount
to 11% of the glyptic objects (as against 4% of the total number
of objects listed from Gaul). This significant figure is not
however restricted to Gaul as at the same time many cameos
are known from the ‘middle’ Danube.19 As in the Rhine or Italy,
glyptics declined from the 3rd century ad, and from then on
intaglios are only present as re-utilised objects in the burials of
the 5th to 7th centuries ad.
Notes
1 See, H. Guiraud, Intailles et camées de l’époque romaine en Gaule
(48e suppl. à Gallia), Paris, CNRS, 1988; eadem, ‘Intailles de la
Comédie à Lons-le-Saunier (Jura)’, Gallia 52 (1995), 359–406;
eadem, Intailles et camées de l’époque romaine en Gaule, II (48e
suppl. à Gallia), Paris, CNRS, 2008.
2 The treasure from Eauze (Pls 4–7) is dated by the latest coins to
ad 261: D. Schaad, Le trésor d’Eauze. Bijoux et monnaies du IIIe siècle
après J.-C., Toulouse, 1992.
3 Destruction of a site: for a cameo (Pl. 1) from the destruction level
of a living area of an estate active during the 2nd to 3rd centuries
ad, see: C. Dorion-Peyronnet (ed.), Les Gaulois face à Rome. La
Normandie entre deux mondes (exh. cat., Musée des Antiquités,
Intaglios and Cameos from Gaul in the 3rd and 4th Centuries AD
Rouen (Seine-Maritime), Mai-Septembre. 2009), Rouen, 2009,
86–91.
4 Pls 2–3, rings of Types 2a and 2d: Guiraud 1988 (n. 1), 78–81;
Guiraud 2008 (n. 1), 76–7.
5 H. Guiraud, ‘Intailles découvertes en France et Aquilée’, in G. Sena
Chiesa and E. Gagetti (eds), Aquileia e la glittica di età ellenistica e
romana (Atti del Convegno ‘Il fulgore delle gemme’ Aquilée, Juin
2008), Trieste, 2009, 119–27, at 121 [map 1], 124 [map 2]).
6 Guiraud 2008 (n. 1), no. 1209 (hairstyle of the time of Julia Domna)
and no. 1462 (hairstyle of the middle of the 4th century ad).
7 Guiraud 2008 (n. 1), 78–80.
8 For hoards containing jewellery including intaglios see: Schaad
(n. 2), 63–4; for one hoard only (Beaurains), excavated at the
beginning of the 4th century ad, see: P. Bastien and C. Metzger, Le
trésor de Beaurains (dit d’Arras), Wetteren, 1977, 170–5.
9 For the north-eastern cemeteries, see: À l’aube de la France. La
Gaule de Constantin à Childéric (exh. cat., Paris Palais du
Luxembourg, Février-Mai 1981), Paris, 1981. By contrast, for a very
poor cemetery in the south, see: Y. Solier, ‘La nécropole gallo-
romaine des Aspres à Sigean (Aude)’, Bulletin de la société des
études scientifiques de l’Aude 65 (1964–5), 213–46.
10 Guiraud 1988 (n. 1), 57 (map 21a showing the wealth of the
Narbonne region at the end of the Republic and the beginning of
the Empire) and Guiraud 2008 (n. 1), 81, fig. 6.
11 Guiraud 1988 (n. 1), no. 922; Guiraud 2008 (n. 1), nos 1426–7.
12 Guiraud 1988 (n. 1), no. 652; Guiraud 2008 (n. 1), no. 1325.
13 For the chi-rho, see: Guiraud 1988 (n. 1), no. 878 (from a female
tomb with coins of ad 364–5 from Vermand, Aisne). It is too early
for the transformation of Victory into an angel.
14 For the richness of hoards at Eauze, see: Schaad (n. 2), 48–58; at
St-Boil (Saône-et-Loire), see: Guiraud 2008 (n. 1), no. 1088 (with
bibliography). For the relative poverty of hoards at the villa of
Chiragnan (Haute-Garonne), see: Guiraud 1988 (n. 1), no. 988;
Guiraud 2008 (n. 1), no. 1370; the villa of Seviac (Gers), see:
Guiraud 1988 (n. 1), no. 984; Guiraud 2008 (n. 1), nos 1149, 1457.
15 For the various cemeteries, see: À l’aube de la France (n. 9). For
Homblières, see: Guiraud 1998 (n. 1), no. 14 (female burial with
coins of Gratian from the years ad 367–83) and no. 13 (male burial,
same period).
16 Guiraud 1988 (n. 1), no. 621 (burial 27, Vert-la-Gravelle, Marne; in
the cemetery coins of Constantine to Valens). For references to
identical glass pastes, see: Guiraud (n. 5) n. 35.
17 Guiraud 2008 (n. 1), nos 1114, 1118, 1132, 1341, 1379.
18 For comparisons between jewellery found in hoards and others
from the Rhine, see: Schaad (n. 2), 68, maps 1,2 (decoration,
forms). For glass-nicolo pastes, see: Guiraud 2009, maps 1–3.
19 I. Popović, Les camées romains au Musée national de Belgrade,
Belgrade, Musée national de Belgrade, 1989; L. Ruseva-Slokoska,
Roman Jewellery. A Collection of the National Archaeological
Museum, Sofia, London, 1991.
‘Gems of Heaven’ | 129
Late Roman Gems from Tilurium in Croatia
Bruna Nardelli
Within the great glyptic collection in the Archaeological
Museum of Split, which numbers more than 2,600 items,1 there
is a large group of gemstones from Tilurium.2 This collection of
300 gems, including both intaglios and cameos, assumes an
important documentary value due to the paucity of historical
evidence for the presence of engraved gems in Dalmatia’s
military settlements. There are only a few gemstones known
Map Roman Dalmatia
130 | ‘Gems of Heaven’
from two important military camps, from Andenterium and
from Burnum; these are kept in Croatian museums and in the
Ashmolean Museum in Oxford.3 In fact, the legionary
fortresses on the territory of the former Roman province of
Dalmatia have never been systematically studied: this is why
the archaeological excavations at Tilurium represent an
important contribution to the research of that territory.4

Plate 1 Red jasper intaglio with Plate 2 Red jasper intaglio with Mars
Bonus Eventus. Split, Archaeo- Ultor. Split, Archaeological Museum
logical Museum
The castrum Tilurium,5 situated within the modern village
of Gardun, was placed on a very prominent site, utilising a
prehistoric hillfort. Situated 20km inland from Salona, it lies
where the river Cetina (the ancient Hippus flumen) can be
easily forded and flows into the fertile Cetina valley, before
cutting its way through the mountains towards the sea; this is
also near cross-roads, close enough for rapid communication
with coastal cities. Tilurium long retained its military
character; it was the only Roman military settlement in the
interior of Dalmatia which kept its leading role during
antiquity. It lasted until the Migration period, when it was
destroyed.6
The engraved gems from Tilurium came into the possession
of the Archaeological Museum of Split from the early years of
its establishment in 1820: they were mainly the result of chance
finds that emerged during the course of agricultural activities.
In this collection one finds glyptic material of various stylistic
trends executed in different techniques, dating from the 2nd
century bc to the 6th century ad. The gemstones belonging to
the Roman period do not show evident artistic or technical
peculiarities, nor any iconographic innovation. This shows that
they have been thoroughly assimilated within the mainstream
Roman production.7
Engraved gems of the Republican period are well
represented in the collection; within an extensive iconographic
range those gemstones depicting Graeco-Roman divinities
show a substantial formal closeness to Roman production.8
Only a few intaglios depicting military life seem to show a
slight difference from those of the same subject found in
famous gem collections.9 The more numerous intaglios in the
collection, however, are those of the Imperial age that depict,
through different artistic trends, Graeco-Roman divinities,
followed by animals, symbols, attributes, all of which are
particularly close to military life.10
The engraved gems belonging to the Late Imperial period
are a minority, and in fact the gemstones from Tilurium, which
reflect Roman glyptic production of the period from the 3rd
and 4th centuries ad, are definitely less frequent in Tilurium,
as in Rome. Intaglios from Late Antiquity match the cultural
and artistic standards of Roman glyptic production. Some
items that illustrate this are shown here, particularly those
gemstones with male divinities that find precise analogies with
intaglios coming from well-known archaeological contexts,
and most notably with those gemstones discovered in legionary
fortresses.11
Late Roman Gems from Tilurium in Croatia
Plate 3 Red jasper intaglio with Plate 4 Red jasper intaglio with
Asclepius. Split, Archaeological Mercury. Split, Archaeological
Museum Museum
Among the gemstones made of opaque stone the most
numerous are jaspers with classical iconography depicting the
divinities of the Roman pantheon, particularly those associated
with prosperity. One may mention here, briefly, some red
jaspers with similar iconography in various collections as, for
example, a jasper with Bonus Eventus shown in profile to the
left with a chlamys falling off the shoulders, and a bent right leg
(Pl. 1).12 The god has the usual attributes: in his raised right
hand he balances a dish with offerings, in his left, two ears of
wheat. The peculiarity of the style finds a direct parallel with
other jaspers of well-known origin, as, for example, one from
Brigetio in Hungary.13
Direct parallels with intaglios in several collections have
also been found for another red jasper depicting Mars Ultor (Pl.
2).14 One can mention a jasper intaglio found in Xanten, whose
iconographic forms and details correspond to this one: the god
is attired as a legionary with a short chlamys draped behind his
shoulders.15 His head is portrayed in profile towards the right
with a helmet adorned by a high lophos. He holds a spear in his
left hand, which is level with the head, while in the right hand
he supports a shield embellished with bossed decorations.
Only a small number of intaglios from Tilurium can be said
to represent simplified work, such as another red jasper
depicting Asclepius standing full face with head turned to the
right (Pl. 3).16 The form of the chest gives the impression of
volume, but the schematic, disproportionate limbs lack
definition. The serpent-wreathed rod is schematic with broken,
vigorous lines, a rendition which finds a direct parallel in a
carnelian intaglio in Carnuntum, Austria.17
The god Mercury is one of the most commonly depicted
figures on intaglios found in Tilurium, although these differ
considerably in style and iconography. A red jasper set in a
fragmentary iron ring presents Mercury with basic features,
without internal details and from multiple perspectives (Pl.
4).18 However, it is designed along the usual pattern: the god is
standing with head in profile towards the right, and a chlamys
draped over his right arm, which is lowered to hold the
caduceus. He holds a pouch in his outstretched left hand. The
gem can be paralleled with some intaglios found in Aquileia,
particularly with respect to the shape of the caduceus wreathed
with flattened serpents.19 Mercury is also depicted on
carnelians (Pl. 5);20 on two of these the god is engraved in a
sketchy way, standing with a chlamys draped over his right arm
that holds a caduceus. The engraving with straight lines closely
resembles two intaglios found in the legionary fortresses of
‘Gems of Heaven’ | 131
Nardelli
Plate 5 Carnelian intaglio with Plate 6 Green jasper intaglio with
Mercury. Split, Archaeological cynocephalus. Split, Archaeo-
Museum logical Museum
Xanten and Carnuntum.21 These schematic figures reveal the
signs of the hasty mass production of well-known images.22
Only one magical intaglio was found in Tilurium (Pl. 6).23
This green jasper depicts an ithyphallic figure in profile
towards the left, standing on its toes with a crowned canine
head. The macrophallic cynocephalus in adoration, depicted
on apotropaic amulets, was almost always accompanied by
magical inscriptions, unlike the piece from Tilurium that is
anepigraphic. The motif is associated with the Egyptian god
Thoth and symbolises fertility and renewed life.
Far fewer cameos than intaglios have been found in
Tilurium; this is typical of all known glyptic collections,24 even
Croatian ones.25 The cameos, which can be dated to the Late
Antique period, are remarkably interesting since they have
provided information on a hitherto unknown period in the
history of Tilurium. They represent portraits whose stylistic
idiosyncracies are often hard to parallel in other gems
belonging to known Dalmatian collections.
The black glass cameo with a bust of a young man, full face,
turned slightly to the right, is not easy to place either
historically or stylistically (Pls 7–8).26 But, judging by the
plastic workmanship of the hair with its curly locks, one must
say that, at a first glance, the item recalls portraits of the
Antonine tradition, which was the reference for portraits
depicting princes in their early youth. The bust is slightly
reminiscent of the marble portrait of Lucius Verus ‘Bambino’,
Plate 7 Black glass cameo with a bust of a Plate 8 Profile drawing of
young man. Split, Archaeological Museum Plate 7
132 | ‘Gems of Heaven’
preserved in the Civic Museums in Trieste; a stylistic
correspondence can be found in the plump youthful face and
the hair, which is arranged in orderly locks that fall on the
forehead in thick curls.27 The physiognomy, especially the facial
features, does not correspond to that of a known personage. It
is presumably a private bust designed stylistically along the
lines of official Roman portraits. The typical characteristics of
official portraits are generalised and repeated for a certain
period.
Comparison with a yellow glass cameo found on Pharos
(Island of Hvar), and today preserved at the Museum of
Contemporary Art in Zagreb28 is very interesting and
significant for the close iconographic correspondence. The two
items have identical iconographic and stylistic details, differing
only in the colour of the glass. The precisely moulded head and
hair, and the cloak’s details indicate that they are
contemporary items, produced from the same matrix made by
the same workshop. But we must mention that the cameo from
Pharos is cast from a much worn matrix (Pls 9 and 10). This
can be deduced from the details of both hair and face (which
are more angular and flattened) and from the cameo’s
thickness: it is just a few decimillimetres thinner than the one
found in Tilurium. This is precious information because the
worn matrix leads us to surmise that production was rather
extensive.
A pale blue glass cameo with a frontally facing male bust
illustrates the typical difficulties one faces when studying this
class of materials (Pl. 11).29 Direct glyptic parallels are
unknown, but one may mention a glass cameo found in Sens,
France, with which we can find a likeness in the general
structure of the face and in the hair style, particularly in the
rendering of the locks on the forehead and in the style of the
dress.30 The gem is thus hard to place within a precise artistic
trend. As a matter of fact, although the cameo from Tilurium is
rather significant from a formal standpoint, it has no specific
elements to indicate its exact classification. As it is not easy to
establish a precise date for the cameo, one has to turn to
numismatics to try and give the cameo a chronological context.
Coins portraying the last western Roman emperors offer
suggestive evidence. The bust is full face with an idealised
expression, and hair combed forward in smooth locks to form a
fluffed fringe on the forehead. One finds a convincing
Plate 9 Yellow glass cameo from Pharos. Plate 10 Profile drawing
Zagreb, Museum of Contemporary Art of Plate 9

Plate 11 Pale blue glass cameo with frontally Plate 12 Ruby glass cameo with male bust. Split,
facing male bust. Split, Archaeological Archaeological Museum
Museum
comparison with coins of Iulius Nepos, the last legitimate
Roman emperor in the West and governor of Dalmatia who was
killed in Diocletian’s palace in ad 480.31 Though the cameo
from Tilurium does not directly copy the coin, it bears a certain
resemblance to the governor’s inspired expression and the
general facial features with the small round eyes and small
fleshy mouth. A more precise parallel can be found in the hair,
which is arranged in locks on the forehead with evenly-placed
fluffed curls.
Interpreting two glass cameos that depict a male bust with
head in profile to the right, has proven to be rather complex
and mostly theoretical.32 A useful chronological clue could be
the type of headdress worn by the personage depicted on a
ruby cameo: a heavy bowl-shaped cap (Pl. 12). Similar
headgear can be seen on portraits of Gothic kings, who wore
bowl-shaped helmets or crowns adorned with precious stones.
One can, for instance, mention some silver and bronze coins
minted in Rome and preserved in the Civic Numismatic
Collection in Milan, and a follis in the British Museum
portraying Theodahad (r. ad 534–6) in profile.33 The Dalmatian
piece resembles them, along with the shape of the headdress
and features of the profile: oval face, small chin and, especially,
the shape of the short moustaches.
The glass cameo has a direct parallel in a purple cameo
found in Tilurium about 20 years after the ‘ruby’ artefact was
found in 1917 (Pl. 13). The two cameos corresponding
iconographic and stylistic features lead us to judge them
contemporary items, produced from the same matrix. The two
cameos are also the same size, although the purple cameo is a
few decimillimeters thinner. This could have been caused by a
worn matrix, and such data is important as it is likely evidence
that similar cameos were locally produced.
In conclusion, this is an important collection of gems
because their known provenance bears more direct and
significant evidence than museum collections of diverse and
complicated origin. In fact, the gems from Tilurium are also
particularly interesting for the information they provide on the
little known local, historical frameworks. The intaglios
presented here are not as significant as those dating to earlier
periods, but the prolonged use of gems confirms that life
continued for a long time at the Tilurium settlement. We must
also say that the study of cameos has shown how the overall
picture still escapes us: in fact if we consider both the
heterogeneous features of the items and data provided by
Late Roman Gems from Tilurium in Croatia
Plate 13 Purple glass cameo with male
bust. Split, Archaeological Museum
individual analyses we see how they are hard to compare. The
question is still open, although an attempt was made to achieve
some clarification by ordering and presenting glass cameos,
which supply interesting information on the historical phase of
Tilurium about which virtually nothing is known.
Lastly our analysis of some significant items found in
Tilurium and presented here for the first time is designed to
underscore both their cultural value and our hope that studies
centring on these precious gems might contribute towards the
overall progress of research on glyptic production in the former
Roman province of Dalmatia, as well as in other Roman
provinces. From this standpoint work is merely at its earliest
stage.
Acknowledgements
I am grateful to Mrs sc. Zrinka Buljević, Director of the Archaeological
Museum of Split, for her kindness in allowing me to study the gems
from Tilurium. Many thanks to Prof. Maja Bonačić Mandinić for her
great help which I was lucky enough to be able to use in the writing of
this paper. I should particularly like to thank Dr. sc. Mirjana Sanader of
the Department of Archaeology, University of Zagreb, for having
proposed and encouraged the study of the glyptic collection from
Tilurium. I am especially grateful to Ardythe Ashley, Cristiana Fusco
and Davide Trame, Remza and Želljko Koščević for their help with
many valuable and helpful suggestions. The photographs were taken
by T. Sesel of the Archaeological Museum of Split; warm thanks also to
I. Prpa Stojanac, from the Archaeological Museum of Split, for kindly
providing me with drawings of two of the glass cameos.
Notes
1 The collection has an ancient history: the gems were acquired by
the Archaeological Museum, as deliberate policy, from the time of
its foundation in 1820. From the second issue of the Bullettino di
Archeologia e Storia Dalmata onwards (from 1879 until 1926) a list
of acquired gems (without photos) was published by Mons. Frane
Bulić, who dedicated particular attention to this kind of
archaeological material: B. Nardelli, ‘Sulle gemme di Salona’, in E.
Marin (ed.), Longae Salonae, Split, 2002, 205–14, at 205.
2 The group of gemstones kept in the Ashmolean Museum in Oxford
were acquired by Sir Arthur J. Evans during his visit to Dalmatia:
S.H. Middleton, Engraved Gems from Dalmatia. From the
Collections of Sir John Gardner Wilkinson and Sir Arthur Evans in
Harrow School, at Oxford and Elsewhere, Oxford, 1991, 61, no. 68;
82, no. 121; 94, no. 153; 116, no. 209; 120, no. 217; A. Brown, Before
Knossos... Arthur Evans Travels in the Balkans and Crete, Oxford,
1993, 19–29.
3 Andenterium (present day Muć) was the home of the Cohors VIII
Voluntariorum: J.J. Wilkes, Dalmatia, London, 1969, 139, 170, 176,
184, 221, 347, 453. For gems displayed in the Ashmolean Museum
see: Middleton (n. 2), nos 129, 166, 171. Burnum (present day
‘Gems of Heaven’ | 133
Nardelli
Ivoševci) was the Legio XI military camp; from ad 42 the Legion
held the honorary title Claudia Pia Fidelis: Wilkes (ibid.), 97, 392,
469. For gems kept in the Archaeological Museum of Zadar: B.
Nedved, Parures dans la Dalmatie du Nord depuis la Préhistoire
jusqu’à nos jours (exh. cat., Š. Batović [ed.]), Zadar, 1981, 161–4, nos
129–48; and in the Ashmolean Museum: Middleton (n. 2), nos 23,
43, 65, 91, 132, 167, 214, 222, 262.
4 Systematic archaeological excavations in Tilurium, conducted by
the Department of Archaeology of the University of Zagreb, have
been ongoing from 1997: M. Sanader, Tilurium I, Istraživanja-
Forschungen 1997–2001, Zagreb, 2003; idem, ‘The Roman legionary
Fortress at Tillurium-State of Research’, in D. Davidson, F. Gaffney
and E. Marin (eds), Dalmatia. Research in the Roman Province
1970–2001. Papers in Honour of J.J. Wilkes (BAR 1576), Oxford,
2006, 59–66. Archaeological research at the legionary camp of
Burnum, led by the University of Zadar and the University of
Bologna, started in 2005: N. Cambi et al. (eds), L’esercito romano a
Burnum (Katalozi i monografije Burnuma II), Drniš-Šibenik –
Zadar, 2007.
5 The military fortress Tilurium was occupied by Legio VII, very
likely in ad 8 when it was included in the military forces in
Dalmatia. Legio VII, from ad 42 Claudia Pia Fidelis, left the castrum
between ad 45 and 61; after that the Cohort VIII Voluntariourum,
which was stationed there until the middle of the 3rd century ad,
occupied the settlement and the province: Wilkes (n. 3), 97, 139,
176, 347, 392, 470; M. Zaninović, ‘Military characteristic of
Tilurium in Antiquity’, Od Helena do Hrvata, Zagreb, 1996, 272–91.
6 In the late Empire the fortress Tilurium was situated on a broad
strip of territory provided for the military stations – the prata
legionis, ‘lands of the legion’ – which had suffered from almost
every kind of invasion. Dalmatia as a frontier zone between the
eastern and western halves of the Empire was constantly exposed
to barbarian attacks: Wilkes (n. 3), 392, 416–27; Zaninović (n. 5),
290.
7 The catalogue of the engraved gemstones found in Tilurium and
kept in the Archaeological Museum in Split is forthcoming:
B. Nardelli, Gemme da Tilurium nel Museo Archeologico di Spalato.
8 We can mention one carnelian with Telephus, as well as a
chalcedony, an agate depicting Rhea Silvia, an iconography known
only from glyptics: Nardelli (n. 7), cat. nos 64, 110.
9 Only a small number of intaglios show evidence of iconographic
idiosyncracies, for example, one red jasper showing Mercury in
the temple; or two carnelians, one with Mercury’s boot, the other
depicting a helmet and a fly (Nardelli [n. 7], cat. nos 16, 176, 178),
as well as one very peculiar carnelian depicting Capaneus
(B. Nardelli, ‘Gemme e iconografia: appunti da Catoro e da
Tilurium’, Pallas 83 (2010), 157–65, at 161–2).
10 We can mention some intaglios with a very richly patterned style:
one carnelian showing Mercury seated (Nardelli [n. 7], cat. no. 22);
one chalcedony with a horse of the Dioscuri (eadem, cat. no. 144).
From the group depicting symbols there is one carnelian with a
cock surround by symbols (eadem, cat. no. 140), and one carnelian
with group of various symbols (eadem, cat. no. 183).
11 They are similar in their general figurative schemes and are close
iconographically with gems found in legionary camps and other
military places all over the Empire: from Caerleon to Xanten, from
Brigetio to Carnuntum: J.D. Zienkiewicz, The Legionary Fortress
Baths at Caerleon. II: The Finds, Cardiff, 1986; G. Platz-Horster, Die
antiken Gemmen aus Xanten I, Bonn, 1987; T. Gesztelyi, Gemstones
and Finger Rings from Brigetio, Collections of the Kuny Domokos
Museum of Tata 6, Tata, 2001; D. Dembski, Die antiken Gemmen und
Kameen aus Carnuntum, Vienna, 2005.
12 Archaeological Museum of Split (hereafter: AMS): Inv. no. I 853;
13.7 x 10.7 x 3.0mm.
13 Of the numerous direct parallels much importance is placed on
those pieces with a well-known provenience: T. Gesztelyi, Antike
Gemmen im Ungarischen Nationalmuseum, Budapest, 2000, 64, no.
159, pl. 140.
134 | ‘Gems of Heaven’
14 AMS, Inv. no. I 2197; 11.2 x 12.8 x 2.3mm.
15 Platz-Horster (n. 11), 62, no. 113, pl. 21; eadem, Die antiken Gemmen
aus Xanten im Besitz des Archäologischen Parks-Regionalmuseums
Xanten, Cologne, 1994, 73, no. 6, pl. 1.
16 AMS, Inv. no. I 792; 8.6 x 7.8 x 1.7mm.
17 Dembski (n. 11), 65, no. 111, pl. 11.
18 AMS, Inv. no. I 1475; with bezel: 10.5 x 6.8 x 2.9mm.
19 G. Sena Chiesa, Gemme del Museo Nazionale di Aquileia, Padua,
1966, 140, nos 176–9, pl. IX; Dembski (n. 11), 71, no. 185, pl. 18.
20 AMS, Inv. no. 38347; 12 x 8.2 x 2.9mm; Inv. no. I 2353; 8.9 x 9.4 x
2.2mm.
21 Platz-Horster (n. 11), 91, n. 158, pl. 32; Dembski (n. 11), 71, no. 179,
pl. 18.
22 Ibid.
23 AMS, Inv. no. I 1665; 12.3 x 8.9 x 3mm. Cf.: H. Philipp, Mira et
Magica. Gemmen im Ägyptischen Museum der Staatlichen Museen.
Preussicher Kulturbesitz Berlin - Charlottenburg, Mainz am Rhein,
1986, 98, no. 148, pl. 39; A. Mastrocinque (ed.), Sylloge Gemmarum
Gnosticarum, I (Bollettino di Numismatica, Monogr. 8.2.I), Rome,
2003, 198, nos 84–5.
24 For cameos belonging to the Late Imperial period: M. Henig, The
Content Family Collection of Ancient Cameos, Oxford, 1990, xiii; E.
Gagetti, ‘Intagli e cammei: catalogo topografico della Croce del
Desiderio’, in G. Sena Chiesa (ed.), Gemme dalla corte imperiale
alla corte celeste, Milan, 2002, 181–229; G. Sena Chiesa, ‘La glittica e
la produzione suntuaria nell’età di Ambrogio: arte e potere fra
classicità e cristianesimo’, in P. Pasini (ed.), 378 d.C. Ambrogio e
Agostino: le sorgenti dell’Europa (exh. cat., Milan), Milan, 2003,
189– 94, at 191–4.
25 In the Tilurium collection numbering 300 gems, there are only
nine cameos. Of the more than 2,000 gemstones found in Salona
and displayed in the Archaeological Museum in Split, there are
only 40 cameos: Nardelli (n. 1), 206. Cameos are likewise fewer
than intaglios in other Croatian glyptic collections: only a small
number have been published: A. Larese and B. Nardelli (eds), Arte e
cultura in Croazia. Dalle collezioni del Museo Archeologico di
Zagabria (exh. cat., Turin), Turin, 1993, 161, 168, nos 217, 243, 245;
R. Koščević, Arheološka zbirka Benko Horvat, Zagreb, 2000, 89–92,
nos 112, 115–20; N. Cambi, Imago animi. Antički portret u Hrvatskoj,
Split, 2000, 200, no. 307; B. Nardelli, ‘Sulle gemme di Spalato: dal
territorio al Museo’, Vjesnik za arheologiju i povijest dalmatinsku
100 (2007), 79–104, at 85–6, nos 1, 2.
26 AMS, Inv. no. I 1300; 12.5 x 10 x 4.8mm.
27 M. Verzár-Bass, 'Ritratti', in M. Verzár-Bass (ed.), Trieste. Raccolte
dei Civici Musei di Storia ed Arte e rilievi del Propileo, 1, Trieste,
2003, 100, no. SR 16, pl. XVIII, 60, 63.
28 Koščević (n. 25), 92, no. 120.
29 AMS, Inv. no. I 1502; 14.4 x 10.5 x 4mm.
30 The two cameos are alike also in relation to colour and glass size:
H. Guiraud, ‘Intailles et camées du Musée Rolin d’Autun’,
Mémoires-Société Éduenne des Lettres, Sciences et Arts 56/2 (1997–
1998), 137–69, at 166, no. 108.
31 Iulius Nepos (ad 475–80): Wilkes (n. 3), 422, 423. One can mention,
for example, one solidus, dated to ad 474–5: M. Radnoti Alfoldi,
‘La monetazione romana in età tardo – antica (284/476 d.C.)’, in
D. Panvini Rosati (ed.), La moneta greca e romana, Rome, 2000, 151,
pl. IX, X, 22b.
32 Cameo in ruby glass: AMS, Inv. no. 1262; 13.4 x 12 x 4.6mm. Cameo
in purple glass: AMS, Inv. no. 2507; 13.3 x 11.8 x 4mm.
33 Theodahad (ad 534–6): E.A. Arslan, ‘La monetazione’, in G.
Pugliese Carratelli (ed.), Magistra Barbaritas. I Barbari in Italia,
Milan, 1984, 413–53; E.A. Arslan, Le monete di Ostrogoti, Lombardi
e Vandali, Catalogo Civiche Raccolte di Milano, Milan, 1978, 47, nos
159–60; P. Grierson and M. Blackburn, Medieval European Coinage.
1. The Early Middle Ages (5th–10th centuries), Cambridge, 1986,
432, no. 142, pl. 8. I am especially grateful to Professors Armanno
Arslan, Tadeusz Baranowski and Murizio Buora for their valuable
and helpful suggestions.
Three Degrees of Separation: Detail Reworking, Type
Updating and Identity
Transformation in Roman Imperial Glyptic Portraits in the Round

Elisabetta Gagetti
The ‘portraits in the round’ in the title belong to the glyptic
class we can call ‘precious sculptures’.1
These ‘precious sculptures’ were made in – more or less –
any usual glyptic material as well as in other materials of
organic origin such as ivory and amber (Table 1). The five
portraits that I am discussing here are: three in chalcedony,
one in agate and one in a much rarer and harder stone,
aquamarine.2 All five pieces are of unknown provenance,
except for the bust of Trajan (Pl. 1), said to have been found in
North Africa.3 A distribution map of the ‘precious sculptures’ on
the whole4 shows their presence throughout the Roman
Empire, as usual for many classes of luxury objects which
travelled with their owners.
Finally, these ‘precious sculptures’ are in miniature: the
criterion for admission into the class is the distance between
the chin (gnathion) and the top of the head (vertex), which
must be shorter than or equal to 12.5cm, more or less half of a
life-size sculpture (that is 24–22cm for male heads, 22–20cm for
female ones).5 The largest ‘precious sculpture’ known at
present is a sardonyx head of Augustus (‘Prima Porta type’),
reworked from a portrait of Domitian (Type III),6 whose
gnathion-vertex distance measures 11.9cm. Rounding off such
height, 12cm is the longest gnathion-vertex admitted for a
‘precious sculpture’.7

chalcedonies
63
agate
12 chalcedony
carneol
carnelian
6
onyx
93 plasma
18 sardonyx
quartzes
aquamarine
1 138
alabaster 10
amber 1 9
2 9
aragonite
ivory 14
1
coral
olivine 45
jasper
fluorite 26
jet 5
4
garnet macrocrystalline quartzes
lapis lazuli 22
steatite 69 amethyst
78 rock crystal
turquoise 4
chalcedonies

Table 1 Glyptic materials attested for the ‘precious sculptures’

200
174
180
160
140 122
120
100
80
60 48
41
40
14
20 3 3
0
portraits divinities images intellectuals 'opera genre incerta
related to nobilia' scenes
Table 2 Subjects attested for the military
‘precious sculptures’ victory

‘Gems of Heaven’ | 135

Gagetti
Hellenistic Roman Portraits
rulers emperors ‘with no name’
Male
portraits 17 29 24
Female
portraits 4 13 13
Infantile
portraits 1 4 17
Table 3 Portrait subjects attested among the ‘precious sculptures’
In the corpus of 458 precious statuettes that I collected (but
now they are somewhat more), 122 are portraits (Table 2),8 that
is 26.6% of the total, divided between dynastic (Hellenistic and
Roman) and unknown portraits as in Table 3. Eleven of these
‘precious portraits’ of Roman emperors and princesses have
been reworked. We know this because their types are well-
known from other media. ‘Precious sculptures’ in the round,
like cameos, can be modified, even to a substantial degree ‘per
via di togliere’ (‘cutting the stone off’). Palinglypt is the
evocative term created by Platz-Horster to define the reworked
glyptic object.9 In fact, like in a palimpsest, the overwriting
sometimes allows one to ‘read’ the underlying, scraped off
glyptic text.
There can be three degrees of separation from the original
version of a ‘precious portrait’. The first degree is a simple
reworking of minor details. The identity of the portrait is not
altered. The piece can have been reworked at any moment of its
life. The second degree involves the reworking of
characterising details, usually hairstyle (often the key element
in determining the type of an imperial portrait), but with no
alteration of the facial features. This kind of reworking is
usually contemporary with the life of the portrayed person. In
the third degree, the reworking affects the most characterising
details and/or facial features. The identity of the portrait is
transformed into another one. There is not a typical moment
for this kind of intervention: it can take place immediately after
the death of the original portrayed person, or centuries later.
The following short case studies will illustrate such progressive
degrees of separation.
First degree
A chalcedony bust of Trajan in Berlin (Pl. 1)10 is a perfect
example of the reworking of minor details, leaving the facial
features of the portrait unaltered, even if its later interpretation
Plate 1 Chalcedony bust of Trajan (Berlin, Antikensammlung)
136 | ‘Gems of Heaven’
changes substantially. The emperor, in his II portrait type (or
‘civic-crown type’),11 is depicted en buste: he wears a cuirass
with shoulder straps decorated by a fulmen (visible on the right
one only), and two rows of fringeless pteryges; in the middle of
the breastplate is a carefully rendered gorgoneion. The
paludamentum is gathered into a Schulterbausch on the left
shoulder, clasped by a round fibula.
At a certain moment, the Schulterbausch could have been
flattened12 and the round fibula could have ‘slid’ towards its
lower border,13 but what is certain is that 11 cylindrical blind
holes have been drilled on the breastplate of the cuirass,
obliterating its surface. They measure 8mm in diameter, and
almost 10mm in depth. All the scholars who have studied the
object have supposed that these blind holes were for the inlay
of precious stones (an intervention that has been generically
dated between Late Antiquity and the Early Middle Ages).
From the 4th century ad onwards, cuirasses embellished with
gems do exist, but the disposition of the gems appears
substantially different, so that a different ‘model’ must be
looked for.
Zwierlein-Diehl,14 taking into account both the holes on the
breastplate and the gorgoneion in relief, realised that they
amounted to 12, a number she immediately connected with the
stones on the breastplate (choschen) of the High Priest of
Exodus 28.15–21.15 She then proposed a reconstruction of the
inlay of these stones into the holes on Trajan’s breastplate (of
course, the rows are to be imagined right to left, according to
the direction of writing and reading in Jewish culture).16 The
disposition of the 3rd and 4th rows along the lower border of
the bust is due to the lack of space under the former two rows.17
Two among the many interesting results of Zwierlein-Diehl’s
study are of particular interest here. First: even in this
christiana interpretatio, or christiana mutatio, the trans-
formation of Trajan into Aaron, the gorgoneion was not erased,
but re-used, notwithstanding its figuration.18 Secondly,
particularly relevant to the theory ‘of the three degrees’ is the
drilling technique using a hollow cylindrical drill bit, the traces
of which can be clearly seen inside the holes on the breastplate.
Such drills came into use in Paris not before the 13th century,
possibly during its second half.19 This implies that the
reworking of the breastplate should be placed in western
Europe, in the central centuries of the Middle Ages. The
question of complete medieval reworkings of ‘precious
portraits’ will be discussed again at the end of this paper.
Second degree
With respect to the reworking of characterising details, usually
the hairstyle, often the key element in determining the type of
an imperial portrait, but with no alteration of the facial
features, a portrait of Tiberius in the Cabinet des médailles,
Paris, is of particular relevance (Pl. 2a–d).20 The head, in grey
chalcedony, slightly inclined towards the right, is broken at the
top of the neck, immediately under the jaw, and today is
mounted on a modern gilt bronze bust wearing lorica and
paludamentum. The back of the head has been cut away and
then a circular area has been hollowed out (Pl. 2b): these
interventions are no doubt modern, as they obliterate the locks
in this area of the head. The face has sharp eyebrows in light
relief but with no engraved etching. In the large eyes, under the
upper eyelid – a little longer then the lower one, affected by

puffy ‘bags’ – the iris is outlined by a carved line and the pupil
is rendered by a very small drill hole.
If we look at the front of the head, there can be little doubt
that we are facing a well-known portrait type (Pl. 3: 34.Le):
over the mid-forehead, but slightly to the left of the exact
centre, there is a fork, and two small ‘pincer’ locks are over the
temple. On the left and right of the fork, the hair is parted into
three main, further parted locks, curved toward the temple. All
the locks are of the same length, so that they form a kind of
upper rectilinear delimitation of the forehead. The outer arms
of the ‘pincer’ locks are longer than the inner ones. The cluster
of all these features can be found in one of the latest types of
Tiberius’ portrait, referred to differently in the typologies,21
and named respectively by Dietrich Boschung and John Pollini
as the ‘Chiaramonti type’22 and the ‘Second Princeps type’.23
The former is considered to have been created in the late
Augustan age24 and replicated throughout his life as well as in
posthumous portraits;25 the latter type dates from ad 31–34,
and likewise was used long after Tiberius’ death.26 In the
corpus of Tiberian portraits, the closest parallel to the front
view of our small chalcedony head, owing to the wideness of
the forehead, seems to be the portrait at Woburn Abbey (Pl. 4).27
But two features do not match the ‘Chiaramonti/Second
Princeps type’. First, the youthful look of the face of the small
head in the Cabinet des médailles, as already pointed out:28 it is
very smooth with no trace of nasolabial folds, which can be
found in the portrait type in question. Secondly, from the side
view, the hair too does not match the ‘Chiaramonti/Second
Princeps type’: in fact the locks are not all combed forward, but
between the temple and the ear, the hair is arranged in locks
running one into another from the face toward the back of the
head. Only in front of the ear are the locks combed forward.
Such an element is to be found in another portrait type of
Tiberius (Pl. 3: 30.La), characterised by a fork over the inner
Three Degrees of Separation
Plate 2a–d Chalcedony
head of Tiberius. Paris,
Cabinet des médailles
Plate 3 Scheme of Tiberius’ portrait
types (after Boschung [n. 22})
corner of the left eye, from which two ‘pincer locks’ depart:
one, on the left, reaches the temple; the other, on the right side,
comes over the mid-eye. But what is interesting here is the
profile view: on both sides of the face the hair is combed
backwards, but the short side-whiskers are formed by hooked
locks arranged toward the cheeks. Moreover, over the ears and
on the nape the hair is done in wavy locks combed forward (in
a double layer on the nape). Both Boschung and Pollini agree
on the fact that this is the first of Tiberius’ portrait types,
created in the ’20s of the 1st century bc,29 or, more precisely, in
19 bc:30 the early dating (Tiberius, born in 42 bc, would not
have been portrayed in this type older than a 23-year-old)
would be, in this case, the actual reason for Tiberius’ youthful
look. Among the known sculptural replicas, the best
comparison for the small chalcedony head is offered by a
marble bust from Luni (Pl. 5):31 here, the locks on the temples
are long enough to allow the reworking of the hair over the
forehead into a short, rectilinear fringe, without ‘holes’ in its
corners.
Plate 4 Marble bust of Tiberius. Plate 5 Marble bust of Tiberius from
Woburn Abbey Collection Luni. La Spezia, Museo Civico
Archeologico
‘Gems of Heaven’ | 137
Gagetti
Plates 6a–d Aquamarine head of Sabina. Venice, Museo Archeologico Nazionale
Anyway, it is important to stress that the small head in Paris
shows the distinctive elements of two different portrait types
of Tiberius, created at a distance of three to five decades,
according to the typology one prefers to follow. It seems, then,
that such precious portraits, created in court ateliers for the
court elite, at the changing of the official portrait of a prince or
emperor were updated to the new, current portrait type.
Third degree
Reworking can also affect the most characterising details of a
‘precious sculpture’ and/or its facial features. The identity of
the portrait is, hence, transformed into another one. Let us
examine three case studies.
3.1. The princess in Venice
In the Museo Archeologico Nazionale in Venice32 there is a
charming small head (3cm) in aquamarine (Pls 6a–d),
identified at different times and by different scholars mainly as
Cleopatra VII,33 although Sabina too has also been suggested.34
Aquamarine is quite rare in Roman glyptic both for its high
value and for its hardeness (7.5–8 on Moh’s scale) and among
the ‘precious sculptures’ it amounts to less than 1% (Table 1).
It is noteworthy that another famous specimen is the intaglio
portrait of another princess who lived a generation earlier:
(Flavia) Julia Augusta, daughter to Titus and Domitian’s
mistress, signed by the engraver Euodos.35 The coiffure of the
Venice head, on the top and on the back, is clearly not finished
or reworked. Where it is better preserved, over the forehead
and on the temples, it is characterised by a central, rather
accentuated parting of the hair, which is arranged, at each side
of it, in at least four rows of waved locks, combed diagonally to
the back of the head. The forehead is framed by the beginning
of the lower part of each lock, worked with a small hole. The
locks framing the face are six at each side of the central
partition. In front of the ear there is a curl, better preserved on
the right side of the face. Beyond this wavy section, there is a
deep groove; and, beyond again, the skullcap is simply rough-
shaped, with circular depressions, meant as a fixing system for
a headgear in a different material. On the nape, the locks are
twisted (the direction of the twisting is always toward the
central axis of the head) and combed up to the smooth
skullcap.
138 | ‘Gems of Heaven’
As a first step, we can exclude Cleopatra VII. Cleopatra’s
famous nose was markedly convex throughout its length, from
the eyes to the tip (Pl. 7); while the Venice head shows a truly
aquiline nose, slightly convex but with a more prominent
‘bridge’. Cleopatra’s coiffure is also different: both on coins and
in the few sculptural portraits attributed to her,36 she wears a
typical Melonenfrisur: the thick locks are twisted (not waved)
and combed parallel from the forehead and the temples to the
nape, where they are arranged in a sort of ‘open’ chignon.
Moreover, Cleopatra, as a Hellenistic ruler, is represented with
the diadema, in the fashion of the late Hellenistic age. No trace
of it can be seen on the aquamarine head, even on the nape,
which still preserves its locks, but no sign of the knot of the
diadema.
Comparison with another small head, in the same material
and of almost the same height (2.8cm), in the Museo
Archeologico Nazionale in Florence (Pls 8a–d),37 makes clear
that the portrayed lady in Venice is Sabina, the wife of Hadrian.
On the Florence head, she wears her ‘turban’ hair dress (Type
Carandini VA3a)38 with a stephane and a wide-nested chignon
formed by loose hair (not by thin plaits), which, according to
coin types of the mint of Rome, seems to have been worn by
Sabina between ad 134 and 137 (iconographic period VIII).39
Reducing the two aquamarine heads of Sabina to the same
scale, the result is that the profiles of the Venice and Florence
Sabinas can be perfectly superimposed,40 although the mass of
the Venice coiffure has been consistently reduced.
The problem, at this stage, is the coiffure of the Venice
Sabina. The original look of this ‘new’ Sabina coiffure can be
Plate 7 Tetradrachm of Ascalon: Cleopatra VII

Plates 8a-d Aquamarine head of Sabina (Florence, Museo Archeologico Nazionale)
Plates 9a-b Marble bust of Sabina said to have come from an aristocratic villa
in the ager Tiburtinus
seen in some other marble heads. The first to consider (Pls
9a–b)41 is supposed to come from an aristocratic villa in the
ager Tiburtinus,42 not far from Hadrian’s Villa. Behind the wavy
frontal section, the hair is combed in thick, plaited locks
toward the nape, where they join, in a large knot, with the
twisted locks from the top of the neck. A nested chignon,
formed by three superimposed plaits, surrounds the head,
looking like a diadem over the wavy frontal part. At first, the
head was attributed to a member of the imperial family, likely
Sabina herself, owing to the resemblance of the facial features
and for the general hairstyle.43 Nonetheless, this identification
was abandoned, precisely because of the coiffure, which finds
no parallels in the known portraiture of Sabina.44 A solution to
the problem seemed to come from the excavations of the
Roman theatre of the Campanian town of Suessa Aurunca.45 A
Julio-Claudian building, the theatre was destroyed by an
earthquake, perhaps in the Flavian period, and was restored,
together with its annexes, by Matidia the Younger, Sabina’s
sister, during the reign of Antoninus Pius, as suggested by a
fragmentary monumental inscription, the original location of
which must have been near the entrance, between the southern
parodos and a hall called ‘southern basilica’.46 As Matidia’s
financial investment was considerable, involving the luxurious
reconstruction of the scenae frons with three orders of columns
in coloured marbles, decorated with numerous statues
depicting the members of the Ulpian-Aelian ‘dynasty’, the
restoration of the porticus post scenam, and the remaking of
waterworks related to the theatre,47 it seemed obvious to
attribute to Matidia the Younger the two heads, found in the
excavations, wearing the same hairstyle as the marble head
Three Degrees of Separation
Plate 10 Marble bust attributed to Plate 11 Marble bust attributed to
Matidia the Younger. Sessa Aurunca, Matidia the Younger. Sessa Aurunca,
Castello Ducale Castello Ducale
from Tivoli. Particularly important is a portrait (Pl. 10)
belonging to a much larger than life-size statue, whose clothes
are in bigio morato, which, from the position of its fragments,
should have been on the second stage of the scaenae frons48 over
the valva regia, depicting the portrayed woman divinised as
Aura.49 Who else could have she have been but the building
patroness, Matidia the Younger? The same identification, then,
followed for the second head (Pl. 11), which, although in a bad
state of conservation and smaller, is identical to the former, and
must represent the same person.50 So, the context gave a face to
Sabina’s mysterious sister,51 of whom no certain portrait
existed,52 as she never appeared on a coin type.53
We could wonder whether it was really possible that the
sister of an Augusta (Sabina received the title in ad 128),
already dead (ad 136–7) and consecrated, could put a larger
than life-size statue of herself in the place of honour on a
scaenae frons where the statues of other members of the
emperor’s family were also displayed, her sister included.54
Anyway, all the scholars who have dealt with the so-called
Matidia-Aura considered the reason strong enough, beyond her
euergetism towards the town of Suessa Aurunca, that Matidia
the Younger, who was never granted the title of Augusta, was
the daughter, niece and sister of divinized Augustae, and also
matertera (‘maternal aunt’, even if no blood tie existed between
the two) of the reigning Emperor Antoninus Pius.55
Anyway, the attribution of the two portraits from Suessa to
Matidia the Younger involved also the head from Tivoli
(formerly Sabina),56 whose ‘family look’ should have been
further stressed by the sculptor. In fact, the application of a
geometrical technique of relief and superimposition of
‘Gems of Heaven’ | 139
Gagetti
Plates 12a-d Female marble head. Paris Musée du Louvre
triangular patterns to some anatomical marker points in the
faces of Matidia-Aura, of Matidia’s head from Tivoli (then still
believed to be Sabina) and to a certain portrait of Sabina from
Piazza d’Oro at Hadrian’s Villa57 gave a result of ‘high
somatometric affinity’, as the patterns can be almost perfectly
superimposed in all three cases.58
In addition three other portraits were grouped under the
name of Matidia the Younger:59 a head in Paris (Pls 12a–d),60
another one in New York (Pls 13a–c)61 and a third one in
Warsaw (Pls 14a–c),62 the latter two pieces seemingly
reworked in their faces. As far as the coiffure is concerned, ‘la
similitude est pratiquement totale, mèche à mèche’.63 As to the
faces, ‘l’identité ... de la structure des visages est également
nette ... et il s’agit bien ... de la même personne’.64 As to the
attribution of the Louvre head (and hence of the other two) to
Sabina, Baratte was fascinated by such a possibility, but he was
doubtful too because of the coiffure, close to those exhibited by
Sabina in her later life, but without any real comparison with
the empress’ sculptural portraits.65 But so many are the
variations on the theme of Sabina, that sometimes, in front of
one of her non-canonical portraits, some scholars have
140 | ‘Gems of Heaven’
Plates 13a–c Female marble
head. New York, Metropolitan
Museum of Art
Plates 14a–c Female marble
head. Warsaw, Muzeum Narodowe
w Warszawie
proposed a different subject, even outside the imperial family.66
So, we return to this particular hairstyle again. It is
noteworthy that all of the considered portraits showing it are
missing their noses, which is no doubt the most typical of
Sabina’s facial features. All that is but one: the aquamarine
head in Venice, even if in its non-original condition. Perhaps,
then, we have that happy and rare circumstance in glyptic,
which is to identify a ‘new’ imperial portrait.
It remains to identify the lady into whom the Venice Sabina
was transformed, completed with a sort of golden wig, now
lost, as we can infer from the 1593 inventory of the gems
adorning the ‘Studiolo’ which Giovanni Grimani, Patriarch of
Venice, bequeathed to the Statuario Pubblico della
Serenissima: ‘Una testa incoronata di oro con il petto dorato’
(‘A head crowned with gold, with a golden breast’).67 The most
likely candidate is Helena, the mother of Constantine.
First of all, the two ladies seem to have had a very similar
nose, with the typical prominent ‘bridge’. Then, one of the
hairstyles worn by Helena on coins (‘Haarkranzfrisur’) (Pl. 15)
was characterised by wavy locks, parted in the centre over the
forehead, combed towards the ears. On the back, the mass of

Plate 15 Constans’ half-centennionalis: Helena
the long hair was parted into two plaits, superimposed onto
each other and surrounding (in opposite directions) the head.
The remaining long, straight hair under the ‘crown’ formed by
the two plaits was bent upward in an S-shape, leaving visible
the earlobe only, and fixed on the nape, passing under the two-
plait ‘crown’. The coiffure seems to have been adorned by
jewelled pins (?) piercing the hair ‘crown’.68
Such a hairstyle was very voluminous and to achieve it
from an older portrait involved the addition of a stone toupet.
This kind of reworking is attested, for instance, on one of
Helena’s most famous portraits, the seated statue in the Museo
Capitolino,69 reworked, according to various scholars, from a
portrait of an Antonine empress, Faustina the Younger or
Lucilla.70 The lower part of the wavy frontal locks was
maintained, while the two surrounding plaits were created in
the upper part; all the back, worked apart and now lost, was
added, as we can infer from the rough surface of the skullcap.
The placement of a hairpiece under the ‘crown’ is clearly
revealed by two holes on the neck, under the ears, still
preserving parts of the ancient iron studs, clearly the traces of
the obliteration of the original coiffure.71
The Venice Sabina, or better Helena, should have worn a
toupée, but of gold, partly due to the high value of the material
of the precious portrait, but largely because an addition in such
a transparent hardstone would have been simply impossible.
The scheme of reworking was different, as different as was the
starting coiffure. The scheme of the transformation can be
exemplified by another head, as the series of Sabina’s portraits
wearing the ‘Venice hair-dress’ is not yet complete. A portrait
(Pls 16a–b),72 surely reworked into the portrait of a young
woman,73 shows over the forehead, immediately behind the
wavy frontal and on the nape traces of the hairstyle under
discussion, while the skullcap is very similar in its present
Plate 17 Agate bust of Claudius, reworked in the Plate 18 Scheme of Claudius’ portrait types
4th century AD. H. 9.5cm, W. 8.8cm. London,
British Museum, GR 1872,0604.1308
Three Degrees of Separation
Plates 16a–b Marble head of a young woman. Formerly Marshall Collection
condition to the Venice Sabina.74 In this latter case, the glyptic
master did not need to ‘create’ the earlobes,75 because, in
contrast to the compact, wavy cap of hair worn by Faustina the
Younger or Lucilla, the ‘Venice coiffure’ left the ears totally
free. What were needed were the plait ‘crown’ and the S-shape
under it. Both must have been realised in a gold sheet, after the
elimination of the existing plait ‘crown’, also shown as receding
on the Sabina version. The gold wig, then, anchored in the
deep groove surrounding three quarters of the aquamarine
head,76 and filled with plaster or pitch in order not to be dented,
must have reproduced the ‘crown’, the hair on the skullcap
inside it77 and the S-shape hairpiece too. The locks on the nape,
rather thick in the marble versions, are in fact reduced into an
almost flat surface furrowed by the deepest levels of the
original carvings, because this area was to be covered by the
golden hair bent up to the ‘crown’.
3.2. Variations on a chronological theme
An agate head in the British Museum (Pl. 17) has,78 since it was
first published by Walters in 1926, been identified as the
emperor Claudius. Only the face with the frontal hair survives
today. The face, which tapers to a point in its lower section, has
a narrow forehead, slightly arched eyebrows rendered by
parallel diagonal incisions, cheeks in low relief, and a thin
straight nose. The eyes, looking upright, are wide open, with
extremely thin eyelids: both the iris and the pupil are outlined
by a carved groove. Some asymmetries allow us to suppose
that the head was originally turned to the left.
Among the three portrait types of Claudius (Pl. 18),79 if we
reduce the agate head and the one depicting the emperor to the
same scale, it is undeniable that the agate head could only be
‘contained’ by the Kassel head, (Pls 17 and 19a–b)80 the
eponymous portrait of Claudius in his ‘Kassel type’ (Pl. 18:56.
Va),81 which is higher on top, and larger, particularly in the
‘Gems of Heaven’ | 141
Gagetti
Plates 19a–b Marble bust of Claudius in his ‘Kassel type’. Kassel, Staatliche
Kunstsammlung
Plates 21a–b Detail of the Emperor
Honorius on the Rothschild Cameo
temporal region. In the ‘Kassel type’ the hair over the forehead,
rather voluminous and compact, shows a fork over the mid-eye
on the right, and another one in the left corner of the forehead,
following immediately a closed ‘pincer lock’. The upper layer of
locks is brushed wholly towards the right. In almost all of the
replicas of the ‘Kassel type’, Claudius’ face appears youthful
and smooth. A comparison with coin portraits of the emperor
finds a perfect parallel in the second type of his first year as
ruler, in ad 41.82
Returning to our comparison, the main facial features, like
the lower edge of the hair over the forehead, the eyebrows, the
outer corners of the eyes and the gap between the lower lip and
the chin, are in the same position on both faces. Also some of
the details are very similar, such as the lines of the lower eyelid
and the design of the mouth. Moreover, the disposition of the
lowest layer of the locks over the forehead, with a fork over the
centre of the right eye and another one on the left temple (a
slight ‘movement’ of the locks over the outer corner of the left
eye can be the trace of a previous ‘pincer lock’) is similar. Also
the ears of the agate head, traces of which are on the right side
of the head, must have been rather large, like those of Claudius.
But the style is quite different. As it appears today, the head
in London reminds us of Claudius, but it is not Claudius. The
London face appears ‘dried’ and further ‘smoothed’. ‘Dried’ in
the treatment of the eyelids, the cheeks and the lips, not so
fleshy as in the Kassel head; the nose is quite straight, while
Claudius’ usually slightly widens in the middle. ‘Smoothed’
because, even if in the ‘Kassel type’ Claudius is young looking,
the London face has no furrows at all, and looks like an icy,
sharp-featured mask. The most interesting stylistic detail is the
142 | ‘Gems of Heaven’
Plate 20 Marble bust of a Theodosian prince. Berlin, Staatliche Museen
Plate 22 Craniometric landmarks of: Theodosian prince (a), Claudius, ‘Kassel
type’ (b, e), agate head of Claudius (c), Honorius on the Rothschild Cameo (d, f)
almost geometrical pattern of the three layers of hair locks.
While in the lowest layer, the direction of the locks is identical
to the ‘Kassel type’, in the second row the carving of a second
fork, in the same position as the one over the right eye (not to
be found in Claudius’ portraits), creates a lozenge and seems to
be connected with the change in the direction of the locks on
its left, now towards the left of the head. Also the direction of
the curved locks of the third rows is opposite in comparison
with Claudius’ hair.
It seems very likely then, that the origin of the London head
was really a portrait of Claudius, in his ‘Kassel type’, which has
been reworked in the stylistic climate of the end of the 4th
century ad. On the agate head the levels of the hair on the top
and at the sides of the head have been lowered; the disposition
of the locks of the two upper layers has been varied with a
geometrical taste; the fringe itself has acquired an arched
design; the nose has lost its bump; any fleshy detail has been
deleted. The first comparison for the hair, organised in three
clearly designated rows of curved locks, is to be found in the
late Constantinian period, as exemplified by the colossal
portrait in the Palazzo dei Conservatori depicting one of
Constantine’s sons as a boy.83 This still retains its plasticity. But
the sharp style, the graphic treatment of the hair and the
smoothness of the face appear fully developed in the
Theodosian age, as on another colossal head of a boy, the
so-called ‘Honorius’, in Rome84 and, above all, on the
Theodosian prince in Berlin (Pl. 20).85
I have suggested that the agate head in London must have
been reworked from a portrait of Claudius at the very end of
the 4th century ad, yet it is very difficult to conjecture whom it

could represent, as no trace of a diadem is visible on it. Other
Late Antique portraits of princes, such as the two mentioned
above, whose more than twice life-size dimensions assure us
that they are members of the imperial family, pose the same
problem. One solution could be that the representation derives
from a family group, such as depicted on the marble pedestal of
the obelisk of Theodosius in Istanbul.86
Now, I would like to consider another glyptic portrait of
Claudius (together with one of his wives, either Valeria
Messalina, married in ad 39, or Agrippina the Younger,
married in ad 49), even if it is not in the round. On the
Rothschild Cameo87 this was reworked into the portrait of a
Theodosian prince, Honorius,88 on the occasion of his wedding
(Pls 21a–b).89 The comparison is very instructive as to the
reworking strategies and to the skill of the court engravers.
First of all it must be noted that in none of his portrait types is
Claudius depicted with such a low forehead. So, even if the
fringe resembles the ‘Turin Type’90 (Pl. 18:58.Vc), it cannot have
been realised in the lowest layer of locks. It is rather likely that
the compact, slightly convex fringe has been re-cut into the
upper region of the forehead and the wreath in the lowest layer
of locks. This resulted in all the facial planes being lowered, as
the side view of the head shows very well: the forehead, profile
of the nose, lips and chin are all almost flattened to the same
level.
Unfortunately, owing to the lack of the diadem and the
different treatment of the rows of locks, which is unparalleled,
it is impossible to attribute the London head to a (young)
Honorius. But the idea remains seductive given that lines
drawn through both craniometric landmarks (dashed: nasion,
acanthion and gnathion) and simply empiric marker points
(dotted: border of the lock layers, horizontal axis of the eyes,
gap between the lower lip and the chin) give the agate head,
Claudius’ head, the Berlin head and the Rothschild Cameo
male head a rather impressively uniformity (Pl. 22).
3.3. A second life: from sceptre to sceptre
One of the best known ‘precious portraits’, now in the Cabinet
des médailles, Paris,91 is a chalcedony cuirassed bust which has
already been studied by Marianne Bergmann and Paul Zanker
(Pls 23a–c).92 According to these two scholars, it represents a
4th-century ad emperor, reworked from a portrait of Domitian
in his II or III type,93 or in his II updated to his III type, as in the
case of the chalcedony head of Tiberius (Pl. 2a–d). The shape
of the bust in Paris is not far different from the bust of Trajan/
Aaron (Pl. 1) and it is very interesting to note here that, unlike
the latter, on which the gorgoneion was preserved and re-used
Three Degrees of Separation
Plates 23a–c Chalcedony bust of a 4th-century
AD emperor. Paris, Cabinet des médailles
as one of the gems of the breastplate of the High Priest, on the
former the gorgoneion at the centre of the cuirass breastplate,
worked with a pattern of radiating scales suggesting the aegis,
was erased and replaced by a Latin cross inside a circle (Pl.
24). In this case, our interest does not lie in the identity of the
emperor into whom Domitian was transformed at a certain
moment in Late Antiquity (even if the traditional identification
with Constantine seems out of the question),94 but in the
astonishing continuity of the function of ‘precious sculptures’
through the centuries.
The most likely function for imperial glyptic busts like that
of Trajan/Aaron and the remarkable bust of Augustus on a
sphere in the Louvre,95 is the crowning of a sceptre. As the
emperor was portrayed in the round on the top of the sceptre
itself, the owners of these can be identified as magistrates of
different rank. The best known from iconographic sources are
the consuls portrayed on ivory diptychs.96 That the crowning of
these sceptres could be not only in precious metal, but also in
precious stones has been recently shown by the extraordinary
discovery on the slopes of the Palatine Hill of a group of imperii
insignia, among which was a chalcedony sphere with a drilled
hole into which one could imagine a ‘precious sculpture’ (such
as a bust or an eagle) on its top.97
During the Middle Ages, a very important dignitary in
some French cathedral chapters was the grand-chantre (in
Nôtre-Dame in Paris, the second in rank after the bishop), a
clergyman whose main task was to preside over the chant, but
who also had a number of other privileges. While on duty, the
grand-chantre held a ceremonial mace called a bâton cantoral:
in other words, a sceptre. Almost all of these maces
disappeared during the French Revolution, yet, in some
treasury inventories we can find some interesting descriptions.
In the treasury of the Cathedral of St-Etienne of Auxerre there
was ‘Un baston couvert d’argent ayant dessus un angle [=
aigle] d’argent doré, lequel porte le chantre d’Auxerre ...’
(inventory of 1531). The cathedral of Chartres also owned a
bâton cantoral with an eagle on the top. In the treasury of the
royal cathedral of St Denis the gold ‘sceptre of Dagobert’, used
by the deacon during the annual feasts and for the consecration
of queens, had an Early Medieval eagle on its top and a small
bust, maybe a cameo. The 1343 inventory of the treasury of the
cathedral of Nôtre-Dame in Paris recorded a ‘baculus cantoris
in III peciis argenteis deauratus et bene operatus cum
manubrio esmailliato et pomo de lapide camahu et ymaginibus
de filiis Israel’ (‘the staff of the choirmaster, in three pieces of
silver, gilded and well worked, with an enamelled handle and a
globe in a stone cameo and images of the sons of Israel).
‘Gems of Heaven’ | 143
Gagetti
Plate 24 Etching (1790) of the chalcedony bust of a 4th-century AD emperor,
mounted on the top of the bâton cantoral of the Sainte-Chapelle, Paris
Finally, in the treasury of the Sainte-Chapelle of Vincennes
there was a bâton cantoral with an agate head on the top
believed in 1792 to be the portrait of Charles V.98
The chalcedony bust in the Cabinet des médailles is today
still mounted on the surviving top of the bâton cantoral of the
Sainte-Chapelle in Paris, whose treasury was almost entirely
robbed and destroyed in 1791:99 its original appearance is
attested by a 1790 etching (Pl. 24).100 Mounted on a precious
ebony staff, it seems to have been made for Charles V between
1368 and 1377,101 as it is first recorded in an inventory of 1480:
... unus camahyeu insculptus sive intailliatus in factione unius
grossi hominis, tenens in manu ejus dextrâ unam coronam
spineam argenti esmaillatam ... et in manu sinistrâ una duplex
crux ....102
one cameo, sculpted or carved as to resemble a large man, holding
in the right hand a crown of thorns in enamelled silver...and in his
left hand a double-barred cross.
So, less than one century after the transformation of Trajan
into Aaron, the cuirassed bust of Domitian was re-used and
transformed into St Louis of France, seated on a cloud, with the
symbols of the Passion in his hands: the crown of thorns and
the cross.
If as seems likely, a certain number of Roman sceptres, like
ivory diptychs, entered church treasuries, certainly the most
surprising depiction of a sceptre crowned by three busts, like
those held in their hands by many Late Antique consuls, is not a
real object, but a detail of a fresco painting. The eastern wall in
the refectory of the Abbaye de Lavaudieu (Haute-Loire) is
decorated by a Maiestas Domini dated to the beginning of the
13th century.103 The seated Christ holds in his veiled right hand
a short sceptre, with a conical staff on the top of which,
connected by a globe, is a parallelepiped base supporting three
144 | ‘Gems of Heaven’
elements. Although now partly faded, these seem to be three
busts, an unusual but not unattested number: three busts are,
for example, reproduced on the sceptre on the diptychs of the
Emperor Anastasius (Constantinople, ad 517).104 The three-bust
sceptre in the hand of Christ at Lavaudieu can be interpreted as
the symbol of the power exerted by the Son in the name of the
Trinity, one and trine.105 This is, I suppose, the best
demonstration of the more general iconographic meaning of
the idiom created by Sena Chiesa to denominate the passage of
glyptic specimens from the pagan to the Christian world: ‘from
the imperial to the heavenly court’.106
Conclusion
The case studies examined here point out different strategies
and different epochs in the reworking of imperial glyptic
portraits in the round. With regard to the time of the
reworking, it can be done either in the same cultural climate in
which the portrait was first realised, that is not after Late
Antiquity, or in a wholly different epoch such as the Medieval
period.
In the first case, the strategies of reworking are three.
(1) Updating a portrait type of the same individual, possibly
during his lifetime. This is the case of the Paris head of Tiberius
(second degree of separation).
(2) Transformation of the portrayed person into another
one, very similar in his/her facial features which then remain
untouched, changing other characterising details, such as the
hairstyle. This strategy seems to have been applied at a
distance of at least some generations. Within this category
comes the Venice head of Sabina transformed into Helena
(third degree of separation).
(3) Transformation of the original portrait into another,
due to the similarity of macro-characters only, like the
triangular shape of Claudius’ face, fitting both the London
‘prince’ and, in the field of cameos, the Rothschild Honorius
(again, the third degree of separation). The source of ‘precious
sculptures’ would have been the imperial treasury,107 at least if
one wants to read à la lettre Claudian’s verses:
Iam munera nuptae / praeparat et pulchros, Mariae sed luce
minores, / eligit ornatus, quidquid venerabilis olim / Livia
divorumque nurus gessere superbae.108
wore the venerable Livia and the haughty daughters-in-law of the
divine emperors.
In the second case, at a chronological distance of many
centuries, i.e. not only in the Early Middle Ages, but also during
the 13th and 14th centuries, the portrayed individuals are no
longer recognised and their valuable portraits were re-used in
the function of a new (Christian) identity. Hence, Trajan
becomes Aaron and Domitian/Late Antique emperor becomes
St Louis (its third-degree transformation had already taken
place in Late Antiquity). In both cases, the reworking at this
stage affects only minor details, such as the insertion of the
twelve gems into the breastplate and the obliteration of the
gorgoneion with a cross. This stage then, is the first degree of
separation. As the case studies show the three degrees of
separation are not determined by the time passed in between,
but depend on the will to reach a particular purpose, i.e.,
ultimately, on the full understanding, or not, of the starting
portrait.

Notes
1 For a survey on the class, see E. Gagetti, Preziose sculture di età
ellenistica e romana (Il Filarete. Collana di studi e testi. Università
degli Studi di Milano. Pubblicazioni della Facoltà di Lettere e
Filosofia, 240), Milan, 2006.
2 Gagetti (n. 1), nos A23 (grey chalcedony), A27 (grey chalcedony),
A42 (white chalcedony), A44 (agate) and B16 (aquamarine).
3 H. Jucker, ‘Trajanstudien zu einem Chalzedonbüstchen im
Antikenmuseum’, Jahrbuch der Berliner Museen 26 (1984), 18–78,
at 51.
4 Gagetti (n. 1), App. 3.
5 These values have been calculated in K. Dahmen, Untersuchungen
zu Form und Funktion kleinformatiger Porträts der römischen
Kaiserzeit, Münster, 2001, 4–5. Also according to Bertha Schneider
(Studien zu den kleinformatigen Kaiserporträts von den Anfängen
der Kaiserzeit bis ins dritte Jahrhundert, Inaugural-Dissertation
zur Erlangung des Doktorgrades der Philosophie an der Ludwig-
Maximilians-Universität zu München, Munich, 1976, 1) a
miniature portrait must be smaller than half-life size. Elizabeth
Bartman (Ancient Sculptural Copies in Miniature [Columbia
studies in the classical tradition, 19], Leiden-New York-Cologne,
1992, 9), instead, states that ‘no single ratio for proportional
reduction prevailed in the making of a miniature copy in antiquity.
[...] the miniature copy could reduce its model by as much as one
half [...] to as little as one hundredth’ (yet, it must be said that
Bartman’s study is concerned with miniature copies of Idealplastik
specimens, and that she sets life size measure = 1.70m).
6 Zaragoza, Museo Provincial, Seccion de Arqueología, inv. no. 80-5-
1. Gagetti (n. 1), no. A22 (with literature, cui adde F. Paolucci,
Piccole sculture preziose dell’Impero romano, Modena, 2006, 79–80,
no. 5). The first identification of the starting portrait, then
reworked, is in M. Beltrán Lloris, ‘El retrato de «Divus Augustus»
del municipium Turiaso (Tarazona, Zaragoza). Un palimpsesto de
época trajanea’, Madrider Mitteilungen 25 (1984), 103–34, pls 21–6.
7 Even if we can suppose that a ‘precious sculpture’ could hardly
exceed such dimensions, the definition of the longest distance
gnathion-vertex is due to two different requirements. First, it is
needed as far as alabaster – available also in large sizes – is
concerned. The status of alabaster in antiquity, in fact, was two-
fold: both lapis and gemma (cf. Pliny Nat. Hist. XXXVI.59–61 and
37.73). Secondly, the chance of discovering a ‘precious sculpture’
whose distance gnathion-vertex is longer than 12cm is remote but
not impossible. Pliny (Nat. Hist. XXXVII.107–8) talks about a
statue, five cubits high (= 1.76m) portraying Arsinoë III realised in
a stone called topazos (also known to the ancients as chrysolithos),
perhaps to be identified as olivine or peridot, from an island in the
Red Sea. Today the island can be identified as Zabargad (St-John’s
Island), where, close to cape Ras Benas, a deposit of olivine with
traces of exploitation in antiquity is still extant. The olivine from
St-John’s Island can be found in blocks of such dimensions that it
lends credence to Pliny’s report (Gagetti [n. 1], 91–2, with
literature).
8 The table must be completed with three post-classical specimens
(Gagetti [n. 1], nos O1-O3 and 51 dubitanda (eadem, P1–P51).
9 G. Platz-Horster, ‘Ein Trajan-Porträt als «Palinglypt». Zu einem neu
erworbenen Kameo-Fragment im Antikenmuseum’, Jahrbuch der
Berliner Museen 26 (1984), 5–15.
10 Berlin, Antikensammlung, inv. no. 1979.5. H. 8.8cm. Gagetti (n. 1),
no. A27 (with literature, cui adde Paolucci [n. 6], 86–7, no. 16).
11 For the type, see Jucker (n. 3), 38, 77, sketch II (c. ad 105): the hair
over the forehead is parted up into locks which are all directed
towards the right; a fork-like lock is over the outer corner of the left
eye.
12 Even without taking into consideration the possibility that the
chalcedony nucleus did not allow the gem cutter to realise a ‘real’
Schulterbausch, examples of a rather flat disposition of the folds of
the paludamentum (referring perhaps to a different arrangement
of the mantle on the right shoulder) also do exist in official marble
sculpture (e.g., the fragmentary cuirassed trunk from Segusium,
dated to the period of Claudius or even of Caligula: M. Cadario, La
corazza di Alessandro. Loricati di tipo ellenistico dal IV secolo a.C. al
II d.C. [Il Filarete. Collana di studi e testi. Università degli Studi di
Milano. Pubblicazioni della Facoltà di Lettere e Filosofia, 218],
Milan, 2004, 190–1 and pl. XXVI 2.
13 Such changes are supposed in Jucker (n. 3), 66.
14 E. Zwierlein-Diehl, ‘Kameen und Chalcedonplastik mit
Three Degrees of Separation
mittelalterlichem Edelsteinschmuck’, in G. Platz-Horster (ed.),
Mythos und Macht. Erhabene Bilder in Edelstein, Berlin, 2008,
38–53, col. pl. 4.
15 On the costume of the High Priest, the stones on his breastplate
and their symbolic meanings, see the useful syntheses by
W. Zwickel, ‘Der Hochpriester in der nachexilischen Zeit, seine
Stellung und seine Kleidung’, in W. Zwickel (ed), Edelsteine in der
Bibel, Mainz, 2002, 24–40; idem, ‘Die Edelsteine im Brustschild des
Hohenpriesters und beim himmlischen Jerusalem’, in ibid., 50–70.
16 A synoptic table of the stone disposition according to our four main
sources (Exodus, in the Septuagint and Vulgate versions, 28.15–21;
and Flavius Josephus: Bellum Judaicum, 5.231–5 and Antiquitates
Judaicae, 3.159–78) is in Zwierlein-Diehl (n. 14), 43. 1st row:
sardius, topazius, smaragdus; 2nd row: carbunculus, sapphyrus,
iaspis; 3rd row: lygurius, achates, amethistus; 4th row:
chrysolithus, onychinus, berillus.
17 In the frame of the new reading of Trajan as Aaron, the shoulder
straps of the cuirass breastplate could have been intended as the
two stones (emeralds in the Septuagint, onyxes in the Vulgate and
sardonyxes according to Flavius Josephus) inlaid in gold on the
shoulder elements of the breastplate, each of them engraved with
the names of six of the twelve sons of Jacob (Exodus 28.9–12):
Zwierlein-Diehl (n. 14), 44.
18 Simply, as it is of course realised in the same chalcedony of the
whole bust, for its colour it could only represent stone no. 11
(onychinus, ‘of the colour of nails’), and this forced a shifting of
stones nos 11 (onychinus) and 5 (sapphyrus, that is lapis lazuli):
Zwierlein-Diehl (n. 14), 52.
19 Ibid., 52.
20 Paris, Bibliothèque Nationale de France, Cabinet des médailles,
inv. no. 72 B 61225. H. 3.1cm. Gagetti (n. 1), no. A23 (with literature,
cui adde Paolucci [n. 6], 82–3, no. 10).
21 No other monograph on Tiberius’ portrait follows L. Polacco,
Il volto di Tiberio. Saggio di critica iconografica, Rome, 1955, now
outdated. For a brief overview of systematic studies of Tiberius’
portraiture see, J. Pollini, ‘A new marble head of Tiberius’, Antike
Kunst 48 (2005), 55–72, pls 7–13, at 57, n. 9.
22 D. Boschung, ‘Die Bildnistypen der iulisch-claudischen
Kaiserfamilie: ein kritischer Forschungsbericht’, Journal of Roman
Archaeology 6 (1993), 39–79, at 58, Le, Skizze 34.Le.
23 Pollini (n. 21), 60, 66 and fig. 2.VI.
24 It already appears on the ‘Gemma Augustea’ (see the detail in
E. Zwierlein-Diehl, Magie der Steine. Die antiken Prunkkameen im
Kunsthistorischen Museum, with contributions by A. Bernhard-
Walcher and P. Rainer, Vienna, 2008, 104, pl. 48). Contra: Pollini
(n. 21), 62.
25 Boschung (n. 22), 58 (with reference to the Claudian groups from
Leptis Magna and Cerveteri: n. 98; see more widely: D. Boschung,
Gens Augusta. Untersuchungen zu Aufstellung, Wirkung, und
Bedeutung der Statuengruppen des julisch-claudischen
Kaiserhauses (Monumenta artis Romanae, 32), Mainz am Rhein,
2002, 18–23 and 85–9). The extensive series of copies would assure
this portrait the status of ‘type’. Contra: K. Fittschen and P. Zanker,
Katalog der römischen Porträts in den Capitolinischen Museen und
den anderen kommunalen Sammlungen der Stadt Rom, I, Kaiser-
und Prinzenbildnisse (Beiträge zur Erschliessung hellenistischer
und kaiserzeitlicher Skulptur und Architektur, 3), 2 vols, Mainz
am Rhein (2nd rev. edn), 1994, I 13, comment to no. 12
(‘Überlieferungs-variante’ of the ‘Berlin-Naples-Sorrento type’; for
this type see also Boschung (n. 22), 57–8, Ld, Skizze 33.Ld).
26 Pollini (n. 21), 66 and notes. The occasion for the creation of the
new type could have been the fall of Seianus or Tiberius’
vicennalia.
27 Mounted on an unrelated alabaster bust: E. Angelicoussis, The
Woburn Abbey Collection of Classical Antiquities (Corpus signorum
imperii Romani, Great Britain, III 3; Monumenta artis romanae,
20), Mainz am Rhein, 1992, 55–6, no. 23, pls 104, 106–7, 113–16;
Fittschen and Zanker (n. 25), I 14, n. 10 to no. 12; app. 12.
28 W.R. Megow, Kameen von Augustus bis Alexander Severus (Antike
Münzen und geschnittene Steine, 11), Berlin, 1987, 181; Dahmen (n.
5), 168.
29 Boschung (n. 22), 56–7, La, ‘Typus Basel’, Skizze 30.La.
30 Pollini (n. 21), 64: ‘First Military type’, created after Tiberius’
military success in the East and, possibly, also on the occasion of
his marriage to Vipsania.
31 La Spezia, Civico Museo Archeologico ‘Ubaldo Formentini’, inv.
‘Gems of Heaven’ | 145
Gagetti
no. 54. On this bust, see also G. Sansica, ‘Ritratto di Tiberio’, in G.
Sena Chiesa (ed.), Augusto in Cisalpina. Ritratti augustei e giulio-
claudi in Italia Settentrionale (Quaderni di Acme, 22), Milan, 1995,
187–93, with another dating option, which can be traced back to
Polacco (n. 21) and is still in Fittschen and Zanker (n. 25), 10–12, o.
10: the type would have been created on the occasion of his
adoption by Augustus in ad 4. In this case, even if at that time
Tiberius was 46 years old, his youthful look would have been be
due to Augustus’ intention to create for his heir a portrait not too
far from those of his dead grandsons C. Caesar and L. Caesar (ibid.,
12).
32 Inv. no. G 150. Gagetti (n. 1), 246–7, no. B16 (with literature).
Already in the Grimani collection, perhaps bought in the Levant.
33 E.g. C. Anti, Il Regio Museo Archeologico nel Palazzo Reale di
Venezia, Rome, 1930, 145; G. Traversari, ‘Nuovo ritratto di
Cleopatra VII Philopator e rivisitazione critica dell’iconografia
dell’ultima regina d’Egitto’, Rivista di Archeologia 21 (1998), 45–8,
pls 18–20; B. Nardelli, I cammei del Museo Archeologico Nazionale
di Venezia (Collezioni e Musei Archeologici del Veneto, 43), Rome,
1999, 29–30, no. 2.
34 For example, G. Valentinelli, Museo Archeologico della R. Biblioteca
Marciana di Venezia, Venice, 1872, 63, no. 131; S. Bosticco et al.,
‘Glittica’, in Enciclopedia Universale dell’arte VI (1958), 267–366, pls
177–96, at 285 (M.-L. Vollenweider); G. Platz­-Horster, review of
B. Nardelli, I cammei del Museo Archeologico Nazionale di Venezia
in Gnomon 74 (2002), 57–62, at 58. Not in agreement was
A. Furtwängler, Die antiken Gemmen. Geschichte der
Steinschneidekunst im klassischen Altertum, Leipzig-Berlin, 1900,
III, 334 (‘eine fälschlich auf Sabina gedeuteter Kopf von Beryll in
Venedig’).
35 Aquamarine: see Gagetti (n. 1), 85–7. Portrait of Julia Titi (Paris,
Bibliothèque Nationale de France, Cabinet des médailles): M.-L.
Vollenweider and M. Avisseau-Broustet, Camées et intailles. II. Les
portraits romains du Cabinet des Médailles. Catalogue raisonné, 2
vols, Paris, 2003, I, 128–9, no. 145, col. pl. 145 and pls 88–9 (with
literature). In the 9th century the intaglio was re-used on the top of
the reliquary known as ‘Escrain de Charlemagne’, formerly in the
treasury of the royal abbey of St-Denis in Paris and destroyed
during the French Revolution, except for the top element with the
aquamarine; transferred to the Cabinet des Antiquités in 1791: Le
trésor de Saint-Denis (exh. cat., Paris), Paris, 1991, 92–9, no. 13
(with literature). I thank Noël Adams very much for having
reminded me of this gem.
36 The portrait of Cleopatra VII is a much debated question, not to be
entered into here. For a survey see: S. Walker and P. Higgs (eds),
Cleopatra regina d’Egitto (exh. cat., Rome), Milan, 2000; S. Walker
and S.A. Ashton (eds), Cleopatra Reassessed (British Museum
Occasional Paper no. 103), London, 2003; B. Andreae et al. (eds),
Kleopatra und die Caesaren (exh. cat., Hamburg), Munich, 2006.
37 Inv. no. 14547. Gagetti (n. 1), 244–6, no. B15 (with literature, cui
adde Paolucci [n. 6], 89–90, no. 22).
38 A. Carandini, Vibia Sabina. Funzione politica, iconografia e il
problema del classicismo adrianeo (Accademia Toscana di Scienze
e Lettere ‘La Colombaria’. Studi, 13), Florence, 1969, 228–31, figs
liv–lviii. Marble portraits, idem., 80–7, nos 42–52bis, pls 206–28,
among which the most similar to the aquamarine head is no. 44
(then New York, Klejman Collection).
39 Ibid., 133 and 175–93.
40 See Gagetti (n. 1), 183, fig. 4.
41 Tivoli, Hadrian’s Villa: Greek marble; H. 28cm. The head, formerly
in a private collection, was sequestrated by the Guardia di Finanza
– Tutela Patrimonio Archeologico.
42 It could be the so-called ‘Villa dei Vibii Vari’, once rising at Colli di
Santo Stefano: see, Z. Mari, Tibur, pars quarta (Forma Italiae, 35),
Firenze, 1999, 237–44; idem, ‘Vibia Sabina e Villa Adriana’, in
B. Adembri and R.M. Nicolai (eds), Vibia Sabina. Da Augusta a Diva
(exh. cat., Tivoli), Milan, 2007, 51–65 (with further literature).
43 B. Adembri, ‘Ritratto femminile’, in Mirabilia recepta. Le Forze
dell’Ordine a difesa dei Beni Culturali (exh. cat., Rome), Rome,
1999, 73–4, no. 27; eadem., ‘Ritratto femminile’, in Adriano.
Architettura e progetto (exh. cat., Tivoli), Rome, 2000, 248, no. 27.
44 A.M. Reggiani, ‘Un ritratto di Matidia Minore da Tivoli e la galleria
celebrativa della Domus Augusta’, in G. Ghini (ed.), Lazio & Sabina
2. Secondo Incontro di Studi sul Lazio e la Sabina (confer. proc.,
Rome, 2003), Rome, 2004, 11–22, at 11; eadem, ‘Ritratto di Matidia
Minore’, in M. Sapelli Ragni (ed.), Villa Adriana. Una storia mai
146 | ‘Gems of Heaven’
finita. Novità e prospettive della ricerca (exh. cat., Tivoli), Milan,
2010, 189, no. 15.
45 For the results of the excavations (1994–9) see: S. Cascella, Il teatro
romano di Sessa Aurunca, Marina di Minturno, 2002.
46 For a description and proposals of integration: F. Chausson, ‘Une
dédicace monumentale provenant du théâtre de Suessa Aurunca,
due à Matidie la Jeune, belle-soeur de l’emperor Hadrien’, Journal
des Savants (2008), 233–59 (fig. 2 for the plan of the theatre and the
location of the inscription; figs 7 and 8 for the ultimate restitution
of the inscription).
47 For all these works and others involving the town of Suessa
Aurunca see: Chausson ibid., 237–44.
48 Its original location has been deduced from the position of its
fragments after the collapse of the scenae frons: see Cascella (n.
45), 75–6.
49 Sessa Aurunca, Castello Ducale, inv. no. 297084; H. 2.60m. See:
Cascella (n. 45), 71–2 (no. 6), pls 36–7; M.G. Ruggi d’Aragona and
P. Pensabene, ‘Statua di Matidia/Aura’, in M. De Nuccio and L.
Ungaro (eds), I marmi colorati della Roma imperiale (exh. cat.,
Rome), Venice, 2002, 325–6, no. 23; Reggiani 2004 (n. 44), 12–14,
pl. 2; C. Valeri and F. Zevi, ‘La statua di Matidia Minore e il teatro di
Sessa Aurunca’, in A.M. Reggiani (ed.), Adriano. Le Memorie al
femminile (exh. cat., Tivoli), 2004a, Milan, 128–33; T. Opper,
Hadrian. Empire and Conflict (exh. cat., London), London, 2008,
203, pl. 189; D. Attanasio et al., ‘Villa Adriana e l’uso dei marmi
afrodisiensi dalle cave di Göktepe’, in Sapelli Ragni (n. 44), 81–90,
at 85, pl. 9.
50 Sessa Aurunca, Castello Ducale, inv. no. 297044: Cascella (n. 45),
70–1, no. 3 (Faustina the Elder); C. Valeri, ‘Ritratto di Matidia
Minore’, in Reggiani 2004a (n. 49), 138–9 (Matidia the Younger).
51 Concerning the life of Matidia the Younger (Prosopographia
imperii Romani saec. I. II. III., Berlin–Leipzig, editio altera, 1933–
99, V, 228, no. 368; M.-T. Raepsaet-Charlier, Prosopographie des
femmes de l’ordre sénatorial (1er-2e siècles) (Académie royale de
Belgique. Classe des lettres. Fonds René Draguet, 4), Louvain,
1987, I, 446–7, no. 533, and II, stemma IX; M.T. Boatwright, ‘Matidia
the Younger’, Echos du Monde Classique 36 (1992), 19–32) many
questions remain open. For a survey, with the vast literature and
the collection of the literary and epigraphic sources, see Chausson
(n. 46), 233–40.
52 Formerly, the portrait of Matidia the Younger was seen in the
‘Fonseca head’ and in a replica of it in Fiesole (U. Hausmann,
‘Bildnisse zweier junger Römerinnen in Fiesole’, Jahrbuch des
Deutschen archäologischen Instituts 74 (1959), 164–202, at 200–02)
and, just hypothetically, in a head from a Roman villa in Milreu
(Portugal), of which a series of at least nine replicas exists
(K. Fittschen, ‘Bildnis einer Frau Trajanischer Zeit aus Milreu’,
Madrider Mitteilungen 34 (1993), 202–9, pls 16–25).
53 Chausson (n. 46), 243.
54 See Cascella (n. 45), 67–76; and C. Valeri, ‘Ritratto di imperatrice a
capo velato (Matidia Diva?)’, and ‘Busto di Sabina’, in Reggiani
2004a (n. 49), 134–5 and 136–7, for the portraits of an empress
(Matidia the Elder Diva?) and Sabina.
55 All these titles – fil(ia), [neptis], (sor)or and [matertera] – can be
inferred from the inscription: see Chausson (n. 46), 252.
56 Reggiani 2004a (n. 49); Opper (n. 49), 203, pl. 188.
57 Tivoli, Hadrian’s Villa. H. 28cm. A.M. Reggiani, ‘Ritratto di Vibia
Sabina’, in Adembri and Nicolai (n. 42), 168–71 (with literature).
58 M. Rubini, ‘Fisionomie imperiali. Indagine antropologica sulle
donne della famiglia di Adriano’, in Reggiani 2004a (n. 49), 79–83;
see also Reggiani 2004 (n. 44), 20, pls 10–13 (with an updating of
the attribution of the Tivoli head to Matidia the Younger).
59 Reggiani 2004 (n. 44), 14.
60 Musée du Louvre, inv. no. MNE 784. H. 27.5cm. Provenance: from
the antiquarian market (1980): F. Baratte, ‘Un portrait féminin des
collections du Louvre’, Revue Archéologique 2 (1984), 301–12.
61 The Metropolitan Museum of Art, inv. no. 21.88.55, Rogers Fund.
H. 37.5cm (it preserved the whole neck): G.M.A. Richter, The
Metropolitan Museum of Art. Roman Portraits, New York, 1948, no.
84; Baratte (n. 60), 303, n. 6 (with literature), pls 5–7.
62 National Museum, mounted on an unrelated bust: A. Sadurska, Les
portraits romains dans les collections polonaises (Corpus signorum
imperii Romani, Pologne, I), Varsovie, 1972, 39; Baratte (n. 60),
303, n. 7 (with literature), pls 8–11.
63 Baratte (n. 60), 303.
64 Ibid., 309. I only saw the Louvre head.

65 Ibid., 309–10.
66 A. Carandini, ‘Introduzione’, in Adembri and Nicolai (n. 42), 19–21,
at 19.
67 P. Pellegrin, Camei di diverse sorti accomodati attorno il predetto
studiolo numero Settantadoi come qui sotto et prima, 1593, 61
(quoted in Nardelli [n. 33], 30).
68 Also the seated statue of Helena in the Museo Capitolino (see
following note) has in the lower plait of the ‘crown’ nine holes, to
be connected to a similar adornment (or to a jewelled diadem):
F.P. Arata, ‘La statua seduta dell’imperatrice Elena nel Museo
Capitolino’, Römische Mitteilungen 100 (1993), 185–200, at 193–4.
69 Stanza degli Imperatori, inv. no. 496: R. Calza, ‘Cronologia ed
identificazione dell’“Agrippina” Capitolina’, Atti della Pontificia
Accademia Romana di Archeologia, serie III, Memorie IX/2 (1955),
107–36; K. Fittschen and P. Zanker, Katalog der römischen Porträts
in den Capitolinischen Museen und den anderen kommunalen
Sammlungen der Stadt Rom, III, Kaiserinnen- und
Prinzessinnenbildnisse. Frauenporträts (Beiträge zur
Erschliessung hellenistischer und kaiserzeitlicher Skulptur und
Architektur, 5), 2 vols, Mainz am Rhein, 1983, 35–6, no. 38, pls
47–8; H.P. L’Orange et al., Das spätantike Herrscherbild von
Diokletian bis des Konstantin-Söhnen 284–361 n. Chr. Die Bildnisse
der Frauen und des Julian (Das römische Herrscherbild, III, 4),
Berlin, 1984, 146, pls 74b–c (M. Wegner); Arata (n. 68); E.R.
Varner, Mutilation and Transformation. Damnatio memoriae and
Roman Imperial Portraiture (Monumenta graeca et romana, 10),
Leiden – Boston, 2004, 273, no. 6.12, pl. 150a; C. Parisi-Presicce,
‘Statua di Elena’, in A. Donati and G. Gentili (eds), Costantino il
Grande. La civiltà antica al bivio tra Occidente e Oriente (exh. cat.,
Rome), Cinisello Balsamo, 2005, 212–13 (all with further
literature).
70 Arata (n. 68), 195–7; Varner (n. 69), 97, 150–1, 154, 273, no. 6.12, pl.
150a. On the reworking of this type of Helena’s hair-dress, see also:
M. Bergmann, ‘Fälschung, Umarbeitung oder einiger Stil.
Beobachtungen zu einer “Konstantinischen” Porträtbuste’, Städel
Jahrbuch 10 (1985), 45–54.
71 The same reworking of the coiffure shows another replica from the
same original, depicting Helena too, now in Florence: Galleria
degli Uffizi, inv. no. 1914.171 (Calza [n. 68], 115–16, no. X, pl. 11; G.A.
Mansuelli, Galleria degli Uffizi. Le sculture. Parte II, Rome, 1961,
131, no. 171, pls 168a–c; details in Arata [n. 68], pl. 49:1–4).
72 H. von Heintze, ‘Ein spätantikes Mädchenporträt in Bonn. Zur
stilistischen Entwicklung des Frauenbildnisses im 4. und 5.
Jahrhundert’, Jahrbuch für Antike und Christentum 14 (1971),
61–91, pls 1–19, at 83, no. VI.3, pls 16b and d (formerly Marshall
Collection).
73 Anyway, it could have been perfectly ‘contained’ in a supposed
Sabina’s original face: Gagetti (n. 1), fig. 4.
74 That the Marshall head in its original state could have been a
portrait of Sabina, was already stated by Helga von Heintze 40
years ago (von Heintze [n. 72], 84; the reworking is there dated to
between the 4th to 5th century ad: 83).
75 The ears are simply rough-hewn on the head of the Capitoline
Helena: Arata (n. 68), 189 and pls 45:3–4.
76 There is no trace of it on the reverse, where it would have been
useless.
77 The depressions on the skullcap seem to be concoidal fractures,
typical of beryls, caused by the removal of the unwanted elements
of the original coiffure.
78 Inv. no. GR 1872,0604.1308 (= H.B. Walters, Catalogue of the
Engraved Gems and Cameos, Greek, Etruscan and Roman in the
British Museum, London, 1926, 369, no. 3947, fig. 95, formerly (ante
1872) the Castellani collection. H. 9.5cm; Gagetti (n. 1), 223–4, no.
A44, pl. XIX (with literature, cui adde Paolucci [n. 6], 102–3, no. 41,
pl. 41).
79 See Boschung (n. 22), 70–1, V.
80 Kassel, Staatliche Kunstsammlung, inv. no. SK 116: Boschung (n.
22), 70, Va, Skizze 56.Va.
81 Lists of replicas are in K. Fittschen, Katalog der Antiken Skulpturen
in Schloß Erbach (Archäologische Forschungen, III), Berlin, 1977,
50–1, n. 22; and in A.-K. Massner, Bildnisangleichung.
Untersuchungen zur Entstehungs- und Wirkungsgeschichte der
Augustusporträts (43 v. Chr.– 68 n. Chr.) (Das römische
Herrscherbild, IV), Berlin, 1982, 136, n. 754.
82 D. Salzmann, ‘Beobachtungen zu Münzprägung und Ikonographie
des Claudius’, Archäologischer Anzeiger (1976), 252–64, at 262–3
Three Degrees of Separation
and pls 3–4. The adoption of a youthful portrait on coins from the
beginning of his reign by the 50-year-old Claudius (born 10 bc)
follows Tiberius’ portrait strategy (ibid., 263).
83 Cortile, inv. no. 2882: Fittschen and Zanker (n. 25), 1, 156–7, no.
125, pl. 156 (with literature and discussion of chronology and
identification).
84 Museo Capitolino, Galleria, inv. no. 494: Fittschen and Zanker (n.
25), 159–61, no. 127, pl. 158 (with literature and a discussion of the
chronology and identification with Honorius ante ad 393, as he
wears no diadem; this was reworked from a more ancient portrait).
85 Berlin, Staatliche Museen, inv. no. R 122: Fittschen and Zanker (n.
25), 161, no. 3 to no. 127 (with literature); J. Meischner, ‘Studien zur
spätantike Kaiserikonographie’, Jahrbuch des Deutschen
Archäologischen Instituts (1995), 431–46, at 441, 444–6 (Honorius,
on the occasion of his first wedding in ad 398), pls 11 and 15 (with
further literature).
86 Fittschen and Zanker (n. 25), 160–1 and n. 9.
87 The cameo still belongs to the Rothschild family. Diam. 16cm. Its
provenance is unknown, though when it was purchased in Paris in
1889 it was supplied with the following data: ‘un camée antique en
pierre dure entouré d’un cadre byzantin, travail hispano-arabe en
argent doré’ (E. Coche de la Ferté, Le camée Rothschild. Un chef
d’oeuvre du IVe siècle après J.-C., Paris, 1957, 57, n. 1); this is the
reason why the cameo is said to have come from Spain. The
literature on this glyptic masterpiece, first mentioned in
E. Babelon, La gravure en pierres fines, Paris, 1894, and first
published by S. Reinach (Gazette des beaux-arts 1 [1926], 185 ff.), is
extensive. See: Coche de la Ferté ibid., 57–8, n. 1; J. Meischner, ‘Der
Hochzeitkameo des Honorius’, Archäologischer Anzeiger (1993),
612–19, at 613, n. 1; S. Sande, ‘The iconography and style of the
Rothschild Cameo’, in J. Fleischer, J. Lund and M. Nelsen (eds),
Late Antiquity. Art in Context [Acta hyperborea. Danish Studies in
Classical Archaeology 8], Copenhagen, 2001, 145–58, at 153, n. 1;
E. Zwierlein-Diehl, Antike Gemmen und ihr Nachleben, Berlin –
New York, 2007, 455, no. 756.
88 The portrait of Honorius is still problematic. His only other certain
representation is his depiction on both leaves of the ivory diptych
of the consul Probus (ad 406), now in the Cathedral of Aosta (R.
Delbrueck, Die Consulardiptychen und verwandte Denkmäler
[Studien zur spätantike Kunstgeschichte, 2], Berlin–Leipzig, 1929,
no. 1), where Honorius, cuirassed and wearing a diadem, appears
with an oblong, triangular face, a compact fringe over the low
forehead, and side-whiskers. All these characteristics can be found
on a marble head in Trier, also considered to be a portrait of
Honorius in his adult years (A. Giuliano, ‘Ritratti di Onorio’, in G.
Sena Chiesa and E.A. Arslan (eds), Felix temporis reparatio, Milan,
1992, 73–86, at 76; Meischner (n. 85), 441, 444–6, pls 12 and 16). On
Honorius’ portrait in general: Giuliano ibid.
89 Different identifications have been proposed: Constantius II with
his cousin, whose name is unknown; Theodosius I with Aelia
Flaccilla; Honorius with his first wife Maria (ad 398) or his second
wife Thermantia (ad 404), Justinian with Theodora.
90 See Boschung (n. 22), 71, Vc, Skizze 58.Vb. The chronology is
debated: see also, on numismatic grounds, Salzmann (n. 82), 263.
91 Paris, Bibliothèque Nationale de France, Cabinet des médailles et
des antiques. H. 9.5cm. Gagetti (n. 1), 218–21, no. A42 (with
literature, cui adde Paolucci (n. 6), 93–4, no. 28).
92 M. Bergmann and P. Zanker, ‘“Damnatio memoriae”.
Umgearbeitete Nero- und Domitians-porträts. Zur Ikonographie
der flavischen Kaiser und Nerva’, Jahrbuch des Deutschen
Archäologischen Instituts 96 (1981), 317–412, at 409–10, no. 48, pls
65a–b.
93 To Type III (see Bergmann and Zanker [n. 92], 366–8. It was in use
from the beginning of Domitian’s reign until his death [ad 81–96])
belong the two rows of curls on the nape, oriented from the centre
of the head towards the ears; typical of Type II (see Bergmann and
Zanker ibid., 360–6. Its chronological range is much narrower and
the type was created before Domitian became emperor [ad
75–80]) are instead the pattern of the vortex at the top of the head
and the row of falcated locks below.
94 None of the typical facial features of Constantine, which also
always occur on those of his portraits which have been reworked
from other ‘ancient’ emperors, appears here. Instead, we have a
straight line forehead-nose; the nose is not as aquiline as
Constantine’s; the upper lip is too straight; the lower one is too
fleshy; the chin does not project enough. Finally, the curly locks
‘Gems of Heaven’ | 147
Gagetti
over the forehead, anyway visible, even in the case of a large
diadem obscuring most of the head, are never to be found in
Constantine’s portrait, also in those surely reworked starting from
imperial faces framed by flowing locks, as in the case of the heads
on Hadrian’s tondi on the Arch of Constantine (reworked from
Hadrian’s ‘Chiaramonti 392’ portrait type: C. Evers, Les portraits
d’Hadrien. Typologie et ateliers [Académie royale de Belgique.
Mémoire de la Classe des Beaux-Arts, VII], Brussels, 1994, 63–71);
on the type, ibid., 225–32.
95 Paris, Musée du Louvre, Département des Antiquités Classiques:
mounted on the ‘monument’ by Valadier, inv. no. MR6. (the latter
carved in the same block of chalcedony, but into a blue vein, while
the portrait is fleshy pink). The chalcedony sculpture was found in
the Catacombs of Priscilla in Rome. H. 14cm: Gagetti (n. 1), 191–3,
no. A20, with literature, cui adde Paolucci (n. 6), 78, no. 3.
96 For the owners of sceptres crowned with imperial busts see:
Gagetti (n. 1), 454–63 and tables 5–6.
97 C. Panella et al., ‘Le insegne imperiali dal Palatino’, Scienze
dell’antichità. Storia, archeologia, antropologia 13 (2006), 701–45.
The objects possibly belonged to Maxentius and were hidden after
his defeat by Constantine at the Milvian Bridge in ad 312. They are
now in Rome in the Museo Nazionale Romano, Palazzo Massimo
alle Terme.
98 Auxerre: M. Quantin, Inventaire du trésor de la cathédrale d’Auxerre
en 1531, Auxerre, 1887, 9, no. 32 (the same cathedral in ad 566 was
presented by the royal couple with a chalice ‘ex lapide onychino’
depicting the story of Aeneas, with ‘captions’ in Greek letters: E.
Babelon, Catalogue des camées antiques et modernes de la
Bibliothèque Nationale, 2 vols, Paris, 1897, LXX); Chartres: Babelon
ibid., 169; St-Denis: Le trésor de Saint-Denis (exh. cat., Paris), Paris,
1977, 69, no. 2 (D. Gaborit-Chopin); Nôtre-Dame in Paris: G.
Faigniez, ‘Inventaires du trésor de Notre-Dame de Paris (suite) (1)’,
Revue Archéologique 7/4 (1874), 249–59, at 254, no. 43; Vincennes:
Le trésor de la Sainte-Chapelle (exh. cat., Paris), Paris, 2001, 208 (D.
Gaborit-Chopin);
148 | ‘Gems of Heaven’
99 Babelon (n. 98), 163.
100 S.J. Morand, Histoire de la Sainte-Chapelle royale du Palais,
enrichie de planches, par M. Sauveur-Jérôme Morand, chanoine de
ladite église …, Paris, 1790, 56.
101 In the 1368–77 inventory the ebony is recorded in the atelier of
‘Hennequin l’orfevre de par le Roy’ (A. Vidier, Le trésor de la Sainte-
Chapelle. Inventaires et documents, Paris, 1911, inv. I, no. 92), to be
identified with the famous goldsmith Hennequin du Vivier. On the
bâton cantoral of the Sainte-Chapelle as a medieval objet d’art see:
G. Gaborit-Chopin, ‘Le bâton cantoral de la Sainte-Chapelle’,
Bulletin Monumental 132 (1974), 67–81 (with extensive
bibliography).
102 Paris, Bibliothèque Nationale, Département des Manuscripts, ms.
lat. 9941, ff. 29–30 (quoted in Babelon [n0. 98]).
103 Y. Christe, ‘Le sceptre du Christ de Lavaudieu’, Cahiers
Archéologiques 26 (1977), 163–7, at 164.
104 Delbrueck (no. 88), nos 19 (Verona, Biblioteca Capitolare) and 20
(divided between Berlin, Antiquarium, and London, Victoria &
Albert Museum), pls 19 and 20. Even if misunderstood, a sceptre
with three busts is also depicted on a sculpted jamb of the
9th-century ad church of San Miguel de Lillo (Oviedo). The stone
relief is in the form of a consular diptych in two registers. In the
upper one, there is the ‘consul’, standing on the sella curulis, with
the mappa in his right hand, and flanked by two men: the man on
his left holds the sceptre in his left hand; in the lower one, a scene
of acrobatics with a lion. See at http://commons.wikimedia.org/
wiki/File:San_miguel_de_lillo_ jamba. jpg (last accessed 31 July
2010).
105 Christe (no. 103), 166.
106 G. Sena Chiesa, ‘Introduzione. Il prestigio dell’antico e il riuso
glittico tra IV e X secolo’, in eadem (ed.), Gemme. Dalla corte
imperiale alla corte celeste, Milan, 2002, 1–16, at 8.
107 Ibid., 2.
108 Claud. Epith. 10.10–3.
Gem Portraits of Soldier-Emperors
Erika Zwierlein-Diehl
Maximin Thrax and Maximus
In ad 235 the Pannonian recruits who had been collected on the
upper Rhine in the previous year revolted against Severus
Alexander and proclaimed their commander Maximin
emperor.1 Severus and his mother Iulia Mamaea were killed in
Vicus Britannicus (modern Mainz-Bretzenheim). Maximin,
later called Thrax after his Thracian origin, was the first
soldier-emperor (February/March ad 23–mid-April[?] ad 238).
During the three years of his reign his coins show three portrait
types as Richard Delbrueck has shown.2 A carnelian intaglio in
Vienna portrays him in the first type, called ‘field-portrait’
(Feld-Porträt) (Cat. no. 1; Pls 1–2). It renders perfectly his
characteristic features: the hooked nose, dropping somewhat
at the tip, the energetic chin and the cropped hair and beard.
The apparently contemporary reverse has Hercules struggling
with the lion, which means that the emperor is compared to the
hero. There is also a medallion of Maximin where the emperor
is being crowned by Hercules.3 The cropped hair, very practical
for a warrior, would become typical for soldier-emperors. The
engraver worked in a linear style, which in several versions, is
attested from Republican times onwards. With regard to the
artistic effect the short, hard cuts of the wheel contribute to
expressing the soldierly qualities of the emperor.
A small jasper in Florence has the confronted busts of
Maximin and his son Maximus, Caesar from 236 (Cat. no. 2; Pl.
3). A glass paste in the Martin-von-Wagner-Museum of the
University of Würzburg is taken from this gem (Pl. 3). These
glass pastes were impressions taken from sealings from the
original gems, and thus were replicas of the originals. The
emperor’s portrait belongs to the third type, the ‘triumphal
portrait’ (Triumphalporträt), which appears on coins from the
summer of ad 236, when the emperor and his son accepted the
triumphal title of Germanicus. The forehead is straight, the
chin protruding. Medallions of the year ad 238 show the
Plate 1 Cat. no. 1A, Maximin Thrax, carnelian. Plate 2 Cat. no. 1B, Hercules and the lion.
Vienna, Kunsthistorisches Museum Vienna, Kunsthistorisches Museum
emperor and Caesar in a similar fashion.4 The portrait of a
young bareheaded man in paludamentum in a clear, brownish
grey quartz is plausibly named Maximus Caesar by Cornelius
Vermeule (Cat. no. 3).
Balbinus, Pubienus and Gordian III Caesar
In ad 238 in the course of a rebellion against taxes in Africa
Proconsularis Gordian I and II, father and son having the same
name, were proclaimed emperors, but the younger Gordian
was killed in battle by the forces of the governor of Numidia,
who had stayed faithful to Maximin, whereupon the older
Gordian committed suicide. The Senate, who had confirmed
the imperial titles of the noble Gordiani, then immediately
elected two emperors from among its ranks: Pupienus
(January/February [?]–the beginning of May ad 238) and
Balbinus (January/February [?]–May [?] ad 238) and the 13
year-old Gordian III, grandson of Gordian I, as Caesar.
Maximin conducted his army from Pannonia to Italy but could
not conquer Aquileia and was murdered together with his son
by the soldiers. A post-antique gem portrait of Maximin is C no.
1 (p. 159 below).
A red jasper in Munich shows a bust of Balbinus facing the
capita jugata of Pupienus and Gordian III (Cat. no. 4; Pls 4–5).
The portraits have been cut on the reverse of an earlier intaglio
with the Capitoline Trias dating from the 1st century ad.
Regardless of the cutting down of this side, the stone may have
been chosen with the intention that the divine Trias might
protect the three rulers. The stone was once in the possession
of Francesco Ficoroni and was first published by Rossi in 1707.
Gem portraits of emperors were sought after from the
Renaissance to the 18th/19th centuries which is why we are
confronted with the problem of copies which – if taken for
ancient – turn into forgeries. A chalcedony in Naples has been
erroneously identified by Adolf Furtwängler and others with
Plate 3 Cat. no. 2, Maximin and Maximus, glass paste in
Würzburg of jasper gem in Florence, Museo Archeologico
‘Gems of Heaven’ | 149
Zwierlein-Diehl

Plate 4 Cat. no. 4B, Pupienus, Balbinus and Gordian III, red jasper. Munich, Plate 5 Cat. no. 4A, the Capitoline Trias. Munich, Staatliche Münzsammlung
Staatliche Münzsammlung

Plate 6 Copy of Cat. no. 4, Pl. 4, glass paste, Würzburg, Plate 7 Cat. no. 5, Pupienus, Balbinus and Gordian III, Plate 8 Cat. no. 6, Balbinus, sardonyx,
of chalcedony. Naples, Museo Archeologico Nazionale sardonyx, impression. St Petersburg, Hermitage impression. Berlin, Antikensammlung

Plate 9 Cat. no. 8, Gordian III, carnelian, Plate 10 Cat. no. 9B, Gordian III, carnelian, Plate 11 Cat. no. 9A, Fortuna and Victoria,
impression, lost impression, Międzyrzecz impression

the Ficoroni stone (C no. 2, see p. 159 below; Pl. 6), but it is a
17th-century copy, probably based on a cast of the Ficoroni gem
and coin portraits. Balbinus is beardless, so the beard may not
have been clear on the model. As is usual in this century the
engraving is not very close to the antique style.
A sardonyx in St Petersburg engraved in a cursory style
shows the facing busts of the same two emperors and the
frontal bust of Gordian III Caesar between them (Cat. no. 5; Pl.
7). A horizontally striped sardonyx from the Stosch collection
in Berlin has a bust of Balbinus with a somewhat slenderer
head than on the coins and the Ficoroni jasper, but with the
typical fat cheeks and double chin (Cat. no. 6; Pl. 8). A bust in
the round of blue chalcedony in the Museo degli Argenti in
Florence (Cat. no. 7), according to Elisabetta Gagetti, resembles
the portrait of Balbinus on his sarcophagus.5 This person wears
a corselet with a Medusa head, but no wreath. The precious
material seems to indicate an imperial portrait. The validity of
the tentative identification cannot be judged from the
published photographs.
Gordian III Augustus
In the middle of the year ad 238 the two rival emperors were
murdered by the praetorians and Gordian III was proclaimed
Augustus (b. ad 225 or 226, Caesar: January/February ad 238,
Augustus: 9 (?) May or 6/7 June ad 238–beginning of ad 244).
The impression from a lost carnelian intaglio has a portrait of
the young laureate emperor in cuirass and paludamentum (Cat.
no. 8; Pl. 9). He is beardless as on the coins of ad 238–40.6 A
carnelian excavated in 1954 in the medieval castle of
Międzyrzecz (Poland) is engraved on both sides (Cat. no. 9; Pls
10–11). The larger slightly convex side has Victoria holding a
palm leaf and presenting a wreath to Fortuna in a fine linear,

150 | ‘Gems of Heaven’

 Gem Portraits of Soldier-Emperors

almost calligraphical style. Ruxerówna is of the opinion that

both sides are contemporary,7 whereas Kolendo,8 thinks the
image of the goddesses is later than the portrait. But as Jeffrey
Spier remarks the image can be dated to the 1st century ad.9 We
may compare gems from Herculaneum and Pompeii, which
means a terminus post quem non of ad 79.10 The rim of the stone
is bevelled so that the slightly concave reverse is somewhat
smaller than the obverse. On the reverse of the earlier intaglio
is engraved the laureate portrait bust of Gordian III wearing
cuirass and paludamentum, in a good linear style. He has
sideburns, but no beard. This corresponds to coins of ad 242
when he was 16 to 17 years old.11 As with the red jasper of ad
238 in Munich (Pl. 4) the image of the obverse has been chosen
Plate 12 Cat. no. 11, Gordian III, nicolo, impression, St Petersburg, Hermitage
as a lucky charm for the emperor. The same portrait type, with
sideburns only, occurs on an onyx in St Petersburg, if we may
trust the drawing in the Catalogue of the Orléans collection
(Cat. no. 10). A nicolo fragment also in St Petersburg shows the
portrait of the young emperor with a small moustache and a bit
of down in front of the ear (Cat. no. 11, Pl. 12). This style of
beard also occurs on coins of ad 242.12 In the spring of this year
the emperor set out for a campaign against the Persians under
the command of Timesitheus.
Timesitheus, who had been appointed praetorian prefect in
ad 241, was the real controller of the Empire and a faithful
advisor of the young emperor, who was married to his
daughter, Furia Sabinia Tranquillina. He died during the
campaign against the Persians in ad 243 and was replaced by
Philip the Arab (so-called because he was born in south Syria, Plate 13 Cat. no. 12, Philip I and II, carnelian, impression, lost
then part of the province of Arabia), who was about 40 years
old at the time. By deliberately creating a famine Philip
provoked a mutiny of the troops during which the 19 year-old
emperor was killed.

Philip the Arab and his family

As expected the army chose Philip as emperor (beginning of ad
244–September/October ad 249). He sent a note to the Senate
that Gordian had died of a disease. The Senate accepted Philip’s
version, recognised the new Augustus and confirmed the
appointment of his son M. Iulius Philip as Caesar (b. ad 237 or
238, Caesar ad 244, Augustus ad 247, murdered ad 249 in
Rome).13
A carnelian known only from the Tassie cast shows the
facing busts of Philip I and his son (Cat. no. 12, Pl. 13). Both
emperors wear the laurel wreath, so the date is probably after Plate 14 Cat. no. 13, Philip I and II (?), red jasper. Cologne, Römisch-
Germanisches Museum
ad 247 when Philip iunior was raised to Augustus. At that time
he was nine or ten years old. The profile of the father shows the
characteristic furrowed forehead, a slightly hooked nose, full
somewhat protruding lips and a beard reaching the upper part
of the neck. Richard Delbrueck14 and Willi Real15 describe the
distinctive features of the emperor. The busts of father and son
also appear on medallions.16
A red jasper in Cologne has the laureate portrait busts of an
emperor with a short full beard and his son in a rough, linear
style (Cat. no. 13, Pls 14–15). The sketchy portraits are not real
liknesses. They are in the tradition of red jasper intaglios of the
Severan family, executed in a cursory linear style.17 The long
chin might suggest that the emperor is Maximin Thrax of the
‘triumphal portrait’ type and the bust opposite him
consequently Maximus Caesar, as Antje Krug has suggested.18
Jeffrey Spier, however, remarks that this is ‘unlikely, since Plate 15 Cat. no. 13, impression

‘Gems of Heaven’ | 151

Zwierlein-Diehl
Plate 16 Septimius Severus and sons, bronze ring, impression. Derek J. Content
Collection
Plate 18 Philip I and II and Otacilia, bronze Plate 19 Cat. no. 16, Philip I and II and Otacilia,
medallion, ad 247, Paris, Cabinet des médailles carnelian. Derek J. Content Collection
Maximus was never elevated to the rank of Augustus and
should not be depicted laureate’.19 He therefore proposes the
identification with Philip I and Philip iunior at the time when
his son was Augustus (ad 247–49). That is possible and as for
Cat. no. 12 (Pl. 13) it is probable that the laurel wreath refers to
the rank of Augustus, but there remains a small uncertainty.
The convention that the Caesar is bareheaded and the
Augustus laureate is well known from Severan family
portraits.20 A new example, a bronze ring in the collection of
Derek J. Content showing the busts of Septimius Severus, Geta
at his back and Caracalla in front of him all with laurel
wreaths, is therefore datable to after the promotion of Geta to
Augustus, that is to ad 209–11 (Pl. 16). But this convention is no
longer inflexible in the 3rd century ad. On a coin from Miletus,
Gordian III, designated ‘Caesar’ by the inscription, is laureate
as well as the emperors Pupienus and Balbinus.21
A now lost intaglio of carnelian or chalcedony has Philip I
with Otacilia Severa and Philip iunior (Cat. no. 14, Pl. 17). As all
scholars from Lippert onwards have noticed, the image is very
much like the family group on medallions, especially a bronze
medallion celebrating the second consulate of Philip I and the
first consulate of his son in ad 247 which is similar to the gem in
most details (Pl. 18).22 There are only slight differences in the
measurements. On the gem Otacilia’s nose is a bit shorter and it
has no hanging tip. Philip iunior has a more rounded nose tip
and fuller lips compared with the medallion. But the similarity
is so great that both the gem and the coin die must have come
from the same hand.
Given the particular difficulties in this field we must ask:
can we exclude the possibility that the intaglio was a modern
copy of the medallion? As the intaglio is lost, we cannot look at
the original, which would make things easier. The earliest
152 | ‘Gems of Heaven’
Plate 17 Cat. no. 14, Philip I and II and Otacilia, carnelian (?), impression, lost
Plate 20 Cat. no. 16, impression from Pl. 19
impression was published by Lippert in 1756. This delimits the
time of possible production. Attempts to make close copies of
antique gems begin in the second quarter of the 18th century.
But if we have the opportunity to compare the original and its
copy, in this initial period, there is always some minor
alteration to be found even in the best copies. Most of the time
it is an embellishment in the eyes of the gem cutter that
distinguishes a copy from the original. The best documented
series of gem copies of portraits of triumvirs and emperors
after coins and gems is that of 40 intagli by Lorenz Natter
(1705–63), once in the Marlborough collection, recently
reconstructed by John Boardman.23 The degree of similarity to
antique gem portraits varies; sometimes Natter even copied
17th-century imitations. One of the best copies in this series is a
Philip I24 copied after a medallion which was a larger and
therefore better model than most gems or coins.25 He
embellished the model making the head slenderer and by
adding a double elegant folding behind and in front of the
neck; he made a wreath with larger, pointed leaves after his
own manner, left out the second ribbon of the loop in the neck,
made the eye and brow look more dramatic and instead of a
vertical furrow he gives a double line of the lower forehead. I
cannot see anything like this in the family group of Philip I and
thus we may continue to take the lost intaglio for ancient.
Incidentally, the authenticity of the gem has never been
questioned. Closer copies were made in the second half of the
18th century by a master like Giovanni Pichler (1734–91), and in
the first half of the 19th century Giovanni Calandrelli (1784–
1853) made perfect copies with the aid of a grid laid over an
impression of the original, as we know thanks to Gertrud Platz-
Horster.26 But even then there is no problem if we have the
original gem. There might be a problem if the original is lost or

Plate 21 Cat. no. 18, Trajan Decius, carnelian. Plate 22 Cat. no. 18, impression
London, British Museum, GR 1872,0314.59
if the model was a coin.
A carnelian in a private collection shows the opposed busts
of Philip I and Otacilia with a frontal bust of Philip iunior
between them, executed in a cursory style (Cat. no. 15). Philip
iunior is bareheaded, so Spier plausibly dates the gem to the
years ad 244–7, before his promotion to Augustus.27 Another
carnelian with a family group in linear style is in the collection
of Derek J. Content and has been identified by the owner (Cat.
no. 16, Pls 19–20). Here the bust of Philip iunior is seen in
profile looking at his father, as is Otacilia. As he wears the
laurel wreath, the date is probably ad 247–9. The star in the
middle, above the head of Philip iunior is very likely intended
as a symbol of the sun, as it often is when combined with a
crescent moon, and so refers to the relationship of the
emperors to the sun god.28 Carina Weiss identifies a cameo
portrait in private possession as that of Philip I, cut from an
older portrait, possibly of Severus Alexander (Cat. no. 17).
Traianus Decius
In ad 249 Decius, who had been appointed supreme
commander in Moesia and Pannonia by Philip I, was
proclaimed emperor by the troops (September/October ad
249–June ad 251). Philip I met his death in a battle against the
legions of Decius near Verona and the praetorians murdered
his young son. A carnelian mounted in an iron setting at the
British Museum has the bust of Decius in cuirass and
paludamentum (Cat. no. 18, Pls 21–2). The gem has two cracks;
the back of the setting is closed and covered by an irregular
area of corrosion. The object may have been a brooch, but there
is no scar from a pin. The head corresponds to coins which
show an elongated contour, a slanting, furrowed forehead, a
long aquiline nose with a somewhat hanging tip and a
pronounced chin.29 The inscription reads xx hea. The three
letters h e a are the initials of the tria nomina of the owner, who
thus was a Roman citizen. The praenomen was probably Hĕrius
or Hērennus;30 for the nomen and cognomen there are several
possibilities. The number 20 is best explained by the
assumption that the owner was a soldier of the Legio XX Valeria
Victrix, even if there is no l or leg before the number as in
other cases.31 This legion was part of Claudius’ invasion army to
Britain and had Chester as its base. Detachments of the
Twentieth Legion are attested during the reign of Gallienus, in
the years ad 260–8, at Sirmium in Pannonia Inferior;32 they
may have been there earlier with Decius. However that may be,
Gem Portraits of Soldier-Emperors
Plate 23 Cat. no. 19, Volusian, green agate,
impression. Baltimore, Walters Art Museum
the gem must have been worn by its owner as a sign of loyalty
to this emperor.
The authenticity of another portrait of Decius has been
rightly suspected by Furtwängler (C no. 3, see below p. 159). It
is a bust with radiate crown on a lost stone (nicolo or
carnelian). Its close similarity to the bust on double sestertii33
and the fact that the bust is larger than usual for ringstones but
exactly as high as the one on the sestertii, as well as the
addition of a senseless loop at the right shoulder, lead one to
the conclusion that it is a modern copy.
Trebonianus Gallus, his son Volusian and Aemilianus
Decius and his son Herennius Etruscus fell in battle against the
Goths near the mouth of the Danube, probably betrayed by
their general, the governor of Moesia, Trebonianus Gallus. As
expected, Gallus was proclaimed emperor by the troops (June
[?] ad 251–August [?] ad 253). Having made a humiliating
peace with the Goths, Gallus went to Rome where the Senate
confirmed the army’s choice. His son Volusian (June [?] ad 251–
August [?] ad 253), who was about 21 years old, was called
imperator together with his father. He was appointed Caesar
and Augustus in the same year ad 251. The characteristics of a
portrait in green agate in the Walters Art Museum, Baltimore
(Cat. no. 19, Pl. 23), correspond with coins showing the young
Augustus with a long slightly curved nose under a more or less
protruding brow, with a rounded angle between the frontal
and temporal hairline and a short full beard.34 He wears tunica
and paludamentum, the dress of an emperor or prince who is
Plate 24 Cat. no. 20, Aemilian, carnelian (?), impression, lost
‘Gems of Heaven’ | 153
Zwierlein-Diehl
Plate 25 Cat. no. 21, Valerian, jasper, Plate 26 Cat. no. 22, Gallienus, Valerian I, Saloninus,
lost, glass paste. Würzburg carnelian, impression. Naples, Museo Archeologico
Nazionale
not at war.35 In ad 253 M. Aemilius Aemilianus (July/August–
September/October ad 253), the governor of Moesia, managed
to defeat the barbarians who were still invading the country
and was promptly raised to the throne by his soldiers. He set off
for Italy with his victorious troops. Gallus sent for his general
P. Licinius Valerianus, commander in Raetia. But help came too
late and Gallus and his son Volusian were killed by their own
soldiers.
An emperor portrait in cuirass and paludamentum on a lost
chalcedony or carnelian has been identified by most authorities
with Trebonianus Gallus, but he does not have the sharp
pointed nose as documented by his coins (Cat. no. 20, Pl. 24).
This emperor instead has an aquiline nose with a slightly
hanging tip, a high, somewhat rounded, wrinkled forehead, a
deep set eye, a thick lower lip and a furrow slanting down-
wards from the wing of the nose. These are characteristics of
good coin portraits of Aemilianus.36 There is no apparent
mistake in the intaglio portrait. But a certain over-perfection in
the rendering of the drapery, and the exaggeration of the
gloomy look, make me suspect that it might be a copy after a
coin. But it is better to leave the discussion open until we know
more of the oeuvres of 18th-century gem engravers. It is very
easy to condemn a piece but terribly difficult to reclaim it for
antiquity in case of error. Aemilianus’ reign lasted for three to
four months only; he was murdered when Valerian was
approaching.
Valerian, Gallienus and Saloninus
Valerian (June/August ad 253–June [?] ad 260) had been
proclaimed emperor by his troops while still in Raetia. Being a
member of an aristocratic family, he had held high positions
under the Gordiani and Decius. The Senate confirmed his
imperial titles and named his son Gallienus co-emperor
(September/October ad 253–c. September ad 268). A lost
intaglio in jasper has a bust of Valerian with laurel wreath,
tunica and paludamentum (Cat. no. 21, Pl. 25). He has a broad
head with a sloping occiput, a slightly curved nose, full lips and
a marked double chin. Apart from sideburns growing down to
the chin, he is shaved. The hair of the forehead is combed in an
even curve from the middle to the temples. Most coins show
the hair on the forehead of the emperor combed to the temples,
forming there a more or less sharp angle. This corresponds to a
sculpture head in Copenhagen, where the hair forks a bit to the
154 | ‘Gems of Heaven’
Plate 27 Cat. no. 23, Valerian I and Shapur I, cameo,
sardonyx. Paris, Cabinet des médailles
left from the middle of the forehead and from there is combed
to the angles at both temples.37 But there are coins where the
hairline goes in a gentle curve38 and there are tetradrachmai
minted in Alexandria in the first and second year of Valerian’s
reign (ad 253–5) showing a full row of hair fringes in a
continuous curve.39 So it seems probable that the gem dates
from the beginning of Valerian’s reign.
The facing portrait busts of two emperors on a carnelian in
Naples, in spite of the rough linear style, show the
characteristics of Gallienus with a rectangular head and a
beard growing down the neck in the type of co-emperor,40 and
Valerian with a broad face without a beard.41 The frontal bust of
the boy between them is Saloninus the younger son of
Gallienus (Cat. no. 22, Pl. 26). The gem can be dated between
ad 258 and 260, that is after the death of Gallienus’ older son
Valerian II and before that of Saloninus. Gallienus overtook the
command at the Danube and the Rhine, while his father led his
army against Persia as had Gordian III and Philip I before him.
Valerian was defeated by the Sasanid king Shapur I (ad 241–
72), who in his victory inscription at Naqsh-e Rustam boasts of
having captured the Roman emperor with his own hands.42 The
capture is shown on a large cameo in the Cabinet des
médailles, Paris, made perhaps by a Roman in Sasanian service
(Cat. no. 23, Pl. 27). The two rulers ride at a flying gallop
towards each other, Shapur in rich Persian dress, Valerian in
Roman cuirass and paludamentum. In a victorious gesture
Shapur grasps the wrist of his enemy.
A large amethyst intaglio in the British Museum has the
portrait of Gallienus with radiate crown in cuirass and
paludamentum (Cat. no. 24, Pls 28–30). It was acquired in 1925
from the Cook Collection, so too late for Walters to include it in
his catalogue. The surface is convex on both sides, somewhat
higher on the back than on the front (Pl. 30). The form is
similar to the biconvex amethyst with a portrait of Constantine
the Great in Berlin, with an even higher convex back.43 The fine
violet colour is darkest in the area of the neck, where the stone
is thickest. Amethyst, especially dark amethyst, has the colour
of purple and therefore is a favourite stone for ruler portraits.44
Gallienus has long hair and a beard growing down the neck but
leaving the chin free. It is the portrait type of his sole
rulership.45 The crown is a circlet with thin rays and a ribbon
wrapped crosswise around it which ends in a loop and floating
ends at the nape of the neck. On coins the radiate crown has a

Plate 28 Cat. no. 24, Gallienus, amethyst. Plate 29 Cat. no. 24, impression
London, British Museum, GR 1925,0715.3
broad band and broad rays as usual in this time. The thin rays
may refer to Gallienus’ assimilation with Sol, which is
documented by a double aureus with Gallienus-Sol and
Salonina-Luna as well as sestertii with Gallienus-Genius PR.46
The hair on the gem is combed downwards from the crown of
the hair as on coins with a laurel wreath, not parallel to the
band in the unnatural but easy-to-make way of most coins with
radiate crowns. A circlet similar to that on the gem, with seven
holes into which thin rays were inserted, occurs on a marble
portrait of Gallienus in Copenhagen; there is no ribbon, but it
could have been painted on.47
Saloninus, the younger son of Gallienus, was Caesar in ad
258 and took the title of Augustus in Cologne in the autumn of
ad 260. We do not know when he was born; coins show him as
a child. In ad 259 his father had set off for the Danube frontier
and left him in Cologne under the guardianship of the
praetorian prefect Silvanus. When in ad 260 the usurper
Postumus marched towards Cologne, the city surrendered and
handed over the young Augustus and Silvanus to Postumus,
who had them killed. Postumus founded the Gallic Empire
which lasted from ad 260 to 274.
A bust of a boy in toga contabulata of opaque blue, lapis
lazuli-coloured glass has been found in Cologne (Cat. no. 25,
Pls 31–2). As the modelling of the back shows, it was originally
fastened to an object, perhaps a ceremonial bowl, a clipeus or a
priestly crown. In any case such a setting points to the fact that
Plate 31 Cat. no. 25, Saloninus, blue glass, bust, Plate 32 Cat. no. 25, right profile
Cologne, Römisch-Germanisches Museum
Gem Portraits of Soldier-Emperors
Plate 30 Cat. no. 24, profile
the person is of princely rank. Most authors date the bust to the
early 4th century ad and, following Jörgen Bracker, identify it
with Constantius II (b. ad 317, Caesar ad 324, Augustus ad 337)
But Hans Jucker in a short note argued that the style of the bust
and the treatment of the hair contradict a Constantinian date.
Dieter Salzmann explained in detail that the form of the toga
and the hairstyle belong to the 3rd century ad and recognised
Saloninus Caesar on the basis of his characteristic double chin,
the round head, roundish cheeks, small projecting nose, large
eyes with heavy eyelids and full lips. He remarks that Tetricus
II (ad 273–4) and Nigrinianus (ad +284/85), who could be
taken into consideration, have no double chin and longer hair.48
In contrast to the coins the hair on the forehead is combed to
the left. Salzmann supposes that this may be due to the
provincial manufacture of the glass bust. It may also be that,
working in the round, the artisan did not feel bound by the
coins, which always show the profile from the right.49
A cameo portrait of white and grey sardonyx in Munich
shows a very young emperor with laurel wreath and
paludamentum (Cat. no. 26, Pl. 33). The tip of the nose and the
forepart of the wreath are missing. He has been tentatively
identified as Philip iunior; but Philip was nine to ten years old
when he was proclaimed Augustus and represented laureate on
coins. He has lean cheeks and a firm chin.50 The characteristics
described above allow one to identify this boy as Saloninus.
The chin is less baby-like than on the coins of the Caesar; this
Plate 33 Cat. no. 26, Saloninus, cameo. Munich, Staatliche
Münzsammlung
‘Gems of Heaven’ | 155
Zwierlein-Diehl
Plate 34 Cat. no. 27, Claudius
Gothicus, carnelian. London, British
Museum, GR 1867,0507.322
Plate 36 Cat. no. 28, Aurelian and Severina, red Plate 37 Cat. no. 28, impression
jasper. London, British Museum, GR 1867,0507.541
might point to a date in the autumn of ad 260 when a few coins
of the young Augustus show a rounded but not a double chin.51
On one type of these coins, an aureus, Saloninus wears the
laurel wreath,52 whereas earlier he is bareheaded or he wears a
radiate crown. In this case the change of attribute marks the
change of title, notwithstanding the fact that in this century
laurel wreath and radiate crown are worn likewise by both
emperor and Caesar.53
Claudius II Gothicus
Gallienus defeated the Heruls in the Balkans, but was
murdered in ad 268 by his officers while besieging the usurper
Aureolus at Milan. His successor was Claudius II (September/
October ad 268–September ad 270), second-in-command to
Gallienus, called Gothicus after his victory over the Goths at
Naissus (Niš). A small carnelian in the British Museum has the
portrait of an emperor as Hercules engraved in a fine linear
style produced by flat bouterolle strokes (Cat. no. 27, Pls 34–5).
The features with the straight forehead, the protruding nose,
the small deep set eye, the hair combed to the forehead
forming a sharp angle at the temple correspond to those of
Claudius Gothicus.54 Stories of Claudius’ great strength were
told (Historia Augusta, Divus Claudius Trebelli Pollionis 13,5–7).
Thus the assimilation with Hercules is suitable for him.
Claudius died of the plague in a camp at Sirmium in September
ad 270 and was succeeded by his brother Quintillus for a short
time (September ad 270).
Aurelian and Ulpia Severina
The troops however proclaimed Aurelian emperor (September
ad 270–September/October ad 275). Like Claudius he had been
one of Gallienus’ high officers. A red jasper in the British
156 | ‘Gems of Heaven’
Plate 35 Cat. no. 27, impression
Plate 38 Cat. no. 29, Aurelian and Severina, lost,
nicolo, impression
Museum has his bust facing that of his wife Ulpia Severina
(Cat. no. 28, Pls 36–37). The iconographic type of the capita
opposita has been studied by Gertrud Platz-Horster.55 The
emperor wears a lamellar cuirass. The furrowed forehead,
straight nose and slim neck are typical of him. For the evenly
concave curve of the front hair we may compare coins of his
last issue of ad 274/5.56 Severina wears the stola of the Roman
matron and the stephane of the empress. The hair is combed to
the neck, leaving the ear visible, plaits are laid from the nape
over the top of the head to the forehead and turned back there;
this ‘Scheitelzopf’ projects beyond the contour of the forehead
as it does on coins.57 The plaits are rendered by strokes of a
small flat bouterolle. In ad 274 Aurelian was consul and
celebrated his triumph over Tetricus, the last emperor of the
Gallic empire. In the same year Severina was raised to Augusta.
These events may have been a motif to engrave on gems too.
The same pair with inverted positions can be seen on a lost
nicolo (Cat. no. 29, Pl. 38). Aurelian wears the lamellar cuirass
and a piece of paludamentum on the left shoulder. Severina is
clothed with tunica and pallium; her hair is more carefully
represented than on the red jasper. Behind the bust appears a
crescent moon, assimilating her to Luna as on several coins.
The comparison of the empress with the moon goddess is
analogous to that of the emperor with the sun god.58 The pair
on the red jasper was identified as Carinus and Magnia Urbica
by Walters, but Carinus has a thicker neck, a fuller beard
growing down the neck and mostly a smooth forehead.59 It has
also been suggested that the pair represent Diocletian and his
wife Prisca or his daughter Galeria Valeria, but the curve of the
hair differs from the angular course of the hairline on
Diocletian’s portraits.60 A carnelian in the Hermitage, St
Petersburg, with busts identified as Galerius and Valeria is a

Plate 39 Cat. no. 30, Aurelian, carnelian. Plate 40 Cat. no. 32, Florian, nicolo. Paris, Cabinet
Copenhagen, Thorvaldsens Museum des médailles
copy of the lost nicolo Cat. no. 29 (C no. 4, see p. 159 below).
The hair coming down to the nape is not connected with the
plait. The drapery of Ulpia Severina’s bust shows knowledge of
Cat. no. 28 too, but the stola has been misunderstood for a male
paludamentum. A carnelian in the Thorvaldsens Museum,
Copenhagen, has a bust of Aurelian, laureate and wearing
cuirass and paludamentum (Cat. no. 30, Pl. 39). Ivana Popović
suggests that an emperor’s portrait with radiate crown on a
yellow glass cameo from Viminacium in Belgrade is probably
Aurelian (Cat. no. 31). This cannot be judged from the
published photograph.
Florian
Aurelian was murdered near Byzantium in the autumn of ad
275 on his march to Persia, not because the army was
discontent with its successful leader but because a member of
his staff wanted to escape severe punishment. As the troops
declined to nominate an emperor, the Senate elected the 75
year-old Tacitus (end of ad 275–mid-ad 276), who was followed
by his brother or half-brother Florian (mid-autumn ad 276). A
nicolo in the Cabinet des médailles, Paris, has the portrait of an
emperor with a square head, a short nose with rounded tip, and
a beard growing deep down the neck (Cat. no. 32, Pl. 40). He is
probably Florian, who reigned for 88 days in ad 276. Aurei from
the mint of Rome show a similar nose with round tip and the
beard growing down the neck.61 There is a marked angle in the
hairline at the left temple, but as intagli are made to be looked
at in the impression we should compare coins with the right
profile, where the line goes in a more gentle curve.62 The
emperor on the nicolo has also been called Marius, who
reigned in the Gallic Empire in ad 269, but his head is more
elongated. A third suggestion is the identification with
Maximian Herculius (October/December ad 285–c. July ad
310), co-emperor with Diocletian. But his beard, if growing
down the neck, has only a neat single row of hair and there is a
sharp angle between the hair of the forehead and at the
temples on both sides.63
Probus
The troops of the East did not accept Florian and chose their
general Probus instead. The two armies met at Tarsus, but it
did not come to a battle because Florian was murdered. Probus
was 44 years old when he became emperor of the Roman
Empire (summer ad 276–autumn ad 282). A Roman medallion
of ad 281 shows the characteristics of his features:64 a vertical,
Gem Portraits of Soldier-Emperors
Plate 41 Cat. no. 33, Probus, carnelian,
impression, lost
double furrowed forehead, a strong almost straight nose, long
whiskers and a short beard. The hairline at the forehead forms
a rounded angle at the temple. The close resemblance of the
portrait on a lost carnelian with the coin type leads us to the
conclusion that it represents Probus (Cat. no. 33, Pl. 41). A
second suggestion, however, can not be excluded with
certainty: the identification with Diocletian (ad 284–305), that
is with his first portrait, which shows him in the type of a
soldier-emperor, and which is strongly influenced by the
portrait of Probus.65 Probus, who had celebrated his military
successes in Germany, Asia Minor and Egypt by a triumph at
the end of ad 281 (?) was killed by discontented soldiers in the
following year.
Numerian
The praetorian prefect Carus, who was born in Narbo in Gallia,
succeeded him. Carus (August/September ad 282–July/August
ad 283) tried to found a dynasty by raising his sons Carinus and
Numerian to Caesars. In the spring of ad 283 he set out for the
Persian war which Probus had prepared. He left his elder son
Carinus, now in the rank of Augustus, in Gaul, to guard the
provinces of the West and took the younger Numerian with
him. Carus penetrated deep into Persia but then died in ad 283
of unknown causes. Numerian became Augustus, co-emperor
with his brother. With the assistance of his father-in law, the
Praetorian Prefect Aper, he began to lead the army back to
Europe. Numerian suffered from an eye disease which made
sunlight and wind very painful for him, so he travelled in a
closed litter, where one day the soldiers found him dead. The
blame was laid on Aper. He was brought before a military
tribunal. The officers chose from among their ranks Diokles as
the new emperor. And he, the later Diocletian, killed Aper by
his own hand, thus fulfilling the prophecy of a Gallic druidess
that he would become emperor after killing a boar (aper in
Latin) (Historia Augusta, Flavi Vopisci Syracusii Carus et
Carinus et Numerianus 14). In ad 285 Carinus fell in battle
against Diocletian or was betrayed by his army. He was the last
of the so-called soldier-emperors.
The first coins of the Caesar Numerian (Caesar ad 282,
Augustus July/August[?] ad 283–November ad 284) show him
as a very young man with only a slight down in front of the ear.
The youthful appearance of the Caesar in ad 282 contradicts
his date of birth of c. ad 253 deduced from the historical
sources.66 A date of birth of c. ad 260 as given by Wegner, or
even later, seems more likely. On coins of him as Augustus we
‘Gems of Heaven’ | 157
Zwierlein-Diehl
Plate 42 Cat. no. 34, Numerian, cameo, chromium-bearing chalcedony.
Vienna, Kunsthistorisches Museum
observe the beard reaching the middle of the chin. Finally, he
wears whiskers and a short full beard.67 He has a low forehead,
a long, sometimes straight, sometimes slightly bent nose; the
contour line of the hair is continuously curved from the
forehead to the temples, passing into the beard.
A cameo in Vienna shows his bust in cuirass and
paludamentum seen from the back, the head turned in profile
(Cat. no. 34, Pl. 42). The material of the base plate and the
portrait is plasma, now identified as green chromium-bearing
chalcedony (see Lüle, this volume).68 Parts of the laurel wreath
and the round fibula are carved out of the upper brownish-
white layer. The so-called Rückenporträt was invented in
Hellenistic times. Its high pathos makes it appropriate for ruler
portraits. The very rich setting with enamel and inscriptions
was made for Rudolph II around 1600. The Christogram and
the inscription on the reverse refer to Constantine the Great, to
whom Rudolph is compared in an inscription on the front side.
So at that time the portrait evidently was identified with
Constantine the Great. Given the current state of research this
identification can be excluded. The features correspond with
the portrait of Numerian Augustus: low forehead, slightly
projecting almost straight nose, the forehead, mouth and chin
lying in a vertical contour line and the short beard leaving the
chin bare; there are no whiskers as yet. Plasma, a green variety
of quartz, was probably not distinguished from emerald in
antiquity. True emerald, namely green beryl, and other fine
green gemstones were all regarded as emerald. It was told that
emerald was good for the eyes: gem engravers, for instance,
relaxed their eyes by looking at emeralds; and so did Nero
during the gladiatorial games.69 Thus the cameo might have
been an appropriate present for the young emperor suffering as
he did from eye disease. Apart from this, emerald was a stone
loved by rulers from Polykrates to Alexander and Claudius.70
The portrait on a carnelian in Munich, cut with swift and
skilled strokes, has the characteristics described above, a
slightly bent nose, whiskers and a beard growing to the mid
chin (Cat. no. 35, Pl. 43).
Finally there is the portrait of Tiberius on the Grand Camée
reworked into that of a soldier-emperor (Cat. no. 36). The hair
of the head has been so to speak shaven and replaced by short
158 | ‘Gems of Heaven’
Plate 43 Cat. no. 35, Numerian, carnelian. Munich, Staatliche Münzsammlung
cropped hair cut with the strokes of a small sharp wheel. In the
same manner a beard is engraved on the cheek. As fortunately
the profile of Tiberius was left untouched, the later emperor
cannot be identified.
Catalogue
No. 1 (Pls 1–2): Maximin Thrax (February/March ad 235–mid-April
[?] ad 238).
Intaglio, carnelian, 23.6 x 18mm. Vienna, Kunsthistorisches Museum.71
No. 2 (Pl. 3): Maximin and Maximus Caesar.
Intaglio, jasper, 15.0 x 12.8mm. Florence, Museo Archeologico.72
No. 3: Maximus Caesar.
Intaglio, grey quartz, 14mm. Boston, Museum of Fine Arts. 62.1158.73
No. 4 (Pls 4–5): Pupienus (January/February [?]–beginning of May ad
238), Balbinus (January/February[?]-May [?] ad 238) and Gordian III
Caesar.
Intaglio, red jasper, 20.5 x 14mm, side B (side A: Capitoline Trias).
Munich, Staatliche Münzsammlung, from the collection of Francesco
Ficoroni.74
No. 5 (Pl. 7): Pupienus, Balbinus and Gordian III Caesar.
Intaglio, sardonyx, 18 x 12mm, St Petersburg, Hermitage.75
No. 6 (Pl. 8): Balbinus.
Intaglio, sardonyx, 16 x 13mm, Berlin, Antikensammlung.76
No. 7: Balbinus?
Bust, chalcedony, H. 60mm. Florence, Museo degli Argenti.77
No. 8 (Pl. 9): Gordian III (9 [?] May or 6/7 June ad 238–early ad 244).
Intaglio, carnelian, 12.1 x 14mm, lost.78
No. 9 (Pls 10–11): Gordian III.
Intaglio, carnelian, 28 x 25mm, side B. (side A: Victoria and Fortuna).
Międzyrzecz Museum (Międzyrzecz, Poland).79
No. 10: Gordian III.
Intaglio, onyx, 12 x 11mm. St Petersburg, Hermitage.80
No. 11 (Pl. 12): Gordian III.
Intaglio, nicolo, lower part missing, diam. 14mm. St Petersburg,
Hermitage.81
No. 12 (Pl. 13): Philip I (beginning of ad 244–September/October ad
249) and Philip iunior (b. ad 237 or 238, Caesar ad 244, Augustus ad
247, murdered ad 249 in Rome).
Intaglio, carnelian, lost.82
No. 13 (Pls 14–15): Philip I and Philip iunior?
Intaglio, red jasper, 19.5 x 13mm. Cologne, Römisch-Germanisches
Museum.83
No. 14 (Pl. 17): Philip I with Otacilia Severa and Philip iunior.
Intaglio, chalcedony (Lippert, Tassie) or carnelian (Furtwängler),
31 x 23mm, lost.84

No. 15: Philip I with Otacilia Severa and Philip iunior.
Intaglio, carnelian, 17.6 x 14mm, private collection.85
No. 16 (Pls 19–20): Philip I with Otacilia Severa and Philip iunior.
Intaglio, carnelian, 17 x 12.5mm. Collection of Derek J. Content.
No. 17: Philip I ? reworked out of a portrait of Severus Alexander?
Cameo, fragment, whitish on grey to black agate, 39.2 x 27.5mm,
private collection.86
No. 18 (Pls 21–2): Decius (September/October ad 249–June ad 251).
Intaglio, carnelian, 23 x 15mm, setting: 27 x 21.2 x 7.4mm. London,
British Museum, from the Hamilton Collection.87
No. 19 (Pl. 23): Volusian (Caesar June [?] ad 251, Augustus August(?)
ad 251–August [?] ad 253), son of Trebonianus Gallus.
Intaglio, green agate, length 11/16 inch. Baltimore, Walters Art
Museum, purchased by Henry Walters at the Carmichael Sale in 1926.88
No. 20 (Pl. 24): M. Aemilius Aemilianus (July/August–September/
October ad 253)
Intaglio, carnelian?, 23 x 19mm, lost.89
No. 21 (Pl. 25): Valerian (June/August ad 253–June (?) ad 260)
Intaglio, jasper, image 19 x 14.8mm, lost.90
No. 22 (Pl. 26): Valerian, Gallienus (September/October ad 253–c.
September ad 268) and Saloninus.
Intaglio, carnelian, 20.9 x 16.7mm. Naples, Museo Archeologico
Nazionale.91
No. 23 (Pl. 27): Valerian and Shapur I (ad 241–72).
Cameo, sardonyx, 103 x 68mm. Paris, Cabinet des médailles.92
No. 24 (Pls 28–30): Gallienus (September/October ad 253–c.
September ad 268).
Intaglio, amethyst, 44 x 36 x 14.8mm. Slight chippings at the occiput,
the lower lip, the 2nd, 3rd and 4th ray from above. London, British
Museum (Reg. no. GR 1925,0715.3), from the Cook collection.93
No. 25 (Pls 31–2): Saloninus (Caesar ad 258, Augustus ad 260, killed
ad 260).
Bust, blue glass, H. 83mm. Cologne, Römisch-Germanisches
Museum.94
No. 26 (Pl. 33): Saloninus.
Cameo, agate, white on grey, 32.9 x 20.9mm. Munich, Staatliche
Münzsammlung.95
No. 27 (Pls 34–5): Claudius II Gothicus (September/October ad 268–
September ad 270).
Intaglio, carnelian, 12 x 9.5 x 2.8mm, small chips at top and bottom.
London, British Museum.96
No. 28 (Pls 36–7): Aurelian (September ad 270–September/October
ad 275) and Ulpia Severina.
Intaglio, red jasper, 15 x 12 x 3.2mm, small triangular amendments in
gold at the lower parts of both busts. London, British Museum.97
No. 29 (Pl. 38): Aurelian and Ulpia Severina.
Intaglio, nicolo, 16 x 13mm, picture 12.5 x 9mm, lost.98
No. 30 (Pl. 39): Aurelian.
Intaglio, carnelian, 11 x 11mm, a piece missing at the bottom.
Copenhagen, Thorvaldsens Museum.99
No. 31: Aurelian?
Cameo, yellow glass, 17 x 13mm. Belgrade, Narodni Muzej, from
Kostolac (Viminacium). 100
No. 32 (Pl. 40): Florian (mid- autumn ad 276).
Intaglio, nicolo, 21 x 18mm. Paris, Cabinet des médailles.101
No. 33 (Pl. 41): Probus (summer ad 276–autumn ad 282), possibly
Diocletian (20 November ad 284–1 May ad 305).
Intaglio, carnelian, 19 x 14mm, lost.102
No. 34 (Pl. 42): Numerian (July/August[?] ad 283–November ad 284).
Cameo, plasma (green, chromium-bearing chalcedony) with thin
brownish white upper layer, 50 x 36mm. Vienna, Kunsthistorisches
Museum.103
No. 35 (Pl. 43): Numerian.
Intaglio, carnelian, 18.3 x 14.6mm. Munich, Staatliche
Münzsammlung, said to come from Tralleis, bought in Izmir.104
No. 36 : The portrait of Tiberius on the Grand Camée reworked into
that of a soldier-emperor. Paris, Cabinet des médailles. 105
Gem Portraits of Soldier-Emperors
Misidentified
Domitian not Gallienus:
‘The cameo from St Castor in Koblenz’, now in the Cabinet des
médailles, Paris, has been erroneously identified as Gallienus.106
Copies and Imitations mentioned (= C)
C no. 1: Maximin Thrax.
Sculpture head, gray chalcedony, H. 34mm. Paris, Cabinet des
médailles.107
C no. 2 (Pl. 6): Pupienus, Balbinus, Gordian III.
Intaglio, chalcedony, 28 x 20mm. Naples, Museo Archeologico
Nazionale.108
C no. 3: Decius.
Intaglio, nicolo or carnelian, 32 x 26mm; lost.109
C no. 4: Aurelian and Ulpia Severina.
Intaglio, carnelian, 12 x 10mm. St Petersburg, Hermitage.110
As Gertud Platz-Horster has remarked111 the authenticity of some
portraits of soldier-emperors ascribed to the 3rd century ad by myself
in 1986 has to be questioned.112 After reconsidering the material I agree
and add four more whose genuineness is dubious.113
Acknowledgements
I am indebted to Brenda Breed (Museum of Fine Arts, Boston),
Gabriella Campini (Museo Archeologico, Florence), Kay Ehling and
Martin Hirsch (Staatliche Münzsammlung, Munich), Friederike
Naumann-Steckner (Römisch-Germanisches Museum, Cologne), Paul
Roberts, Alex Reid and Alex Truscott (The British Museum, London)
for giving me access to gems in their respective collections, as well as to
Derek J. Content for allowing the publication of two pieces in his
collection (Pls 16, 19, 20). My thanks go to Ruth Bowler (The Walters
Art Museum, Baltimore), Martin Hirsch (Staatliche Münzsammlung,
Munich), Karen Nystrøm Simonsen (Thorvaldsens Museum,
Copenhagen), Gertrud Platz-Horster (Staatliche Museen, Berlin) and
Claudia Wagner (Beazley Archive, Oxford) who provided new
photographs. I gratefully acknowledge the help of Ursula Tröger who
corrected my English in the version for the conference lecture and of
Chris Entwistle and Noël Adams who did this for the final stage. I
would also like to thank the anonymous referee for their comments
and Lisbet Thoresen.
Abbreviations
Cades, T. Auswahl Bonn Selection from Cades Rom, 1504
impressions in 3 x 11 boxes, Bonn, Akademisches Kunstmuseum.
Cades, T. Rom Collezione di impronte in stucco cavate dalle più
celebri gemme incise conosciute che esistono nei principali musei e
collezioni particolari di Europa/Collection de 5000 empreintes
[Rome, 1836]. Complete edition, 78 vols, Deutsches
Archäologisches Institut, Rome.
Lippert1 I 1755 Dactyliothecae Universalis signorum exemplis
nitidis redditae Chilias..., Leipzig. (For the full titles of Lippert see,
Zwierlein-Diehl 2007 [n. 43], 530).
Lippert1 II 1756 Dactyliothecae Universalis signorum exemplis
nitidis redditae Chilias altera..., Leipzig.
Lippert1 III 1762 Dactyliothecae Universalis signorum exemplis
nitidis redditae Chilias tertia ..., Leipzig.
Lippert2 I, II 1767 Dactyliothec. Das ist Sammlung geschnittener
Steine der Alten aus denen vornehmsten Museis in Europa ...,
Leipzig.
Lippert2 III 1776 Supplement zu Philipp Daniel Lipperts
Dacktyliothek bestehend in Tausend und Neun und Vierzig
Abdrücken, Leipzig.
RIC The Roman Imperial Coinage by H. Mattingly and E. A. Sydenham
(eds), London, 1923–
Notes
1 For the historical background I refer to: The Cambridge Ancient
History vol. XII. The Imperial Crisis and Recovery A.D. 192–324,
Cambridge, 1939; G.C. Brauer, The Age of the Soldier Emperors.
Imperial Rome, ad 244–84, New Jersey, 1975; M. Sommer, Die
Soldatenkaiser, Darmstadt, 2004; K.-P. Johne (ed.), Die Zeit der
‘Gems of Heaven’ | 159
Zwierlein-Diehl
Soldatenkaiser. Krise und Transformation des Römischen Reiches im
3. Jahrhundert n. Chr. (235–84), vols I, II, Berlin, 2008; For the dates
see: D. Kienast, Römische Kaisertabelle, Darmstadt, 21996.
2 R. Delbrueck, Die Münzbildnisse von Maximinus bis Carinus. Das
römische Herrscherbild, III, 2, Berlin, 1940, 66–7.
3 F. Gnecchi, I medaglioni romani, I–III, Milan, 1912, III, pl. 102,7; RIC
IV 2, 151, no. 119.
4 Delbrueck (n. 2), 65, pl. 1,14.
5 M. Gütschow, ‘Das Museum der Prätextat-Katakombe. 3
Sarkophag mit Darstellung von Dextrarum iunctio und Opfer’,
Memorie IV. Atti della Pontificia Accademia Romana di Archeologia
3 (1934–8), 77–109, pls 11–12; B.M. Felletti Maj, Iconografia romana
imperiale da Severo Alessandro a M. Aurelio Carino (222–85 d. C.),
Rome, 1958, 142, no. 136, pl. 15,50.
6 Delbrueck (n. 2), 77, pls 4, 17–24.
7 M. Ruxerówa, ‘Gemma Międrzyzecka’, Fontes archaeologici
Posnanienses 8–9 (1957–58), 443–7 (with French abstract).
8 J. Kolendo, ‘L’intaille antique découverte au château Międzyrzecz
(Pologne). Contribution aux études sur le portrait impérial dans la
glyptique’, Archaeologia Polona 14 (1973), 359–67.
9 J. Spier, Late Antique and Early Christian Gems, Wiesbaden, 2007,
17, with n. 19.
10 U. Pannuti, Museo Archeologico Nazionale di Napoli. Catalogo della
Collezione Glittica I, Rome, 1983, nos 79, 98, 105.
11 Delbrueck (n. 2), 77, 214, pl. 5, 31; J.P.C. Kent, B. Overbeck and M.
and A. Hirmer, Die römische Münze, Munich, 1973, pl. 105, 455 (c.
ad 243).
12 Delbrueck (n. 2), 77, 214, pls 5, 32.
13 For a detailed monograph of Philip’s reign, see: C. Körner, Philip I
Arabs. Ein Soldatenkaiser in der Tradition des antoninisch-
severischen Prinzipats, Berlin, 2002.
14 Delbrueck (n. 2), 82–4.
15 In: M. Wegner, J. Bracker and W. Real, Gordianus III. bis Carinus.
Das römische Herrscherbild III 3, Berlin, 1979, 31–2.
16 Gnecchi (n. 3), II, 99–101, pls 109, 9.10 and 110, 1–5, and III, 48f., pl.
153,19.20.
17 Cf. E. Zwierlein-Diehl, Die Gemmen und Kameen des
Dreikönigenschreines, Cologne, 1998, cat. no. 261 and examples
cited there; M.-L. Vollenweider and M. Avisseau-Broustet, Camées
et intailles. II Les Portraits romains du Cabinet des Médailles, Paris,
2003, nos 204–5.
18 See n. 83.
19 Spier (n. 9), 18, n. 21.
20 G. Richter, The Engraved Gems of the Greeks, Etruscans and
Romans. II: Engraved Gems of the Romans, London, 1971, no. 577;
Vollenweider and Avisseau-Broustet (n. 17), no. 223.
21 H. Weber, ‘Rare or unpublished Coins in my Collection’, Corolla
Numismatica. Numismatic Essays in Honour of Barclay V. Head,
London, 1906, 296–300, at 298, pl. 15,5.
22 Gnecchi (n. 3), II, 98, no. 10, pl. 109,4; Delbrueck (n. 2), 81, 215, pl.
8, 23.
23 J. Boardman, The Marlborough Gems formerly at Blenheim Palace,
Oxfordshire, Oxford, 2009, no. 461, 1–40.
24 Ibid., no. 461, 36.
25 Delbrueck (n. 2), 215, pls 7, 20.
26 G. Platz-Horster, ‘Zeichnungen und Gemmen des Giovanni
Calandrelli’, in D. Willers and L. Raselli-Nydegger (eds), Im Glanz
der Götter und Heroen. Meisterwerke antiker Glyptik aus der
Stiftung Leo Merz, Mainz, 2003, 49–62, 204–11, at 208, no. 220;
eadem, L’antica maniera. Zeichnungen und Gemmen des Giovanni
Calandrelli in der Antikensammlung Berlin, Berlin and Cologne,
2005, 125, cat. no. C4.
27 Spier (n. 9), 17.
28 G. Dembski, Die antiken Gemmen und Kameen aus Carnuntum,
Vienna, 2005, no. 660; Vollenweider and Avisseau-Broustet (n. 17),
no. 205; and see n. 58.
29 Cf. Delbrueck (n. 2), 89–90, pl. 9, 2.3; Wegner et al. (n. 15), 63–4, pl.
28, a.d.
30 O. Salomies, Die römischen Vornamen (Societas Scientiarum
Fennica. Commentationes Humanarum Litterarum 82), Helsinki,
1987, 72–5.
31 E. Zwierlein-Diehl, Glaspasten im Martin-von-Wagner-Museum der
Universität Würzburg I, Munich, 1986, no. 427; eadem, Die antiken
Gemmen des Kunsthistorischen Museums in Wien III, Munich, 1991,
nos 1918, 1924, 2120; M. Henig, D. Scarisbrick and M. Whiting,
Classical Gems. Ancient and Modern Intaglios and Cameos in the
160 | ‘Gems of Heaven’
Fitzwilliam Museum, Cambridge, Cambridge, 1994, no. 156;
Dembski (n. 28), no. 890 (LEGIO without number).
32 St. J. Malone, Legio XX Valeria Victrix: Prosopography, Archaeology
and History (BAR Int. Ser. 1491), Oxford, 2006, 68–70.
33 Delbrueck (n. 2), pl. 9,4.
34 Sestertius: Wegner et al. (n. 15), pl. 36e.
35 Delbrueck (n. 2), 13, 27.
36 Wegner et al. (n. 15), 97, pl. 37 a–d.
37 V. Poulsen, Les portraits romains II, Copenhagen, 1974, no. 174, pls
279–80.
38 Delbrueck (n. 2), pl. 12,2 = Gnecchi (n. 3), I, 51, pl. 26,5, II, pl. 113,3.
39 A. Geißen, Katalog der Alexandrinischen Kaisermünzen der
Sammlung des Instituts für Altertumskunde der Universität zu Köln,
vol. 3: Marcus Aurelius–Gallienus, Kleve, 1982, nos 2850, 2852–3,
2855.
40 Wegner et al. (n. 15), 107, pl. 40.
41 Delbrueck (n. 2), 98, pl. 12,6.9; Wegner et al. (n. 15), 101, pl. 39a.
42 Ph. Huyse, ‘Die sāsānidischen Inschriften und Felsreliefs’, in Johne
(n. 1), I, 109–23, 110–12, 118–20.
43 37.1 x 27.7 x 14.8mm: E. Zwierlein-Diehl, Antike Gemmen in
Deutschen Sammlungen II. Staatliche Museen, Preußischer
Kulturbesitz. Antikenabteilung Berlin, Munich, 1969, 140, 194, no.
545, pl. 94; E. Zwierlein-Diehl, Antike Gemmen und ihr Nachleben,
Berlin/New York, 2007, 183–4, 445 (bibliography), pl. 672.
44 G. Platz-Horster, ‘Agrippina, die obsolete Mutter. Neue Gemmen
aus Xanten’, Bonner Jahrbücher 201 (2001), 53–68, at 57–8 with n.
17.
45 Delbrueck (n. 2), 121–2, pl. 15,54–6; 16,59–61; 17,67–8; Wegner et al.
(n. 15), 109, pl. 44.
46 Delbrueck (n. 2), pl. 16,64; 17,74; M. Bergmann, Studien zum
römischen Porträt des 3. Jahrhunderts n. Chr., Bonn, 1977, 50, n.
170; eadem, Die Strahlen der Herrscher, Mainz, 1998, 277–8;
P. Bastien, Le buste monétaire des empereurs romains, vols I–III,
Wetteren, 1992–4, I, 109.
47 V. Poulsen, ‘Den fangne kejser og andre romerske ansigter’,
Meddelelser 24 (1967), 1–30, at 4–7, figs 3,4; C.C. Vermeule, Roman
Imperial Art in Greece and Asia Minor, Cambridge (Mass.), 1968,
fig. 168; K. Fittschen, ‘Zwei römische Bildnisse in Kassel’, Römische
Mitteilungen 77 (1970), 134–45, at 138, pl. 65; Poulsen (n. 37), no.
175, pls 281–2; Bergmann 1977 (n. 46), 51, no. I 1; eadem 1998 (n. 46),
122, 282, pl. 55,3. For the type of circlet with a ribbon, see: A. Krug,
Binden in der griechischen Kunst, Mainz, 1968, 45–7, 104–6, Typ 12
II, pl. 3.
48 D. Salzmann, Antike Porträts im Römisch-Germanischen Museum
Köln, Cologne, 1990, 212, n. 13.
49 Delbrueck (n. 2), pls 13, 23–4; Salzmann (n. 48), 212, fig. 131, n. 17
and 19; R. Göbl, Die Münzprägung der Kaiser Valerianus I./
Gallienus / Saloninus (253/268), Regalianus (260) und Macrianus/
Quietus (260/262), Vienna, 2000, Table 6, nos 266–76, pl. 25
[Rome], Table 8, nos 320–3, pl. 29 [medallions, Rome], Table 26,
nos 913–7, pl. 70 [Cologne], Table 27, nos 936–42, pl. 72 [Milan],
Table 50, no. 1696, pl. 121 [Samosata]).
50 Wegner et al. (n. 15), pl. 16d–f.
51 RIC V 1, 123, no. 1, 124, no. 14; Delbrueck (n. 2), pl. 14,43 = Kent et al.
(n. 11), pl. 111, no. 491; G. Elmer, ‘Die Münzprägung der gallischen
Kaiser in Köln, Trier und Mailand’, Bonner Jahrbücher 146 (1941),
1–106, at 17, 26, no. 108, 109, 113, 114; Salzmann (n. 48), 212, n. 18;
Göbl (n. 49), Table 26, 916c, pl. 70 [Cologne]).
52 RIC V 1, 123, n. 1; Elmer (n. 51), no. 113, 114.
53 Bergmann 1998 (n. 46), 277–81.
54 Delbrueck (n. 2), 147, pl. 22; especially 22,12; Wegner et al. (n. 15),
135, pl. 52.
55 G. Platz-Horster ‘Capita opposita. Zur Antoninen-Bulla in Neapel’,
in H. v. Steuben (ed.), Antike Porträts, Zum Gedächtnis von Helga
von Heintze, Möhnesee, 1999, 215–23.
56 Delbrueck (n. 2), 156–60, pls 24,31–4.
57 Wegner et al. (n. 15), pl. 55b (uncertain mint); Kent et al. (n. 11), pl.
118,536 (Rome); Bergmann 1977 (n. 46), 182, Münztafel 5,1.2
(Rome).
58 Bastien II (n. 46), 645; Bergmann 1998 (n. 46), 271–4.
59 Delbrueck (n. 2), pl. 29,1.3.5.8.10.12.14.16; pl. 30.18–20. 26.31–2;
Wegner et al. (n. 15), pl. 60.
60 H.P. L’Orange, Das spätantike Herrscherbild von Diokletian bis zu
den Konstantin-Söhnen 284–361 n. Chr. Unter Mithilfe von R. Unger
mit einem Nachtrag von M. Wegner, Die Bildnisse der Frauen des
Julian. Das römische Herrscherbild III/4, Berlin, 1984, pl. 64.

61 RIC V 1, 351f., nos 20, 21, 23, pl. 10,157, no. 24; Wegner et al. (n. 15),
pl. 55c; A.S. Robertson, Roman Imperial Coins in the Hunter Coin
Cabinet IV, Oxford, 1978, 154, no. 1, pl. 38,1.
62 Delbrueck (n. 2), pl. 25,2–4; Wegner et al. (n. 15), pl. 55d–f.
63 L’Orange (n. 60), pl. 65a–d.
64 Delbrueck (n. 2), 231, pl. 28,32; Felletti Maj (n. 5), pl. 54, fig. 190.
65 L’Orange (n. 60), 14–16, pl. 64b.
66 Kienast (n. 1), 260; Johne (n. 1), I, 417 (G. Kreucher).
67 Wegner et al. (n. 15), 161–2, pl. 59a–e.
68 G. Platz-Horster, ‘Kleine Praser and Chromium-bearing
Chalcedonies. About a small group of engraved gems’, Pallas 83
(2010), 177–200. For the term ‘plasma’ see: E. Zwierlein-Diehl,
Magische Amulette und andere Gemmen des Instituts für
Altertumskunde zu Köln, Opladen, 1992, 43–6.
69 Theophrastus, De Lapidibus 23; Pliny, Nat. Hist. XXXVII,62–3; A.
Krug, ‘Nero’s Augenglas. Realia zu einer Anekdote’, in Archéologie
et Médecine. VIIémes rencontres internationales d’archéologie et
d’histoire d‘Antibes, Juan-les-Pins, 1987, 459–75, at 467–71.
70 E. Zwierlein-Diehl et al., Magie der Steine. Die antiken
Prunkkameen im Kunsthistorischen Museum, Vienna, 2008, 181, 327.
71 Felletti Maj (n. 5), no. 90, pl. 8,30; Zwierlein-Diehl 1991 (n. 31), no.
1730; A.B. Marsden, ‘Imperial Portrait Gems, Medallions and
Mounted Coins: Changes in Imperial donativa in the 3rd century
ad’, in M. Henig and D. Plantzos (eds), Classicism to Neo-classicism.
Essays dedicated to Gertrud Seidmann (BAR International Series
793), Oxford, 1999, 89–103, intaglios no. 7; Zwierlein-Diehl 2007
(n. 43), 183, 445, fig. 669 a,b; Spier (n. 9), 17, n. 15; C. Weiss, ‘Ein
Soldatenkaiser auf einem Kameo in Privatbesitz’, in G. Platz-
Horster (ed.), Mythos und Macht. Erhabene Bilder in Edelstein,
Berlin, 2008, 29–37, at 33, n. 22.
72 A.-F. Gori, Museum Florentium I, Florence, 1731, 16,1; S. Reinach,
Pierres gravées des collections Marlborough et d’Orléans, des recueils
d’Eckhel, Gori, Lévesque de Gravelle, Mariette, Millin, Stosch,
réunies et rééditées avec un texte nouveau, Paris, 1895, pl. 10;
Lippert2 III 1776, 2, 335; J. Tassie and R.E. Raspe, A Descriptive
Catalogue of a General Collection of Ancient and Modern Engraved
Gems, Cameos as well as Intaglios, taken from the most Celebrated
Cabinets in Europe..., London, 1791, no. 12038 (all as Macrinus and
Diadumenianus); (The Tassie casts of the Victoria & Albert
Museum are accessible via The Beazley Archive, http://www.
beazley.ox.ac.uk.); Zwierlein-Diehl 1986 (n. 31), no. 786
(Maximinus and Maximus); Marsden (n. 71), intaglios no. 13
(Maximinus and Maximus). It is not Macrinus, whose portrait has
been studied by: D Salzmann, ‘Die Bildnisse des Macrinus’,
Jahrbuch des Deutschen Archäologischen Instituts 98 (1983), 351–81;
idem, ‘Macrinus mit Caracalla und Iulia Domna’, Archäologischer
Anzeiger (1989), 559–68.
73 C.C. Vermeule, ‘A Collection of Greek and Roman Gems’, Museum
of Fine Arts, Boston, Bulletin 61, no. 323 (1963), 4–19, at 15–16, fig.
16 (‘topaz or yellow, smoky quartz’); quartz, according to an
analysis conducted by Michele Derrick at the request of Lisbet
Thoresen, see the corrected record: http://www.mfa.org/search/
collections?accessionnumber=62.1158. Spier (n. 9), 18, n. 16.
74 D. de Rossi and A. Maffei, Gemme antiche figurate I, Rome, 1707,
61f., pl. 52 (‘Balbino Pupieno e Gordiano. In diaspro rosso del
Signor Francesco Ficoroni’: cf. Zwierlein-Diehl 1986 (n. 31), pl. 182,
fig. 76); F. de’ Ficoroni, Le Vestigia e rarità di Roma antica ricercate,
e spiegate, Rome, 1744, 185f. with pl. (‘Dentro la predetta collina fra
le due vie [1 mile off Porta Portese] l’anno 1709 fu ritrovata una
gemma in Diaspro rosso anulare con tre teste d’Imperadori in essa
incise, ed io la comprai dal cercatore di cose antiche Domenico
Gaudenzi ...[description of both sides, identification of the
emperors]’. Ficoroni erroneously takes the date of the attached
copper plate by Hieronymus Odam for that of the discovery and
acquisition, which as documented by Rossi and Maffei took place
earlier); Lippert1 II 1756, 2, 439; Lippert2 II 1767, 846
(‘Chalcedonier,’ ‘War ehemals dem Ficoroni’); Tassie and Raspe (n.
72), no. 12076 (Ficoroni); A. Furtwängler, Antike Gemmen, I–III,
Leipzig and Berlin, 1900, pl. 50,44; Gütschow (n. 5), 86 and 89, no.
2; E. Brandt, A. Krug, W. Gercke and E. Schmidt, Antike Gemmen in
deutschen Sammlungen, vol 1: Staatliche Münzsammlung,
München, vol. 3, Munich, 1972, no. 2459 (A. Krug); P. Zazoff, Die
antiken Gemmen. Handbuch der Archäologie, Munich, 1983, 327, n.
147, pl. 100,1; Zwierlein-Diehl 1986 (n. 31), no. 787 (with parallels
for the symmetrical arrangement of the paludamenta); Marsden
(n. 71), intaglios no. 24; Platz-Horster (n. 55), 220, with n. 31; Weiss
Gem Portraits of Soldier-Emperors
(n. 71), 33, n. 26.
75 O. Neverov, ‘Concordia Augustorum’, Wissenschaftliche Zeitschrift
der Universität Rostock 19 (1970), 605–612, at 609, with n. 67, pl.
28,5,6; idem, ‘The semantics and function of some large portrait
gems of the Late Empire’, Soobščenija Gosudarstvennogo Ėrmitaža
35 (1972), 46–50 and 90, at 48, figs 7 and 8, 49 with n. 18 (Russian
with English abstract); J. Kagan and O. Neverov, Le destin d’une
collection. 500 pierres gravées du cabinet du Duc d’Orléans. Musée de
l’Ermitage, St Petersburg, 2001, 62, no. 42/14; Spier (n. 9), 17 with n.
17.
76 Lippert1 II 1756, 2, 434; Lippert2 II 1767, 841 (‘Carnelian’, Balbinus);
Tassie and Raspe (n. 72), no. 12071 (‘Carnelian’, Balbinus); Cades
Rom IV C 592 (Balbinus); Cades Auswahl Bonn cl. 13 III 403;
Furtwängler (n. 74), pl. 48,26 (Balbinus?); Gütschow (n. 5), 86 and
89, no. 1, 89, fig. 20 (Balbinus); Felletti Maj (n. 5), 146, no. 145
(maybe Balbinus somewhat idealised); Richter (n. 20), no. 587
(perhaps Balbinus); Zwierlein-Diehl 1969 (n. 43), no. 544
(Balbinus); Marsden (n. 71), intaglios no. 19, pl. 11 (Balbinus).
77 E. Gagetti, Preziose sculture di età ellenistica e romana, Milan,
2006, 212–13, A 39, pl. 15.
78 Lippert1 II 1756, 2, 435; Lippert2 II 1767, 847; Tassie and Raspe (n.
72), no. 12079; Marsden (n. 71), intaglios no. 26 (all: Gordian III).
79 Ruxerówa (n. 7); Kolendo (n. 8); Spier (n. 9), 17 with n. 19.
80 Reinach (n. 72), 144, Orleans II, 50, pl. 130; Felletti Maj (n. 5), 163,
no. 184; Kolendo (n. 8), 362, no. 2, fig. 4 (small, shadowy
photography of the impression).
81 Furtwängler (n. 74), pl. 48,23; Felletti Maj (n. 5), 163, no. 183; O.
Neverov, Antique Intaglios in the Hermitage Collection, St
Petersburg, 1976, no. 141; Marsden (n. 71), intaglios no. 25, pl. 3;
Kagan and Neverov (n. 75), 125, no. 214/32.
82 Tassie and Raspe (n. 72), no. 12081 (Philip, father and son);
Marsden (n. 71), intaglios no. 28 (Philip I and Philip II).
83 A. Krug, Antike Gemmen im Römisch-Germanischen Museum Köln,
Mainz, 1981, no. 389, pl. 121 (Maximin and Maximus); Marsden (n.
71), intaglios no. 12 (Maximin and Maximus); Spier (n. 9), 17–18
with n. 21 (Philip I and Philip II).
84 Lippert1 II 1756, 442; Lippert2 II 1767, 848; L. Pirzio Biroli Stefanelli,
La collezione Paoletti. Stampi in vetro per impronte di intagli e
cammei I, Rome, 2007, III, no. 246; Tassie and Raspe (n. 72), no.
12082; Cades Rom IV C 596; Cades Auswahl Bonn cl. 13 III 31, 407;
Furtwängler (n. 74), pl. 48,31; Felletti Maj (n. 5), 176, no. 211, pl.
26,80; Neverov 1970 (n. 75), 609, pl. 28,7; Bastien (n. 46), II, 670;
III, pl. 95,2 (after Cades); Marsden (n. 71), intaglios no. 29.
85 Spier (n. 9), 17, pl. 137, fig. 1.
86 Weiss (n. 71).
87 H.B. Walters, Catalogue of Engraved Gems and Cameos, Greek
Etruscan and Roman in the British Museum, London, 1926, no.
2028, pl. 25 (‘sard’, according to actual nomenclature a carnelian);
Felletti Maj (n. 5), 192, no. 243; Richter (n. 20), no. 588; Marsden (n.
71), intaglios no. 31; Spier (n. 9), 18, n. 23 (all: Trajan Decius, Feletti
Maj with a question mark).
88 Tassie and Raspe (n. 72), no. 12063 (root of emerald, Maximin);
Cades Rom IV C 583 (‘Corniola’, Maximin); Cades Auswahl Bonn
cl. 13 III 30, 397 (18 x 15mm); Felletti Maj (n. 5), 177, no. 212 (‘forse
Filippo l’Arabo’); D. Kent Hill, ‘Gem Pictures’, Archaeology 15
(1962), 121–5, 125, fig. 9 (perhaps Maximian); Marsden (n. 71),
intaglios no.11 (Maximin).
89 Tassie and Raspe (n. 72), no. 12094 (Aemilian, chalcedony); Cades
Rom IV C 604 (C. Giulio Emiliano, coronato d’alloro. Calcedonio);
Cades Auswahl Bonn cl. 13 III 31, 414; Pirzio Biroli Stefanelli (n.
84), III, 254 (Volusian); Furtwängler (n. 74), pl. 48,25 (carnelian,
probably Trebonianus Gallus); R. Delbrueck, Antike Porträts,
Bonn, 1912, pl. 59,13 (probably Trajan Decius); G. Lippold, Gemmen
und Kameen des Altertums und der Neuzeit, Stuttgart, 1922, pl. 74,7
(Trebonianus Gallus?); A. Minto, ‘Un nuovo ritratto di C. Vibius
Trebonianus Gallus’, Critica d’arte 2 (1937), 49–54, at 52–3, n. 17
(Trebonianus Gallus); P. Ducati, L’Arte in Roma dalle origini al sec.
VIII, Bologna, 1938, 310, pl. 213,4 (Trebonianus Gallus,
erroneously: British Museum); Felletti Maj (n. 5), 207, no. 268
(‘British Museum’, Trebonianus Gallus or Maximian); Marsden (n.
71), intaglios no. 9 (Maximin).
90 Lippert1 II 1756, 2,424; Lippert2 II 1767, 816 (Caracalla); Zwierlein-
Diehl 1986 (n. 31), no. 792.
91 U. Pannuti, La Collezione Glittica II, Rome, 1994, no. 214 (Pupienus,
Balbinus and Gordian III?); Marsden (n. 71), intaglios no. 23
(Balbinus, Pupienus and Gordian III).
‘Gems of Heaven’ | 161
Zwierlein-Diehl
92 R. Gyselen, Catalogue des sceaux, camées et bulles sassanides de la
Bibliothèque nationale I, Paris, 1993, 198; Vollenweider and
Avisseau-Broustet (n. 17), no. 257; Zwierlein-Diehl 2007 (n. 43),
205–6, 455, pl. 758.
93 C.H. Smith and C.A. Hutton, Catalogue of the Antiquities ... in the
Collection of the late Wyndham Francis Cook, Esquire, London,
1908, 23, no. 80, pl. 4, from the A. Morrison (no. 200) and Robinson
Collections (Gallienus not Postumus); Delbrueck (n. 2), App. 4 to
110, fig. 11, 236; Felletti Maj (n. 5), 234, no. 316, pl. 43,142.
94 J. Bracker, ‘Zur Ikonographie Constantins und seiner Söhne’,
Kölner Jahrbuch für Vor- und Frühgeschichte 8 (1965/6), 12–23, at
13, fig. 1, 18–20, pl. 7,9; H. Jucker, ‘Trajanstudien. Zu einem
Chalzedonbüstchen im Antikenmuseum’, Jahrbuch der Berliner
Museen 26 (1984), 17–78, at 69, n. 103 (‘Zu erwägen wäre z. B. Philip
iunior’); Salzmann (n. 48), 209–12, no. 19 (Saloninus); K. Dahmen,
Untersuchungen zu Form und Funktion kleinformatiger Porträts der
römischen Kaiserzeit, Paderborn, 2001, Anh. 10,6 (Saloninus?); F.
Paolucci, Piccole sculture preziose dell’ Impero romano, Modena,
2006, 92–3, no. 27, pl. 27 (principe tardo antico); H. Hellenkemper
in A. Demandt and J. Engemann (eds), Konstantin der Grosse (exh.
cat., Trier), 2007, Mainz, no. 1.9.28 (imperial prince, second half of
3rd/first half of 4th century ad).
95 H. Küthmann, ‘Staatliche Münzsammlung’, Münchner Jahrbuch
3.F.20 (1969), 241–2, no. 11 (probably Philip iunior); Brandt et al. (n.
74), no. 2815 (Philip iunior?); Marsden (n. 71), cameos no. 8 (Philip
II); Spier (n. 9), 218, n. 22 (Philip II as emperor? Caracalla or
Geta?).
96 Walters (n. 87), no. 2029 (Claudius Gothicus as Heracles); Felletti
Maj (n. 5), 263, no. 355 (Claudius Gothicus?); Richter (n. 20), no.
590 (Claudius Gothicus? or Postumus); Marsden (n. 71), intaglios
no. 35 (Postumus).
97 Cades Rom IV C 614 (Aurelian with Severina); Cades Auswahl
Bonn cl. 13, III 31, 423; Walters (n. 87), no. 2031, pl. 25 (Carinus and
Magnia Urbica); R. Delbrueck, Antike Porphyrwerke, Berlin, 1932,
121, pl. 57b (Diocletian and his wife Prisca or his daughter Galeria
Valeria); Felletti Maj (n. 5), 284–5, no. 384 (not Carinus and
Magnia Urbica); Richter (n. 20), no. 592 (Carinus and Magnia
Urbica?); R. Calza, Iconografia romana imperiale da Carausio a
Giuliano (287‑363 d. C), Rome, 1972, 116, no. 23 = 118, no. 27, pl.
20,49 (Diocletian and Prisca); Marsden (n. 71), intaglios no. 44
(Carinus and Magnia Urbica); Spier (n. 9), 18 with n. 25 (Aurelian
and Severina).
98 Tassie and Raspe (n. 72), no. 12104 (nicolo, Aurelian and Severina);
Cades Rom IV C 615 (‘onice’ [sc. nicolo], Aurelian with Severina);
Cades Auswahl Bonn cl. 13, III 31, 424; J.J. Bernoulli, Römische
Ikonographie II/3, Berlin, 1894, 184, n. 1 (Aurelian and Severina);
Delbrueck (n. 97), 121, pl. 57c (Diocletian and his wife Prisca or his
daughter Galeria Valeria); Felletti Maj (n. 5), 268, no. 361
(Acquamarina [erroneously], collezione privata? Aureliano);
Marsden (n. 71), intaglios no. 45 (Carinus and Magnia Urbica).
99 P.M.A. Fossing, Catalogue of the Antique Engraved Gems and
Cameos. The Thorvaldsen Museum, Copenhagen, 1929, no. 1787, pl.
20; Marsden (n. 71), intaglios no. 41.
100 I. Popović, Les camées romains au musée national de Beograd,
Belgrade, 1989, no. 51 (probably Aurelian); Marsden (n. 71),
cameos no. 10 (Aurelian).
101 Lippert1 I 2, 1755, 443; Lippert2 II 1767, 859 (Flavius Valerius
Severus); Tassie and Raspe (n. 72), no. 12115 (Constantius Chlorus,
erroneously identified with Lippert2 II 858); Richter (n. 20), no. 589
(Marius); Zwierlein-Diehl 1986 (n. 31), no. 798 (Florian); Marsden
(n. 71), intaglios no. 37 (Marius); Vollenweider and Avisseau-
Broustet (n. 17), no. 258 (Maximian Herculius).
102 Cades Rom IV C 625 (Diocletian); Cades Auswahl Bonn cl. 13, III 31,
432; Delbrueck (n. 97), 121, pl. 57a (Diocletian); Felleti Maj (n. 5),
279, no. 375, pl. 55,197 (Probus); Calza (n. 97), 116, no. 22, pl. 20,50
(Diocletian); Richter (n. 20), no. 593 (Diocletian). For the triple
leaf-like loop, see Gnecchi (n. 3), II, pls 10,4 and 121,5.6.8.9, with
corkscrew ribbons on pl. 121,1.3 (Probus), earlier e.g. pls 106,8
(Gordian III), 107,3 (Philip I), 118,10 (Florian).
162 | ‘Gems of Heaven’
103 F. Eichler and E. Kris, Die Kameen im Kunsthistorischen Museum,
Vienna, 1927, no. 81, pl. 16 (emperor of mid-3rd century ad); Felletti
Maj (n. 5), 192, no. 244 (‘Pasta vitrea’, Trajan Decius);
W. Oberleitner, Geschnittene Steine. Die Prunkkameen der Wiener
Antikensammlung, Vienna, 1985, 62–3, pls 46–7 (Decius?);
W.-R. Megow, Kameen von Augustus bis Alexander Severus, Berlin,
1987, A 169, pl. 51,3 (Alexander Severus?); W. Oberleitner, ‘Ein
unbekannter Kameoring – Zur Vergoldung antiker Kameen’,
Jahrbuch der Kunsthistorischen Sammlungen in Wien 87 (1991),
59–79, at 74 with n. 88 (Alexander Severus); Marsden (n. 71),
cameos no. 5 (‘Onyx’, Maximin?); Zwierlein-Diehl et al. (n. 70),
178–81, 325–7, no. 17 (Numerian).
104 Brandt et al. (n. 74), no. 2816 (emperor, c. mid-3rd century ad);
Marsden (n. 71), intaglios no. 32 (Trajan Decius[?]).
105 Vollenweider and Avisseau-Broustet (n. 17), no. 275, pl. 30;
L. Giuliani and G. Schmidt, Ein Geschenk für den Kaiser. Das
Geheimnis des grossen Kameo, Munich, 2010, 16–18, pls 11–12.
106 Gallienus by: Delbrueck (n. 2), App. 4 to 110, fig. 12 and Felletti Maj
(n. 5), 234, no. 317, pl. 43,143. But see: Megow (n. 103), A 110, pl. 37
(Domitian as Minerva); Bergmann 1977 (n. 46), 48, n. 161 (as
Megow); D. Boschung, Gnomon 63 (1991), 258 (ideal type, sc.
Minerva); Zwierlein-Diehl (n. 17), 83–5, 87, pl. 50a,b (as Megow);
Vollenweider and Avisseau-Broustet (n. 17), no. 132 (Nero as
Minerva).
107 Vollenweider and Avisseau-Broustet (n. 17), no. 255 (Maximin
Thrax, towards ad 236); Weiss (n. 71), 33, n. 22. It has been studied
and qualified as ‘imitazione all’ antica’ by: Gagetti (n. 77), 562, P 10,
pl. 81.
108 Lippert1 II 1756, 2, 440; Lippert2 II 1767, 845 (‘Carneol’); Tassie and
Raspe (n. 72), no. 12075 (Chalcedony, King of Naples). For further
impressions see: Pirzio Biroli Stefanelli (n. 84), 225, III 4, 242.
Judged as (probably) identical with the Ficoroni jasper (no. 4):
Furtwängler (n. 74), pl. 48,30; Felletti Maj (n. 5), 139–40, no. 132,
pl. 13,44; Brandt et al. (n. 74), ad no. 2459 (A. Krug). Judged as
ancient but not identified with the Ficoroni gem: Neverov 1970 (n.
75), 609, pl. 28, fig. 4; Pannuti II (n. 91), no. 213; Gütschow (n. 5), 86
and 89, no. 3 (not Balbinus); Marsden (n. 71), intaglios no. 22, pl. 10.
Judged as a 17th-century copy of No. 4: Zwierlein-Diehl 1986 (n.
31), no. 919; C. Gasparri (ed.), Le Gemme Farnese, Naples, 1994, 33,
fig. 30, 146, no. 344; Platz-Horster (n. 55), 220, with n. 31; Spier (n.
9), 17, n. 17 (expresses doubts about the authenticity of the jasper in
Munich [No. 4] as well, which in view of the style and the early
publication may be excluded).
109 Tassie and Raspe (n. 72), no. 12084 (nicolo); Cades Rom IV C 597
(‘onice’, sc. nicolo); Cades Auswahl Bonn cl. 13, III 31, 408; Pirzio
Biroli Stefanelli (n. 84), 225, III 5, no. 250; Furtwängler (n. 74), pl.
48,36 (‘Karneol unbek. Besitzers. Trajanus Decius ganz wie auf den
Münzen, nur von noch besserer Ausführung. Der antike Ursprung
des Steines, dessen Original ich nicht kenne, ist indes gerade
wegen der genauen Übereinstimmung mit der Münze
zweifelhaft’); Felletti Maj (n. 5), 192, no. 242, pl. 28,96 (corniola,
Decius).
110 Copy of no. 29. Cades Rom IV C 616 (Aquamarina, Aurelian and
Severina); Cades Auswahl Bonn cl. 13, III 31, 425; Bernoulli (n. 98),
184, n. 1 (Aurelian and Severina); Delbrueck (n. 97), 121f., pl. 58a
(Galerius and Galeria Valeria); Felletti Maj (n. 5), 268, no. 362
(‘Onyce’ [mistaking the material for that of No. 29] collezione
privata?’ s.v. Aurelianus, same woman as on Cades 615 [no. 29], the
man different, perhaps because of a coarser execution); Neverov
1970 (n. 75), 609–10, pl. 29,4.5; Neverov (n. 81), no. 142 (Galerius
and Valeria); Marsden (n. 71), intaglios no. 43 (Carinus and Magnia
Urbica).
111 In Bonner Jahrbücher 189 (1989), 639.
112 She names nos 791, 796 (a replica in the British Museum, Walters
[n. 87], no. 2030), 797.
113 Nos 793 (now in Bern University, from the Leo Merz collection:
D. Willers and L. Raselli-Nydegger [eds], Im Glanz der Götter und
Heroen. Meisterwerke antiker Glyptik aus der Stiftung Leo Merz,
Mainz, 2003, 156–7, no. 160 [R. v. Kaenel]), 794 (16th century, from
the Praun collection), 795, 799.
Overtones of Olympus
Roman Imperial Portrait Gems, Medallions and Coins in the 3rd Century AD
Adrian Marsden
Introduction
In a previous paper published some ten years ago I briefly
discussed 3rd-century ad imperial portrait gems and their
relationship to the numismatic media. This paper seeks to
expand on some of the points addressed there and is perhaps
above all a corrective to some of the attributions made in that
article, since I now believe a number of the gems listed there as
authentic are, in fact, more recent fabrications. This is an
important question, deserving of discussion, for without
serious consideration of what are genuine Roman gems and
what are not, many questions of why these gems were
produced and how they related to objects in other artistic
media cannot be answered.
Coins and medallions offer, of course, a safe starting point.
They bear portraits allied with inscriptions and, with the
exception of forgeries which are generally easy to spot, there is
no doubt of their authenticity. Their portraits form a corpus of
images by which it should be easy to identify and validate
portraits in other media, particularly portraits on engraved
gems which are, on many levels, so similar. However, there are
a number of difficulties.
Engraving a gem is a very different art from cutting a coin
die. Although good engraving in either medium will produce a
likeness that is clearly recognisable, the more schematic
treatment that one can encounter on some gems renders
certain portraits rather awkward. Just as the so-called
‘barbarous radiates’, irregular copies of the 3rd-century ad
antoninianus coinage, can carry portraits that are not
recognisable as those of particular emperors, so certain glyptic
portraits can provide problems.
There are also many gems that are clearly, on the basis of
their subject matter and stylistic considerations, not Roman.
Here, it is irrelevant whether a portrait resembles those
encountered on the coins and medallions of a particular
emperor if the gem is a modern creation. Admittedly, a group
of intaglios from Pompeii offer a rather worrying corrective to
arguments centring on stylistic analysis of gem engraving as a
method of distinguishing genuine Roman intaglios from later
examples. Four gems published by Professor John Boardman in
an illuminating paper appear at first sight to be late copies after
classical subjects.1 That all were found in stratified contexts
proves beyond reasonable doubt their authenticity. However,
despite this warning, many gems are false and we should not
allow the Pompeii specimens to push these more modern
examples into the corpus of authentic intaglios.
Portrait gems
Portrait gems depicting rulers had a long history, their origins
going back long before the Roman Empire to the days of the
Hellenistic kingdoms. Many magnificent examples of Roman
date exist from the 1st and 2nd centuries ad and new specimens
continue to appear. Cameo portraits, usually in sardonyx, with
relief engraving of the pale upper layers of the stone against a
dark background, seem likely to have been produced as gifts
for those close to the emperor. The large size of some, the
so-called Staatskameen or ‘State Cameos’, would have made
them suitable for display in a house interior; smaller examples
would probably have been worn as settings in jewellery.
Intaglio portraits were engraved into a large variety of
stones but nicolo, carnelian and red jasper were generally the
most popular for imperial portraits. These would usually have
been mounted in rings to serve as signets; the use of an
imperial portrait as a seal stone implies an official function.
Some of these are set into medieval mounts, evidence that their
wonderful appearance or subject matter was appreciated many
centuries later. For example, a red jasper intaglio engraved
with an excellent portrait of Antoninus Pius and set into a
13th-century silver seal matrix, was recovered at Swanley
(Kent) in 2005 and is now in the British Museum collection (Pl.
1).2 However, it must be said that many other gems set into such
matrices exhibit workmanship of a far lower order.
The early 3rd century ad saw this engraving of gems with
imperial portraits continuing for a few decades but the
tradition seems to have quickly died around the middle years of
the century. I have speculated elsewhere that this is due to
several factors, both economic and aesthetic.3 The collapse of
the imperial monetary system in the mid-3rd century ad
compounded a situation in which the nature of the payments
made by the emperor to those in his service, both the military
stipendium and imperial donativa, had changed enormously.
The stipendium, the soldiers’ regular salary, had come to have a
negligible value so that the army relied on the periodic
donatives handed out to celebrate imperial anniversaries and
victories to supply their needs. Changes in fashion, in
Plate 1 Red jasper intaglio depicting Antoninus Pius mounted in a medieval
seal matrix. Matrix: 25mm x 19mm. London, British Museum, PE 2006,1004.1
‘Gems of Heaven’ | 163
Marsden
Plate 2 Sardonyx cameo of Julia Domna as the Dea Caelestis, 136 x
102mm. London, British Museum, GR 1956,0517.1
particular a moving away from the wearing of signets to the
display of brightly-coloured settings in rings and brooches
most likely also played a part.
Prior to considering these reasons, however, a survey of the
material and a discussion of the large number of gems that are
modern (or relatively modern) fabrications is necessary. By
discounting what is not genuine and concentrating upon what
is, the resulting corpus of material, together with other
categories of object carrying imperial images such as coins and
medallions, can then be used to speculate on the causes behind
the rapid decline in imperial portrait gems in the 3rd century.
A number of earlier 3rd-century ad portrait gems are
indubitably genuine. It is perhaps most instructive to discuss
these and their features before considering the examples which
purport to represent imperial personages of the later 3rd
century ad.
Third-century ad cameos later than the Severan dynasty
are almost unknown. Several fine examples depicting
members of the Severan dynasty can be seen as continuing the
tradition of the Staatskameen of the Julio-Claudian period.
Most notable are two cameos, both in sardonyx and each
depicting Julia Domna, wife of Septimius Severus. One of
these, in the British Museum (Pl. 2), figures the empress as the
Dea Caelestis, riding a biga of bulls and recalls a metrical
inscription from Carvoran identifying the empress with this
divinity.4 The second, in Kassel, represents her as a
combination of Pax and Victory, winged, holding a wreath and
palm branch and seated on a pile of arms.5 Other cameos of the
period exist but few apply such overtly divine connotations to
their subjects. One, a sardonyx cameo in the Bibliothèque
nationale, depicts a radiate bust of Severus with the rest of his
family, including a laureate bust of Caracalla with an aegis
draped across his shoulders.6
A very fine sardonyx cameo in the British Museum may
represent Diadumenian, son of Macrinus (Pl. 3); the ‘crew cut’
hairstyle certainly places the very young male bust in the 3rd
century ad.7 Other examples portray Severus Alexander, the
last emperor of the Severan dynasty. Perhaps most notable, in
sardonyx, is a cut down bust in the Hermitage showing the
emperor’s laureate head to the left. It represents one of the last
imperial portrait cameos that can be confidently assigned to a
particular emperor or empress.8
Some cameos may belong to the generation following the
164 | ‘Gems of Heaven’
Plate 3 Sardonyx cameo of Plate 4 Sardonyx cameo depicting a female
Diadumenian (?), 18 x 20mm. London, bust, cameo 15 x 10mm. Private collection
British Museum, GR 1867,0507.357
fall of the dynasty in ad 235 when Severus Alexander was
murdered on the Rhine frontier. The overwhelming majority
need not detain us here since they portray not imperial
personages but privati or generic representations of women
(Pl. 4); these were produced as gifts from lovers to their
sweethearts and thus do not fall into the ambit of imperial
portraits. They are discussed in the paper written by Martin
Henig and Helen Molesworth in this volume. It is interesting
that these colourful objects, well suited to the 3rd-century ad
aesthetic as their mounting in the elaborate necklaces of the
period demonstrates, survive in such numbers; it is equally
instructive that cameos with distinctively imperial portraits do
not. That they do not survive is surely testimony that very few
were produced in the first place.
One or two of these cameos do remain as possible imperial
portraits. A fragment, approximately half of a confronted
group, offered for sale on the German antiquities market,
portrays a bareheaded bust of a youth with drapery about his
shoulders.9 This could have been a portrait of Philip II as Caesar
facing his father Philip I (perhaps in company with the prince’s
mother Otacilia Severa) but the missing half makes this
identification impossible to verify. It is equally probable that it
is another portrait of a privatus.
Another, in Munich, is more convincing and its laureate,
draped portrait of a boy is probably a likeness of the same
prince, commissioned to celebrate his elevation to the rank of
Augustus in ad 247.10 If this is the case then it represents the last
sardonyx cameo that the author can confidently ascribe to the
3rd century ad.
Glass paste cameos and intaglios
One group of cameo gems was mass-produced in moulds using
coloured glass (paste) and despite their bright and eye-
catching appearance, they should be set apart from cameos
individually carved in natural hardstones. Examples are
known from the earlier Empire such as the series of glass
phalerae which probably depict Drusus and his children,
Germanicus, Claudius and Livilla or other members of the
imperial family (Pl. 5).11 All surviving examples have been
moulded in either dark blue or turquoise glass and mounted in
settings that they might be worn. Other examples depict
different members of the Julio-Claudian family, for example
Agrippina.12 These phalerae would presumably have been

Plate 5 Glass cameo
depicting Drusus
and his children, D.
38mm. Colchester
Castle Museum
distributed to soldiers as dona militara; it is only natural that
such gifts should have been linked to the imperial family under
whose auspices they were awarded.
Similar moulded glass objects do survive from the 3rd
century ad which bear close comparison with these phalera
settings. One of these pieces, a moulded cameo portrait in
variegated green glass, shows Julia Mamaea, the mother of
Severus Alexander; this paste has been worn as a pendant.13
Another, in amber glass, is moulded with a facing bust of Julia
Domna and likewise mounted in a pendant.14 A head of Severus
Alexander is featured on an example in sapphire blue glass in
the Musée d’Art et d’Histoire in Geneva,15 whilst the confronted
busts of an imperial couple from the mid-3rd century ad,
possibly Philip I and Otacilia Severa, are stamped on a
medallion of brown glass in the same collection.16 All serve to
signal the increasing emphasis on colour evident in the 3rd
century ad although it is also true that glass had been used to
imitate brightly-coloured gemstones such as amethyst and
emerald since the beginning of the Imperial period. They were
probably not necessarily produced as imperial gifts but the
portraits often recall those used on medallions and on gold
coins.
A turquoise glass medallion with a tongue attachment for
suspension, also in the Musée d’Art et d’Histoire in Geneva, is
impressed with a scene that appears to represent Caracalla
seated on a folding camp stool delivering an adlocutio to his
troops.17 It is interesting that it thus reproduces a scene that is
often figured on medallions in metal struck at the imperial
mint.
Other small glass settings recall coins and medallions even
more closely. An example in Vienna produced in black glass,
perhaps imitating jet or sard, carries the likeness of a mid-3rd-
century ad empress, perhaps Salonina.18 The high relief of the
portrait finds parallels on medallic issues whilst the
fragmentary remains of an inscription again recall the
products of the mint.
An example in Belgrade of Aurelian is also symptomatic of
the species, a small oval of yellow glass that would have been
suited for setting in a ring.19 It is stylistically similar to portraits
produced at the mint of Rome during that emperor’s reign and
was probably produced for distribution at the festivities
celebrating his triple triumph and quinquennalia in the spring
of ad 274. Given the similarity of the portrait to those employed
on the coinage at Rome an origin at the mint may be speculated
in this case.
Overtones of Olympus
Plate 6 Orange glass
intaglio with a portrait of
Claudius II, 16 x 13mm.
Private collection
More intaglio portraits do survive but even here the corpus
is not large. As with the paste cameos just mentioned, a number
of gems bearing intaglio portraits of imperial figures are
moulded in glass. Two examples are typical and instructive in
different ways with regard to the problems they raise.
A blue glass paste offered for sale in Germany is, like many
of these glass gems, rather difficult to confidently assign to a
particular emperor.20 The laureate portrait, with draped
shoulder viewed from behind, has a prominent nose and chin
with a fairly short beard. It is most similar to portraits of
Maximinus I and should probably be assigned to that emperor.
However, it is stylistically very different to his coin portraits
and serves as a reminder that intaglio portraits do not always
resemble those found on the numismatic media.
Another, bright orange in colour and imitating carnelian,
bought in the Sudan and now in a private collection, is easier to
interpret at least in terms of its subject.21 It depicts a laureate
bust that shows striking affinities with the portraits
encountered on billon tetradrachms produced at Alexandria in
the name of Claudius II (Pl. 6). If glass cameos were, by the
middle years of the 3rd century ad, being produced at the same
workshops that were responsible for striking coinage, then it is
quite likely that similar intaglios also had their origins there.
Not only is the style almost identical with that of the very
distinctive tetradrachms but the piece also has the remains of a
coin-like inscription running around the bust. Corrosion and
the use of what seems to have been too small a piece of glass for
the stamp used to impress the design has rendered this
inscription illegible but it is clearly in Greek. The dimensions of
the bust are similar to those on the coin portraits of
Alexandrian tetradrachms, making it very likely that a coin has
been used to impress the design upon this piece. Thus it may
represent either a signet stone made at the workshops of the
mint in Alexandria or a privately produced piece.
Given the small number of impressed glass gems bearing
imperial portraits, many of which could have been produced in
a short time, it is not surprising that engraved intaglio gems,
the production of which was far more time-consuming, are
even rarer. Those that have come down to us merit particular
consideration.
Portrait gems: intaglios
An interesting group consists of a fine series of nicolos dating to
ad 238 and perhaps the years immediately following. I have
previously commented that the accession of Balbinus and
‘Gems of Heaven’ | 165
Marsden
Plate 7 Sardonyx intaglio (cast) of Plate 8 Nicolo (?) intaglio of Balbinus from Da Sanzeno,
Balbinus, 16 x 13mm. Berlin. Italy, 19 x 15mm
Plate 10 Nicolo intaglio (cast) of Plate 11 Carnelian intaglio (cast) of
Maximus. Tassie corpus Maximus, 12 x 15mm. Ex-Velay
Collection
Pupienus and the raising of Gordian III to the purple furnished
an occasion for the production of intaglios intended as gifts for
those close to the three emperors.22
A sardonyx intaglio in Berlin features a laureate and draped
bust of Balbinus (Pl. 7).23 The exquisite portrait places this in
the first rank of Roman portrait intaglios and an origin outside
the ambit of the imperial court is unlikely. Another intaglio of
Balbinus from Da Senzeno, although listed as a glass paste, has
every appearance of being a very finely engraved nicolo (Pl.
8).24 It depicts a bust of the emperor wearing no laurels but clad
in a paludamentum and must have been produced in the
summer of ad 238, as part of a programme to celebrate the
inception of Balbinus and Pupienus’ short reign. The modelling
is of excellent quality and and very similar to that encountered
on the Berlin gem. It is probable that it was engraved in the
same studio.
Another nicolo, also of excellent quality and now in the
Hermitage, has at some point been cut down to a circular shape
(Pl. 9).25 This depicts a laureate portrait of Gordian III, dating
the piece to the period after the murders of Balbinus and
Pupienus. It was presumably produced in the days after
Gordian’s elevation by the Senate to the rank of full Augustus,
probably in the same workshop as the Balbinus gems.
A cast of a nicolo in the Tassie collection represents
Maximus Caesar, son of Maximinus and is stylistically very
close to the three gems mentioned above (Pl. 10).26 It may thus
represent an earlier product of the workshop to which those
gems have been tentatively attributed and indeed it is not
unreasonable to assign it to the same engraver.
166 | ‘Gems of Heaven’
Plate 9 Nicolo intaglio (cast) of Gordian III, D.14mm.
St Petersburg, Hermitage
Plate 12a Carnelian intaglio of Trajan Pl. 12b Cast of Plate 12a
Decius, 23 x 15mm. London, British
Museum, GR 1872,0314.59
Some intaglio gems depicting these emperors in other
materials are also probably genuine. A carnelian surviving as a
cast in the Tassie corpus features a laureate bust of Maximinus,
seen draped and cuirassed from behind the shoulder.27 Another
carnelian from Dalmatia depicts a young bareheaded male
with the facial physiognomy of Maximus (Pl. 11).28 The drapery,
seen from the back, is inexpertly cut, but this portrait with the
shoulders seen from behind is the conventional way in which
imperial princes were depicted. The portrait is of good style
and is likely to have been intended to represent the young
emperor.
A carnelian intaglio in the British Museum is engraved with
a bust of Trajan Decius (Pls 12a–b).29 The appearance of the
portrait is very similar to coin images of Decius but the gem is
engraved with a strange inscription that runs upwards from
the front of the emperor’s chest. The letters read xx.he a and,
although there remains the possibility that the inscription is a
later addition, their cutting and placement have every
appearance of being contemporaneous with the imperial
portrait they accompany. Perhaps the inscription has some
votive significance; Decius issued radiate antoniniani with the
legend votis decennalibvs and it may be that the xx of the
gem’s legend refers to the imperial vota although, of course,
this would signify a vota vicennalia and not a decennalia.30
Obviously, given the short length of Decius’ reign, one would
expect a decennial vota but this inscription may have been
looking forward, somewhat prematurely, to vicennalian
celebrations. The Emperor Tacitus issued medallions with
vicennalian vota during his six-month reign.31 The he a could

Plate 13a (left) Carnelian intaglio of
Septimius Severus, 31 x 14mm.
Hanover, Kestner-Museum
Plate 13b (right) Reverse of Plate
13a: confronted busts of Julia Domna
and Caracalla
then represent a formula such as herennius etruscus
augustus or perhaps herennia etruscilla augusta.
Etruscilla was Decius’ wife and Etruscus their son who was
raised to the rank of Augustus in ad 250. The initials of
Herennius Etruscus in particular would provide a plausible
explanation for these letters which thus may have celebrated
his acclamation as Augustus alongside his father. Such a
situation would also have provided a fitting occasion for the
undertaking of imperial vota. Alternatively, the inscription
could have a completely different meaning but an
interpretation which places the letters in the context of an
imperial gift, perhaps connected to the elevation of Etruscus to
partnership in the Empire, is most likely.
Other gems appear perfectly genuine but, like the glass
paste that may depict Maximinus discussed above, raise
difficulties when it comes to accurately identifying their
subjects. A red jasper in Cologne shows two confronting
laureate and cuirassed male busts viewed from behind.32 The
fact that both subjects wear laurels, together with the fact that
one portrait wears a short beard and the other appears
beardless, would seem to place them in the late ad 240s or early
250s. Physiognomically the portraits resemble Maximinus I
and Maximus but the fact that both are laureate renders this
identification impossible. Most likely another father and son
group was intended, perhaps Philip I and Philip II after his
elevation, or perhaps Trajan Decius and Herennius Etruscus.
Double-sided intaglios
The group of so-called biface intaglios where engraving
occupies both faces of the gem deserves especial mention.
These integrate, on a basic level, features of both medallions
and intaglios in that both sides carry an image. This pairing of
what might be termed obverse and reverse types has its nearest
parallel in coins and medallions where a portrait bust of an
imperial personage is coupled with a reverse showing another
design. There are four examples from the 3rd century ad; in
each case the gem is of an exceptional standard in terms of its
quality and style of engraving.
The earliest 3rd-century ad gem in this category is a
carnelian fragment in Hanover.33 The obverse is carved with a
likeness of Septimius Severus (Pl. 13a) whilst the reverse is
engraved with the confronted busts of Julia Domna and
Overtones of Olympus
Caracalla (Pl. 13b). This was clearly a dynastic piece and its
iconography is strikingly comparable to certain coin and
medallion issues of the period.34 The apparent ages of the
portraits puts its production in the last years of Severus’ life
and the absence of any bust of Geta suggests a date of perhaps
ad 208–9. The size of the piece would have made it too large for
use as a ringstone; it would have been more suitable as a
pendant with the consequent display of both sides being thus
made possible.
Another gem, a carnelian in Vienna, combines a bust of
Maximinus on one side (Pl. 14a) with a full-length depiction of
a Hercules, who appears to have the facial features of the
emperor, wrestling with the Nemean lion on the other (Pl.
14b).35 The third example, engraved in red jasper and in
Munich (see Zwierlein-Diehl, this volume, Pls 4 and 5), shows
the busts of Balbinus, Pupienus and Gordian III; the reverse
face features the Capitoline Triad, Jupiter, Juno and Minerva.36
On both of these examples, the reverse images form an
iconographic programme of their own, extending and building
upon the obverse portrait in the same manner as occurs on
coins and medallions. Hercules strangling the Nemean lion is
an apt artistic metaphor for Maximinus’ defeat and subjugation
of the barbarians who threaten the Empire’s stability; this use
of Hercules as a model for the emperor’s struggles against
chaos is very common on coins although none occur on those of
Maximinus. The image on the Munich gem implies the renewal
Plate 14a Carnelian intaglio of Plate 14b Reverse of Plate 14a:
Maximinus, 20 x 15mm. Vienna, Hercules and the Nemean lion
Kunsthistorisches Museum
‘Gems of Heaven’ | 167
Marsden
Plates 15a–b Carnelian with a helmeted male bust on one side, a diademed
female on the other, 14 x 8mm. Geneva, Musée d’Art et d’Histoire
of the Roman state religion through the restoration of the
position of Rome’s three greatest deities. The sentiment is
echoed slightly by the iovi conservatori and pietas mvtva
avg[ustorum] types that occur on denarii, radiates and aes
issues of Balbinus and his colleague Pupienus.37 The reverse
images of these two gems are most akin to those found on gold
coins and particularly medallions; the relative rarity of aurei
and medallions of this period perhaps accounts for the lack of
comparanda in this medium.
A fourth example, of carnelian, from Międzyrzecz
Wielkopolski in Poland and now in the collection of the Poznań
Archaeological Museum, combines an intaglio head of Gordian
III on one side with a composition of Fortuna being crowned by
a Victory on the other (see Zwierlein-Diehl this volume, Pls
10–11).38 Although Victory and Fortuna both appear with
regularity on Gordian’s coinage they are rarely associated with
one another. However, one unusual antoninianus issue pairs a
depiction of Victoria with a legend of fort[una] redvx and
may be linked to the same programme that saw the distribution
of the gem.39
Another carnelian, in Geneva, pairs a helmeted male bust
on one side with that of a diademed female on the other (Pls
15a–b).40 It may be intended to have represented the Palmyrene
prince Vabalathus and his mother Zenobia or an imperial
couple of around the same date, but it is of very dubious
authenticity and can be effectively discounted from a corpus of
genuine 3rd-century ad imperial portrait gems.
These pieces are interesting in that they seem to represent
an attempt at integrating certain features of gems and
medallions. They are restricted to the first few decades of the
3rd century ad and do not occur before or after this period. The
cutting of the two faces is contemporary on the two earlier
examples but on the two later gems the reverse images appear
to represent a primary engraving, the imperial portraits having
been carved later on the unused face of the stone. Indeed, the
edge of the Munich gem has been ground down, perhaps to
remove chipping, and this has resulted in the figures of the
three gods losing their feet; this operation was most likely
carried out during the engraving of the portraits on the other
side of the gem. In any case, the images that became the
reverse types of these gems were clearly selected as apt
accompaniments to the obverse portrait busts and suggest that
these objects were fashioned to provide their recipients with
168 | ‘Gems of Heaven’
both a signet for use on official documents and with a
medallion-like gift that could be worn as a mark of imperial
favour. Incidentally, this re-use of earlier gems may provide
further evidence for a decline in the glyptics industry, the
adaptation of existing intaglios proving easier than the
alternative, that is the engraving of new ones.
It is also interesting to note that these objects appear during
the last phase of widespread imperial portrait engraving on
intaglios and it seems likely that they never became established
in a Roman context, perhaps as a result of this late appearance.
In terms of conception they have similarities both with cylinder
seals and scarabs but, above all, they are best seen as
combining the functions of a portable stamp for impressing
sealings and a colourful and wearable imperial portrait blazon.
Thus, they are as akin to the dies used to produce medallions as
they are to the medallions themselves.
What is, rather ironically, most striking about all of these
authentic intaglio gems is the plain imperial images they carry.
There is no ‘dressing up’ of emperors and empresses such as is
found on medallions and on some coins, in particular the issues
in precious metal. This dressing up on the numismatic media is
discussed below; for now it is enough to highlight the fact that
it does not intrude into the realm of genuine intaglio gems.
Indeed, the fact that an imperial portrait intaglio does not
carry plain images is enough to render its authenticity suspect.
Intaglio gems were intended to seal public documents and
those bearing imperial portraits were presumably produced to
function as signets for men conducting imperial business; the
appearance of such outré images on this medium would be
surprising.
The sealings made by these gems that survive attached to
documents attest this usage such as one depicting a bust of
Gordian III from Dura Europos still fixed to a papyrus
recording the purchase of a female slave.41 This was plainly the
signet of the official before whom this document was drawn
up, one Marcus Aurelius Antiochus.42
Some gems (or possibly metal stamps), like the Dura
example, survive only in the lead sealings they were used to
produce. One example, offered for sale a few years ago, depicts
the confronted busts of Trajan Decius and Herennius
Etruscus.43 These images are far more in keeping with
3rd-century ad imperial intaglio iconography. Like the
authentic intaglios discussed above, the busts are simple, with
the addition of only drapery, cuirass and laurel wreaths. They
have none of the artistic flourishes that distinguish some of the
dubious gems discussed below.
Imperial portrait gems: dubitanda
Many imperial portrait gems are of dubious authenticity. A
large number have been in collections for many years and this
has often meant that they have been unchallenged, researchers
assuming that their age equates to authenticity. However,
simply because a gem has been known for 200 years does not
mean that it is of Roman date; collectors in the 18th century
were avid for gems with which to embellish their cabinets and
it should come as no surprise that there were artisans willing to
satisfy this desire. Thus, these gems, often accepted as being
ancient, are a reflection not of genuine 3rd-century ad
workshops but rather of the post-medieval desire for imperial
glyptic portraits to stand alongside other representations of

Plate 16 Glass paste of an emperor of post-medieval date, 20 x 16mm.
University of Würzburg, Martin-von-Wagner Museum (hereafter MVWM)
emperors and empresses encountered on coins and medallions.
Indeed, identifications have often been forced on some
neo-classical specimens, in an attempt to fit them within the
corpus of imperial gem portraits, when they were only ever
intended to represent the idea of an emperor and not portray a
specific individual. These classicising busts were probably not
even intended to deceive at the time they were engraved; we
should not allow them to deceive us now. Some assemblages,
such as the Raspe-Tassie corpus, are littered with neo-classical
gems.44
A rather brutal-looking head in the Würzburg University
collection is a case in point (Pl. 16).45 The piece does not bear
any real resemblance to any Roman emperor although its squat
head can be said to convey very well what might be termed the
idea of a Roman emperor.
Many other gems are so obviously later fabrications that
they need not detain us for long. Some can be instantly
dismissed on account of the material from which they were
produced. Stones such as lapis lazuli were not used in the
Roman world to create ringstones. Thus, a lapis purporting to
show Laelian or Tetricus I in the British Museum may be safely
disregarded (Pl. 17).46 In any case, the style of the piece is
unconvincing to say the least.
Plate 17a Lapis lazuli intaglio of Laelian or Tetricus I, Plate 17b Impression of Plate 17a
H. 17mm. London, British Museum, GR
1923,0401.248
Overtones of Olympus
These falsi include a large number of glass paste
impressions in the Würzburg collection. Some of these are of
utterly unconvincing style as well as being immediately suspect
on the basis that they portray emperors whose coins are
exceedingly rare. An impression of a jasper purporting to
represent the short-lived Gallic usurper Laelian, for example,
would deceive none but the most naïve student of Roman
imperial glyptics (Pl. 18).47
Other gems can also be discounted on the basis of who they
portray. It is very unlikely, for example, that intaglios would
have been produced showing the features of Diva Paulina. Very
little memorial coinage was issued in her honour and the
portrait of a dead empress would not have been overly
appropriate as an official signet. A paste in Würzburg with
Paulina’s veiled bust clearly denoting a posthumous portrait
appears to be from the same gem as a cast in the Cades and
Lippert collections. All are dubious to say the least.48
A chalcedony in Naples showing a group of Pupienus,
Balbinus and Gordian III betrays itself by its large size and also
on stylistic grounds (Pl. 19).49 The strange, rather linear,
engraving and the unusual treatment of the drapery about the
subjects’ shoulders are not convincing. Chalcedonies were also
not generally used for intaglio gems at this date.
A number of intaglio gems remain that deserve critical
scrutiny. A large carnelian from the Cades cast collection
carries a radiate and cuirassed bust of Trajan Decius.50 Genuine
intaglios do not depict radiate crowned imperial busts nor do
they generally occur this large; rather the creator of this piece
has taken one of Decius’ large double sestertii, an impressive
though short-lived denomination, and used it as a model for
this oversized intaglio.
Another large intaglio, also in carnelian, features Philip I,
Otacilia Severa and Philip II (Pl. 20).51 Like the carnelian of
Trajan Decius mentioned above it is immediately suspect on the
grounds of its size. It is too large to have been used as a
ringstone. Even more troublesome is the fact that it replicates
exactly the medallion images produced at the mint of Rome.
This, together with the superb engraving might be said to lull
the viewer into a false sense of security, leading them to believe
the piece was produced at the same time as the medallions and
Plate 18 Glass paste from a jasper of Laelian of post-
medieval date, 17 x 16mm. MVWM
‘Gems of Heaven’ | 169
Marsden
Plate 19 Chalcedony intaglio (cast) showing Balbinus, Pupienus and Gordian III of
post-medieval date, 28mm x 19.6mm. Naples, Museo Archeologico Nazionale
for a similar purpose. It is surely more realistic to interpret the
gem as a post-medieval construct, engraved as a replica of the
medallion to fill a position in a collector’s cabinet alongside an
example of the object which it so faithfully imitates.
A large amethyst in the British Museum depicting Gallienus
is also problematic (Pl. 21).52 The physiognomy of the subject is
unmistakable and amethysts were used for several rather
oversized intaglio gems from the late 2nd and early 3rd century
ad.53 It was also used, as was rock crystal, for a small number of
intaglios of Constantius II and his close relatives in the middle
years of the 4th century ad. However, we should probably not
view this large and impressive gem as either a successor to the
earlier gems or a forerunner to the later ones. Gallienus, with
his outlandish, quasi-divine style, may have anticipated the 4th
century ad in some respects but there are good reasons to
believe this gem was not one of his creations.
The draped and cuirassed bust, with the shoulders viewed
neither from the rear nor the front, gives some cause for
concern as does the treatment of the hair. Most difficult is the
strange radiate crown he wears. Authentic intaglios, intended
to have a public function as signets, do not show emperors with
a radiate crown. Admittedly, this gem, were it genuine, cannot
be considered strictly as a signet given its large size. However,
the radiate crown, with its straight base combined with ties at
the back of the neck, is a strange hybrid form and looks more
like the misunderstood creation of an 18th-century gem
Plate 21a Amethyst intaglio of Gallienus, Plate 21b Cast of Plate 21a
H. 44mm. London, British Museum, GR
1925,07015.3
170 | ‘Gems of Heaven’
Plate 20 Carnelian intaglio (cast) showing Philip I, Philip II and Otacilia Severa of
post-medieval date, 31.8mm x 24mm. Formerly in Berlin
engraver. Radiate crowns without neck ties are encountered
very rarely on the minor arts and have a straight base; where
this item of headgear does have neck ties, as it does on all of the
antoninianus and aurelianianus coin issues of the 3rd century
ad, the base of the crown curves (Pl. 22). On a gem of this type
the addition of ties at the back of the neck to a radiate crown
with a straight base strongly implies that the gem engraver was
intent only on producing a large and eye-catching gem for a
collector’s cabinet and had misunderstood the conventions of
3rd-century ad imperial portraiture.
Smaller gems, mainly intaglios, also offer problems. An
amethyst from the Lippert cast collection and also represented
as a glass paste in Würzburg shows a draped and cuirassed
(seen from the rear) bust of Gallienus (Pl. 23).54 The addition of
berries to the laurel wreath, whilst a pleasing touch
aesthetically, does not do anything to make this gem appear
authentic.
There are many gems with even more unusual portrait
types and these are invariably of dubious authenticity. A
carnelian in the Bibliothèque nationale, Paris, shows a bust of
Septimius Severus, with bare chest, an aegis fixed on one
shoulder by a balteus and with a spear pointing forward (Pl.
24).55 The bust type is taken from that emperor’s medallions
and, whilst the medallions are genuine, the intaglio is certainly
not.56
A three-quarter facing and radiate bust of the Emperor
Plate 22 Base silver issue of Gallienus (mint of Antioch)
with radiate crown as shown on the coinage, AD 260–8.
Private collection
 Overtones of Olympus

Plate 23 Glass paste of Gallienus of post- Plate 24 Carnelian intaglio of Septimius Severus of Plate 25 Red jasper intaglio of Postumus of
medieval date, 16 x 13mm. MVWM post-medieval date. Paris, Bibliothèque nationale post-medieval date. Ex-Christie’s, New York

Postumus on a red jasper intaglio offered for sale in America
some ten years ago (Pl. 25),57 is a facsimile of a portrait
surviving on two aurei, one in Munich and the other in Paris.58
Once more, the piece is not ancient but rather comparatively
modern work.
A nicolo engraved with a fine portrait of the Gallic Emperor
Marius is surely out of place (Pl. 26).59 No other authentic gems
survive from this period and it would be strange if the only one
that did depicts an ephemeral emperor whose brief bid for the
purple occurred in an era where imperial intaglios appear to
have died out. The laureate and draped bust, although a
perfectly acceptable image for an intaglio gem, has been copied
from Marius’ coinage.
A carnelian in the British Museum has been variously
identified as Postumus or Claudius Gothicus (Pl. 27).60 The
lionskin hood would suggest an identification with Postumus
as would the rather retroussé nose. The portrait appears to
derive from various lionskin-hooded and lionskin-draped coin
issues struck for Postumus and Gallienus as each waged a form
of propaganda war against the other in the ad 260s.61 However,
the style of cutting is unconvincing and, as noted above, the
quasi-divine image is strongly at odds with what would be
expected for a genuine imperial portrait intaglio. Whilst it is
possible that this crude intaglio was carved for use as a signet
by someone in the emperor’s service, it is more likely that it is a
much later adaptation of one of the coins themselves.
A group of very similar intaglios offer an interesting subject
for discussion. These carry confronting portraits that appear to
represent Carinus and his wife Magnia Urbica. The author has
speculated elsewhere that this series of gems were produced
for distribution at the celebration of Carinus’ marriage with
Urbica in August ad 283.62 One, a red jasper in the British
Museum, shows a cuirassed and laureate bust of an emperor of
the late 3rd century ad facing a draped female bust with a
hairstyle placing her in the same period (Pl. 28).63 The other
two gems, however, one a carnelian now in the Hermitage,
St Petersburg (Pl. 29),64 and the other a nicolo last recorded as
being in the Kunstmuseum, Bonn (Pl. 30),65 raise questions
about the authenticity of the group as a whole.
Both gems have the female bust resting on a crescent. This,
the attribute of the goddess Luna, the counterpart of Sol the
sun god, is a standard portrait feature of empresses as depicted
on the antoninianus coinage and, where encountered, the
corresponding male busts might be expected to wear the
radiate crown. As has been noted above, these features do not
occur on genuine 3rd-century ad intaglios and the presence of
the crescent here is a cause for concern. It is most probable that
this group, engraved in a very similar style despite the different
gemstones used, was the product of a post-Roman workshop.
The lack of any close numismatic comparanda, is telling. There
is a small gold medallion in Gnecchi’s medallion corpus with a
cuirassed bust of Carinus on one side and a draped bust of

Plate 26 Nicolo intaglio with a portrait of Marius, Plates 27a-b Carnelian intaglio of Postumus (with cast) of post-medieval date, 15mm x 12mm. London,
20mm x 17mm. Paris, Bibliothèque nationale British Museum, GR 1867,0507.322

‘Gems of Heaven’ | 171

Marsden
Plate 28 Red jasper intaglio (cast) depicting Plate 29 Carnelian intaglio of Carinus and Magnia
Carinus and Magnia Urbica but probably of post- Urbica of post-medieval date. St Petersburg,
medieval date, 15 x 12mm. London, British Hermitage
Museum, GR 1867,0507.541
Urbica on the other, but this is not the same iconographic
composition and indeed is the sort of piece that might have
encouraged a neoclassical gem engraver to place obverse and
reverse sides of the medallion on the one side of an engraved
gemstone.66
Few cameos, false or genuine, exist that depict emperors
from the period after the collapse of the Severan dynasty. A
large cameo in Paris, described as representing Gallienus as
Minerva is surely utterly wrong as a 3rd-century ad product
(Pl. 31; see also Sena Chiesa this volume, Pl. 8).67 Various
arguments have been made, mainly on the basis of a strange
gold issue with the obverse legend gallienae avgvstae,
produced at Rome around ad 266 (and later adapted at the
mint of Siscia), that Gallienus sought to identify himself with
Kore/Demeter (Pl. 32).68 These arguments have now been
generally refuted. It was perhaps a logical step, once Gallienus
had been seen as having identified himself with one female
point of reference, to look for (or invent) other cases where he
did the same. Gallienus, with his unusual coin types, often
enigmatic to say the least, may have been more than usually
Plate 31 Cameo of Gallienus as Minerva of post-medieval date. Paris,
Bibliothèque nationale
172 | ‘Gems of Heaven’
Plate 30 Nicolo intaglio of Carinus and Magnia
Urbica of post-medieval date. Bonn,
Kunstmuseum
aware of his feminine side but it is going too far, surely, to
believe that he would have depicted himself in the guise of
Minerva on an object that is, in view of its size, best viewed as a
Staatskameen. The scale-armoured bikini, worn by a Gallienus
who appears to have strenuously applied a modern Gillette
razor to his chin, owe more to the fantasies of a post-medieval
cameo engraver wishing to invent something unusual for his
latest client. Even when it was supposed that Gallienus was
playing the part of Kore/Demeter on the aurei mentioned
above, he never wore an armoured bikini nor divested himself
of his beard. Why do so for the engraver of a large cameo that
would have provoked, even with a recipient who was most
aware of Gallienus’ obscure propaganda, at best titters and at
worst outright laughter?
It is best, having considered the extant glyptic material,
both false and genuine, to return to a consideration of the
reasons behind the decline of imperial gem portraits. A number
of changes may have led to this decline. One of the most
important concerns the sort of gifts that the emperor gave to
imperial servants. It is most probable that the cameos of the 1st
and 2nd centuries ad were produced as tokens of imperial
favour, luxury objects that could be worn or displayed to
advertise that favour in the same way as the large precious
metal crossbow brooches of the 4th century ad were
distributed to those highly placed in the emperor’s service.
That imperial portrait cameos die out in the early 3rd century
ad must be significant. It was not the case that emperors
stopped giving gifts, only that the nature of those gifts
changed.
Plate 32 Aureus of Gallienus with obverse legend of GALLIENAE AVGVSTAE,
c. AD 266. London, British Museum, CM 1864,1128.131

Plate 33 Bronze medallions of Antoninus Pius (AD138–61), one framed the
other unframed, D. 72mm and 42mm. Trier and Vienna
Plate 35 Bronze medallion of Julia Mamaea (AD 222–35), D. 37mm. London,
British Museum, CM 1872,0709.413
Medallions
A moving towards imperial gifts with an intrinsic value is
surely implied by the production of gold and silver medallions
and it is most noteworthy that these objects both began to carry
images of the imperial family that had previously been
confined to the realms of cameos and silver plate and began to
appear in the period when imperial portrait cameos were dying
out.
Indeed, an increasing blurring of the boundaries between
certain categories of the minor arts is suggested by particular
objects such as two thin silver discs in the Bibliothèque
nationale, Paris, embossed with repoussé portraits of Gordian
III and set within bronze frames.69 These strange items
represent what is best described as a hybrid between
medallions and silver plate. Examples of identical 2nd-century
ad bronze medallions are known where one medallion has
been set into a large bronze frame and the other has not, such
as two specimens of Antoninus Pius struck from the same pair
of dies (Pl. 33).70 True medallions from the 3rd century ad blur
the boundaries even further.
Many of these medallions are in bronze but with elaborate,
quasi-divine portraits which implies they were intended to
function as gifts, albeit gifts with no great intrinsic value. Their
images recall those on the earlier Staatskameen whilst the fact
that they were struck from the same dies used to produce
examples in gold and silver strongly suggests that they were
distributed at the same ceremonies where their more valuable
counterparts were handed out. Here we have what might be
Overtones of Olympus
Plate 34 Silver medallion of Julia Domna (AD 193–217), D. 37mm. Berlin,
Staatliche Museen, Münzkabinett
seen as an extension of imperial gift-giving. Those of the
emperor’s servants who would, in an earlier era, have been
given a cameo could now be given a gold medallion with
similar imagery whilst identical gifts in bronze would serve as
marks of imperial recognition at little real cost. The
psychological effect of the gift of what was in effect a bronze
cameo should not be underestimated; the bond it was intended
to form between emperor and subject, in an era where imperial
presents were openly displayed, was hardly meaningless.
Whilst the larger gold medallions were surely struck to
provide gifts to those high in the emperor’s service, examples
in silver and base metal presumably formed similar gifts to
those of lower status. The trend towards the appearance of
quasi-divine images on silver and copper-alloy medallions is
exemplified by two specimens dating to the years of the
Severan dynasty.
A large silver medallion in the name of Julia Domna in
Berlin is instructive (Pl. 34).71 Here a half-length portrait of
Julia Domna holds a cornucopia and a statuette of Concordia.
The large size of the medallion would almost have placed this
piece in the realm of small objects of silver plate and its
imagery is certainly more in keeping with the iconography
encountered on that medium than with the relatively plain
imperial busts which had hitherto been the norm for
medallions. Similar examples survive in bronze, evidence for a
distribution of these pieces in more than one metal.72
A copper-alloy medallion in the British Museum struck for
Julia’s niece, Julia Mamaea, mother of the Emperor Severus
‘Gems of Heaven’ | 173
Marsden
Plate 36 Silver medallion of Plate 37 Bronze medallion of
Gallienus, mint of Milan (AD 253– Gallienus (AD 253–68), D. 41mm.
68), D. 38mm. Milan Rome
Alexander, demonstrates the increasingly exaggerated images
being applied to members of the imperial family (Pl. 35). Her
half-length bust is figured with the wings of Victory, the lotus
of Isis, the crescent of Diana (or Luna), the cornucopia of
Abundantia, the caduceus of Felicitas and the torch of Ceres, as
Toynbee notes ‘a veritable epitome of the syncretistic
tendencies of third-century religion’.73 This is most likely a
survivor in bronze of a medallion issue that would also have
included examples in gold and probably silver. The examples in
precious metal would have been powerful statements of the
way the later Severan dynasty viewed themselves; as it is, the
bronze medallions are not much less eloquent.
Later in the 3rd century ad imperial medallions continue
the trend exhibited by these two pieces. It is not the place to list
every specimen but a few examples will serve by way of
illustration.
A silver medallion of Gallienus struck at Milan around the
end of his joint reign with his father Valerian shows his
upward-gazing portrait wearing a cuirass and with a fillet,
symbol of Macedonian royalty, bound about his head (Pl. 36).74
The armour he wears, evidenced by the pteruges at his
shoulders, appears unusual, as if there has been a conscious
attempt at making it look deliberately archaic. Here the
emperor is portrayed as a new Alexander, an exemplar of
military glory but also an autocrat and tyrant.
Many medallions portray either Gallienus or his Gallic rival
Postumus as Hercules, another demigod famed for his courage.
Commodus had employed similar comparisons on his
medallions and coins but whereas such outré behaviour can be
ascribed in his case to a madness that led to his assassination,
in the case of Gallienus and Postumus it is best viewed as a
symptom of the way imperial iconography was heading. A
range of examples in gold and bronze survive.75
A half-length bust of Gallienus on a bronze medallion, has
the emperor clad in a himation and holding a caduceus over his
shoulder (Pl. 37).76 Whilst the caduceus, when carried by
empresses, signifies equation of the bearer with Felicitas, when
it is grasped by an emperor it surely means an association with
Mercury.
174 | ‘Gems of Heaven’
Plate 38 Bronze medallion (cast) of Probus (AD 276–82), D. 37mm. Paris,
Bibliothèque nationale
What is striking about medallions from the later 3rd
century ad in general is the proliferation of portraits that,
whilst not explicitly quasi-divine in nature, are certainly
overtly martial or monarchical in the images they disseminate.
Those of Probus are particularly noteworthy here. Examples
feature the emperor clad in various ways, for example
heroically nude wearing only an aegis and grasping a spear or
wearing a cuirass and holding a shield and spear (Pl. 38).
Other specimens depict his half-length portrait standing jugate
with the sun god Sol. These are the sorts of images that were
earlier confined to cameos and their widespread appearance
on base metal medallions must surely be significant if not least
for what this appearance says about the spread of these portrait
types across the range of the numismatic and semi-numismatic
media.
An interesting class of medallions which came into
prominence in the 3rd century ad is that comprising examples
with a Moneta reverse depicting the Tres Monetae, the Three
Moneys. It seems likely that these were connected with the
imperial donativa, the depiction of the Tres Monetae
underlining that function. A number of these carry explicitly
military portraits, particularly those from the reign of Probus,
perhaps implying that by his time the appearance of these
images was standard on Moneta medallions.
Coins
Having considered medallions, the state coinage should also be
subjected to study. The years covering the sole reign of
Gallienus, his successor Claudius II and their Gallic rivals,
Postumus, Victorinus and the Tetrici, have often been decried
as representing the nadir of Roman coinage. However, it is in
precisely this period that we see the imagery previously
confined to cameos being transferred to these objects. It is
most appropriate to begin with the gold coinage since this is
the medium in which we might expect to see the earliest
intrusion of cameo-like images.
Plate 39 Silvered bronze coin with Janiform portrait of Gallienus. London,
British Museum, CM 2000,0402.1

Plate 40 Light aureus (cast) of Gallienus (mint of Milan) with club-bearing
portrait. Paris, Bibliothèque nationale
A piece discovered a few years ago in Hertfordshire, which
surely represents a striking in silvered bronze using the dies
produced to create gold aurei, shows a Janiform portrait in
which each of the god’s two faces has the features of Gallienus
(Pl. 39).77 The typology of this interesting object is taken from a
coin of the mid-Republic of some five centuries before, a
fascinating example of a very contemporary imperial dressing
up combined with a remarkably antiquarian re-use of the
images from one of Rome’s earliest coins. The placement of
Gallienus’ physiognomy on the bust of Janus offers a striking
statement about the emperor’s quasi-divine pretensions. When
there was peace across the Empire the doors of Janus’ temple
were closed. Whilst there was certainly not peace across the
Empire when this coin was struck the message is clear; that
Gallienus/Janus confidently believed that peace could and
would be brought about.
Other gold coins displayed other images that made explicit
statements about the emperor’s quasi-divine pretensions. Here
the emphasis was often on the heroic courage that would bring
about a state of peace and prosperity in the increasingly
troubled years of the ad 260s. A so-called light aureus struck
for Gallienus at Milan displays the emperor in laurels and
carrying the club of Hercules over his shoulder (Pl. 40). The
imagery is identical to that encountered on the medallions just
discussed and represents a reproduction in miniature of the
sort of iconography that had previously been confined to the
realm of cameos and silver plate.
The aurei of Postumus, Gallienus’ rival in the western
provinces, and the associated strikings from the same dies in
bronze, termed abschlage, with their jugate portraits showing
the emperor alongside one of his divine Comites, represent a
Plate 41 Aureus of Postumus (cast) with jugate portraits of Postumus and
Hercules on one side and Pax and Victory on the other. Oxford, Ashmolean
Museum
Overtones of Olympus
Plate 42 Abschlage of an aureus of Postumus (cast) with jugate portraits of
Postumus and Hercules on one side and reference to the labours of Hercules on
the other. Oxford, Ashmolean Museum
similar case (Pl. 41). Once again, the viewer is reminded of the
cameo portraits of an earlier age.
Some of the reverses of these pieces depict Postumus/
Hercules in the course of the 12 labours; coins, unlike a cameo,
have two sides and can extend upon the message of one face in
a way that most engraved gems cannot (Pl. 42). It might be
speculated that this was yet another reason for the death of the
imperial portrait cameo; quite apart from the cost of producing
engraved gemstones, discs of metal stamped between two dies
could not only be used as bullion but could tell a much longer
tale. The juxtaposition of Postumus with Hercules on the
obverse or Postumus dressed as Hercules was then combined
with a reiteration of one of the labours of Hercules on the
reverse. The implicit statement on one side of the coin was
reinforced by an explicit image on the other.
What need, indeed, for finely engraved cameos as gifts to
loyal supporters of the Gallic emperors when their gold
coinages and accompanying medallic issues were so gem-like
in their imagery and had two sides to boot?
Nor was this trend confined to the gold coinage. We have
seen how medallions began to usurp the images hitherto
confined to cameos. It was not long before the radiate coinages
of Gallienus, Postumus and their successors began in their turn
to usurp the iconography that had relatively recently appeared
on medallions and issues in gold.
Early examples of this phenomenon appeared on the base
silver issues of Gallienus and Postumus where the lionskin-clad
and club-wielding portraits that had not long before been seen
on medallions began to filter down onto the coins that formed
the lowest value issues produced by official mints (Pl. 43).78 By
the ad 270s the coinages of Probus were not only showing him
Plate 43 Base silver radiate (cast) of Postumus with a club-wielding portrait.
Oxford, Ashmolean Museum
‘Gems of Heaven’ | 175
Marsden
Plate 44 Gold necklace with coin pendant of Plate 45 Gold aureus of Postumus mounted in a
Gordian III. Ortiz Collection gold ring. D. (bezel) 21mm. London, British
Museum, CM 1998,1202.1
in a wide range of overtly militaristic and monarchical poses,
but some examples were even styling him Deus and Dominus,
‘God’ and ‘Lord’. When portrait types that had previously
appeared on cameos, imperial gifts of great prestige, were now
being disseminated on base metal coin issues and which
carried, furthermore, legends that explained in words their
visual images, it is surely plain that a great change had taken
place.
Conclusion
This paper has tried to investigate the reasons behind this
change. Plainly, part of this development is only to be expected
given the increasingly autocratic style of emperors in the
3rd century ad, and is merely a symptom of the change from
Principate to Dominate. Other reasons are more intimately
connected with the types of gifts that emperors gave to their
supporters as well as the very use of signet rings themselves.
It is perhaps apt here to comment on the numbers of
gemstones, some very finely-engraved indeed, that have been
reset into disc brooches of the mid-or late 3rd century ad. It is
surely the case that if these objects were still desired as
ringstones then such treatment would not have taken place.
Signet use had plainly become a thing of the past. Imperial
portrait intaglios were part of this whole system and so it
would seem that their use too became redundant.
The 3rd-century ad aesthetic is again surely of relevance
here as is also the economic situation of the period. When
emphasis was placed not on the artistic virtuosity inherent in
engraved gemstones but on the display of gold jewellery set
with brightly coloured gems, pieces of glass or indeed gold
coins, it should not surprise us that intaglios engraved with
imperial portraits became obsolete. Almost without exception
those intaglios that do survive are either mass-produced
objects in moulded glass or else are carved in such a crude and
schematic style that it is difficult to believe them to be the
products of imperially-sponsored workshops. The death of the
imperial portrait gem in the later years of the 3rd century ad is
a tragedy for students of imperial glyptics; we should not,
however, allow this tragedy to blind our judgement to what is
surely staring us in the face. Those surrounding the emperor
and those who carried out the emperor’s business far away
from his person did not want gifts of gemstones bearing his
portrait. What the emperor’s servants wanted, when emperors
came and went so quickly, they generally got. Titus Sennius
176 | ‘Gems of Heaven’
Plate 46 Gold aureus of Licinius II. Private
collection
Sollemnis advertised his good fortune by being paid in gold in
the ad 220s; that he should do so proves it was an aspiration
held by many.
Here we should consider the mounting of coins and
medallions in precious metal. These were often old coins that
had been placed into mounts in what we might see as nothing
more than a rather vulgar means of displaying personal
wealth. Certainly, as I have argued previously, the resulting
products of this rather haphazard setting of coins issued many
years before where ‘good’ emperors are indiscriminately mixed
with ‘bad’ ones can hardly be described as ideological. Usually
there are no coins that were recently issued at the time of
mounting; even in cases where there are then the placement of
a coin of the current emperor alongside one of a reviled tyrant
cannot be understood as imperially-commissioned propaganda
or even as the behaviour of people close to the emperor wishing
to make a propagandistic point.
Many of these chains belong, furthermore, to the mundus
muliebris, the realm of women. This is borne out by the fact
that the only painted representations of these rather gaudy
objects appear on the portraits of females found painted on
sarcophagi in Egypt; these women were themselves unlikely to
have received the coins as gifts from emperors. Yeroulanou
seeks to link two of Postumus’ aurei, mounted in openwork
pendants of different style, as ‘part of a necklace made in ad
263, probably as a gift for the emperor’s 50th birthday’.79 This is
not likely to have been the case; if it was, then one would
expect the settings to have been identical, before even
beginning to speculate on why Postumus might have
considered wearing such an item of jewellery in the first place.
Sometimes, however, the medallions or coins were plainly
new at the time of mounting and were mounted in settings
more appropriate to men. A large gold medallion of Gordian III,
set into a pendant and now in the George Ortiz collection,
could feasibly have functioned as a male ornament (Pl. 44).80
An aureus of Postumus mounted in a gold ring found at
Poringland in Norfolk in 1998 is also of note here (Pl. 45). The
ring is attractive but not of the best quality. The impression one
forms is of a recipient of imperial largesse having the gold coin
he had received as a mark of imperial favour mounted as a way
of displaying that favour. Most significantly, one might
consider the series of rings of the very early 4th century ad
where gold coins were mounted into gold rings and carry
inscriptions that strongly imply they were imperial gifts. The

Poringland ring and others like it, such as an example in the
British Museum set with an aureus of Diocletian,81 might be
seen as privately-commissioned anticipations of the later
system where rings did form one sort of gift within the realm of
imperial donativa.
Here we might also mention other gifts that contain coin-
like settings such as the series of dishes produced for the Licinii
on the occasion of particular anniversaries.82 Those carrying
portraits of Licinius II, for example, issued to celebrate the
prince’s Quinquennalia, bear facing, coin-like images that are
identical to those found on that emperor’s gold coins (Pl. 46).
They were produced in the same workshops as those gold coins
and for a very similar group of recipients.
To return to the Poringland ring. Displaying a gold coin in
one’s ring was an understandable feature of the time and one
which now, when the price of gold is higher (at the time of
writing this article) than ever, we should certainly appreciate.
It seems that, by the second half of the 3rd century ad, this was
a more appropriate way of displaying the imperial image than
by carrying a signet ring bearing the emperor’s portrait
engraved upon it. Gold had a value that pieces of semi-precious
stone did not; by setting a gold coin into a ring one could not
only continually appreciate the fact that one was important
enough, like Titus Sennius Sollemnis, to be paid in gold, but
could also advertise that fact to everyone one came across. In
the increasingly showy epoch of the later Roman Empire this
was surely what mattered. That the trend was recognised and
was soon crystallised in the issue of strikingly similar objects a
few years later, objects that were part of a rigidly-produced
system of imperial gift-giving, must surely demonstrate the fact
that emperors had got the message. There are no 4th-century
ad ringstones because they were no longer appropriate. There
are so few, if any, from the second half of the 3rd century ad
because, in this instance, there was an anticipation of the
coming era when fashions and cold realities would run hand in
hand. The age of the imperial portrait gem, whether executed
in cameo or intaglio, had gone; imperial portraits now came
stamped on discs of metal, to furnish the pay of those whose
support the emperor sought.
Notes
1 J. Boardman, ‘Roman gems: problems of date and identity’, in
L. Gilmour (ed.), Pagans and Christians – from Antiquity to the
Middle Ages, BAR International Series 1610, Oxford, 2007, 75–7.
2 J.P. Robinson and T. Opper, ‘Swanley, Kent: medieval silver seal
matrix’, Treasure Annual Report 2005/6, 2005, 128, no. 536; see
also, M.E. Henig, ‘The re-use and copying of ancient intaglios set in
medieval personal seals, mainly found in England: an aspect of the
renaissance of the 12th century’, in N. Adams, J. Cherry and J.
Robinson (eds), Good Impressions: Image and Authority in Medieval
Seals, British Museum Research Publication no. 168, London,
2008, 27–8.
3 A.B. Marsden, ‘Imperial portrait gems, medallions and mounted
coins: changes in imperial donativa in the 3rd century ad’, in
M. Henig and D. Plantzos (eds), Classicism to Neo-classicism, Essays
dedicated to Gertrud Seidmann, BAR International Series 793,
Oxford, 1999, 89–103.
4 A.B. Marsden, ‘Between Principate and Dominate: imperial styles
under the Severan dynasty and the divine iconography of the
imperial house on coins, medallions, and engraved gemstones ad
193–235’, Journal of the British Archaeological Association 150
(1997), 1–16.
5 W.R. Megow, Kameen von Augustus bis Alexander Severus, Berlin,
1987, 270, no. B52, pls 46–8.
6 Ibid., 239–40, no. A143, pl. 48.
Overtones of Olympus
7 H.B. Walters, Catalogue of the Engraved Gems and Cameos, Greek,
Etruscan and Roman in the British Museum, London, 1926, 340, no.
3165, pl. XLII.
8 O. Neverov, Antichnye kamei v sobranii Ermitaza, Leningrad, 1988,
no. 270.
9 Sternberg Fixed Price List no. 10, June 1998, no. 723.
10 E. Zwierlein-Diehl, Antike Gemmen in Deutschen Sammlungen.
München, I.3, Munich, 1972, no. 2815.
11 M.E. Henig, A Corpus of Roman Engraved Gemstones from British
Sites (3rd edition), BAR British Series 8, 2007, 182, no. 747, and 201,
no. 70 (Appendix).
12 Ibid., 182, no. 748.
13 F.H. Marshall, Catalogue of Jewellery, Greek, Etruscan and Roman
in the Department of Antiquities, British Museum, London, 1911,
350–1, no. 2944.
14 Bonham’s, Knightsbridge, 10/6/1997, 9, no. 36.
15 M.L. Vollenweider, Musée d’Art et d’Histoire de Genève: Catalogue
raisonné des sceaux, cylindres, intailles et camées, vol. II, Mainz,
1979, 247, no. 259, pl. 80.
16 Ibid., 250, no. 262, pl. 80.
17 Ibid., 248–9, no. 261, pl. 80.
18 E. Zwierlein-Diehl, Die Antiken Gemmen des Kunsthistorischen
Museums in Wien III, Munich, 1991, 229, no. 2515, pl. 160.
19 I. Popović, Les camées romaines au Musée Nationale de Beograd,
Belgrade, 1989, 78–9, no. 51.
20 Sternberg (n. 9), no. 689.
21 Unpublished; pers. comm.
22 Marsden (n. 3), 89–103.
23 G.M.A. Richter, Engraved Gems of the Romans, London, 1971, 119,
no. 587.
24 L.E. Marzatico and F. Marzatico, Ori delle Alpi, Trento, 1997, 494,
no. 1183, fig. 111.
25 O. Neverov, Antique Intaglios in the Hermitage Collection,
Leningrad, 1976, 79, no. 141.
26 J. Tassie and E. Raspe, A Descriptive Catalogue of a General
Collection of Ancient and Modern Engraved Gems, Cameos and
Intaglios, Taken from the Most Celebrated Cabinets in Europe; and
Cast in Coloured Pastes, White Enamel, and Sulphur, London, 1791,
no. 12067.
27 A. Furtwängler, Die Antiken Gemmen, Leipzig and Berlin, 1900,
no. 25, pl. XLVIII. Tassie and Raspe (n. 26), no. 12094.
28 S.H. Middleton, Engraved Gems from Dalmatia, Oxford, 1991, 116,
no. 208; Richter (n. 23), no. 586.
29 Richter (n. 23), 119, no. 588; see also, Walters (n. 7), 212, no. 2028,
pl. XXV.
30 H. Mattingly, E.A. Sydenham and C.H.V. Sutherland, The Roman
Imperial Coinage IV.3, London, 1949, 123, no. 30.
31 J.M.C. Toynbee, Roman Medallions, New York, 1944, 81.
32 A. Krug, ‘Antike Gemmen im Römisch-Germanischen Museum,
Köln’, Berichte der Romische-Germanischen Kommission 61 (1980),
244, no. 389.
33 M. Schlüter, G. Platz-Horster and P. Zazoff, Antiken Gemmen in
deutschen Sammlungen, Band IV, Hannover Kestner-Museum,
Hamburg Museum für Kunst und Gewerbe, Weisbaden, 1975, 293–4,
no. 1599; also, P. Zazoff, Die Antiken Gemmen, Munich, 1983, 327,
pl. 99,4.
34 For example, H. Mattingly and E.A. Sydenham, The Roman
Imperial Coinage Volume IV. I, London, 1936, 120–2, nos 226–47;
see also, Marsden (n. 4), 5–10.
35 Zwierlein-Diehl (n. 18), 71–2, no. 1730, pl. 20.
36 Zwierlein-Diehl (n. 10), 59, no. 2459, pl. 223.
37 H. Mattingly, E.A. Sydenham and C.H.V. Sutherland, The Roman
Imperial Coinage Volume IV.2, London, 1938, 169–76, nos 2, 12–13
(Balbinus) and 2, 12 (Pupienus).
38 K. Jazdzewski, Poland, London, 1965, 231, no. 54, pl. 54.
39 Mattingly et al. (n. 30), 41, no. 247A.
40 Vollenweider (n. 15), 260–1, no. 272, pl. 84.
41 C.B. Welles, R.O. Fink and J.F. Gilliam, Excavations at Dura-
Europos V/1, New Haven, 1958, no. 28, pl. LXXI.
42 M. Henig, ‘Roman Sealstones’, in D. Collon (ed.), 7000 Years of
Seals, London, 1997, 95.
43 Classical Numismatic Group, mail bid sale no. 53, 15/3/2000, 200,
no. 1675.
44 See n. 26.
45 E. Zwierlein-Diehl, Glaspasten im Martin-von-Wagner Museum der
Universität Würzburg, Munich, 1986, no. 13; P.D. Lippert,
‘Gems of Heaven’ | 177
Marsden
Dactyliotheca III, Leipzig, 1787, no. 343.
46 Walters (n. 7), no. 2039, pl. XXVI, and B.M. Felleti-Maj,
Iconographia Romana Imperiale 222–285, Rome, 1958, 255.
47 Zwierlein-Diehl (n. 45), 264, no. 794.
48 Lippert (n. 45), II, no. 838.
49 U. Pannuti, Museo Archeologico Nazionale di Napoli. La Collezione
Glittica, Rome, 1994, 252–3, no. 213; Tassie and Raspe (n. 26), no.
12075 and Lippert (n. 45), II, no. 845.
50 Cades no. 597 (The Cades Collection of casts in the German
Archaeological Institute, Rome).
51 Furtwängler (n. 27), 231, no. 31, pl. XLVIII; Tassie and Raspe (n. 26),
no. 12082; Lippert (n. 45), II, no. 848.
52 R. Delbrueck, Die Munzbildnisse von Maximinus bis Carinus,
Berlin, 1940, 110.
53 M.-L. Vollenweider and M. Avisseau-Broustet, Camées et intailles.
Tome II : Les portraits romains du Cabinet des médailles: catalogue
raisonné, Paris, 2003, no. 2097 (Commodus) and no. 2101
(Caracalla).
54 Lippert (n. 45), III, no. 333; Zwierlein-Diehl (n. 45), 263, no. 793.
55 Vollenweider and Avisseau-Broustet (n. 53), no. 206 (Septimius
Severus).
56 F. Gnecchi, I Medaglioni Romani II, Milan, 1912, 75, nos 20–1, pls
93.9–10.
57 Christie’s, Ancient Jewelry, New York, 8/12/1999, 52, no. 119.
58 B. Schulte, Die Goldprägung der Gallischen Kaiser von Postumus bis
Tetricus, Aarau, Frankfurt and Salzburg, 1983, 101, no. 104a, and
112, no. 138a.
59 Richter (n. 23), 119, no. 589.
60 Walters (n. 7), 212, no. 2029; Richter (n. 23), 119, no. 590. For a
differing opinion on this intaglio,see Zwierlein-Diehl, this volume,
Pl. 34, who considers it to be genuine.
61 A.B. Marsden, ‘Some sing of Alexander and some of Hercules:
178 | ‘Gems of Heaven’
artistic echoes of Hercules and Alexander the Great on coins and
medallions, ad 260–269’, in Gilmour (n. 1), 65–74.
62 Marsden (n. 4), 89–103.
63 Walters (n. 7), 213, no. 2031, pl. XXV; R. Delbrueck, Antike
Porphyrwerke, Berlin and Leipzig, 1932, 124, pl. 57b; Richter (n. 23),
no. 592.
64 Neverov (n. 25), 80, no. 142; Delbrueck (n. 63), 126, pl. 58a.
65 Delbrueck (n. 63), 124, pl. 57c; Tassie and Raspe (n. 26), no. 12104.
66 F. Gnecchi, I Medaglioni Romani I, Milan, 1912, 10, pl. 4.
67 Delbrueck (n. 52), 110, pl. 4. For a different opinion, see Sena
Chiesa, this volume, Pl. 8 who considers it to be 1st century ad.
68 J.P.C. Kent, ‘Gallienae Augustae’, Numismatic Chronicle Series 7,
vol. XIII (1973), 64–8.
69 Bibliothèque nationale, Paris, Medallions, Gordian III, 30.
70 Toynbee (n. 31), 93, pl. XV, nos 5–6.
71 Ibid., 148, pl. XLIV, 1.
72 Gnecchi (n. 66), 76–7, nos 6–8, pls 94–5, 9–11.
73 Toynbee (n. 31), 158.
74 Marsden (n. 61), 66.
75 Ibid., 65–74.
76 Toynbee (n. 31), pl. XLVI, 5.
77 R. Abdy, ‘A new coin type of Gallienus found in Hertfordshire’,
Numismatic Chronicle 162 (2002), 346–50.
78 Marsden (n. 61), 65–74.
79 A. Yeroulanou, Diatrita, Athens, 1999, 221, no. 103.
80 In Pursuit of the Absolute. Art of the Ancient World from the George
Ortiz Collection, Royal Academy of Arts, London, 20 January–6
April 1994, London, 1994, no. 238.
81 J.P.C. Kent and K.S. Painter, Wealth of the Roman World, ad 300–
700, London, 1977, 27, no. 16.
82 Ibid., 20, nos 1–3.
Love and Passion
Personal Cameos in Late Antiquity from the Content Collection
Helen Molesworth and Martin Henig
The majority of engraved gems in collections or, indeed,
recovered from Roman sites are intaglios intended for the
practical purpose of sealing correspondence or confirming
wills and contracts of all kinds. These are mainly datable down
to the beginning of the 3rd century ad when a definite decline
in seal usage appears to have taken place; this decline affected
the quality of the intaglios produced as well as the quantity.
The rather rare high quality intaglios recorded, for instance
those bearing the images of emperors which must have served
official purposes (see Zwierlein-Diehl and Marsden, this
volume), do not change the general profile of intaglio usage.
Cameos were invented only comparatively late in the
history of glyptics, during the Hellenistic age and, from the
beginning, they were designed for display either as miniature
objets d’art or as settings in jewellery including rings, bracelets,
earrings, brooches and pendants. They continued to be cut
right through the Roman period and appear to have been at
their most numerous in later antiquity, especially in the 3rd
century ad. The only cameos considered in most general
histories of Roman art, and even in some books about gems,
are those which refer, frequently allegorically, to the political
history of the Empire. These so-called Staatskameen include:
the Augustan Gemma Augustea depicting Augustus with Roma
triumphant over the barbarians; the Severan cameos depicting
Julia Domna as a winged Victory in Kassel, and as Juno
Caelestis in London; the Constantinian cameo in the Dutch
Royal Collection showing the victorious emperor in a chariot
pulled by centaurs trampling down the supporters of
Maxentius; and another cameo in Belgrade depicting a battle
scene (see Krug this volume). Such gems record, sometimes in
metaphorical form, the activities of the ruler.1
However, very much more numerous and of equal interest
for the history of society are the private cameos generally given
by men to girlfriends, mistresses, fiancées and wives. There are
certainly a few cameos related to Staatskameen including fine
portraits of the young Nero (Pl. 1)2 and part of a large gem
Plate 1 Three-layer sardonyx cameo of the young Emperor Nero, laureate. 1st
century ad, 30.3 x 28.0 x 5.2mm. Content I, no. 59
which may portray two Tetrarchic emperors greeting one
another,3 but most are private, the majority being love-tokens.
Such cameos did not begin to be employed at the self-same
moment that intaglios went out of favour nor were they
confined to glyptic art. Love-tokens in various materials can be
dated both before this watershed and later and a selection from
the province of Britannia has already been assembled by
Martin Henig.4 We have been working for some time on a
private collection of cameos assembled by Mr Derek Content.
This is probably the most important collection of its kind in
private hands and examples from it will be the basis for this
paper.5
The 3rd- and 4th-century ad cameos which can be
described as love tokens of one sort or another comprise a
discrete group. There is, however, much that we cannot know,
such as where they were manufactured. A majority of those
which bear legends are inscribed in Greek; those in Latin are
rarer. This would suggest that they were mostly made in the
Greek-speaking provinces although Greek was the more
refined language and conveyed nuances better. Relatively few
cameos have been recovered from Gaul and Britain but some of
these were surely significant to their owners. From Britain
there are only three inscribed cameos, two of them (North
Wraxall, Wiltshire, and Keynsham, Somerset) from villas; two
(Keynsham and Bradwell, near Maldon, Essex) are still set in
gold rings (the North Wraxall gem had lost its setting but one
suspects it was also set in a gold ring).6 From Gaul two fine gold
rings hold similar gems both from the rich Eauze (Gers)
Treasure, again suggesting they were the possessions of
wealthy land-owners (see Guiraud, this volume).7 However, it is
worth indicating that jet medallions cut with Medusa heads,
with portraits and, in one instance at least, clasped-hands,
were produced in northern Britain (Yorkshire) and the
Rhineland at the same period.8 The range of motifs might have
been narrower but the purpose was the same.
Plate 2 Two-layer sardonyx cameo of Leda and the swan. 3rd/4th century ad,
24.3 x 18.0 x 11.5mm. Content II
‘Gems of Heaven’ | 179
Molesworth and Henig

Plate 3 Two-layer sardonyx cameo of Apollo, Plate 4 Two-layer sardonyx cameo of a Medusa Plate 5 Two-layer sardonyx cameo of Eros
with attributes of lyre and swan, and Daphne, head. 3rd century AD, in contemporary walking right and playing a lyreor cithara within a
who is metamorphosing into a tree. 6th openwork gold mount, 31.0 x 25.0 x 11.5mm. raised border. 1st century bc/ad, 15.0 x 13.8 x
century AD. Cameo: 12.7 x 11.2mm. Content II Content I, no. 159 2.2mm. Content I, no. 97

Plate 6 Two-layer sardonyx cameo of Plate 7 Two-layer sardonyx cameo of

Eros mourning, leaning left on a female bust facing right. 3rd century ad,
downward pointed torch. 3rd century in contemporary openwork gold and
AD, in contemporary gold pendant pearl brooch mount. Cameo: 23.3 x
mount. Cameo: 12.0 x 6.5mm. Content I, 27.7mm. Content II
no. 109

Plate 8 Two-layer sardonyx cameo of two clasped right hands (dextrarum Plate 9 Two-layer sardonyx cameo of a hand pinching an ear beneath the
iunctio) between a garland and the legend ΟΜΟΝΟΙΑ. 3rd century AD, in legend ΜΝΗΜΟΝΕΥΕ within a raised border. 3rd/4th century AD, in
contemporary gold ring. Cameo: 15.3 x 14.2mm. Content I, no. 50 contemporary gold ring. Cameo: 16.1 x 12.2mm. Content I, no. 52

Inscribed cameos hold a particular interest because, in the
case of the longer examples at any rate, the legends relate to
Greek and Roman literary culture. The same might be said of
some of the gems showing scenes from myth, including some
which recall imagery popularised through Ovid’s
Metamorphoses, including Ganymede and the eagle, Leda and
the swan, and the probably much later (6th century ad) cameo
depicting Apollo and Daphne (Pls 2–3).9
But commoner are the many cameos showing Medusa, or
Gorgon, heads (to deflect the evil eye), erotes (the
personification of love), and mourning erotes (whose rather
strange yet common expression of one aspect of love may have
been regarded rather as mementi mori or protective/
apotropaic), and, motifs that we shall see more of later: female
‘portraits’ (intended to represent the beautiful girl wearing
them); hands clasped (for concord, and more specifically,
marriage); or a single hand pinching an ear (for remembrance)
(Pls 4–9).10 Such cameos, especially when very small and
roughly fashioned for service in ear-rings and the like, are in
danger of being overlooked. However it is their very ubiquity
which gives them interest and demonstrates the central part
played by such charms in the intimate relations of young
women and men.

180 | ‘Gems of Heaven’

 Love and Passion

Plate 10 Two-layer sardonyx cameo of the legend ΕΥΤΥΧΙ Plate 11 Two-layer sardonyx cameo of clasped Plate 12 Two-layer sardonyx cameo of Eros
within a raised border. Early 3rd century ad, in hands above the legend ΕΥΤΥΧΙ within a raised sitting on a rock, having stabbed the two dead
contemporary gold ring. Cameo: 5.0 x 9.5mm. Content II border. 3rd century ad, in contemporary gold Psyche butterflies before him. 1st century ad,
ring. Cameo: 12.0 x 10.5mm. Content II 14.4 x 11.8 x 2.9mm. Content II

It is the inscribed cameos which give the best indication of their
social context or purpose, by being, to some degree,
intrinsically self-explanatory. The existence of such a group of
cameos points to a more coherent purpose than that of simple
adornment: by carrying messages, albeit most frequently
generic stock messages, these gems express an existing
purpose in their own right.
The most common type is the εὐτυχι (‘Good luck’) gem (as a
stand-alone inscription, or sometimes twinned with a similar
blessing), clearly expressing the simple wish for good luck from
the giver to the wearer, more likely than not, from a lover to his
girl (Pl. 10).11 The ὁμονοια (‘Harmony’) gems, almost
exclusively bearing accompanying images of clasped hands,
and known in contemporary 3rd-century ad gold ‘wedding’
rings, signify betrothal or marriage (Pl. 8).12 One example,
however, of the clasped hands and the legend εὐτυχι, conflates
these two sentiments (Pl. 11).13 While the μνημονευε μου
(‘Remember me’) legend, more often than not inscribed
around a hand tweaking an ear, literally explains, by way of
encouragement – perhaps from a distant lover wishing not to
be forgotten – the concept, alluded to by Virgil in his Sixth
Eclogue, of memory being located in the earlobe (Pl. 9).14
Considering the amount of attention paid to love poetry, it
is a pity that the legends on cameos have been neglected by
those who approach the ancient world largely through its
literature. As one of us has already noted,15 the fact that most
inscriptions are in Greek may in part reflect a greater
sophistication among Greeks in the Empire; although equally, a
number of gems coming from villa sites in the western Latin
speaking provinces probably indicate the use of Greek amongst
the more literate upper classes of Latin-speaking society. Just
as French was spoken in Medieval (and indeed Renaissance
and Enlightenment) England by the upper echelons of society
in an official and literary capacity, so too Greek featured even
in the West as the standard language of love, literature and
official documentation on cameos in the 3rd and 4th centuries
ad.16
Aside from the Eclogue ear-tweaking allusion, there are
more substantial indications that these cameos should be read
within an elegiac framework. In imagery alone, cameos
frequently mirror the world of love elegy, from mourning and
playful erotes, sometimes in battle with a butterfly-winged
Psyche (Pl. 12),17 to explicit sexual scenes (Pl. 13)18 and more
pedestrian stock portraits intended to ‘represent’ the giver’s
beloved (Pl. 7).19 This focus on the beloved, impersonal in its
stock representation, chances to recall the literary devices
employed by such Latin poets as Catullus and Propertius,
themselves drawing on Greek prototypes, to frame their love
affairs with Lesbia and Cynthia.20
Although this comparison of gems with elegy is perhaps
somewhat superficial, with the assumption that many cameos
would have been gifts between lovers, nevertheless they are
undoubtedly indebted to the genre, recalling love motifs used
by Anacreon through to Ovid. Bacchic and erotic
representations on gems (Pl. 14)21 recall the Hellenistic
epigrammatists, such as Posidippus encouraging the drinking

Plate 13 Two-layer sardonyx cameo fragment of two lovers in bed together. Plate 14 Two-layer sardonyx cameo of Silenus sitting on a lion-skin in a cart
3rd/4th century ad, 29.6 x 24.9 x 8.3mm. Content I, no. 144 pulled by two Erotes; Psyche stands over him holding a rhyton. 1st century ad,
in 18th-century gold ring. Cameo: 19.0 x 15.0mm. Content I, no. 120

‘Gems of Heaven’ | 181

Molesworth and Henig
Plate 15 Two-layer sardonyx cameo with a Plate 16a Two-layer sardonyx cameo of a hound chasing a hare
six-line Greek inscription. 3rd/4th century over a ground line. 3rd century AD, in contemporary gold ring.
AD. 18.1 x 14.8 x 4.5mm. Content I, no. 40 Cameo: 7.2 x 11.3mm. Content II
of the ‘dewy moisture of Bacchus’ (πολυδροσον ἰκμαδα Вακχου)
so that ‘our concern may be with bittersweet Eros’
(ὁ γλυκυπικρος Єρως).22 The same poet draws on the militia
amoris device, asking love only to carry him off drunk, as if so
sober, he will draw up battle against Eros.23 One gem, carved
with a well-attested four- and, also in this collection, six-line
type teasing the owner for an unrequited love that is deserved/
suited (οὐ φιλω/μη πλανω/νοω δε/και γελω/συ φιλι με/συνφερι σοι
- ‘I do not love you and I don’t care; I know it and I laugh; you
love me and that’s your lot’) is certainly epigrammatic in style,
and in the vein of Catullus’ ‘odi et amo’ (Pl. 15).24
The most explicit literary allusions on the inscribed
cameos, however, are undoubtedly the hunting metaphor and
militia amoris motifs. In these instances, the reading of these
gems within the elegiac framework is particularly important,
as they throw light on interpreting the imagery of their non-
inscribed counterparts. A cameo depicting a hound chasing a
hare might seem to belong with numerous other images in all
sorts of media from mosaics and sculpture to the minor arts, as
illustrating the popularity of hare coursing. Perhaps it does. On
gravestones such hunt scenes have a solemn meaning, the
inevitability of death; but an almost identical cameo bearing
the legend reading ἐλαβες, ‘You are caught’, shows this to be
another instance of love-banter, akin to the hunting and fishing
metaphors in Ovid’s Ars Amatoria (Pls 16a-b).25 A simple
representation of weapons on one cameo, mounted in an
earring, would be hard to understand within the context of
Plate 17 Two-layer sardonyx cameo
of a vexillum between a military
trumpet and sword between the
legend ΠΡΟΚ/ΟΠΤΕ. 3rd century ad, in
contemporary gold and emerald
earring. Cameo: 9.7 x 7.4mm.
Content II
182 | ‘Gems of Heaven’
Plate 16b Two-layer sardonyx cameo of a
hound chasing a hare over a ground line above
the legend ΕΛΑΒΕC. 3rd century AD, 9.5 x 12.0
x 3.2mm. Content II
jewellery if we did not catch the metaphorical meaning of
‘advancing’ in love clarified by a second gem, set within an
earring and inscribed with the legend προκοπτε (‘Onwards’),
showing a certain debt to Propertius, Tibullus and Ovid’s
Amores, in particular (Pl. 17).26
Furthermore, a sleeping hound, whose significance in
funerary art would be chthonic, is here often accompanied by a
legend telling the wearer γρηγορι, to ‘Be wakeful!’ (Pl. 18),27 in a
similar vein to the ‘remember me’ of the hand pinching an ear
(Pl. 9).28 Outside the amatory realm, but nevertheless an
example of inscribed gems clarifying plain depictions, one
cameo in the collection inscribed εὐτυχι γοργονι (‘With the luck
of the Gorgon)’ neatly explains and proves the popularity and
amuletic capacity of the well-attested Gorgon cameo type in
two simple words (Pl. 19).29 Comparison may be made with the
well-known contemporary mosaic from the eponymous House
of the Gorgon at Ostia, where a Medusa head has the
accompanying Latin inscription, gorgoni bita (i.e. a Late Latin
form of ‘vita’).30
In as much as the use of Greek on cameos is explained by
their amatory (literary) or official function, there is also a
potential subtext, itself an amatory device, on the seemingly
entirely official cameos, as in the case of the προκοπτε gems.
One gem in the collection, inscribed το τισει μνημονευε
(‘Remember what you owe me’) at first glance appears to be the
equivalent of an IOU in stone, recalling a financial debt (Pl.
20).31 Yet were we to consider it as one of this group of
Plate 18 Two-layer sardonyx
cameo of a sleeping hound
above the legend ΓΡΗΓΟΡΙ. 3rd
century ad, in 19th-century
gold ring. Cameo: 13.0 x 10mm.
Content I, no. 48
Plate 19 Two-layer sardonyx
cameo of the legend ΕΥΤΥΧΙ/
ΓΟΡΓΟΝΙ within raised border.
3rd century ad, 12.9 x 9.0 x
3.1mm. Content I, no. 34

Plate 20 Two-layer sardonyx cameo of the legend ΤΟ ΤΙCΕΙ ΜΝΗΜΟΝΕΥΕ within
raised border. 3rd century ad, 14.0 x 9.7 x 4.3mm. Content II
sophisticated love inscriptions, it may be word-play recalling a
lover’s debt. Although not to be read within quite the same
context as the ὁμονοια rings, which explicitly express, both in
image and in word, a lover’s contract, nevertheless it might still
recall the merces annua,32 the annual fee paid by a man for his
rights to a courtesan, as in the lover-courtesan relationship of
Catullus and Lesbia, and alluded to by Propertius in his
description of the marriage contract.33 At the very least,
however, within the framework of ‘official’ gems, this is likely
an amatory device humorously playing on owing a debt that is
more sexual than financial.
At this point, one is tempted to include even the most
unassuming and apparently official εὐτυχι ὁ φορων (‘Good luck
to the bearer’) ‘messenger gems’ within this grouping of
personal love gifts, as a possible play on the owner’s ‘official’
capacity as carrying out his duty [in love]. Is one of these gems,
in a tabula ansata border, a seemingly literal official cameo, a
passport tessera, in the hand of a government messenger,
exactly that, or is this another tongue in cheek lover’s gift,
referencing official form and language in a personal context
(Pl. 21)?34 Given the word play and allusions already seen
employed in this medium, an, at least secondary, amatory
interpretation cannot be completely disregarded.
Engraved love tokens can take different forms and be
nuanced in various directions. The better quality portraits with
individualised coiffures and jewellery were presumably
marriage or betrothal gifts or simply presents to a beloved
spouse. This is sometimes made evident in the rich, aristocratic
clothing worn by the subject (Pl. 22).35 Double portraits
emphasise such close bonding as in the case of a jet medallion
Plate 22 Two-layer sardonyx cameo Plate 23 Two-layer sardonyx cameo of double facing male and female busts.
of a detailed draped female bust with 3rd century AD, 24.3 x 12.2 x 7.4mm. Content II
fine Severan helmfrisur facing right.
1st half of the 3rd century ad,
24.3 x 16.9 x 9.2mm. Content II
Love and Passion
Plate 21 Two-layer sardonyx cameo of the legend ΕΥΤΥΧΙ Ο ΦΟΡWN within
tabula ansata border. 3rd century ad, 16.7 x 10.9 x 2.5mm. Content II
from York and though less well delineated, a cameo in the
Content collection (Pl. 23).36 On other occasions we can suspect
that cheaper stones were given to mistresses, courtesans and
even prostitutes.
The portraits are fascinating and, because hair-styles
changed during the Middle Empire, as shown by the differing
coiffures of imperial women on contemporary coinage, we can
date them, at least approximately.37 The best are fairly
individualised down to details of jewellery and personal hair-
embellishments though many are standardised and clearly
bought ‘off the peg’.38 In very few instances can there be any
suggestion that the women, and they are predominantly
women, were anything other than private citizens. A recent
study has emphasised how common cameos of this type are on
the Middle and Lower Danube compared with their rarity in
the western provinces so perhaps their main centres of
production lay in this region and perhaps further east.39
Inscriptions are comparatively rare in the case of portraits,
although one cameo in the Content Collection is addressed τῃ
καλῃ, ‘To the beautiful girl’ (Pl. 24).40
The language of love in word and image can be seen in
three gems, each in a pendant setting and belonging to the
same necklace. The inscribed example, a hand tweaking an ear
with surrounding legend χαρις τωη φορουντι ἐστωη, ‘Grace to
the bearer’ sets the tone; the others depict the goddess/
personification Hygeia and Eros: here we have beauty, health
and love all extolled though words are only employed in the
first instance (Pls 25a–c).41
Other images remain to be described, showing something
of the range of cameos as love tokens. The physical side of love-
Plate 24 Two-layer sardonyx cameo
of right facing female bust above the
legend ΤΗ ΚΑΛΗ. 3rd century ad,
24.0 x 15.0 x 8.0mm. Content I, no. 45
‘Gems of Heaven’ | 183
Molesworth and Henig

Plate 25a Two-layer sardonyx cameo of a hand
pinching an ear surrounded by the legend ΧΑΡΙC ΤWΗ
ΦΟΡΟΥΝΤΙ ΕCΤWΗ. 3rd century AD, in contemporary
gold pendant, 20.0 x 15.1 x 4.8mm. Found together
with Pls 25b and 25c. Content I, no. 56
Plate 25b Two-layer sardonyx cameo of Hygeia Plate 25c Two-layer sardonyx cameo of Eros,
holding a serpent and vessel, her foot on a raised one arm raised, standing above a goose. 3rd
plinth. 3rd century AD, in contemporary gold century AD, in contemporary gold pendant,
pendant, 16.0 x 11.5 x 5mm. Content I, no. 91 16.0 x 13.0 x 4.5mm. Content I, no. 98

making is presented in the Content Collection by a couple in
bed. The cameo is well-cut and expensive and may have been
worn by a courtesan or at least was a very personal lover’s gift
(Pl. 13).42 Another cameo of a flagrantly erotic sort, wrongly
categorised as a satire on Elagabalus, is in Paris and shows two
nude women harnessed to a cart and being driven by a man
with a pronounced erection inscribed ἐπιζενι νεικας 43 Not
uncommon are archaeological finds of low quality rings
inscribed on the metal with blatant sexual innuendoes perhaps
best left in the decent obscurity of a learned language: veni
futuue or misce me.44
However, recalling the tradition of the Priapeia, that is of
poems written about or invoking the god Priapus,45 even here
we may not have moved too far beyond the literary culture of
the Roman world. More politely one can say the same thing
with myth, specifically Leda and the swan, which presents a
scene not so very different in the way it is portrayed from the
human love-making scene discussed above (Pl. 2).46
The mime actor with his bald head is quasi-phallic in
representation, but is protection against the Evil Eye rather
than an invitation to an orgy. To judge by the small size of the
rings in which some were set, they were worn in at least some
instances by young women (or even by children, like the more
common phallus rings), perhaps indicating possible lover’s gifts
(Pl. 26).47 Although heads of mime actors are common, not so
the complete figure, apart from a splendid example complete
with shrouded phallus found not too many years ago in north
Lincolnshire.48 These were predominantly apotropaic in the
same way as the ubiquitous Medusa head. Relatively early
examples are known, but the majority appear to date to the 3rd
and 4th centuries (Pl. 4).49
Other devices, to the modern eye, might seem to have even
less relevance to love, but in a world where a love affair could
too often end in death, through purely natural causes such as
fevers, or, in the case of women, high mortality in childbirth,
charms against Invidia were very important. The Evil Eye was
confounded by such devices (commoner on intaglios) of
animals performing human actions; the absurdity of a pair of
mice pulling a cart bearing an elephant, for example (Pl. 27).50
Powerful animals, lions, panthers and bears, for example,
might also protect the wearer, just as in funerary art wild
beasts are often portrayed protecting the tomb.51 Of course
even more powerful protection was afforded by deities whose
image can be regarded as enshrined on a ring, pendant or
brooch. Such images would allow one to carry the deity on
one’s person, just as a Christian nowadays might wear the
image of a saint.52 Thus we find figures of Hermes, the healing
deities Asklepios and Hygeia with little Telesphoros and, above
all, Athena and Herakles, though the latter two are generally
only figured as portraits.53

Plate 26 Two-layer sardonyx cameo of the head of a mime in profile to left. Plate 27 Two-layer sardonyx cameo of two mice pulling an elephant in a cart.
3rd century ad, 14.7 x 13.2 x 10.0mm. Content I, no. 153 3rd century ad, 12.0 x 9.3 x 3.0mm. Content I, no. 177

184 | ‘Gems of Heaven’


This is by no means an exhaustive tract on the subject of
personal cameos, but it does remind us that the study of
cameos should not be confined exclusively to scholarship on
state propaganda, as so often is the case. It highlights the fact
that the primary purpose of a cameo, even more so than an
intaglio, was to be worn as jewellery, and, within that
decorative function, as an expression of love or protection.
Furthermore, these tokens or charms were often designed with
a sophisticated literary or mythological subtext beyond their
apparently simple decorative beauty. The cameo, so much more
obviously visible and hence impressive and perhaps more
expressive than an intaglio, became the ideal medium for gift-
giving, remaining popular against the declining need to
possess an intaglio seal to authenticate one’s own signature. In
the 3rd century ad, gem-cutters presumably adapted with the
changing needs of the times and produced what the market
demanded. One of the most interesting gem groups in the
British Museum is the cache of jewellery and mass-produced
carnelian intaglios from Snettisham in Norfolk. It is securely
dated to the mid-2nd century ad, but it may be suggested that
had it been a 100 years later (late Severan) and contained
gems, these might well have consisted of the sort of cameos our
paper has been discussing.54
Notes
1 For example, N. Hannestad, Roman Art and Imperial Policy,
Aarhus, 1986, 78–82. G.M.A. Richter, Engraved Gems of the
Romans, London, 1971, includes a good number including the
Gemma Augustea (no. 501), the Grand Camée de France (no. 502)
and the Triumph of Constantine (no. 600).
2 M. Henig, The Content Family Collection of Ancient Cameos, Oxford
and Houlton, Maine, 1990 (hereafter Content I), no. 59 and a
similar cameo, Content II (see n. 5 below).
3 Content I, no. 178. More definitely identified in E. Hartley, J.
Hawkes , M. Henig and F. Mee (eds), Constantine the Great. York’s
Roman Emperor, Yorkshire Museum, 2006, 138, no. 75.
4 M. Henig, ‘The language of love and sexual desire in Roman
Britain: jewellery and the emotions’, in M. Henig (ed.), Roman Art,
Religion and Society. New Studies from the Roman Art Seminar,
Oxford 2005 (BAR Int. ser. 1577), Oxford, 2006, 57–66.
5 These are being recatalogued by M. Henig and H. Molesworth
together with approximately 200 other cameos (Content II),
although catalogue numbers have not yet been assigned. Citations
of Content II are all to, as yet, uncatalogued gems and not to those
already included in Content I. All illustrations in this paper are
from this collection.
6 Ibid., 59–60.
7 H. Guiraud, Intailles et camées de l’époque romaine en Gaule II (48é
supp. Gallia), Paris, 2008, nos 1470–71.
8 L. Allason-Jones, Roman Jet in the Yorkshire Museum, York, 1996.
9 Ovid, Metamorphoses X, 155–61; VI, 106; I, 452–567. One would
have expected Ganymede to be more common; the only example
we can find in a 3rd-century ad cameo is in the Hermitage. See
O. Neverov, Antichnye kamei v sobranii Ermitazha, Leningrad,
1988, 144–5, no. 378. Perhaps the theme was too homo-erotic to
render it a suitable gift for a girl. Leda and the swan, and Apollo
and Daphne, are both found in Content II. The late cameos
(perhaps as late as the 6th century ad) are discussed by J. Spier,
Late Antique and Early Christian Gems, Wiesbaden, 2007, 139–41.
10 Content I, nos 156–64 and five in Content II (Medusa heads);
Content I, nos 97–110, 112–18 and 12 in Content II (erotes); Content I,
nos 101–09 and five in Content II (mourning Erotes); Content I, nos
68–85 and 16 in Content II (female portraits); Content I, nos 50–1
and two in Content II (clasped hands); Content I, nos 52–6 and one
in Content II (hand pinching ear).
11 Content I, nos 29–30, 32–9, Content II.
12 Cf. n. 10 above.
13 Content II.
14 Virgil, Eclogue VI, 3–4. Cf. n. 10 above.
Love and Passion
15 M. Henig, ‘Ancient cameos in the Content Family Collection’, in
M. Henig and M. Vickers (eds), Cameos in Context, Oxford and
Houlton, Maine, 1993, 28.
16 Henig (n. 4), 59.
17 Content II. As noted by both Sir John Boardman and Jeffrey Spier
during the conference, this particular image of Eros stabbing more
than one butterfly, and portrayed after the act, sitting amidst his
victims, would have recalled Ajax slaying the cattle.
18 Content I, no. 144.
19 Cf. n. 10 above.
20 R.O.A.M. Lyne, The Latin Love Poets from Catullus to Horace,
Oxford, 1981, covers many aspects of such literary devices
employed by the Roman elegists.
21 Content I, nos 120, 121 and two in Content II.
22 Posidippus, AP 5.134 = HE 3054 ff = 123 AB.
23 Ibid., AP 12.120 = HE 3078 ff = 138AB.
24 Content I, no. 40, Content II.
25 Content II. Ovid, Ars Amatoria I 45–50; 253–74, for example.
26 Content II. Propertius II.7; Tibullus I.10; Ovid, Amores I.9 and II.12.
For the tradition of militia amoris, cf. A. Spies, Militat Omnis
Amans, Diss. Tubingen, 1930.
27 Content I, nos 47–8.
28 Cf. n. 10 above.
29 Content I, no. 34.
30 For ‘Gorgoni bita’, cf. G. Becatti (ed.), Scavi di Ostia. Mosaici e
Pavimenti Marmorei, Rome, 1961, 25, no. 42, pl. lxxii.
31 Content II.
32 S.L. James, ‘A courtesan’s choreography’, in C. Faraone and
L. McClure (eds), Prostitutes and Courtesans in the Ancient World,
Wisconsin, 2006, 226–7.
33 Propertius, III.20.15–18, where the terms (‘foedera’) of a new
lover’s contract (‘lex’) must be made, and ratified by Love’s own
signet (‘haec Amor ipse suo constringet pignora signo’).
34 Content II.
35 Also Content I, nos 76 and 85.
36 Allason-Jones (n. 8), 25, no. 5; Henig (n. 15), 31, fig. 2.5; Content II.
37 M. Wegner, Datierung römischer Haartrachten (AA 53), Berlin,
1938, 275–327; K. Wessel, Römische Frauenfrisuren von der
severischen bis zur konstantinischen Zeit (AA 61), Berlin, 1946–7,
62–76.
38 Content I, nos 68–85; also W.R. Megow, Kameen von Augustus bis
Alexander Severus, Berlin, 1987, 311–16, nos F10–F41.
39 I. Popović, ‘Roman cameos with female busts from Middle and
Lower Danube’, Pallas 83 (2010), 203–24, pls IX–XIII.
40 Content I, no. 45.
41 Content I, nos 56, 91 and 98.
42 Cf. n. 18 above.
43 Megow (n. 38), 247, no. A164.
44 Britannia 38 (2007), 351; Henig (n. 4), 59.
45 W.H. Parker, Priapea: Poems for a Phallic God, London and Sydney,
1988.
46 Content II; also one in M. Henig and A. McGregor, Catalogue of the
Engraved Gems and Finger-Rings in the Ashmolean Museum II,
Roman, Oxford, 2004, no. 10.34.
47 Content I, nos 152–5.
48 Henig (n. 15), 35, fig 2.9.
49 Cf. n. 10 above.
50 Content I, no. 177.
51 Content I, no. 165 and C. Johns and T. Potter, The Thetford Treasure.
Roman Jewellery and Silver, British Museum, London, 1983, 103,
no. 39; Henig (n. 15), 37, fig. 2.10 for a bear cameo from South
Shields perhaps worn by a lady in the Empress Iulia Domna’s suite
when she and Septimius Severus campaigned in north Britain (ad
208–11); a more demotic example of a bear cameo was found on the
beach some years ago at Caesarea Maritima (inf. Galia Davidson).
52 Cf. M. Henig, ‘A house for Minerva: temples, aedicule shrines, and
signet-rings’, in M. Henig (ed.), Architecture and Architectural
Sculpture in the Roman Empire (OUCA monograph 29), Oxford,
1990, 152–62.
53 Content I, no. 92 and another in Content II (Hermes); Content I, nos
88–91 and two in Content II (healing deities); Content I, nos 129–37
and one in Content II (Athena); and Content I, nos 139–43 and two
in Content II (Herakles).
54 Cf. C. Johns, The Snettisham Roman Jeweller’s Hoard, British
Museum, London, 1997.
‘Gems of Heaven’ | 185
The Belgrade Cameo
Antje Krug

Plate 1 The Belgrade cameo

Around the year 1900 an exceptionally large fragment of a

Roman cameo was found during fieldwork between the small
towns of Mladenovac and Kusadak some 50km south-east of
the Serbian capital Belgrade (Pl. 1). To date no other ancient
remains of importance, such as a large villa or a mausoleum,
have been discovered on that site so the question remains open
as to how and when in antiquity or later this cameo reached the
place where it was eventually found. The cameo became
known after its place of safekeeping as ‘The Belgrade cameo’.1
The find was of such importance that Adolf Furtwängler Plate 2 Marks of cutting, after cast
included the cameo in his publication on ancient engraved
gems and cameos, already in print, and hence introduced it
into scholarly discussion.2
It is obvious that the cameo, notwithstanding its impressive
size of 15 x 19cm and roughly triangular shape, is the fragment
of a far larger glyptic work of art. Part of the original edge is
preserved as an irregularly curved line. The other edges show
traces of cutting with tools (Pl. 2). The cameo is made of
sardonyx in brown-white layers. The depiction represents a
battle scene: a beardless rider in Graeco-Roman armour, Plate 3 Relief of the Belgrade cameo, after cast

186 | ‘Gems of Heaven’


Plate 4 Section and angles of the edge
followed by a smaller warrior equally armoured on foot, is
brandishing a lance and galloping over trouser-clad corpses. A
small relief element - no longer explicable – at the upper edge is
the only indication that the scene continued above the
horseman.
Furtwängler identified the rider on the basis of the sketchy
diadem as the Thracian king Rhoimetalkes and linked the
battle as well as the cameo with the Thracian insurrection of
the years 16–13 bc. Some years later Gerhart Rodenwaldt3
again undertook a comprehensive study of the cameo and
dated it on the basis of its style to the Constantinian era, i.e. the
second quarter of the 4th century ad, a date which has been
universally accepted in the literature. The rider has mainly
been identified with either the Emperor Constantine I – after
his acceptance of the diadem in ad 323/4 – or one of his sons. In
the course of the century since its discovery no convincing
parallel has been found as to the size, representation or style of
the Belgrade cameo. Therefore only slight dating variations are
found in the literature, notwithstanding that almost every
study of the portraiture, glyptics or iconography of the Late
Roman Empire refers to this piece.
In addition to its size the high relief of the cameo is notable;
the figures almost look like a woodcut (Pl. 3). From a frontal
view the varying thickness of the layers which are yet decisive
for the relief are not visible.4 The uppermost or brown layer is
extremely thin. Only the essential parts of the figures such as
heads or clothing are carved in the brown layer, at least
partially. The opaque white layer underneath the brown one is
very thick by comparison. Therefore the gem cutter had to cut
The Belgrade Cameo
as far as 15mm deep until he reached a differently coloured
layer of grayish-blue that provided a contrast and a background
for the figures (Pl. 4). This peculiar feature is responsible for
the steep edges and sharply cut outlines of the figures which
seem to be propped up on a pedestal. The plastic work of the
relief itself is rather shallow but emphasises a technical
constraint easily overlooked when working with plaster casts.
Layered agate – sardonyx – is an unforgiving material. The
colours can be influenced, changed or intensified by dyeing,
also with the intent of fraud. Several methods were known in
antiquity.5 But the course and the thickness of the layers are
determined by the nature of the raw material, and they cannot
be altered. The first of the problems to be solved by the
workshop, and the patron as well, was to procure a piece of
sardonyx with layering appropriate for a cameo. To follow the
course of the layers in the non-transparent stone and expose
them according to the intended picture was no less a challenge.
Then as now the mastery of a gem carver is measured by his
skill in accomplishing that task. The fluctuating availability of
agates suitable for cameos may have caused their apparent
absence from the luxury arts at some periods, and the use of
substitutes.6 How high the relief or how three-dimensional the
figures in a cameo were cut was always, (not only in the
Constantinian era) determined by the layers of the raw
material and not by a ‘style’ or ‘artist’s intention’, however it is
defined. This problem due to the material is obvious, too, on
the reverse side of the Belgrade cameo (Pl. 5). Layers and
colours run in irregular clouds not suitable for cameos.
‘Gems of Heaven’ | 187
Krug
Plate 5 The Belgrade cameo, reverse
Plate 7 The Cologne cameo, layering
Furthermore, the decorated rim, separated by a deep furrow
from the main field of the cameo, is a peculiar feature. Even
those masterpieces among surviving cameos do not possess
such a rim.7 The mostly white, threadlike frame around the
inscriptions on small-size cameos8 is something different.
Basically, the rim along a decorated zone is a feature developed
in silver toreutics, in particular for large sumptuous silver
plates. In this respect there is a parallel for the Belgrade cameo,
although of smaller size, in a cameo in Cologne (Pl. 6)9 also
dating to the Constantinian era. This cameo is supposed to
have been found in Cologne, and it too has traces of cutting
with tools.10 With an original size of 10cm before cutting it is
not small either. The rim is decorated with confronted female
heads while the slightly deepened centre is occupied by a male
head, most likely a portrait. The layers of the cameo are not as
irregularly thick as on the Belgrade cameo, but they
demonstrate similar constraints (Pl. 7). The uppermost layer of
light brown is very thin. The cameo cutter had to cut through
several equally thin layers in white and very pale brown and
red until he reached a brownish layer providing a background.
These layers determined the height of the relief. The
decoration on the rim of the Belgrade cameo is chipped in parts
and not discernible. Clearly recognisable is a rhombus, flanked
by ornaments (Pl. 8). Shua Amorai-Stark11 has suggested a row
of jewels as used in Constantinian and later times to decorate
diadems, thrones, robes or other insignia,12 a very convincing
idea.
Most archaeologists have refrained from making
suggestions as to the original size and shape of the Belgrade
cameo. Their attention has been focused on matters of style,
date and iconography, although those questions are dependent
on each other. Hans Möbius has suggested placing the cameo
188 | ‘Gems of Heaven’
Plate 6 The Cologne cameo
Plate 8 The Belgrade cameo, rim, after cast
on the ‘tip’ of the triangle to give the rider a more level position
and therefore has concluded a pelta-like shape.13 Jeffrey Spier
went a step farther and recognised the fragment as part of a
large oval plate.14 The original size and shape of the cameo
cannot easily be inferred from this irregular fragment. Most
scholars have seen the Grand Camée de France15 as the closest
parallel for its size and shape as an irregular rectangle with
rounded corners. Irregular variations of circular forms are by
no means rare among cameos since the raw material here
again enforced its limitations. But, with the background of the
Grand Camée (Pl. 9) it is immediately clear that the horseman
cannot be the central figure of the completed cameo and
therefore the key figure for the picture; he and the other
warriors are reduced to mere marginal figures. Moreover, the
Belgrade fragment cannot easily be placed in an outline
comparable to the Grand Camée. With a slight tilt the rider is
galloping more level and has the appearance of a central
figure, but the resulting rectangle is standing on one of its
corners. The tiny trace of another object above the horseman is
an indication of an extension of the cameo, but reduces the
rider and battle even more to a marginal frieze. On closer
inspection it becomes clear that the irregularly rounded corner
is formed by a series of short straight lines. The outermost lines
meet at an angle of exactly 96°, i.e. they form a slightly inexact
right angle (cf. Pl. 4). Thereby at least the shape, though not the
size of the cameo before destruction, is given as a rectangle
with slightly rounded corners.
Amongst the large silver plates which offer the best
parallels for the decorated rim, plates with four or more
corners are not very frequent.16 Like the use of the wheel in
pottery the use of the lathe favours circular shapes in metal
work. But traces of the lathe on the reverse of the octagonal
 The Belgrade Cameo

Plate 9 Grand Camée de France, with Belgrade Cameo inset

Plate 10 Risley Park Lanx, after Stukeley

‘Gems of Heaven’ | 189

Krug
Plate 11 Risley Park Lanx, with Belgrade cameo inset
Achilleus plate from the Kaiseraugst silver treasure prove that
it had been worked on the lathe too.17 Named lanx quadrata or
rectiangularia,18 square plates or dishes were used among
luxurious silver table ware. Synonyms such as discus or pinax
point also to flat trays. Dishes of small size and without
figurative ornaments are rather frequent, for instance in the
treasures of Traprain19 and Alder Carr,20 as well as fragments in
silver hoards.
Essentially, there are only three large rectangular silver
plates with relief decoration which are of consequence for the
present considerations. They belong to a class of luxury
tableware and are among the most impressive pieces of silver
toreutics anywhere. The Ariadne plate, also from the
Kaiseraugst treasure,21 with its baroque forms is rather less
suitable for comparison. The Corbridge Lanx, no less
impressive, gives priority to the assembly of gods in the centre
of the tray compared to the vine-tendrils on the rim.22 The
reconstruction of the damaged and distorted Hylas plate, from
the hoard of Groß Bodungen, as a rectangle is questionable.23
Closest to the Belgrade cameo, and the battle scene, too, is the
Risley Park Lanx,24 which was found in fragments in 1729 but
was lost again immediately afterwards. Its appearance is
known from a contemporary publication,25 but the fragments
do not provide reliable measurements for the lanx’s height and
width. The proportions of the completed drawing (Pl. 10) (51 x
38cm) seem reasonable compared with the other plates. The
dramatic boar hunt in the square central picture has a
counterbalance by theme and composition as well in the
surrounding frieze depicting bucolic scenes with animals,
shepherds and again a boar hunt. Parallels for the square form
and the decorated rim are known among the fine sigillata
190 | ‘Gems of Heaven’
wares of imperial times and later African Red Slip ware of the
4th and 5th centuries ad.26 A close parallel to the Risley Park
Lanx is a large silver plate found in Trier in 1628 but melted
down together with the rest of the enormous treasure.
Contemporary reports27 of that find do not give a very clear
description of a very large square silver plate with a rim
decorated with hunting scenes (figuras venationis/venatio) and
a central picture with mythological figures (figuras dearum/
simulacra fabulosa). With a weight of libras XIII28 it should
come close to the Corbridge Lanx (libras XIIII).
Further considerations can only be hypothetical but they
should be mentioned if only summarily. A large plate made of
sardonyx, sketched on the model of the Risley Park Lanx with
one of its corners preserved, produces a lanx of far more than
50cm width (Pl. 11). Through its weight alone such a lanx is an
unwieldy piece of luxury. Maybe its breaking up at a later time
is understandable as an effort to reduce it to usable pieces. A
circular bowl of agate in Vienna has a circular foot made from
the same piece of agate.29 A foot of rectangular shape can be
observed on both the Corbridge Lanx and the sigillata
imitations of a lanx quadrata, and such a foot is to be expected
on the cameo ‘lanx’ of Belgrade. The preserved fragment of the
cameo-plate depicts the detail of a battle. More battle scenes
and warriors who are smaller than the rider may have
continued along the rim on all four sides like the frieze with
scenes of the life of Achilleus on the Achilleus plate or the
bucolic pictures on the Risley Park Lanx. But they are unlikely
to refer to an actual, historically fixed battle. It was the
foremost duty of kings and emperors to protect the people
living in their empire against the menace from outside the
empire of ‘barbarians’.30 As indicated by the small relief

fragment above the horseman, the centre of the lanx was not a
brightly polished surface with a central medallion but a
sweeping picture similar to the Corbridge Lanx. For
chronological reasons and in reference to the battle scenes of
the rim we may compare the reverse of a gold multiplum of
Constantine I of ad 313, showing the emperor’s adventus
flanked by Victoria and Dea Roma.31 Significantly, it complies
with a square composition.
A rectangular plate of sardonyx in cameo technique is a
breathtaking piece of artistry. But its sheer size does not make
it an impossible object. The ‘Bowl of St Helen’ in the Vienna
Treasury has an even greater width of 75cm.32 The close
relationship between agate and silver vessels becomes obvious
in the Hellenistic period when silver vessels with relief scenes
are repeated in the then new cameo technique.33 The
unchallenged masterpiece of that time, the ‘Tazza Farnese’34
has its equivalent in a silver bowl of the late 2nd century bc of
the same shape and with a decorated interior and exterior
bottom as well.35 In equally unchallenged skill the gem cutter
succeeded in following and exposing the layers of the sardonyx
according to the intended picture. The ‘Coupe des Ptolémées’36
abandons the characteristic layering in favour of a strong relief
on its exterior. Relief-decorated silver kantharoi were much
appreciated and produced throughout Imperial times, along
with alabastra in silver and in sardonyx.37 The fragmentary
Nereid bowl completely abandons colour and cameo effect by
using colourless rock crystal as its material.38 A silver bowl with
shell-shaped feet from a 3rd century bc burial offers the
equivalent in toreutic.39 Very large silver plates decorated with
reliefs, niello and partly gilded are documented in Late Antique
silver treasures and they moreover show a certain interest in
tonality comparable to cameos.
The venationes, too, so prominent on the lances of Trier and
Risley Park emphasise the princely character of the
iconography. In particular the lion hunt but also the hardly less
dangerous boar hunt has been a privilege of nobility,
transferring the subject of battle into civilian life. Thus the
tondi with hunting scenes created for the Emperor Hadrian
were prominently re-used on the arch of Constantine.40 Battle
scenes combined with the admittedly hypothetical central
image of the emperor on horseback can hardly be understood
otherwise than as a symbol of triumph. In this respect the
pictures of the Belgrade cameo continue the subject of the
silver missoria donated as meaningful gifts of honour on
special occasions and thereby increasing the significance of
sumptuous table ware. The Clementinus diptych of ad 513
depicts square and round vessels without doubt made of silver
filled with gold coins.41 Likewise the illustrations of the Notitia
Dignitatum display with the headings of Comes Sacrarum
Largitionum and Comes Rerum Privatarum the precious gifts
they had to supervise and to provide.42 The large sums of coins
as the essence of gift giving and the true donativa were
presented in large silver bowls, the largitio or vota bowls. They
were chiefly and in great number manufactured in the later
Empire to provide the imperial gifts with a proper ‘wrapping’.43
The great number of preserved bowls demonstrates that they
were seen and kept as a sign of honour and not instantly melted
down. The large silver trays known as missoria,44 however,
were part of luxury table ware. Moreover, the dimensions of
these imperial gifts were by no means negligible: even the
The Belgrade Cameo
‘Shield of Scipio’45 with a width of 70cm is surpassed by the
missorium of Theodosius46 with a width of 74cm. The value of
the material was supplemented by the associative value of the
images, in many cases the emperor in question amidst his
court. Likewise the imperial jubilees, the decennalia and the
vicennalia offered an opportunity for the display of luxury. The
practical use of such a tray as costly, and as unwieldy and
heavy, was without doubt of secondary importance although
many of the silver plates show traces of use. Luxury tableware
was a visible and tangible part of the social status of its owner.47
Constantine I celebrated both his decennalia (ad 315) and his
vicennalia (ad 325), and perhaps he marked this achievement
with a very special gift.
Notes
1 National Museum of Beograd/ Narodni Muzej u Beogradu Inv. no.
116/IV. I express my thanks to the National Museum for the
photographs. I also owe a very special thank you to Hans
Rupprecht Goette for manifold help with the pictures and
especially for the digital preparation and creation of the montages.
2 A. Furtwängler, Die antiken Gemmen, vol. 3, Berlin, 1900, 453–8,
figs 234–6.
3 G. Rodenwaldt, ‘Der Belgrader Kameo’, Jahrbuch des Instituts 37
(1922), 17–38, figs 1–3.
4 My thanks go to the director of the Akademisches Kunstmuseum
Bonn, Harald Mielsch, for permission to work with the cast in the
museum’s collection and to photograph it.
5 Pliny, Nat. Hist. XXXVII, 79.197; G. Schmidt, ‘Erfahrungen und
Fragen beim Nachschneiden der drei größten Sardonyx-Kameen
der Antike: Tazza Farnese, Gemma Augustea und Grand Camée de
France’, in G. Platz-Horster (ed.), Mythos und Macht. Erhabene
Bilder in Edelstein. Internationales Kolloquium zur gleichnamigen
Ausstellung der Antikensammlung Staatliche Museen zu Berlin im
Alten Museum am Lustgarten 27. Juli 2007, Berlin, 2008, 6–12. His
considerations go far beyond the question of dyeing and need
critical reviewing, but in another context.
6 G. Platz-Horster, Nil und Euthenia: der Kalzitkameo Berlin im
Antikenmuseum Berlin (Berliner Winckelmannsprogramm 133),
Berlin, 1992, 6–9.
7 The stiff astragal frieze around the Minerva cameo in the Cabinet
des Médailles, Paris, looks as non-antique as the cameo itself, the
authenticity of which has already been questioned by W. R.
Megow, Kameen von Augustus bis Alexander Severus (Antike
Münzen und geschnittene Steine, vol. 11), Berlin, 1987, 223 f.,
no. 113, pl. 37, 4.
8 Cf. M. Henig, The Content Family Collection of Ancient Cameos,
Oxford, 1990, nos 27–35, 37–40, and passim.
9 A. Krug, ‘Antike Gemmen im Römisch-Germanischen Museum
Köln’, Bericht Römisch-Germanische Kommission 61 (1980), 184 f.,
no. 61, pl. 75. I want to thank the director of the RGM, Hans-Gerd
Hellenkemper, and Friederike Naumann-Steckner for permission
to see the cameo once more and photograph it outside of the
showcase.
10 J. Bracker, ‘Eine Kölner Kameenwerkstatt im Dienste
konstantinischer Familienpolitik’, Jahrbuch für Antike und
Christentum 17 (1974), 103–8, pl. 8b.
11 In London during discussion on May 29, 2009.
12 Cf. A. Demandt and J. Engemann (eds), Konstantin der Große (exh.
cat. Rheinisches Landesmuseum Trier), Mainz, 2007, 44, no. 1.4.15
and 68, fig. 23 (belts and shoes of tetrarchs); 98, figs 4–5 and 159,
fig. 7 (coin portraits of Constantine I with diadem); 302, no. 1.16.1
(marble portrait of Gratian with diadem); 138, no. 1.12.18 (gilded
and jewelled helmet from Berkasovo). Enthroned togatus
(Diocletian?), Alexandria: R. Delbrueck, Antike Porphyrwerke,
Berlin, 1932, 96–8, pls 40–1, mouldings of the throne.
13 H. Möbius, ‘Zweck und Typen römischer Kaiserkameen’, Aufstieg
und Niedergang der Römischen Welt II Prinzipat, vol. 12,3, Berlin,
1985, 32–88; 74 f. He does not offer references for his statement that
the pelta is ‘Hoheitszeichen der persischen Könige’. His paper was
published sometime after its completion (indeed posthumously)
and it therefore remains open if Möbius would have been more
explicit later.
‘Gems of Heaven’ | 191
Krug
14 J. Spier, ‘Late Antique Cameos c. A.D. 250–600’, in M. Henig and
M. Vickers (eds), Cameos in Context. The Benjamin Zucker Lectures
1990, Oxford and Houlton, 1993, 43–54, at 44; idem, Late Antique
and Early Christian Gems, Wiesbaden, 2007, 130.
15 Paris, Cabinet des Médailles: M.-L. Vollenweider and M. Avisseau-
Broustet, Camées et intailles II. Les Portraits romains du Cabinet des
Médailles, Paris, 2003, no. 275, with comprehensive bibliography.
16 S. Künzl, ‘Römisches Tafelsilber – Formen und Verwendung’, in
H.-H. von Prittwitz und Gaffron and H. Mielsch (eds), Das Haus
lacht vor Silber. Die Prunkplatte von Bizerta und das römische
Tafelgeschirr (Kataloge Rhein. Landesmus. Bonn 8), Cologne-
Bonn, 1997, 9–30, esp. 24.
17 A. Mutz, Die Kunst des Metalldrehens bei den Römern, Basel, 1972,
33, fig. 45: H.A. Cahn and A. Kaufmann-Heinimann (eds), Der
spätrömische Silberschatz von Kaiseraugst, Derendingen, 1984, 282
f. (V. von Gonzenbach); 371 f. (E. Foltz).
18 W. Hilgers, ‘Römische Gefäßnamen’, Beihefte der Bonner
Jahrbücher 31 (1969), 65, 206–09, no. 209.
19 A.O. Curle, The Treasure of Traprain, Glasgow, 1923, 40, no. 34,
pl. 22.
20 L. Pirzio Biroli Stefanelli, ‘I Tesori di argenteria rinvenuti in Gran
Bretagna’, Archeologia Classica 17 (1965), 96, no. 3, pl. 32,1.
21 Cahn and Kaufmann-Heinimann (n. 17), 194–205, no. 61 (F.
Baratte): size 41.5 x 35cm.
22 London, British Museum, property of the Duke of
Northumberland. L. Pirzio Biroli Stefanelli, L’argento dei Romani,
Rome, 1991, no. 177, fig. 244: size 48.26 x 38.1cm.
23 W. Grünhagen, Der Schatzfund von Groß Bodungen (Römisch-
Germanische Forschungen, vol. 21), Berlin, 1954, 41 f. pl. 4; Cahn
and Kaufmann-Heinimann (n. 17), 199.
24 D. E. Strong, Greek and Roman Gold and Silver Plate, London, 1966,
185–6; C. Johns, ‘The Risley Park Silver Lanx: a Lost Antiquity from
Roman Britain’, The Antiquaries Journal 61 (1981), 53–72, pls 7 and
9; G. Fischer-Heetfeld, ‘Studien zu spätantikem Silber: Die Risley
Lanx’, Athenische Mitteilungen 98 (1983), 239–63, pls 47, 49–51;
J.M.C. Toynbee and K.S. Painter, ‘Silver Picture Plates of Late
Antiquity: ad 300 to 700’, Archaeologia 108 (1986), 41 f., no. 50,
pl. 20c; M.A. Guggisberg (ed.) (with A. Kaufmann-Heinimann),
Der spätrömische Silberschatz von Kaiseraugst, die neuen Funde:
Silber im Spannungsfeld von Geschichte, Politik und Gesellschaft der
Spätantike (Forschungen in Augst, Vol. 34), Augst, 2003, 333–46;
341 HF 62 (M. Guggisberg). Found in 1729, tentative size 20 x 15
inches (51 x 38cm).
25 W. Stukeley, An account of a large silver plate of antique basso
relievo, Roman workmanship, found in Derbyshire, 1729, London,
1736.
26 Johns (n. 24), 63, pl. 8; N. Franken, ‘Imitationen römischer
Silbertabletts in Ton’, in von Prittwitz und Gaffron and Mielsch
(n. 16), 31–40, at 34 f., fig. 5.
27 W. Binsfeld, ‘Der 1628 in Trier gefundene römische Silberschatz’,
Trierer Zeitschrift 42 (1979), 113–27; 115, no. 1 / 2; 120; no. 1 / 2. The
Latin of the reports leaves many questions open as to the shape of
the rim: M(asen): ‘paropsides 2 quadratae formae, in crepidine
sine limbis habentes diversas figuras venationis…, W(iltheim):
Nona quadrata et oblonga, omnigenam in orâ venationem …’
28 Binsfeld (n. 27), 114, has reckoned with a pound in the scale of two
Kölner Mark = 467.7g, leading to a weight of close to 6kg, that is
much heavier than the Corbridge Lanx.
29 R. Egger, ‘Die Trierer Achatschale a.d. Weltlichen Schatzkammer
der Wiener Hofburg’, Trierer Zeitschrift 22 (1953), 217–18; R. Noll,
‘Zur Achatschale ("Hl. Gral") in der Wiener Schatzkammer’,
Anzeiger Phil.-Hist. Klasse Österreichische Akademie der Wissen-
schaften 118 (1981), 134–6, pl. 1.2; W. Oberleitner, Geschnittene
Steine. Die Prunkkameen der Wiener Antikensammlung, Vienna,
1985, 64 with col. pl.; H. Fillitz, Die Schatzkammer in Wien, Vienna/
Salzburg, 1986, 202, pl. 42, col. pl. 48.
30 J. Engemann, ‘Konstantins Sicherung der Grenzen des römischen
Reiches’, in Demandt and Engemann (n. 12), 155–9.
31 Demandt and Engemann (n. 12), 205, fig. 14.
32 Cf. n. 29.
33 Cf. also M. Vickers, Skeuomorphismus oder die Kunst, aus wenig viel
zu machen (Trierer Winckelmannsprogramm, 16), Mainz, 1998,
20–6.
34 Naples, Museo Archeologico Nazionale. For the numerous
publications see, E. La Rocca, L’età d’oro di Cleopatra. Indagine
sulla Tazza Farnese (Documenti e Ricerche d’Arte Alessandrina),
192 | ‘Gems of Heaven’
Rome, 1984, with extensive bibliography.
35 Toledo Museum of Art: A. Oliver and K.T. Luckner, Silver for the
Gods, 800 Years of Greek and Roman Silver, Toledo Museum of Art,
Toledo, 1977, 90, no. 53 (A. Oliver).
36 Paris, Cabinet des Médailles: Le Trésor de Saint-Denis (exh. cat.
Musée du Louvre), Paris, 1991, 83–7, no. 11, with bibliography
(D. Alcouffe).
37 Cf. E. Zwierlein-Diehl, Das Onyx-Alabastron aus Stift Nottuln in
Berlin (Berliner Winckelmannsprogramm, 138), Berlin, 1999.
38 A. Krug, Die Berliner Nereidenschale aus Bergkristall (Berliner
Winckelmannsprogramm, 137), Berlin, 1998.
39 From Trichonion (Aetolia): Ph. N. Zapheiropoulou, ‘Τάφοι στο
Τριχόνιο Αιτωλίας’, in Επιστημονκὴ Συνάντηση για την
Ελληνιστική Κεραμική, Athens, 2000, 324f., pl. 164β.
40 W. Martini and E. Schernig, ‘Das Jagdmotiv in der imperialen
Kunst hadrianischer Zeit’, in W. Martini (ed.), Die Jagd der Eliten in
den Erinnerungskulturen von der Antike bis in die frühe Neuzeit,
Göttingen, 2000, 129–55.
41 Liverpool, Merseyside Museum: R.E. Leader-Newby, Silver and
Society in Late Antiquity: Functions and Meanings of Silver Plate in
the 4th to 7th Centuries, Aldershot, 2004, 41–3, fig. 1.16.
42 O. Seeck, Notitia Dignitatum, Frankfurt am Main, 1876 (repr.,
1962), 148–56, nos XI und XII; R. Scharf, ‘Die Diocletianisch-
Constantinischen Reformen’, in M. Geiberger, A. Stute and A.
Hofmann (eds), Imperium Romanum. Römer, Christen, Alamannen
– die Spätantike am Oberrhein (exh. cat. Badisches Landesmuseum
Karlsruhe, 2005), Stuttgart, 2005, vol. 2, 72–9, fig. on 74.
43 J. Szidat, ‘Der Schatz von Kaiseraugst und die Geschenke des
Kaisers (sacrae largitiones)’, in Guggisberg (n. 24), 225–32.
44 M. Guggisberg, ‘Silbergeschirr und Largitio’, in idem (n. 24), 255–
69; Reallexikon zur Byzantinischen Kunst, vol. 6, Stuttgart, 2005,
s.v. Missorium, 378–9 (J. Engemann).
45 Paris, Cabinet des Médailles, found in the river Rhône near
Avignon: Toynbee and Painter (n. 24), 30–2, no. 21, pl. 12 a.
46 Madrid, Real Academia de Historia: Toynbee and Painter (n. 24),
27–8, no. 16, pl. 10a; M. Almagro-Gorbea et al. (eds), El disco de
Teodosio, Madrid, 2000.
47 Cahn and Kaufmann-Heinimann (n. 17), 407–09 (H. Wrede).
Select bibliography of the Belgrade cameo
Brilliant, R. in: K. Weitzmann (ed.), 1977/78, Age of Spirituality. Late
Antique and Early Christian Art, Third to Seventh Century (exh. cat.
Metropolitan Museum of Art), New York, 83, no. 71, pl. II.
Bruns, G. 1948, Staatskameen des 4. Jahrhunderts nach Christi Geburt
(Berliner Winckelmannsprogramm,104), Berlin, 19–20, fig. 14.
Demandt, A. and J. Engemann (eds), 2007, Konstantin der Große (exh.
cat. Rheinisches Landesmuseum Trier), Mainz, cat. I.7.33.
Furtwängler, A. 1900, Die antiken Gemmen, vol. 3, Berlin, 453–8, figs
234–6.
Grbić, M. 1958, Choix de plastiques grecques et romaines au Musée
National de Beograd, Belgrade, 132f., pl. 68.
Krug, A. in: Demandt and Engemann, Konstantin der Große, 135, fig. 5.
Krug, A. 2009, ‘Ein politisches Manifest’, Archäologischer Kalender,
November.
Kuzmanović-Novović, I. 2009, ‘Portraits of the emperor Constantine
and his family members in the glyptic art of Serbia’, in M. Rakocija
(ed.), 3–5 June 2008, Niš and Byzantium. Seventh Symposium, Niš,
The Collection of Scientific Works VII, Niš, 77–86, fig. 1.1a.
Schumacher, K. 1909, ‘Die Germania des Tacitus und die erhaltenen
Denkmäler’, Mainzer Zeitschrift 4, 1–13; 8.11, pls 1, 2.2a.
Stutzinger, D. in: D. Stutzinger (ed.), 1983, Spätantike und frühes
Christentum (exh. cat. Liebieghaus Museum alter Plastik,
Frankfurt am Main), Frankfurt am Main, 434–5, no. 46.
Möbius, H. 1985, ‘Zweck und Typen römischer Kaiserkameen’, Aufstieg
und Niedergang der Römischen Welt, II Prinzipat, vol. 12,3, Berlin,
32–88; 74 f.
Rankov, J. (-Kondić) 1988, in: Antike Porträts aus Jugoslawien (Katalog
zur Ausstellung, Frankfurt am Main, Museum für Vor- und
Frühgeschichte), Frankfurt am Main, 199, no. 233, frontispice.
Rodenwaldt, G. 1922, ‘Der Belgrader Kameo’, Jahrbuch des Instituts 37,
17–38, pls 1–3.
Veličković, M. 1983, in: Archaeological Treasury of Serbia, Belgrade,
94f., no. 75.
Zadoks-Josephus Jitta, A.N. 1966, ‘Imperial messages in agate, II’,
Babesch 41, 91–104; 99, fig. 10.
Late Antique and Early Christian Gems
Some Unpublished Examples
Jeffrey Spier
The study Late Antique and Early Christian Gems was intended
to offer as complete a corpus as possible,1 listing just over 1000
gems and cameos, but in the five years or so that have passed
since its completion nearly 100 additional gems have been
brought to my attention, increasing the number of recorded
examples by nearly 10%. The new discoveries do not alter the
general picture of gem engraving in the Late Antique and Early
Byzantine periods (the mid-3rd to early 7th centuries ad) but
do provide some new and interesting varieties. The present
article will present a brief overview of the various categories
already established with a discussion of some of the new
material and a summary catalogue of the additional gems.
Late Antique portraits
Tens of thousands of Roman gems from the Imperial period
survive, but only a small number of these can be attributed to
the years following the collapse of the Severan dynasty at the
death of Severus Alexander in ad 235 and the onset of political
anarchy in the Roman Empire. It is clear that there was a
dramatic decrease in the production of engraved gems and
cameos of all sorts, although whether for economic reasons or
merely as a result of a changing taste in fashion is uncertain.
Rings of the later 3rd and 4th centuries ad tend to re-use older
intaglios or employ unengraved stones, or to have solid,
engraved bezels, or bezels set with coins. Gems engraved with
imperial or private portraits of the second half of the 3rd
century ad are very rare, as the study of imperial portraits on
gems by Erika Zwierlein-Diehl in this volume suggests,2 and the
lack of such gems is another indication of the rapid decline of
the art of gem engraving. There is a dramatic revival of gem
engraving, including the use of large amethysts and sapphires
of fine style, under Constantine the Great in the 4th century ad,
but such gems are very rare and appear to be the products of a
very small number of workshops associated with the imperial
court.3 Similarly, in the 5th century ad the production of
Plate 1 (Add. 1) Silver disc, diademed busts of Constans and Constantius II (?),
crowned by Victory, DD NN. Ex-auction
engraved gems appears to have been confined largely to a few
ateliers in Constantinople. Also notable is the almost total
cessation of the production of engraved gems in the West, with
the exception of a few extraordinary examples, such as the
famous sapphire seal of the Gothic King Alaric now in Vienna.4
Although superb gem engraving was practised in the
Constantinian court, very few of these gems survive. More
common are engraved silver discs of 4th century ad date that
were set in rings or pendants, reflecting the change of fashion
at the end of the 3rd century ad from gems to engraved rings.
Most of these discs are engraved with portraits, and many are
inscribed in Latin (only a few in Greek) with the names of the
individuals or married couples they depict. To the 35 examples
previously published,5 14 more can be added, along with four
others that are engraved with other images and one lead
sealing impressed by a silver disc. The engraving on these discs
is stylistically very close to that of 4th century ad coins, but
there had been no evidence linking the production of the discs
to imperial workshops until recently. One disc, however, which
appeared in a German auction in 2008 (Add. 1, Pl. 1), displays
the confronted busts of two of the sons of Constantine
(probably Constantius II and Constans, c. ad 340–50), wearing
diadems and each crowned by a small figure of Victory; around
them are inscribed the letters ddnn, for domini nostri. An
object such as this must have served an official purpose,
perhaps as gift-payment (donativum) for a soldier. Another
imperial portrait is seen on a small silver disc that depicts an
emperor with crested helmet in frontal view, a pose that came
into use in the late 4th century ad; the inscription around
includes the acclamation vivas, but the name is unclear (Add.
2, Pl. 2). A disc engraved with the portrait of a slightly bearded
young man wearing military dress is inscribed lic cons,
perhaps denoting the period of joint rule of Licinius and
Constantine (c. ad 313–23) and suggesting this piece, too,
served as a military gift (Add. 3). Other discs bear portraits of
Plate 2 (Add. 2) Silver disc with facing emperor in crested helmet. Munich,
Collection C.S., no. 2175
‘Gems of Heaven’ | 193
Spier

Plate 3 (Add. 4) Silver disc with facing busts of Plate 4 (Add. 5) Silver disc with facing busts of Plate 5 (Add. 6) Silver disc with busts of a
man and woman, viva. Munich, Collection a man and woman, uncertain inscription. man, woman, and child, vivatis in deo.
C.S., no. 2229 Munich, Collection C.S., no. 2751 Munich, Collection C.S., no. 2122

Plate 6 (Add. 10) Silver disc with male bust, Plate 7 (Add. 11) Silver disc with male bust, Plate 8 (Add. 12) Silver disc with facing male
fibula on shoulder, ΕΙC [Θ]ΕΟC. Munich, anastasi vivas. Munich, Collection C.S., no. bust, iovine vivas. Munich, Collection C.S.,
Collection C.S., no. 2169 2171 no. 2172

Plate 9 (Add. 14) Silver disc with male bust, Plate 10 (Add. 15) Lead sealing, 13mm, Plate 11 (Add. 17) Silver disc with standing
dalmativs. Ex-auction impressed from a silver disc or ring of similar Victory and kneeling Genius, vivas in deo.
style. Male bust, … e vivas. eBay Munich, Collection C.S., no. 2182

private individuals, either families, young men, or, rarely,
young women (Add. 4–15, Pls 3–10). Often they are inscribed
with names (the new examples include Anastasius, Cercus,
Dalmatius, Iovinus, Zosimus, Vebeca, and Laurentia), usually
accompanied by the acclamation vivas or vivatis and
sometimes made explicitly Christian by the addition of in Deo.
In one instance (Add. 10, Pl. 6), the Christian phrase in Greek,
ΕΙC [Θ]ΕΟC (‘There is one God’) is used. A tiny lead impression
made by one such disc (found on that invaluable archaeological
resource, eBay) shows the bust of a young man accompanied by
the word vivas and an illegible name (Add. 15, Pl. 10); this
example is notable for demonstrating that the discs were
actually used as seals.
To the small number of silver discs engraved with images
other than portraits, four can be added. The first is engraved
with the chi-rho monogram within a wreath and the inscription
con, suggesting again an imperial association, perhaps
referring to Constantine the Great or one of his sons (Add. 16).
Another disc recalls devices found on coins, with a figure of
Victory standing before a kneeling winged Genius,
accompanied by the Christian acclamation, vivas in deo (Add.
17, Pl. 11), again a suitable subject for a military officer. Two
other discs with Christian images have come to light, a finely
engraved image of Abraham’s Sacrifice of Isaac (Add. 18, Pl.
12) and another combining the Sacrifice of Isaac with Daniel in
the Lions’ Den (Add. 19, Pl. 13).

194 | ‘Gems of Heaven’


Plate 12 (Add. 18) Silver disc with the Sacrifice Plate 13 (Add. 19) Silver disc with Daniel
of Isaac. Munich, Collection C.S., no. 2336 in the Lions’ Den and the Sacrifice of Isaac.
Munich, Collection C.S., no. 2142
Plate 15 (Add. 21) Carnelian in gold Plate 16 (Add. 22) Green chalcedony
mount. Seated Constantinople, with seated Constantinople holding a
holding a cross on globe. Derek cross on globe. Munich, Collection
Content Collection C.S., no. 2671
Gems engraved with portraits dating after the
Constantinian period are very rare. Nothing survives of the
late 4th century ad, but a small number of 5th-century ad
engraved garnets depict aristocratic young men, usually with a
fibula on the shoulder to denote status.6 Related to this group is
a very fine amethyst in a private collection engraved with the
bust of a young bearded man with a crossbow fibula on his
shoulder (Add. 20, Pl. 14). His name in Greek is written
around, CΑΤΟΡΝΙΛΟΥC, Satornilous. His hairstyle, with its
prominent roll above the forehead, is typical of portraits of the
early to mid-5th century ad.
A remarkable series of gems engraved with the
personifications of Constantinople and Rome commemorate
the founding of the new capital in ad 330 and are likely related
to other imperial works, most notably coins and medallions
with similar imagery.7 The finest surviving example is a large
(60mm) nicolo in Paris of mid-4th century ad date depicting an
enthroned Roma,8 while a more modest carnelian in
Cambridge shows Constantinople in a similar pose but
identifiable by the prow of a ship on which her feet rest.9 Two
new examples show that similar gems were still being
engraved in the 5th century ad. A carnelian in a gold mount
(probably from a necklace) depicts Constantinople now
holding a cross on a globe (Add. 21, Pl. 15), an image
introduced by Theodosius II on his coinage sometime in the ad
430s. A second example of this type is engraved on a green
chalcedony gem (Add. 22, Pl. 16).
Late Antique and Early Christian Gems
Plate 14 (Add. 20) Amethyst with portrait
of bearded man, fibula on shoulder, cross
above, CΑΤΟΡΝΙΛΟΥC, Satornilous. Munich,
Collection C.S., no. 2631.
Plates 17a–b (Add. 23) Carnelian, F3, in silver ring. ihcoy. Munich, Collection
C.S., no. 2365
Early Christian gems, 3rd–4th centuries AD
Among the many thousands of engraved gems of the mid- and
late-3rd century ad are a small number of examples with
Christian imagery and inscriptions. The materials and shapes
of these gems, the style of engraving, and the forms of the rings
in which they are set, are identical to those of contemporary
pagan works, and very likely they were produced in the same
workshops. The gems tend to be simple in style, either aniconic
with inscriptions referring to Jesus Christ or with symbolic
images, such as the fish or the Good Shepherd. Only rarely are
biblical scenes depicted. Approximately 400 gems dating from
the 3rd and 4th centuries ad have been recorded previously, a
very small number when one considers the many thousands of
Roman imperial gems which survive today. Less than 30
additional gems are listed below, some of which are notable for
being in their original settings.
A very distinctive group (about 25 examples previously
recorded) names Jesus Christ, engraved in angular Greek
letters, always in the genitive, signifying that the wearer was
‘of Jesus Christ’, likely understood as a ‘servant’ of Christ, a
frequent acclamation in Christian texts. Several gems spell the
name out in full, ihcoy xpictoy,10 while others, all very similar,
name just ihcoy or xpictoy.11 All are carnelian, in the shape
that sits above the bezel like a truncated pyramid (shapes F2 or
F3), and many are cut octagonally. They may well be from a
single workshop, and most have a vague eastern provenance in
Asia Minor or Syria. A few are set in rings, always of typically
mid-3rd-century ad shape. Six new examples have now
appeared, all of which fit well into the previously recorded
group. A carnelian inscribed ihcoy is set in a simple silver ring
of 3rd-century ad shape (Add. 23, Pls 17a–b), and an octagonal
‘Gems of Heaven’ | 195
Spier

Plate 18 (Add. 25) Carnelian, F3. ihcoy. Private Plates 19a–b (Add. 26) Carnelian, F3, cut octagonally, in a gold ring. xpictoy. Munich,
collection Collection C.S., no. 2345

Plate 20 (Add. 28) Green-brown chalcedony, F3, Plate 21 (Add. 30) Carnelian, F3, in gold ring Plate 22 (Add. 31) Red jasper, F3, octagonal.
9.5 x 7.7 x 3.4mm. xpictoy. Private collection with beaded wire hoop. Chi-rho. Munich, Chi-rho-tau monogram. Munich, Collection C.S.,
Collection C.S., no. 1890; from the Balkans? no. 2478

Plate 23 (Add. 32) Banded agate, orange and Plate 24 (Add. 33) Yellow jasper, F2, in an ancient Plate 25 (Add. 34) Carnelian, octagonal, F3.
white. Chi-rho monogram combined with gold ring that may not belong. ΙΧΘΥC and star. Two fish flank an anchor, ΙΧΘΥC. Munich,
staurogram. Munich, Collection C.S., no. 2370 Munich, Collection C.S., no. 2846 Collection C.S., no. 2432

carnelian inscribed xpictoy is in a contemporary gold ring
(Add. 26, Pls 19a–b).
A second group of inscribed Early Christian gems appears
to be roughly contemporary with the previous examples. They
are generally of the same shape, although some are of red or
yellow jasper as well as carnelian. They are engraved with the
earliest known forms of the chi-rho monogram, denoting the
word Christos. The shapes of the gems, as well as the forms of
the rings in which they are set, show they are of mid- or late
3rd-century ad date and likely of eastern (perhaps Syrian)
origin. The gems are especially significant for demonstrating
an early, Christian use of the symbol that would later be
promoted by Constantine the Great as his personal emblem,
one that was said to have been seen by the emperor in a vision
before his decisive battle against Maxentius in ad 312. In
addition to the specimens previously published,12 four more can
be added, including a carnelian in an iron ring from
Carnuntum (Add. 29) and another in a gold ring (Add. 30, Pl.
21).
Some unconventional forms of the chi-rho monogram also
appear on similar gems of no doubt contemporary date.13 They
sometimes combine other letters or incorporate the sign of the
cross. A new example in red jasper combines the chi-rho with
the tau-cross (Add. 31, Pl. 22). More puzzling is a very
unconventional banded agate of very orange colour (Add. 32,
Pl. 23) with an odd form of the monogram, seemingly a chi
combined with the staurogram (a monogram-symbol
composed of tau-rho for the Greek word stauros, ‘cross’). The

196 | ‘Gems of Heaven’


Plates 26a–b (Add. 36) Carnelian, F3, in silver ring. Two fish flank an anchor;
above, IAW IH in wreath. Munich, Collection C.S., no. 2145; from Bulgaria?
Plate 28 (Add. 48) Carnelian, F3, Plate 29 (Add. 49) Carnelian, F2–3,
broken octagonal, the Good Shepherd with the Good Shepherd. Derek Content
with staff. Munich, Collection C.S., no. Collection
2673
staurogram was one of the earliest of all Christian symbols,
appearing already in 3rd-century ad manuscripts of the
Gospels, but it was not widely used until its revival at the end of
the 4th century ad.14 It is unclear if the gem dates to the late 3rd
century ad, when other variants of the chi-rho monogram were
used on gems, or the later 4th or early 5th century ad, when the
staurogram was more current.
Another inscription found on Christian gems of the 3rd
century ad is the word ΙΧΘΥC, ‘fish’ in Greek, but also an
acrostic composed of the first letters of the words ‘Jesus Christ,
son of God, saviour’. A number of gems bear the inscription,
sometimes alone and other times written around an image.15
On a newly discovered yellow jasper, the word is accompanied
by a star (Add. 33, Pl. 24).
The pictorial image of a fish was very popular on Early
Christian gems as well, especially the distinctive composition
adapted from the repertoire of earlier pagan decorative images
of a pair of fish flanking an anchor or a cross-like object.16
Although most examples of this type are not inscribed, a
significant number do bear explicit Christian inscriptions,17
demonstrating they were indeed made for Christian patrons.
Two more gems with this device accompanied by the
inscription ΙΧΘΥC have now come to light (Add. 34, Pl. 25, and
Add. 35), as well as more uninscribed examples (Add. 37–44),
including some with recorded provenance (Carnuntum,
Aquileia, Slovenia, and southern Spain).
Late Antique and Early Christian Gems
Plates 27a–b (Add. 45) Nicolo in a heavy gold ring. The Good Shepherd
stands with a sheep over his shoulders
Plates 30a–b (Add. 50) Black obsidian, F1, fragmentary, with nearly half
lost; length: 34mm. Both sides are engraved: a: the Good Shepherd, sheep
at feet; around, …C ΘΕΟΥ ΥΙΟC. b: two fish flank an anchor; around, …
CWΤΗ. Private collection
A unique variety of the fish-and-anchor motif appears on a
carnelian gem set in a silver ring of typically 3rd-century ad
type (Add. 36, Pls 26a-b). The image is placed below a wreath
containing the Greek words IAW (the transliteration of the
Hebrew name of God, Yahweh), I H (for Iesous), and a form of
christogram. Although the divine name IAω was often
employed on magical gems at this time, this particular usage
appears rather to refer to Jesus, as the son of God.
After the symbol of the fish, the most popular image on
Early Christian gems was the Good Shepherd, who is always
depicted as a young man in a short tunic carrying a sheep on
his shoulders.18 Nearly 100 gems were published in the initial
corpus, about a third of which bear accompanying inscriptions
or symbols that are explicitly Christian. Additions to the corpus
include a nicolo in a heavy gold ring of 3rd-century ad type
(Add. 45, Pls 27a–b) and another in a silver ring of similar date
(Add. 46). The most remarkable new discovery is a very large,
but fragmentary, obsidian gem engraved on both sides (Add.
50, Pls 30a–b). On the front side, the Good Shepherd stands
with a sheep at his feet. The inscription around, in angular
letters, reads, …c ΘΕΟΥ yioc, which can be restored, ihcoyc
xpictoc xpictoc ΘΕΟΥ yioc, ‘Jesus Christ, son of God’. The
reverse side is engraved with the fish-and-anchor motif, with
an inscription ending, CWΤΗ[Ρ], ‘saviour’. A few other gems
engraved with the Good Shepherd bear Christian images on
the reverse side, an otherwise unusual practice.19 Three
‘Gems of Heaven’ | 197
Spier

Plate 31 (Add. 51) Discoloured carnelian, Plate 32 (Add. 53) Garnet, broken in half. An Plate 33 (Add. 54) Garnet. Cross.
octagonal. Cross in wreath. Munich, Collection angel, holding a long cross. Munich, Collection Ex-auction
C.S., no. 2809 C.S., no. 2136

Plate 34 (Add. 55) Garnet. Box-type Greek Plate 35 (Add. 56) Garnet set in a fragmentary gilt-bronze Plate 36 (Add. 57) Garnet. Cross on steps.
monogram, cross above. Munich, Collection collar (from a buckle?). Probably from a Visigothic work- Munich, Collection C.S., no. 2189
C.S., no. 2500 shop. Munich, Collection C.S., no. 2299; said to be from Spain

examples, all now lost,20 had the fish-and-anchor motif on the
reverse. Another, with Jonah engraved on the reverse21 is also
obsidian, a material seldom used for personal seals at this time
but often for magical amulets.
Early Byzantine gems, 5th–7th centuries ad
Very few gems or cameos can be assigned to the late 4th or
early 5th centuries ad, and there can be little doubt that by this
time the art of gem engraving had nearly disappeared. A
revival did, however, occur around the middle of the 5th
century ad, when a few workshops, some associated with the
imperial court in Constantinople, produced gems of an entirely
new style. The most prolific 5th-century ad workshop
specialised in garnets and a few sapphires of a distinctive shape
previously not seen in Roman gem cutting, with sloping sides
and flat top (a cabochon, with its top ground down flat for
engraving). All the gems from this workshop, with over 60
published examples,22 are of this shape, although the sizes and
quality of engraving vary considerably. The earliest and finest
gems include the large and beautifully cut portrait of
Theodosius II, datable to c. ad 440, and other portraits of young
aristocratic men with crossbow fibulae on their shoulders that
denote their high status.23 Other gems are engraved with doves,
eagles, dolphins, and religious images, including the Virgin as
orant, the Prepared Throne, and crosses. The workshop also
produced gems engraved with monograms of the block-type
typical of the late 5th and early 6th centuries ad (Add. 55, Pl.
34). The portraits, the personalised nature of the monograms,
and the presence of some garnets set in gold rings discovered in
hoards in Italy suggest that many of these gems were presented
as imperial gifts.24 Joining a number of other garnets
discovered in the West is a gem engraved with the popular
image of a dove, set in a gold ring from a grave in Slovenia
(Add. 52). Another significant addition is a fragmentary garnet
engraved with an angel holding a long cross (Add. 53, Pl. 32),
an image frequently seen in the years around ad 500 and one
that is found on engraved glass gems that are related to the
garnets.25 Two garnets that have recently come to light appear,
however, to be products of a workshop of significantly later
date. The first gem (Add. 57, Pl. 36), set in a now fragmentary
ring, is of a shape somewhat different from the earlier garnets,
with straight, rather than curving, sides. The engraving is of a
cross, with perpendicular bars on the terminal of each arm,
standing on steps, an image that first appeared on the coinage
of Tiberius II (ad 578–82). It is unlikely that the gem dates
before this time. A small cabochon garnet is engraved with a
similar device, although not so finely executed, and must be of
similar date (Add. 58, Pl. 37).
Closely related to the garnets from the Garnet Workshop is
a group of slightly larger glass intaglios of similar shape and
subject matter (Add. 59–61, Pls 38–9). These gems were
typically set in belt buckles.26 An angel holding a cross is the
most common type, but there are also other Christian images,
including the Virgin as orant and the chi-rho monogram and
crosses, as well as a seated Constantinople similar to the gems
noted above. To this group can be added the remarkable find of

198 | ‘Gems of Heaven’


Plate 37 (Add. 58) Garnet, strongly Plate 38 (Add. 59) Brown glass. The
convex. Cross on steps. Munich, Virgin stands frontally, arms raised in
Collection C.S., no. 2847 prayer. Saltwood, Kent, from Anglo-
Saxon grave 4699 [156]
Plate 40 (Add. 63) The Crucifixion with the Virgin and John flanking the cross,
once set in the Annoschrein, Abbey of St Michael, Siegburg, Germany. Rock
crystal (?), cut down to about half the original size, c. 19 x 15mm
a brown glass intaglio engraved with the Virgin as orant
between two crosses discovered a few years ago in an Anglo-
Saxon grave at Saltwood in Kent (Add. 59, Pl. 38).27 A similar
image appears on another glass intaglio as well as on a garnet.28
A small number of unengraved garnets of this shape, certainly
of Byzantine origin, as well as some engraved Sasanian gems,
have been found in other Anglo-Saxon graves, a notable
indication of the great distance such prized objects could
travel.29
Very few Byzantine engraved gems, however, reached the
West. It is notable that among the hundreds of precious stones
in the 7th-century ad Guarazzar treasure from Spain, famous
for the votive crowns studded with large sapphires and garnets
with names of Visigothic kings (now in Madrid), there was only
one engraved gem, an emerald crystal depicting the
Annunciation, surely a contemporary Byzantine import.30 A
similar find of mostly fragmentary votive crosses and crowns
from Torredonjimeno in Andalusia (most of which are now in
the Museu d’Arqueologia de Catalunya in Barcelona) contained
another engraved gem set in one arm of a fragmentary votive
cross, an amethyst engraved with a cross (Add. 62). This, too,
may be a Byzantine import, although a local work is possible in
view of the existence of some crudely cut garnets of Visigothic
manufacture, imitating those from Constantinople.31 One
Late Antique and Early Christian Gems
Plates 39a–b (Add. 61) Brown glass in a bronze frame, probably from a belt
buckle. Greek box-type monogram between two crosses. Private collection
Plate 41 (Add. 66) Rock crystal. Christ healing the Blind Man: Christ stands,
raising his hand, while the blind man leans on his staff; border of thick lines.
Munich, Collection C.S., no. 2323
additional example of a Visigothic garnet engraved with a cross
has now come to light as well (Add. 56, Pl. 35).
Erika Zwierlein-Diehl has proposed an Early Byzantine
date for a now-lost intaglio (probably rock crystal) depicting
the Crucifixion that was once set in the 12th-century Anno-
schrein in the Abbey of St Michael in Siegburg, Germany (Add.
63, Pl. 40). The engraving is indeed similar to the linear style
seen on a number of Early Byzantine works, including the
emerald from the Guarazzar Treasure. On the lost gem, Christ
is nimbate, with his head tilted to the side, and wears a long
tunic (colobium); on either side of his head are a crescent and a
star (representing the moon and sun). Standing on either side
of the cross are the Virgin and John, a composition that
probably originated in the late 6th century ad. More difficult to
categorise is a garnet cameo set in the late 10th-century Crown
of the Golden Virgin preserved in the treasury of Essen
Cathedral (Add. 64). The engraving is very linear and angular,
showing the frontally facing bust and arms of man with large
eyes and pointed chin. Antje Krug has proposed a late-5th
century ad Gothic origin, but nothing like it is known. There
are vague similarities to portraits on engraved rings of the time
and to the engraving of the sapphire of Alaric, but perhaps a
date contemporary with the setting is more likely.
‘Gems of Heaven’ | 199
Spier
Plate 42 (Add. 67) Rock crystal, rectangular Plate 43 (Add. 68) Sardonyx in modern ring. Apollo
with back convex (cut down?), engraved with a stands before Daphne, who transforms into a tree.
cross; set face down in a gold ring. Private The god leans on a column on which a lyre rests, and
collection a goose stands before him. Derek Content Collection
Returning to Byzantium, two further examples can be
added to the output of the distinctive workshop that produced
engraved rock crystal pendants. These gems are all of similar
shape, flat on the engraved side and slightly convex on the
back; the engraving was inlaid with gold foil, and a second,
unengraved gem was placed on top and enclosed in a gold
frame. The image, reflected in the shining gold, was meant to
be viewed through the crystal. The engravings are very
summary and executed in a linear style without modeling, but
they nevertheless are easily recognisable as scenes from the life
and miracles of Christ (Add. 108, Pl. 57). Nearly 40 examples
survive,32 a surprisingly large number, although most have lost
their mounts. Two intact pendants, now in Dumbarton Oaks,
were discovered in a hoard of Early Byzantine jewellery. This
evidence, along with iconographical considerations, suggests a
late 6th- or early 7th-century ad date for the workshop and a
possible origin in Antioch. The individual hands of the
engravers can usually be recognised by their idiosyncratic
styles. One of the new examples, depicting Christ healing the
blind man (Add. 66, Pl. 41), is surely the work of the engraver
responsible for as many as 11 others, distinctive for his
rectangular heads with triangular, cap-like tops.33 Two other
engravers in the workshop treated the scene very differently,
showing Christ in frontal view.34 The other new rock crystal is
engraved only with a cross and is unique for being set, face
down, in a ring (Add. 67, Pl. 42). The gem does appear to be a
product of this workshop (others engraved with crosses are
known), but it may have been cut down to a rectangular form
and re-used in the ring.
Late Antique and Early Byzantine cameos
Late Antique and Early Byzantine cameos are rare and their
attribution often controversial, especially those purporting to
show imperial portraits. Small cameos, usually with only
inscriptions, are the most common variety of Byzantine cameo
and are more clearly datable. Another group of cameos is
remarkable for its range of pagan mythological imagery (thus
named the ‘Mythological Workshop’).35 The precise date and
location of the workshop are very difficult to establish, but the
distinctively stylised figures and the few surviving settings
200 | ‘Gems of Heaven’
Plate 44 (Add. 69) Sardonyx in gold frame.
Herakles, wearing a lion skin and a quiver over his
shoulder, shoots a bow; to the right a bird (one of
the Stymphalian birds). Derek Content Collection
suggest an Early Byzantine date. A number of important
additions to this group may be noted here, which again
tentatively suggest an Early Byzantine date. Four fine cameos
are in the Content Collection. The first, small cameo, depicts
Apollo standing before Daphne, who transforms into a laurel
tree (Add. 68, Pl. 43). Before the god stands a goose, a bird that
accompanies Apollo on several other cameos from this group.
Another cameo depicting Apollo and Daphne, of related but not
identical style, is set in a Byzantine gold pendant, now in
Dumbarton Oaks, which provides evidence for the date of the
first.36 Two further cameos, both by the same engraver and set
in similar gold frames that may have been strung on a
necklace, portray episodes from the cycle of Labours of
Herakles. One depicts Herakles shooting arrows at the
Stymphalian Birds (Add. 69, Pl. 44), one of which (again of
rather goose-like form) falls before him, and the other shows
Herakles subduing the Cretan Bull (Add. 70, Pl. 45). The
engraving, with its attention to stylised outlining of
musculature, links the cameos to the finer works of the group.
The gold mounts are not distinctive enough to provide a firm
date, but they could well be of Byzantine date. A fourth cameo
in the Content collection is small and simply cut but has the
charming device of Eros, seated on a rock, catching a fish on a
line (Add. 71, Pls 46a–b). The cameo is set in an unusual gold
ring composed of a hollow hoop with dolphin head terminals
and a circular box-bezel which sits on an openwork base. Rings
with hoops in the form of dolphins were present in the late 4th
century ad Thetford Treasure from England, and a ring
especially similar in shape was found in the late 5th century ad
Piazza della Consolazione treasure from Rome.37 This ring,
with its hollow construction and raised bezel, appears to be of
somewhat later date and is likely to be of Early Byzantine date.
Several more cameos, as well as evidence for their Early
Byzantine date, come from an unlikely source – the burial
mounds at Gamla Uppsala in Sweden, which are dated to the
6th or 7th century ad (Add. 72, Pl. 47).38 Four cameos were
discovered in the Västhögen mound, unfortunately all quite
damaged from burning; it is unclear how they had been
mounted or otherwise used. One of the better preserved
examples depicts an Eros blowing a horn, while another

Plate 45 (Add. 70) Sardonyx in gold frame. Plate 46a-b (Add. 71) Sardonyx (white), set in gold ring with dolphin head terminals; the bezel is
Herakles, a lion skin around his neck, grasps the a circular box resting on an openwork base. Eros sits fishing; a fish dangles from his line. Derek
head of the Cretan bull. Derek Content Collection Content Collection
Plate 47 (Add. 72) Four sardonyx cameos, all damaged by burning, from the
Västhögen mound in Gamla Uppsala, Sweden. The tomb is of a 6th–7th
century AD date. The images include an Eros blowing a horn; a fisherman (?);
a bull (part of a sacrificial scene?); and an unidentifiable scene
appears to show a fisherman, not unlike the Content example.
A third cameo, large but badly damaged, preserves only one
side, which depicts a bull very close in style to that on a
previously published cameo in the Content collection
portraying a woman sacrificing a bull39 and certainly a product
of the Mythological Workshop. The fourth cameo is too poorly
preserved to identify the image. Like other imported luxury
objects, such as helmets and swords, the Byzantine cameos
were obtained by the Swedish aristocrats by trade or as gifts
and likely served as signs of high status.
One group of superbly cut cameos was produced in a
workshop very likely located in Constantinople and associated
with the Byzantine imperial court in the 6th century ad.40 Of
the eight surviving examples, six are quite large, each c. 50mm,
and one preserves its original mount, a gold frame ringed with
small pearls suspended from a fine gold chain,41 certainly the
product of one of the finest Early Byzantine jewellery
workshops. Five of the cameos depict the Annunciation,
sometimes with the words of the angel engraved around the
Late Antique and Early Christian Gems
Plate 48 (Add. 73) Sardonyx cameo, bust of Christ, before which two angels
bow; around, ΑΓΙΟC, ΑΓΙΟC, ΑΓΙΟC, kc. 55 x 46mm. St Petersburg, State
Hermitage, inv. no. ω 373
scene, while the three others show two archangels flanking a
cross. To this group should be added a very fine sardonyx
cameo in St Petersburg, from the collection of Catherine the
Great (Add. 73, Pl. 48). On this cameo is engraved the frontally
facing bust of a nimbate, bearded Christ, who is flanked by two
angels, who bow in veneration. Engraved below is the Greek
phrase, ΑΓΙΟC ΑΓΙΟC ΑΓΙΟC kc (‘Holy, holy, holy, is the Lord’), the
words of the angels who sing in the presence of God in Isaiah’s
vision (Isaiah 6:3; and Revelation 4:8). Zalesskaia, who
published the cameo recently, has suggested a 12th-century ad
date, but the style and iconography fit better in the Early
Byzantine period. The same image, sometimes with the
accompanying inscription, is found engraved on a number of
rings of 6th- or early 7th-century ad date, as well as on an
embossed gold plaque, now in Naples.42 A fuller version of the
scene, in which Christ appears in a mandorla supported by
angels, is found on a contemporary tin-lead ampulla made in
the Holy Land.43 The size of the cameo (55mm) and style of
engraving link it closely to the other cameos in the group.
‘Gems of Heaven’ | 201
Spier

Plates 49a–b (Add. 76) Rock crystal stamp seal of conoid shape; pierced with gold mount in Plate 50 (Add. 77) Rock crystal stamp seal, similar to the
the form of a simple hoop with knob handle. Cross within hatched border. Private collection previous but without mount. Munich, Collection C.S., no. 2814

Plate 51 (Add. 81) Rock crystal stamp seal with Plate 52 (Add. 82) Rock crystal stamp seal Plate 53 (Add. 91) Garnet, convex, 14.2 x 11mm.
Syriac inscription (personal names?). Private with with Syriac inscription, ‘The Living The Sacrifice of Isaac, Hebrew inscription, Shmuel
collection Passion’. Private collection bar Yehudah. Derek Content Collection

Christian Sasanian gems
In contrast to the marked decline of gem engraving in the late
Roman Empire, vast numbers of gems were being produced
within Sasanian Persia. Among the many thousands of extant
gems, nearly all of which were the property of Persians of
traditional Zoroastrian faith, is a small number of seals that
belonged to Christians and Jews. These gems are carved in
purely Sasanian style but are identifiable by their devices or
inscriptions. Nearly 200 examples have been published, and no
doubt many more lie unnoticed in unpublished collections of
Sasanian gems. Popular Christian images include the Sacrifice
of Isaac44 and Daniel in the Lions’ Den.45 Less common are
images of the cross, angels, or other biblical scenes. Some gems
are identifiable only by their inscriptions. Hebrew inscriptions
(always personal names) confirm a Jewish origin, as Syriac
inscriptions must be Christian. Many gems, however, are
inscribed with the Middle Persian script used by most
Sasanians, but the names or phrases are specifically Christian.
One distinctive group of seals appears to be the product of a
Christian workshop in Mesopotamia.46 They are rock crystals
of a conoid shape not typical of Sasanian seals and are
engraved with a cross of distinctive form, with arms
terminating in pairs of circular elements. A number are
inscribed in Syriac, the language of Syrian and Mesopotamian
Christians. A relatively large number have been published
(over 20), and seven further examples can be added (Add.
76–82, Pls 49–52), including two with inscriptions (the first a
personal name not yet translated and the other the frequent
phrase ‘The Living Passion’) and the first to survive with its
original gold mount, a simple gold ring with a small knob
handle. The workshop, or one related to it, also produced some
ringstones of haematite and black jasper,47 and two further
examples have now appeared (Add. 83–4).
Jewish seals
Clearly identifiable Jewish seals of Roman and Early Byzantine
date are very rare. The only gems of certain Jewish origin are
ones with inscriptions or those engraved with the menorah
(the seven-branched candlestick) accompanied by the lulav
(bunch of branches) and etrog (citrus fruit), the symbols of the
harvest festival of Sukkoth. These gems, often very skillfully
made and close in style, shape and material to contemporary
pagan Roman examples, are likely of relatively early date,
belonging to the 2nd and 3rd centuries ad. Several very fine
specimens can be added here, including a 3rd-century ad
nicolo from Portugal (Add. 85) and a fragmentary nicolo from
Hungary (Add. 86).
Jewish rings are also known, most of bronze and usually
engraved with the device of the menorah. One example found
in the excavations at Kaiseraugst in Switzerland in 2001 has
prompted an excellent study.48 Also notable is a 3rd-century ad
gold ring of conventional Roman type and very fine quality but
inscribed in Greek with the name of its Jewish owner, ΙΟΥΔΑΖ,
Ioudas.49 Byzantine rings of the 6th and 7th centuries ad
engraved with the menorah or Jewish inscriptions are also
known, as are an increasing number of lead sealings (both one-
sided of 4th-century ad date and double-sided of the 5th–7th
centuries ad); these warrant further study.50

202 | ‘Gems of Heaven’


Plate 54 (Add. 98) Banded agate with Plate 55 (Add. 99) Rock crystal with the apostles standing
the Virgin standing as orant, cross above. around a large cross. Formerly German eBay
Formerly German eBay
The surviving engraved gems of the Jews living in the
Sasanian Empire are more numerous than those from Roman
territory but had been largely overlooked until the excellent
work of Shaul Shaked brought most of them together.51 The
previous corpus of gems raised the total to about 30. The
menorah never appears on Sasanian Jewish gems, and the
most common image is the lulav and etrog. Figural devices are
rare, but the Sacrifice of Isaac does appear on a seal in
Brussels.52 To the previously published group, four more can be
added (Add. 88–91, Pl. 53),53 including two with the Sacrifice
of Isaac, the first a chalcedony with the name Yaakov bar
Yehuda (Add. 89), and the second a garnet in the Content
collection engraved in very simple style with the name Shmuel
bar Yehudah (Add. 91, Pl. 53).
Uncertain gems and forgeries
Finally, the question of forgeries must be considered. Quite a
few gems of doubtful origin are listed in the previous
publication, and no doubt many more can be added.54 It is often
difficult to categorise these gems, some of which are
incorrectly attributed (that is, of uncertain date and purpose)
and other outright forgeries intended to deceive the collector.
Some copies of ancient gems are as old as the 16th century, and
forgeries continue to be engraved today (workshops in Beirut
being particularly prolific, as they have been since the 1950s).
The additional gems listed in the catalogue here are examples
that cause suspicion, but I have seen only photos and cannot
confirm that they are indeed forgeries.
A recently published emerald engraved with the chi-rho
monogram within a wreath was acquired by the National
Museum of Ireland from the collection of the Duke of Leinster
in 1888 (Add. 92). The style and material are not typical of
Early Christian works, and the gem may be one of several
engraved in the early 18th century, such as the example, once
in the Passeri collection, published in 1750.55 Suspicious gems of
presumably recent origin continue to appear at auction (Add.
93–100), including some forgeries that may well be the result
of my own publications. Three gems that appeared on German
eBay are all very likely false, probably made in Beirut and
copied from photos of genuine gems I posted on the Beazley
Archive’s website.56 The first (Add. 98, Pl. 54) is a copy in
banded agate of a standing Virgin as orant from the garnet
workshop;57 the second (Add. 99, Pl. 55) a copy in rock crystal
Late Antique and Early Christian Gems
Plate 56 (Add. 100) Lapis lazuli with SS Peter and
Paul standing on either side of a cross surmounted
by a bust of Christ. Formerly German eBay
of a famous, large banded agate in Munich;58 and the third
(Add. 100, Pl. 56) a copy in lapis lazuli (the favourite stone of
the Lebanese forger) of a famous gem now in Krakow.59 It is
gratifying that someone pays attention to our scholarship.
Catalogue
Additions to J. Spier, Late Antique and Early Christian Gems,
Wiesbaden, 2007
Portraits
Add. 1 (Pl. 1): Gorny & Mosch. Giessener Münzhandlung, Munich,
Auktion 168, 24 June 2008, lot 156. Silver disc, diademed busts of
Constans and Constantius II(?), crowned by Victory, dd nn.
Add. 2 (Pl. 2): Munich, Collection C.S., no. 2175. Silver disc with facing
emperor in crested helmet.
Add. 3: Munich, Collection C.S., no. 2212. Silver disc with male bust
wearing short beard and military dress, lic cons. A. Demandt and J.
Engemann (eds), Konstantin der Grosse, Trier, 2007, cat. no. I.7.10.
Add. 4 (Pl. 3): Munich, Collection C.S., no. 2229. Silver disc with facing
busts of man and woman, viva. Demandt and Engemann ibid., cat. no.
II.1.74.
Add. 5 (Pl. 4): Munich, Collection C.S., no. 2751. Silver disc with facing
busts of a man and woman, uncertain inscription.
Add. 6 (Pl. 5): Munich, Collection C.S., no. 2122. Silver disc with busts of
a man, woman, and child, vivatis in deo. Demandt and Engemann
ibid., cat. no. II.1.53.
Add. 7: Gorny & Mosch. Giessener Münzhandlung, Munich, Auktion 150,
11 July 2006, lot 280. Silver disc with busts of a man and woman,
lavrentialionis.
Add. 8: Gorny & Mosch. Giessener Münzhandlung, Munich, Auktion 163,
14 December 2007, lot 219. Silver disc with bust of a man and woman,
cercvs et vebica v.
Add. 9: Munich, Collection C.S., no. 1369. Silver disc with facing bust of
woman, vivas in deo. Demandt and Engemann ibid., cat. no. II.1.58.
Add. 10 (Pl. 6): Munich, Collection C.S. no. 2169. Silver disc with male
bust, fibula on shoulder, ΕΙC [Θ]ΕΟC. Demandt and Engemann ibid., cat.
no. II.1.69.
Add. 11 (Pl. 7): Munich, Collection C.S., no. 2171. Silver disc with male
bust, anastasi vivas. Demandt and Engemann ibid., cat. no. II.1.71.
Add. 12 (Pl. 8): Munich, Collection C.S., no. 2172. Silver disc with facing
male bust, iovine vivas. Demandt and Engemann ibid., cat. no. II.1.72.
Add. 13: Sirmium excavations, found in the public bath. Silver disc with
frontally facing male bust, zosime vivas. F. Baratte, ‘Un médaillon
d’argent du bas-empire à Sirmium’, Mélanges de l’École Française de
Rome 87 (1975), 413–18.
Add. 14 (Pl. 9): Harlan J. Berk, Ltd., Chicago, Auction 158, February 2008,
lot 371. Silver disc with male bust, dalmativs.
Add. 15 (Pl. 10): Private collection; eBay 200309627340, 19 February
2009. Lead sealing, 13mm, impressed from a silver disc or ring of
similar style. Male bust, … e vivas.
‘Gems of Heaven’ | 203
Spier
Add. 16: Munich, Collection C.S., no. 2237. Silver disc with chi-rho in
wreath, CON. Demandt and Engemann ibid., cat. no. II.1.77.
Add. 17 (Pl. 11): Munich, Collection C.S., no. 2182. Silver disc with
standing Victory and kneeling Genius, vivas in deo. Demandt and
Engemann ibid., cat. no. II.1.73.  35, no. 64.
Add. 18 (Pl. 12): Munich, Collection C.S., no. 2336. Silver disc with the
Sacrifice of Isaac. Demandt and Engemann ibid., cat. no. II.1.76.
Add. 19 (Pl. 13): Munich, Collection C.S., no. 2142. Silver disc with
Daniel in the Lions’ Den and the Sacrifice of Isaac. Demandt and
Engemann ibid., cat. no. II.1.47.
Add. 20 (Pl. 14): Munich, Collection C.S., no. 2631. Amethyst. Portrait of
bearded man, fibula on shoulder, cross above, CΑΤΟΡΝΙΛΟΥC,
Satornilous.
Add. 21 (Pl. 15): Derek Content Collection. Carnelian in gold mount.
Seated Constantinople, holding a cross on globe.
Add. 22 (Pl. 16): Munich, Collection C.S., no. 2671. Green chalcedony.
Seated Constantinople holding a cross on globe.
Gems with Christian inscriptions
Add. 23 (Pls 17a–b): Munich, Collection C.S., no. 2365; from
Auktionshaus H.D. Rauch, Vienna, Kunstobjekte der Antike, 27 January
2007, lot 398. Carnelian, F3, in silver ring. ihcoy.
Add. 24: Rome, Patrimonio Archeologico (confiscated from a private
collection as illegal in 2004 with numerous archaeological objects).
Carnelian, F3,ihcoy. C. Gasparri, ‘Un Tesoro recuperate. Gemme, vetri
e lavori in pietra dura da una collezione privata’, in G. Sena Chiesa and
E. Gagetti (eds), Aquileia e la glittica di età ellenistica e romana, Trieste,
2009, 281–92, pl. 1, fig. 16, no. 99.28.
Add. 25 (Pl. 18): Private collection; formerly USA market (2010).
Carnelian, F3. ihcoy.
Add. 26 (Pls 19a–b): Munich, Collection C.S., no. 2345. Carnelian, F3,
cut octagonally, in a gold ring. xpictoy. Demandt and Engemann ibid.,
cat. no. II.1.124;
Add. 27: Munich, Collection C.S., no. 2429. Carnelian, 6 x 7mm.
xpictoy.
Add. 28 (Pl. 20): Private collection, 2009. Green-brown chalcedony, F3,
9.5 x 7.7 x 3.4mm. xpictoy.
Add. 29: Private collection, from Carnuntum. Carnelian in iron ring.
Chi-rho. G. Dembski, Die antiken Gemmen und Kameen aus Carnuntum,
Vienna, 2005, no. 1136.
Add. 30 (Pl. 21): Munich, Collection C.S., no. 1890; from the Balkans?
Carnelian, F3, in gold ring with beaded wire hoop. Chi-rho. L. Wamser
(ed), Die Welt von Byzanz: Europas östliches Erbe, Munich, 2004, 333,
no. 690.
Add. 31 (Pl. 22): Munich, Collection C.S., no. 2478; formerly eBay
260237075850, 11 May 2008; said to have been purchased from Malter
Galleries, Encino, California. Red jasper, F3, octagonal. Chi-rho-tau
monogram.
Add. 32 (Pl. 23): Munich, Collection C.S., no. 2370; from Auktionshaus
H.D. Rauch, Vienna, Kunstobjekte der Antike, 27 January 2007, lot 417.
Banded agate, orange and white. Chi-rho monogram combined with
staurogram.
Add. 33 (Pl. 24): Munich, Collection C.S., no. 2846; from Bonham’s,
London, Antiquities, 28 April 2010, lot 247. Yellow jasper, F2, in an
ancient gold ring that may not belong. ΙΧΘΥC and star.
Christian symbols (fish, anchor, dove, and ship)
Add. 34 (Pl. 25): Munich, Collection C.S., no. 2432. Carnelian,
octagonal, F3. Two fish flank an anchor, ΙΧΘΥC.
Add. 35: Derek Content Collection. Carnelian, A4, of fine style. Two
fish flank an anchor, ΙΧΘΥC .
Add. 36 (Pl. 26): Munich, Collection C.S., no. 2145; from Bulgaria?
Carnelian, F3, in silver ring. Two fish flank an anchor; above, IAWIH in
wreath. Wamser ibid., 333, no. 686; Demandt and Engemann ibid., cat.
no. II.1.25.
Add. 37: Museum Carnuntinum, inv. no. 17.902. Carnelian. Two fish
flank anchor. Dembski ibid., no. 959.
Add. 38: Museum Carnuntinum (Traun collection), Nicolo. Two fish
flank anchor. Dembski ibid., no. 960.
Add. 39: Traun collection, from Carnuntum. Carnelian. Two fish flank
anchor. Dembski ibid., no. 961.
Add. 40: Aquileia, Museo Archeologico Nazionale. Chalcedony. Two
fish flank cross.
204 | ‘Gems of Heaven’
Add. 41: Regional Museum of Postojna (Slovenia); from a cremation
burial (grave 41) at Cerknica, Svinja Gorica. Brown glass in silver ring.
Two fish flank an anchor (very crude). A. Nestorović, Images of the
World Engraved in Jewels. Roman Gems from Slovenia, Ljubljana, 2005,
Add. 42: Seville, Museo Arqueológico. Carnelian. Two fish flanking a
cross. M.D. López de la Orden, La Glíptica de la Antiguedad en
Andalucía, Cádiz, 1990, 172, no. 194.
Add. 43: Munich, Collection C.S., no. 2642. Carnelian, F3. Two fish
flank a cross.
Add. 44: Private collection; previously with Jean-Philippe Mariaud de
Serres, Paris. Rock crystal, c. 13.5 x 9.8mm. Two fish flank a cross.
The Good Shepherd
Add. 45 (Pls 27a–b): Private collection; from Christie’s, New York,
Ancient Jewelry, 8 December 2005, lot 105. Nicolo in a heavy gold ring.
The Good Shepherd stands with a sheep over his shoulders.
Add. 46: Gorny & Mosch, Giessener Münzhandlung, Munich, Kunst der
Antike, Auction 189, 23 June 2010, lot 141. Nicolo in silver ring. The Good
Shepherd.
Add. 47: Oxford, Ashmolean Museum, inv. no. 2003.108. Red jasper, F1,
re-used and engraved on both sides. Side A is older, engraved with a
ship. Side B: the Good Shepherd, one sheep at feet, with uncertain
Latin inscription, nekasre. M. Henig and A. MacGregor, Catalogue of
the Engraved Gems and Finger-Rings in the Ashmolean Museum, II.
Roman, Oxford, 2004, 131, no. 14.20.
Add. 48 (Pl. 28): Munich, Collection C.S., no. 2673. Carnelian, F3,
octagonal, broken. The Good Shepherd with staff.
Add. 49 (Pl. 29): Derek Content Collection. Carnelian, F2–3. The Good
Shepherd.
Add. 50 (Pls 30a–b): Private collection. Black obsidian, F1, fragmentary,
with nearly half lost; length: 34mm. Both sides are engraved. Side A:
the Good Shepherd, sheep at feet; around, …C ΘΕΟΥ ΥΙΟC. Side B: two
fish flank an anchor; around, …CWΤΗ… .
Other Christian images
Add. 51 (Pl. 31): Munich, Collection C.S., no. 2809. Discoloured
carnelian, octagonal. Cross in wreath.
The Garnet workshop, late 5th century AD
Add. 52: Ljubljana, National Museum of Slovenia; inv. no. S 2543; from
Gradec near Velika Strmica. Garnet. Dove, set in a gold ring with
granulation on the shoulders. Nestorović ibid., 34, no. 58.
Add. 53 (Pl. 32): Munich, Collection C.S., no. 2136.  Garnet, broken in
half. An angel, holding a long cross. M. Fansa and B. Bollmann (eds),
Die Kunst der frühen Christen in Syrien. Zeichen, Bilder und Symbole
vom 4. bis 7. Jahrhundert, Mainz, 2008, no. 152.
Add. 54 (Pl. 33): Münzenhandlung Gerhard Hirsch Nachfolger, Munich,
Auction 262, 22–26 September, 2009, lot 1389. Garnet. Cross.
Add. 55 (Pl. 34): Munich, Collection C.S., no. 2500. Garnet. Box-type
Greek monogram, cross above.
Add. 56 (Pl. 35): Munich, Collection C.S., no. 2299; said to be from
Spain. Garnet set in a fragmentary gilt-bronze collar (from a buckle?).
Probably from a Visigothic workshop.
Add. 57 (Pl. 36): Munich, Collection C.S., no. 2189. Garnet. Cross on
steps. Fansa and Bollmann ibid., no. 151.
Add. 58 (Pl. 37): Munich, Collection C.S., no. 2847. Garnet, strongly
convex. Cross on steps.
Add. 59 (Pl. 38): Saltwood, Kent, from Anglo-Saxon grave 4699 [156].
Brown glass. The Virgin stands frontally, arms raised in prayer.
Publication by Penelope Rogers is forthcoming.
Add. 60: Corning, New York, Corning Museum of Glass, inv. no.
59.1.288; formerly Smith collection. Brown glass. Two doves flank a
cross. D. Whitehouse, Roman Glass in the Corning Museum of Glass, vol.
3, Corning, New York, 2003, 69, no. 1007.
Add. 61 (Pl. 39): Private collection. Brown glass in a bronze frame,
probably from a belt buckle. Greek box-type monogram between two
crosses.
Fine quality gems, 5th–7th centuries ad
Add. 62: Barcelona, Museu d’Arqueologia de Catalunya, inv. 25097;
from the Torredonjimeno treasure. Amethyst engraved with a cross,
set in a fragment of a gold votive cross. A. Casanovas and J. Rovira i

Port (eds), Torredonjimeno. Tesoro, Monarquía y Liturgia, Barcelona,
Córdoba, Madrid, and Jaén, 2003, 130, no. 34; and A. Perea (ed.), El
Tesoro visigodo de Torredonjimeno, Madrid, 2009, 98–100, fig. 62.
Add. 63 (Pl. 40): Once Siegburg, Germany, Abbey of St. Michael; set in
the Annoschrein. Rock crystal (?), cut down to about half the original
size, c. 19 x 15mm. The Crucifixion with the Virgin and John flanking
the cross. E. Zwierlein-Diehl, ‘Ein karolingischer Saphir mit
Dreierkopf, ein frühbyzantinisher Bergkristall mit Kreuzigung und
andere Gemmen vom Annoschrein’, Kölner Domblatt (2007), 27–48,
figs 13–16, as Early Byzantine.
Add. 64: Essen, Cathedral Treasury, set in the Crown of the Golden
Virgin. Garnet cameo with a facing bust and arms in highly linear style.
A. Krug, ‘Antike Gemmen und das Zeitalter Bernwards’, in M. Brandt
(ed.), Bernward von Hildesheim und das Zeitalter der Ottonen: Katalog
der Ausstellung, Hildesheim 1993, vol. 1, Mainz, 1993, 161–72, fig. 47, as
Gothic, late 5th century ad .
Add. 65: Museum Carnuntinum, inv. no. 18.093; purchased in 1966 (but
is this likely to be from Carnuntum?). Glass imitating nicolo. A seated
figure holding a cross. Dembski ibid., no. 628. For the group of glass
intaglios imitating nicolo with similar imagery, see Spier (n. 1), nos
609– 15bis.
Rock crystal pendants, 6th–7th centuries ad
Add. 66 (Pl. 41): Munich, Collection C.S., no. 2323. Rock crystal. Christ
healing the Blind Man: Christ stands, raising his hand, while the blind
man leans on his staff; border of thick lines. Fansa and Bollmann ibid.,
no. 153.
Add. 67 (Pl. 42): Private collection. Rock crystal, rectangular with back
convex (cut down?), engraved with a cross; set face down in a gold ring.
J. Spier, ‘Some Unconventional Early Byzantine Rings’, in C. Entwistle
and N. Adams (eds), ‘Intelligible Beauty’. Recent Research on Byzantine
Jewellery, British Museum Research Publication Series no. 178,
London, 2010, 17–18, pls 18a–b.
Cameos
Add. 68 (Pl. 43): Derek Content Collection. Sardonyx in modern ring.
Apollo stands before Daphne, who transforms into a tree. The god
leans on a column on which a lyre rests, and a goose stands before him.
Add. 69 (Pl. 44): Derek Content Collection. Sardonyx in gold frame.
Herakles, wearing a lion skin and a quiver over his shoulder, shoots a
bow; to the right a bird (one of the Stymphalian birds).
Add. 70 (Pl. 45): Derek Content Collection. Sardonyx in gold frame.
Herakles, a lion skin around his neck, grasps the head of the Cretan
Bull.
Add. 71 (Pls 46a–b): Derek Content Collection. Sardonyx (white), set in
gold ring with dolphin head terminals; the bezel is a circular box
resting on an openwork base. Eros sits fishing; a fish dangles from his
line.
Add 72 (Pl. 47): Four sardonyx cameos, all damaged by burnings, from
the Västhögen mound in Gamla Uppsala, Sweden. The tomb is of a
6th-7th century ad date. The images include an Eros blowing a horn; a
fisherman (?); a bull (part of a sacrificial scene?); and an unidentifiable
scene. B. Arrhenius, ‘Regalia in Svealand in Early Medieval Times’, Tor
27 (1995), 321–5.
Add. 73 (Pl. 48): St Petersburg, State Hermitage, inv. no. ω 373; from the
collection of Catherine the Great. Sardonyx cameo. 55 x 46mm. Bust of
Christ, before which two angels bow; around, ΑΓΙΟC, ΑΓΙΟC, ΑΓΙΟC, KC.
Zalesskaia, in F. Althaus and M. Sutcliffe (eds), The Road to Byzantium,
London, 2006, 165, no. 102 (who suggests a 12th-century ad date).
Add. 74: Jonathan P. Rosen collection. Haematite, 12.8 x 11.3mm. The
Sacrifice of Isaac, with the hand of God above. Friedenberg (n. 53),
35–6, no. 15.
Add. 75: Jonathan P. Rosen collection. Banded agate, 17.4 x 12.7mm.
The Sacrifice of Isaac, abbreviated (without Isaac). Friedenberg (n. 53),
38, no. 20.
Add. 76 (Pls 49a–b): Private collection. Rock crystal stamp seal of
conoid shape; pierced with gold mount in the form of a simple hoop
with knob handle. Cross within hatched border.
Add. 77 (Pl. 50): Munich, C.S. collection, no. 2814. Rock crystal stamp
seal, similar to the previous but without mount.
Add. 78: Christie’s, New York, Ancient Jewelry, 6 December 2007, lot 479.
Another, similar to previous.
Add. 79: Derek Content Collection. Another, similar.
Late Antique and Early Christian Gems
Add. 80: Private collection. Another, similar.
Add. 81 (Pl. 51): Private collection. Another, similar, with Syriac
inscription (personal names?)
Add. 82 (Pl. 52): Private collection. Another, similar, with Syriac
inscription, ‘The Living Passion’.
Add. 83: Marseille, Musée Borély. Haematite, ‘cachet en forme
d’anneau’, 25.5 x 18.9mm. Cross of Mesopotamian type (like that on the
rock crystals). Arts de l’ancien Iran. Musée Borély, Marseille, 1975, 116–
17, no. 319.
Add. 84: Derek Content Collection. Black jasper, rectangular. Cross of
Mesopotamian type.
Jewish seals
Add. 85: Lisbon. Nicolo. Menorah, lulav, etrog, and shofar. G. Cravinho
and S. Amorai-Stark, ‘A Jewish Intaglio from Roman Ammaia,
Lusitania’, Liber Annuus 56 (2006), 521–46.
Add. 86: Once Szombathely (Hungary), Savaria Múzeum, inv. 5476485,
lost after 1944. Fragmentary nicolo, preserving the upper half.
Menorah with flaming lamp on each arm. Berger (n. 48), 87–9, F3, fig.
27 (with earlier references).
Add. 87: Gorny & Mosch, Giessener Münzhandlung, Munich, Kunst der
Antike, Auction 189, 23 June 2010, lot 202. Glass, F1. Menorah, lulav, and
etrog.
Add. 88: Private collection. Garnet, convex, 13 x 10.5mm. Lulav and
etrog, Hebrew inscription, Levi bar Shimon. Friedenberg (n. 53), 32–3,
no. 11.
Add. 89: Private collection. Chalcedony, diam. 15.5mm. The Sacrifice of
Isaac, Hebrew inscription, Yaakov bar Yehuda. Friedenberg (n. 53), 34,
no. 13.
Add. 90: Jonathan P. Rosen collection. Banded agate, 14.8 x 11.6mm.
Lion mask, Hebrew inscription, Ibrami (?). Friedenberg (n. 53), 47, no.
36.
Add. 91 (Pl. 53): Derek Content Collection, Garnet, convex, 14.2 x
11mm. The Sacrifice of Isaac, Hebrew inscription, Shmuel bar Yehudah.
Misattributions, forgeries, and uncertain works
Add. 92: Dublin, National Museum of Ireland, inv. no. 1888:250; from
the collection of the Duke of Leinster. Emerald in modern ring. Chi-rho
monogram within wreath. L. Mulvin, Roman and Byzantine Antiquities
in the National Museum of Ireland: a Select Catalogue, Bray, 2006, 213,
no. 262.
Add. 93: Christie’s, New York, Ancient Jewelry, 6 December 2007, lot 343.
Banded agate. Menorah, lulav, etrog, and shofar. The style is odd, and
the date is uncertain.
Add. 94: Münzenhandlung Gerhard Hirsch Nachfolger, Munich, Auction
257–8, 23–7 September, 2008, lot 1467. Lapis lazuli engraved with
cruciform monogram, set in a gold ring. Probably recent.
Add. 95: Münzenhandlung Gerhard Hirsch Nachfolger, Munich, Auction
257–8, 23–7 September, 2008, lot 1659. Rock crystal. Facing bust of St
Peter. Probably recent.
Add. 96: Münzenhandlung Gerhard Hirsch Nachfolger, Munich, Auction
257–8, 23–7 September, 2008, lot 1678. Carnelian. Menorah. Probably
recent.
Add. 97: Gorny & Mosch. Giessener Münzhandlung, Munich, Auction
174, 16 December 2008, lot 199. Banded agate. Staurogram, alpha and
omega. Probably modern.
Add. 98 (Pl. 54): Formerly German eBay. Banded agate. The Virgin
stands as orant, cross above.
Add. 99 (Pl. 55): Formerly German eBay. Rock crystal. The apostles
stand around a large cross.
Add. 100 (Pl. 56): Formerly German eBay. Lapis lazuli. Peter and Paul
stand on either side of a large cross surmounted by a bust of Christ.
Engraved rings
Numerous rings with Christian images or inscriptions, primarily of
4th- and 5th-century ad date continue to appear. The following are a
selection of special interest:
Add. 101: For bronze rings with chi-rho, probably from Rome, in the
Vettori collection in the late 18th century ad , see F. Vettori, Nummus
aereus veterum christianorum, Rome, 1737, 52.
Add. 102: For a large selection of Early Christian rings, see Demandt
and Engemann ibid., cat. nos II.1.10-II.1.78 and II.1.124–5.
‘Gems of Heaven’ | 205
Spier
Plate 57 (Add. 108) The Adoration. The three Magi with Phrygian caps
approach the seated Virgin, who holds the nimbate Christ child on her lap; a
cross is above. Rock crystal, inlaid with gold and sandwiched with another in a
silver pendant with beaded wire frame. Ex-Gorny & Mosch, Munich
Add. 103: Kaiseraugst (Switzerland) excavations, inv. no. 1999.01.
Bronze ring. Ship and chi-rho. Berger (n. 47), 50–51, fig. 19.
Add. 104: Jet ring engraved with staurogram. T. Graham, ‘A Rho-cross
engraved on a jet finger-ring from Bagshot, Surrey’, Oxford Journal of
Archaeology 21 (2002), 211–16.
Add. 105: Brussels, Musées Royaux d’Art et d’Histoire, inv. B.1766. A
gold ring, the hoop engraved vtere felix and chi-rho, set with an older
nicolo with Bonus Eventus. K. Sas and H. Thoen (eds), Schone Schijn.
Brillance et Prestige, Leuven, 2002, 243, no. 231.
Lead sealings
Add. 106: Trier, Rheinisches Landesmuseum, inv. no. EV 1994,185 e (PK
1306). The Good Shepherd. Demandt and Engemann ibid., cat. no.
I.13.111.
Add. 107: Trier, Rheinisches Landesmuseum, inv. no. EV 1994,257 (PK
809). Noah in the ark. H.-J. Leukel, Römische Plomben aus Trierer
Funden 1995–2001, Trier, 2002, no. 161; Demandt and Engemann ibid.,
cat. no. I.13.112.
Further additions
Add. 108 (Pl. 57): Gorny & Mosch, Giessener Münzhandlung, Munich,
Kunst der Antike, Auction 189, 23 June 2010, lot 161. Rock crystal, inlaid
with gold and sandwiched with another in a silver pendant with
beaded wire frame. The Adoration. The three Magi with Phrygian caps
approach the seated Virgin, who holds the nimbate Christ child on her
lap; a cross is above.
The rock crystal pendant depicting the Adoration is similar to two
others published previously (Spier 2007, nos 667–8). The mount is the
only recorded example in silver of a type well represented in gold.
Add. 109 (Pl. 58): Derek Content Collection. Sardonyx cameo set in a
heavily encrusted silver ring with circular box-bezel and cylindrical
hoop. Aphrodite Kallipygos turns to a swan with raised wings. The
cameo is another deriving from the Mythological Group. Aphrodite
Kallipygos is named for a famous statue type of 5th century bc date.60
Interest in classical mythological themes is characteristic of the
workshop. The cameo cutter again displays his liking for swans and
geese, which appear on several others of his works. The ring in which
the cameo is set is an Early Byzantine variety, providing further
206 | ‘Gems of Heaven’
Plates 58a–b (Add. 109)
Sardonyx cameo set in a
heavily encrusted silver ring
with circular box-bezel and
cylindrical hoop. Aphrodite
Kallipygos turns to a swan with
raised wings. Derek Content
Collection
evidence for the Byzantine date of the cameos.
Add. 110: Sasanian Christian seal impressions with Daniel in the Lions’
Den, from Ak-Depe (Turkmenistan). A. G. Gubaev, S. D. Loginov and A.
B. Nikitin, ‘Sasanian Bullae from the Excavations of Ak-Depe by the
Station of Artyk’, Iran 34 (1996), 55–9.
Add. 111: Munich, C.S. collection. Sasanian clay bulla with multiple
impressions, including a cross with Pahlavi inscription. L. Wamser,
(ed.), Die Welt von Byzanz-Europas östliches Erbe (Exh. cat., Munich,
2004), 348, no. 762.
The clay impressions of Sasanian gems engraved with Daniel are
important for their provenance and for demonstrating the presence of
Christians at Ak-Depe in Turkmenistan.61 The Christian seal on the
bulla in Munich displays a cross of fine style, but its inscription has not
yet been read.
Further information on previously published gems
Spier 2007, no. 83. Vienna, Kunsthistorisches Museum. Sapphire, inv.
no. VII B 23. The sapphire seal of Alaric. G. Kornbluth, ‘The Seal of
Alaric, rex Gothorum’, Early Medieval Europe 16 (2008), 299–332.
Spier 2007, no. 151. Perugia, Museo Archeologico Nazionale
dell’Umbria, inv. no. 1745. Bloodstone. ΙΧΘΥC, monograms of
Christ, and magical symbols. P. Vitellozzi, Gemme e Magia dalle
collezioni del Museo Archeologico Nazionale dell’Umbria, Perugia,
2010, 66–7, A24.
Spier 2007, no. 330. Rome, Campo Santo Teutonico, inv. G4. Garnet.
The Good Shepherd and a dove standing on the ark. Demandt and
Engemann ibid., cat. no. II.3.16
Spier 2007, no. 596. Formerly Campbell Bonner collection, is now
University of Michigan, Ann Arbor, Taubman Medical Library, no.
330.
Spier 2007, no. 753. London, Victoria & Albert Museum, inv. 580–1871.
Cameo inscribed,vibas lvxvrihomobone. Catalogue de la
précieuse collection d’objets d’art…de feu M. Louis Fould, Paris, 4
June 1860, lot 962.
Spier 2007, no. 970. New Haven, Yale University Babylonian
Collection, NCBS 952. Garnet. Lulav and etrog, with Hebrew
inscription, Shmuel. Friedenberg (n. 53), 32, no. 10.
Spier 2007, no. 983. Private collection. Garnet. Ram and Hebrew
inscription, ‘Uqba bar Papa. Friedenberg (n. 53), 49, no. 39.
 Late Antique and Early Christian Gems

Notes one of which is inscribed around the edge of the bezel, ‘Holy, holy,
1 J. Spier, Late Antique and Early Christian Gems, Wiesbaden, 2007. holy is the Lord of Hosts’; see O. M. Dalton, Catalogue of Early
2 Zwierlein-Diehl in this volume. Christian Antiquities and Objects from the Christian East in the
3 Spier (n. 1), 4–14. Department of British and Mediaeval Antiquities and Ethnography
4 Ibid., no. 83. of the British Museum, London, 1901, 19 and 30, nos 120 and 189. For
5 Ibid., nos 35–69. a silver ring, see S. de Ricci, Catalogue of a Collection of Ancient
6 Ibid., nos 72–8. Rings formed by the late E. Guilhou, Paris, 1912, 97, no. 835, pl. 13.
7 E. Zwierlein-Diehl, ‘Constantinopolis et Roma. Intailles du IVe et For the gold plaque in Naples, see Y. Christe, La Vision de Matthieu
du Ve siècle après Jésus-Christ’, in M. Avisseau Broustet (ed.), La (Matth. XXIV–XXV). Origines et développement d’une image de la
glyptique des mondes classiques, Paris, 1997, 83–96; and see also Second Parousie, Paris, 1973, 17, fig. 16.
Spier (n. 1), nos 28–34. 43 A. Grabar, Ampoules de Terre Sainte (Monza–Bobbio), Paris, 1958,
8 Spier (n. 1), no. 28. 33–4, pl. 33.
9 Ibid., no. 32. 44 Spier (n. 1), nos 803–38.
10 Ibid., nos 86–9. 45 Ibid., nos 839–52.
11 Ibid., nos 90–110. 46 Ibid., nos 909–29.
12 Ibid., nos 112–27. 47 Ibid., nos 934–5.
13 Ibid., nos 133–43. 48 L. Berger, Der Menora-Ring von Kaiseraugst. Jüdische Zeugnisse
14 J. Spier, Picturing the Bible. The Earliest Christian Art, New Haven römischer Zeit zwischen Britannien und Pannonien. The Kaiseraugst
and London, 2007, 233–4, no. 58. Menorah Ring. Jewish Evidence from the Roman Period in the
15 Spier (n. 1), nos 149–65. Northern Provinces (Forschungen in Augst 36), Augst, 2005, esp.
16 Ibid., nos 211–96. 87–98 and 245–6 for rings and gems with a menorah.
17 Ibid., nos 198–209. 49 D. Scarisbrick, Rings. Jewelry of Power, Love and Loyalty, London,
18 Ibid., nos 317–409. 2007, 22 and 355, no. 24.
19 Ibid., nos 405–8; see also nos 464–5, which combine the Good 50 For lead seals, see Spier (n. 1), 163, nn. 50–1. Additional examples
Shepherd with magical images and inscriptions. include one in the Schøyen Collection in Oslo that I take to be of
20 Ibid., nos 407–8 and 466, the last with a magical inscription. 4th-century ad date; see http://www.schoyencollection.com/
21 Ibid., no. 406. seals.htm#5160_1. For another, from Mertingen-Burghöfe,
22 Ibid., nos 76–80 and 482–541. Germany, see Berger (n. 48), 84, no. B5, fig. 26; and an unpublished
23 Ibid., nos 76–9. example in a private collection (via eBay). Two-sided Byzantine
24 J. Spier, ‘5th Century ad Gems and Rings: From Constantinople to examples with menorah and various reverses include: Frank
Italy and the West’, in G. Sena Chiesa and E. Gagetti (eds), Aquileia Sternberg AG, Zurich, Auction 23, 30–31 October 1995, lot 961; Triton
e la glittica di età ellenistica e romana, Trieste, 2009, 237–45. II (numismatic auction), New York, December 1–2, 1998, lot 1108;
25 Ibid., nos 551–60. Gorny & Mosch. Giessener Münzhandlung, Munich, Auktion 163, 14
26 Ibid., nos 551–71. December 2007, lot 185; Archaeological Center, Tel Aviv, Auction 44,
27 I am very grateful to Penelope Rogers for bringing this gem to my 13 April, 2009, lot 183; Amphora, New York, List 95 (2009), no. 482;
attention and to Stuart Foreman of the Anglo-Saxon Laboratory Amphora, New York, List 97 (2010), no. 114; and several others in
for permission to reproduce it. A full study of the find by Penelope private collections.
Rogers is forthcoming. 51 See Spier (n. 1), 167–8, nos 966–97.
28 Ibid., nos 521 and 564. 52 Ibid., no. 980.
29 See ibid., 87, for other examples. 53 See the recent study by D. M. Friedenberg, Sasanian Jewry and Its
30 Ibid., no. 584. Culture. A Lexicon of Jewish and Related Seals, Urbana and Chicago,
31 Ibid., nos 547–50. 2009.
32 Ibid., nos 666–701. 54 Spier (n. 1), 171–82, nos X1–144.
33 Ibid., 125. 55 Ibid., no. X34.
34 Ibid., nos 685–6. 56 I am very grateful to Claudia Wagner of the Beazley Archive in
35 Ibid., nos 759–70. Oxford for bringing these gems to my attention. For the originals
36 Ibid., no. 765. from which these were copied, see http://www.beazley.ox.ac.uk/
37 C. Johns and T. Potter, The Thetford Treasure, London, 1983, 83–4, gems/styles/late-antique/agates.htm and http://www.beazley.
nos 5–6. For the ring from the Piazza della Consolazione treasure, ox.ac.uk/gems/styles/late-antique/garnets.htm.
see: Bedeutende Kunstwerke aus dem Nachlass Dr. Jacob Hirsch, 57 Spier (n. 1), no. 520.
Adolf Hess AG and William H. Schab, Auction, Lucerne, 7 58 Ibid., no. 573.
December 1957, lot 93. 59 Ibid., no. 575.
38 I am very grateful to John Ljungkvist for bringing these to my 60 Lexicon Iconographicum Mythologiae Classicae , vol. 2, Zurich,
attention. 1984, 85–6, nos 765–71, s.v. Aphrodite (Angelos Delivorrias, Gratia
39 Spier (n. 1), no. 767. Gerger-Doer, and Anneliese Kossatz-Deissmann).
40 Ibid., nos 771–8. 61 I am grateful to Judith Lerner for bringing this information to my
41 Ibid., no. 774. attention. For her very informative review of Friedenberg (n. 53),
42 Two gold rings engraved with the scene are in the British Museum, see: Journal of the American Oriental Society 129 (2009), 653–64.

‘Gems of Heaven’ | 207

The Argument from Silence
Iconographic Statements of 1981 on Faked Gems Reconsidered

Josef Engemann

Our colleagues Chris Entwistle and Corby Finney can confirm
the fact that only with great reluctance did I accept this
invitation to share with you my thoughts on ‘Iconography as a
criterion for establishing the authenticity of Early Christian
glyptics’. Many of you have long devoted yourselves to the
study of ancient glyptics, and have written significant books
and catalogues, as well as published important papers in this
field. I myself have but once touched on the field of Late
Antique intaglios and cameos, when contributing my article
‘Glyptik’ to the Reallexikon für Antike und Christentum.1
Presumably, I have been invited to this illustrious gathering to
review my statements made over a quarter of a century ago. I
confess at the outset: the accused is partly guilty.
Jeffrey Spier, in his book on gems published in 2007,
assumed correctly that in my 1981 article magical gems
captured my interest more than Early Christian ones.2 Indeed,
in my writings I have repeatedly commented upon Late Antique
magic.3 I regret all the more that in 1981 I classified a group of
early gems depicting the Crucifixion as magical, without due
regard to their iconographic uniqueness in this context. Since
depictions of Christ’s Crucifixion in other areas of Early
Christian art do not appear before the 5th century ad,
I believed their appearance on supposedly earlier gems (for
instance, the example in the British Museum, published by
Philippe Derchain; Pl. 1)4 to be the work of modern forgers.5
Attilio Mastrocinque however, in 1993, explained that this
argument ‘e silentio’ regarding the field of magical gems is not
acceptable.6 His research began with the lost Berlin intaglio
depicting the Crucifixion which includes the inscription
ΟΡΦΕΟC BAKKIKOC (Pl. 2). He pointed out that 18th-century
illustrations show gems (now lost) whose only ornament is just
this inscription.
Professor Mastrocinque has also defended the Late Antique
authenticity of a second group of gems as well, likewise
labelled by me as suspected forgeries in 1981 (Pls 3–6). The two
intagli in question are of a considerably later date than the
British Museum’s magical gem with Crucifixion: on both, the
twelve Apostles appear in a rather unusual arrangement under
the cross.7 But with regards to the lost gem with the inscription
ehco x-pect-oc (‘Jesus Christ’), which formerly belonged to
the extensive Roman collection of George Frederick Nott
dating back to 1820, I disagree with Attilio Mastrocinque that
its representation by Raffaele Garrucci in 1880 as an alleged
catacomb find serves as a criterion for authenticity (Pls 3–4).8
This is why I still doubt the authenticity of all three gems in
question (Pls 1–6). However, a recent publication presents us
with new problems. Jutta Dresken-Weiland has called our
attention to an archaeological find in an ossuary in Jerusalem,
demonstrating that the feet of a buried crucified person had
been pierced by nails. His arms, however, must have been tied

Plate 1 Magical intaglio with the Crucifixion. 30 x Plate 2 Magical intaglio (drawing) with the Plate 3 Impression of an intaglio with the
25mm. London, British Museum, PE 1986,0511.1 Crucifixion. Berlin, now lost Crucifixion. Nott collection, now lost

Plate 4 Drawing of Plate 3 Plate 5 Intaglio with the Crucifixion. 13.5 x 10.5mm. Plate 6 Drawing of Plate 5
London, British Museum, PE 1895,1113.1

208 | “Intelligible Beauty”


Plate 7 Intaglio with conglomeration of scenes, 16 x 16mm. London, British
Museum, PE 1856,0425.10
Plate 9 Intaglio (cast) with conglomeration of Plate 10 Drawing of Plate 9
scenes. Baltimore, Walters Art Museum
to the horizontal crossbeam by ropes, since neither his hands
nor lower arms were pierced.9 This kind of representation, as
on the magical gem in the British Museum (Pl. 1),10 and on the
piece with the inscription IΧΘΥC from the same collection (Pls
5–6),11 seems thus proven to belong to Late Antiquity. Modern
forgers could hardly have been acquainted with this detail, as
for many centuries all representations of the Crucifixion
showed Christ nailed down at the hands and feet.
Furthermore, I do not agree with Mastrocinque’s
fundamental rejection of the application of iconographic
criteria in determining genuineness. It is not a proper method,
he insists, to first establish a developed order of iconography
and a typological sequence only to discard later as fake any
object not fitting into this pattern.12 I believe iconographical
criteria as a rule have not been established at will but derive
their parameters from the evidence of historical monuments.
Should there be an obvious and contradictory change, for
instance through new finds or new research, the criteria have
to be adjusted. Furthermore, a total abandonment of
iconographic arguments when dealing with antique gems
would be premature as up to the 19th century few differences
in production between antique and modern gems are to be
found.13 This situation has resulted in uncertainty among
collectors, as illustrated by Peter and Hilde Zazoff with, among
other documents, a statement from Johann Wolfgang von
Goethe in the year 1823:
Those, who wish to cast doubt on everything, will especially do so
when discussing gems. Might this piece be a classical copy or is it a
modern reproduction? Could it be another version of a known
original or is it a mere imitation? One moment the stone itself raises
doubts, the next moment the inscription – which should otherwise
be of particular interest – is called into question. To engage with
gems is thus even trickier than to get involved with ancient coins,
although the latter also require considerable circumspection... .14
The Argument from Silence
Plate 8 Drawing of Plate 7
Plate 11 Intaglio with conglomeration of scenes,
13 x 10.5mm. London, British Museum, PE 1856,0425.9
Goethe made these remarks with reference to the Hemsterhuis
collection, but his caution has proved correct as:
Today the greater part of the gems, which in the meantime have
reached Den Haag, are classified as works of the Renaissance.15
In view of such uncertainties of judgement one should not
abstain from additional iconographic arguments.
The possibility, therefore, exists that new scientific results
may support a thesis founded on iconographic criteria. This
seems to be the case in the third group of gems with Christian
images, which in 1981 I suspected were forgeries (Pls 7–11).
These are carnelians with a conglomeration of biblical and
symbolic vignettes.16 I will not discuss here the iconographic
curiosities, for instance the impression that the shepherd
(Κριοφόροc) stands on a fish, as seen on an oval gem in the
British Museum (Pls 7–8), or the lack of symmetry between the
shepherd’s sheep on a round gem in the same collection (Pl. 11).
More important is the fact that the artist added to Christ‘s
monogram two rather unusual accompanying letters, I C
(‘Iesous’), to the figurative representation on the oval gem in
the British Museum and an IΧΘΥC inscription on a gem in the
Walters Art Museum in Baltimore (Pls 9–10). These motifs
individually occur in many inscriptions and frequently on
gems, often of very poor quality.17
But in addition to assemblages of biblical vignettes, the
Christogram and IΧΘΥC inscription would be out of place in a
Late Antique context. This we can deduce from their absence in
the numerous arrangements of multiple biblical scenes in
catacomb paintings and on early sarcophagi, e.g. the panel in
Velletri (Pl. 12).18 Although on the Adam and Eve sarcophagus
in Arles (Pl. 13)19 a Christogram is displayed on the tablets
received by Moses on Mount Sinai (Pl. 14), this monogram is
not a mere addition but reveals a typological statement. The
same is valid for the Christograms on the scrolls in the hands of
‘Gems of Heaven’ | 209
Engemann

Plate 12 Panel from a

sarcophagus with a
conglomeration of
scenes. Velletri, Museo
Civico

Plate 13 Sarcophagus with Adam and Eve. Arles, Musée de l’Arles antique Plate 14 Detail from Adam and Eve sarcophagus. Moses receiving the law

Plate 15 Detail from a sarcophagus with Christ holding a scroll. Rome, Plate 16 Detail from the same sarcophagus as Plate 15 with St Peter holding a
St Peter’s scroll

Christ and St Peter on a sarcophagus in St Peter’s in Rome (Pls In 1981 I only had a vague idea of the number of gemstone
15–16).20 cutters active in Rome during the 19th century. In the
Corby Finney likewise strongly suspects that two of the meantime, making use of contemporaneous sources, more has
gems with vignettes in the British Museum’s collection are fakebeen published about their large numbers, mainly by Lucia
(Pls 7–8; 11).21 I quote here his statement regarding the art Pirzio Biroli Stefanelli.23 Some of the works produced by these
trade in the 19th century: artists and bought or ordered by collectors may have been
by the late 1860s the market for all early Christian antiquities labelled ‘antique’. In his catalogue of the 1990 Fake? exhibition
(including cameos and intaglios) was heating up, and it became in the British Museum, Mark Jones entitled one chapter ‘The
hotter as the century drew to an end. One of the by-products was 19th century: the great age of faking’.24 In 1981 I was also
the manufacture and distribution of multiple fakes... .22
unaware of the degree of cooperation between collectors and

210 | ‘Gems of Heaven’


fakers. It was a revelation a few years ago when Gertrud Platz
demonstrated the connection between the collector Stanislas
Poniatowski and gemstone cutters like Giovanni Calandrelli
and his colleagues in Rome.25 Above all, I was unaware in 1981
that the three aforementioned gemstones (Pls 7–8, 11) once
belonged to the collection of the Scotsman Abbé James
Hamilton, who had resided in Rome between 1841 and 1851.
Corby Finney has devoted his attention to this multifaceted
personality who was equally active as collector and art dealer.26
He describes Hamilton, who converted to the Roman Catholic
faith, as an ‘ardent Catholic and a strong supporter of Pius IX’
(1846–78) and enlarges upon his activities of not only
collecting antique originals – among them gems –27 but
commissioning as well works of art by contemporary artists.28
As already mentioned, Hamilton is an example of the
collaboration between collectors of antiquities and
contemporary artists. But until proven otherwise we might
assume he was bona fide: he did not himself sell works of the
19th century as Late Antique originals. When Hamilton
ordered gems ‘alla maniera paleocristiana’ to augment his
collection, these needn’t necessarily be fakes but may have
been normally commissioned works. Their production
reflected actual artistic interests in Rome. It has been
emphasised:
It can be shown that in the 19th century knowledge of Christian art
was widespread, the majority of the public were aware of it, and it
was accessible to them.29
It is noticeable that Hamilton, in the then tense and critical
situation of ecclesiastical politics, placed himself at the side of
Pius IX, the pope who with his Syllabus errorum laid the
foundation for the oath against the modernism of Pius X in
1907. Pius X is relevant today, because it is no coincidence that
the followers of Bishop Levebre named their brotherhood in
honour of this pope. Due to his religious connections, we may
assume that for Hamilton the commissioned gems couldn’t be
pious enough. Hence, we might speculate that he ordered
several biblical scenes to be collectively represented on a gem
together with the shepherd, topping it with a christogram or
IΧΘΥC inscription in order to enhance its Christian character.
Let me emphasise: this is an attempt at a hypothetical
psychological explanation. Basically, for iconographical
Plate 17 ‘Dogmatic’ sarcophagus. Rome, Museo del Vaticano
The Argument from Silence
reasons, I believe the gems in question not to be Early Christian
originals, even though they have been published at a very early
stage as such.
However, the poor artistic quality of these gems presents a
problem. The standard of intagli and camei of the 18th- and
19th-century gemstone cutters is totally different from Late
Antique works and of much greater competence. I derive
support for my statement from the following sources: the
catalogue of the Munich exhibition in 1995,30 six lectures of the
related colloquium published in 1996 in Zeitschrift für
Kunstgeschichte,31 the volume of a 1994 symposium held in the
National Gallery of Art in Washington DC,32 the afore-
mentioned Calandrelli book by Gertrud Platz, and finally the
first volume of the Collezione Paoletti Catalogue.33 Is it
conceivable that Hamilton’s gems surfaced in this ambiance?
How is such a lack of style and composition to be explained? I
believe the reason for this to be obvious: the classicising artists
had masterpieces of antique glyptic for orientation, whereas
the workers in Hamilton’s employ had to rely on models of Late
Antique gems depicting shepherds, Jonas scenes, and Daniel in
the lion’s den, designed to deter every ‘mal occhio’ (Evil Eye)
through their poor artistic quality.34
I conclude with a last reference to ‘argumentum e silentio’.
Its value can be documented for instance by the application of
similar reasoning to the above-mentioned Adam and Eve
sarcophagus in Arles (Pl. 13) and its counterpart in Rome, the
so-called ‘Dogmatic’ sarcophagus (Pl. 17), labelled as such on
the assumption that the three persons in the creation scene
represent the Christian Trinity (Pl. 18).35 This interpretation
has been transferred to the corresponding scene on the
sarcophagus unearthed in 1974 in Arles (Pl. 19). In this case
four persons are engaged in the creation of Adam and Eve. For
iconographic reasons I strongly opposed this interpretation.36
Only from the 12th century onwards do we have images of the
Trinity as human figures in Christian art – that is to say as
anthropomorphic depictions. Before that, typological displays
existed in the Old Testament scene of the three visitors coming
to Abraham, for instance in S. Vitale in Ravenna (Pl. 20),37
symbolic numbers as demonstrated in the triple Christogram in
the baptistery in Albenga (Pl. 21),38 or the figurative symbolism
of Jesus’ Baptism in the Rabbula Codex (Pl. 22).39 From the
‘Gems of Heaven’ | 211
Engemann

Plate 18 Detail of Plate 17. Creation of Adam and Eve Plate 19 Detail of Plate 13. Creation of Adam and Eve

Plate 20 Detail of mosaic with Abraham’s three visitors. Ravenna, S. Vitale Plate 21 Detail of ceiling mosaic with numerical symbolism. Albenga,
Baptistery

corresponding volume of the Repertorium der christlich-antiken

Sarkophage published in 2003 I concluded that my remarks on
representations of the Trinity have generally been accepted by
researchers.40 But later I discovered that the Trinity
interpretation had been repeated again in 2000, without new
arguments to support it.41
After having reinvestigated my glyptic article from 1981, I
close by emphasising the importance of the argument from
silence.

Notes
1 J. Engemann, ‘Glyptik’‚ Reallexikon für Antike und Christentum 11
(1981), 270–313.
2 J. Spier, Late Antique and Early Christian Gems, Wiesbaden, 2007,
9, n. 114.
3 For instance: J. Engemann, ‘Zur Verbreitung magischer
Übelabwehr in der nichtchristlichen und christlichen Spätantike’,
Jahrbuch für Antike und Christentum 18 (1975), 22–48; idem,
Deutung und Bedeutung frühchristlicher Bildwerke, Darmstadt,
1997; idem, ‘Anmerkungen zu philologischen und archäologischen
Studien über spätantike Magie’, Jahrbuch für Antike und
Christentum 43 (2000), 55–70.
4 British Museum, PE 1986,0501.1; P. Derchain, ‘Die älteste Plate 22 Detail of the Rabbula Codex (folio 4v) with the Baptism of Christ.
Florence, Bibiloteca Mediceo Laurenziana

212 | ‘Gems of Heaven’


Darstellung des Gekreuzigten auf einer magischen Gemme des 3.
(?) Jahrhunderts’, in K. Wessel (ed.), Christentum am Nil,
Recklinghausen, 1964, 109–13; S. Michel, Die Magischen Gemmen
im Britischen Museum, London, 2001, 283–4, no. 457; Spier (n. 2),
73, no. 443.
5 Earlier critique by: P. Maser, ‘Die Kreuzigungsdarstellung auf
einem Siegelstein der Staatlichen Museen zu Berlin’, Rivista di
Archeologia Cristiana 52 (1976), 257–75.
6 A. Mastrocinque, ‘Orpheos Bakchikos’, Zeitschrift für Papyrologie
und Epigraphik 97 (1993), 16–24.
7 Engemann (n. 1), 277–8; Mastrocinque (n. 6), 22–3; Spier (n. 2), 73,
nos 444–5.
8 Mastrocinque (n. 6), 22–3, n. 40.
9 J. Dresken-Weiland, ‘Passionsdarstellungen in der
frühchristlichen Kunst’, in T. Nicklas, A. Merkt and J. Verheyden
(eds), Gelitten, gestorben, auferstanden, Tübingen, 2010, 31–46, at
34; A. Donati (ed.), Dalla terra alle genti. La diffusione del
cristianesimo nei primi secoli (exh. cat., Rimini), Milan, 1996, 46–8,
169, no. 5 (J. Zias).
10 Michel (n. 4), 283–4, no. 457; eadem, Die magischen Gemmen. Zu
Bildern und Zauberformeln auf geschnittenen Steinen der Antike
und Neuzeit, Berlin, 2004, 124–6; Spier (n. 2), 73, no. 443.
11 British Museum, PE 1895,1113.1; Spier (n. 2), 73, no. 444.
12 Mastrocinque (n. 6), 18; cf. 22, n. 40.
13 J. Rudoe, ‘The Faking of Gems in the Eighteenth Century’, in
M. Jones (ed.), Why Fakes Matter: Essays on Problems of
Authenticity, Symposium 1990, London, 1992, 23–31; P.C. Finney,
‘Critical Problems in Early Christian Glyptics’, in ...zur Zeit oder
Unzeit, Studien zur spätantiken Theologie-, Geistes- und
Kunstgeschichte und ihrer Nachwirkung, Hans Georg Thümmel zu
Ehren, Mandelbachtal, 2004, 143–60, esp. 145.
14 P. and H. Zazoff, Gemmensammler und Gemmenforscher. Von einer
noblen Passion zur Wissenschaft, Munich, 1983, 190; for the
context: G. Femmel and G. Heres, Die Gemmen aus Goethes
Sammlung, Leipzig, 1977, 268–70, Quotation 269: ‘Nun aber findet
die Zweifelsucht kein reicheres Feld sich zu ergehen als gerade bei
geschnittenen Steinen; bald heißt es eine alte, bald eine moderne
Copie, eine Wiederholung, eine Nachahmung; bald erregt der
Stein Verdacht, bald eine Inschrift, die von besonderem Werth
sein sollte, und so ist es gefährlicher sich auf Gemmen einzulassen,
als auf antike Münzen, obgleich auch hier eine große Umsicht
gefordert wird... ’.
15 Femmel and Heres (n. 14), 41: ‘Heute gelten die meisten der
inzwischen nach Den Haag gelangten Steine für
Renaissancearbeiten’.
16 British Museum, PE 1856,04–25.10: Spier (n. 2), 69, no. 428; British
Museum, PE 1856,0425.9: Spier (n. 2), 69, no. 429; Baltimore,
Walters Art Museum, inv. no. 57.1641: Spier (n. 2), 69, no. 432.
17 For examples, see Spier (n. 2), pls 18–22.
18 J. Dresken-Weiland, Repertorium der christlich-antiken
Sarkophage, Vol. 2: Italien mit einem Nachtrag Rom und Ostia,
Dalmatien, Museen der Welt, Mainz, 1998, 83–4, no. 242, pl. 80,2.
19 C. Christern-Briesenick, Repertorium der christlich-antiken
Sarkophage, Vol. 3: Frankreich, Algerien, Tunesien, Mainz, 2003,
23–5, no. 38.
20 H. Brandenburg, Repertorium der christlich-antiken Sarkophage,
Vol. 1: Rom und Ostia, Wiesbaden, 1967, 271–2, no. 674; E. Stommel,
Beiträge zur Ikonographie der konstantinischen Sarkophagplastik,
The Argument from Silence
Bonn, 1954, 103.
21 P.C. Finney, ‘Abbé James Hamilton: Antiquary, Patron of the Arts,
Victorian Anglo-Catholic’, in C. Entwistle (ed.), Through a Glass
Brightly. Studies in Byzantine and Medieval Art and Archaeology
Presented to David Buckton, Exeter, 2003, 195 (and nos 2 and 3 of his
Appendix).
22 Ibid., 192.
23 L.P. Biroli Stefanelli, ‘Roman Gem Engravers of the Eighteenth and
Nineteenth Centuries: The Present State of Research’, Jewellery
Studies 4 (1990), 53–8; eadem, ‘Del cammeo e dell’ incisione in
pietre dure e tenere nella Roma del XIX secolo’, in L. Biancini and F.
Onorati (eds), Arte e artigianato nella Roma di Belli, Rome, 1998,
13–30; Finney (n. 13), 144–5.
24 M. Jones (ed.), Fake? The Art of Deception (exh. cat., British
Museum), London/Berkeley, 1990, 161.
25 G. Platz-Horster, ‘Zeichnungen und Gemmen des Giovanni
Calandrelli’, in D. Willers and L. Raselli-Nydegger (eds), Im Glanz
der Götter und Heroen. Meisterwerke antiker Glyptik aus der
Stiftung Leo Merz, Mainz, 2003, 49–60; eadem, L’ antica maniera.
Zeichnungen und Gemmen des Giovanni Calandrelli in der
Antikensammlung Berlin (exh. cat., Berlin), Berlin, 2005.
26 Finney (n. 21), 190–8; idem (n. 13), 145.
27 Finney (n. 21), 191; cf. J. Spier, ‘Early Christian Gems and their
Rediscovery’, in C.M. Brown (ed.), Engraved Gems: Survivals and
Revivals (Symposium, Washington DC, 1994), Hanover, 1997,
33–43, at 39.
28 Finney (n. 21), 191.
29 A. Reiß, Rezeption frühchristlicher Kunst im 19. und 20.
Jahrhundert: Ein Beitrag zur Geschichte der Christlichen
Archäologie und zum Historismus, Dettelbach, 2008, 36: ‘Kann
festgestellt werden, dass im 19. Jahrhundert das Wissen um die
Existenz frühchristlicher Kunst bereits in hohem, Maße der
Allgemeinheit zugänglich und bewusst war’.
30 I.S. Weber (ed.), Geschnittene Steine des 18. bis 20. Jahrhunderts.
Vergessene Kostbarkeiten in der Staatlichen Münzsammlung
München, Munich, 1995.
31 Zeitschrift für Kunstgeschichte 59 (1996), 137–243.
32 See Brown (n. 27).
33 L.P. Biroli Stefanelli, Museo di Roma, La Collezione Paoletti, Vol. 1,
Rome, 2007.
34 Survey by Spier (n. 2), pls 39–53.
35 Brandenburg (n. 20), 39–41, no. 43.
36 J. Engemann, ‘Zu den Dreifaltigkeitsdarstellungen der
frühchristlichen Kunst: Gab es im 4. Jh. anthropomorphe
Trinitätsbilder?’, Jahrbuch für Antike und Christentum 19 (1976),
157–72.
37 P. Angiolini Martinelli and P. Robino (eds), La basilica di San Vitale
a Ravenna 2, Modena, 1997, 210, fig. 411.
38 M. Marcenaro, Il battistero di Albenga, Recco, 1994, fig. on 26. For
further 6th-century ad examples of such numerical Trinity
symbolism, see: A. Terry and H. Maguire, Dynamic Splendor. The
Wall Mosaics in the Cathedral of Eufrasius at Poreč, University Park,
PA, 2007, 128–31.
39 C. Cecchelli, J. Furlani and M. Salmi, The Rabbula Gospels, Olten-
Lausanne, 1959, pl. 9.4.
40 Christern-Briesenick (n. 19), 25.
41 U. Utro, ‘Trinity’, in F. Bisconti (ed.), Temi d‘iconografia
paleocristiana, Vatican City, 2000, 294–5.
‘Gems of Heaven’ | 213
The Constanza Carnelian and the Development of
Crucifixion Iconography in Late Antiquity
Felicity Harley-McGowan
In 1895, a small intaglio held in a private collection in London
came to the attention of the archaeologist Cecil Harcourt-
Smith (1859–1944), then working at the British Museum in the
Department of Greek and Roman Antiquities. The transverse
oval carnelian preserved what Smith argued was the earliest
extant representation of the Crucifixion (Pl. 1).1 He thus
brought it to the attention of his colleague, Sir Augustus
Wollaston Franks (1826–97), then Keeper of British and
Mediaeval Antiquities and Ethnography, who subsequently
purchased the intaglio and presented it to the British Museum
in the same year.2
In his account of the gem’s provenance, Smith reported that
it was shown to him as one in a group of 30 or 40 gems
reputedly found on a beach at Constanza, Romania. Classifying
this group as Roman, and dating the gems collectively from
between the 1st to the 3rd century ad, he remarked that ‘the
only one of real importance’ was that which bore the standard
Early Christian formula IXΘYC (the Greek word for ‘fish’ but
also used as an acrostic to signify ‘Jesus Christ, Son of God,
Saviour’)3, with an image of the Crucifixion engraved on its
obverse. Smith proposed a production date for this gem of
around the 2nd/3rd century ad, reasoning that this accorded
with the general style, provenance and date of the gems in the
group as a whole. In addition, following a careful examination
of iconographic evidence for the representation of the
Crucifixion before the 6th century ad, he concluded that with
the exception of a graffito excavated on the Palatine hill in
Rome, taken to be 2nd century ad and discussed further below,
the gem preserved a composition of the subject that was earlier
than other known evidence.
Smith’s placement of the Crucifixion gem from Constanza
at an early point in the development of Crucifixion iconography
(such as it can be documented by a very small body of extant
evidence) was acute, especially given the general reticence to
incorporate the evidence of engraved gems into the study of
Early Christian art. With a much enlarged knowledge and thus
clearer understanding of the production of Christian engraved
gems in Late Antiquity, it can now be contended with
confidence that the iconography, in conjunction with the
inscription, the form of the letters and carving style, indicate
an early or mid-4th-century ad date of production in the
eastern part of the Roman Empire.4 Focusing on the image, this
paper will demonstrate the ways in which the iconography can
be seen to clarify the gem’s date and thereby reaffirm the
importance originally bestowed on it by Smith in his
assessment of the history of the representation of the
Crucifixion in Christian art.5
Crucifixion Iconography and Early Christian gems
Beneath the acrostic, the design of the Constanza gem is
dominated by the figure of a man presented in rigid frontality
214 | ‘Gems of Heaven’
against the upright shaft of a tau cross, with his arms shown
outstretched and tied at the wrists to the patibulum or cross
bar. Given the inscription and specific iconographic features
outlined below, the identity of the man is without question:
Jesus. While his head and feet are turned to the viewer’s right
(and so shown in rather flat, two dimensional profile), some
modelling is attempted by the carver to indicate Jesus’
anatomy. This is most notable in the rendering of the knees, the
demarcation of the waist and abdomen in the torso, along with
the shaping of the shoulders and the neckline. The sensitivity
shown in the articulation of these physical features makes it
clear that although not explicit, the figure is also
unambiguously nude. This fact is underscored by the carver’s
very careful attempt to clothe the 12 male figures represented
half the size of Jesus and shown processing toward him, six on
either side. Diagonal cuts made at regular intervals across their
upright bodies indicate that they wear close-fitting mantles, or
pallia. As will be maintained here, in comparison with
established iconographic formulae for the depiction of the
Apostles either side of Jesus in the 4th century ad, a version of
which is directly replicated on this gem, the figures are clearly
representative of the twelve Apostles.
The use of semi-precious stones engraved with images or
monograms, and used as seals in finger-rings, was an integral
part of daily life through to the 3rd century ad when the use of
gems as seals began to diminish. Within this cultural and
economic context, Christianity struggled with a variety of
philosophical and theological issues pertaining to the
fundamental question of adornment of the body, as well as
with the thorny question of image-use upon such personal
items as jewellery. On facing the first issue, they were not
alone: in matters relating to clothing, physical embellishment
and even care of the body, males and females in the Roman
Plate 1 Constanza gem. Carnelian, flat, 13.5 x 10.5mm. Syria (?), mid-4th
century AD. Said to have been found in Constanza, Romania. London, British
Museum, PE 1895,1113.1
 The Constanza Carnelian and the Development of Crucifixion Iconography in Late Antiquity
world negotiated a tenuous balance between the observation of
moral strictures, the necessary practicalities of daily life, and
the general pursuit of fashion. On facing the second issue,
Christians were in distinctively hazardous territory.
Clement of Alexandria (c. ad 150–c. ad 215) provides clear
evidence that as early as the 2nd century ad, Christians had to
navigate their way very carefully through the already
established use of images to decorate finger-rings. In his oft-
cited Paedagogus, he indicates a small range of images
appropriate for Christian use upon rings. This includes a dove,
a fish, a ship in sail, a musical lyre, a ship’s anchor, and a
fisherman (Paed. 3.59).6 For although Clement is against luxury
and the ornamentation of the body (it is the Christian soul, he
writes, that is to be decorated with the ornament of goodness,
Paed. 2.3), he is explicit in permitting one finger-ring of gold.
For a woman, this ring is to be worn in the fulfilment of
domestic tasks only, namely protecting household goods (Paed.
3.57). For a man, this is to be worn at the base of the little finger
so that his hand is free to conduct business (Paed 3.58–9).
Hence the seal, or signet ring, is expressly permitted for
security purposes only, both commercial and domestic – that
is, in order to mark ownership of property.
When gems produced specifically for Christian clients first
began to appear in the eastern Mediterranean – soon after
Clement’s time, around the middle of the 3rd century ad – they
are largely identical to other gems produced at the period in
shape, material (usually carnelian, agate or jasper) and
engraving style. The key features that distinguish them as
Christian are the inscriptions they carry, and the symbols they
bear, which are appropriate to Christian use and accord
directly with Clement’s specifications.7 Among the many extant
examples attesting to the use of these symbols in conjunction
with Christian inscriptions, the iconography of the Constanza
gem is highly unusual relative to Clement’s small catalogue of
approved images. Nonetheless, as Smith observed, a second
known example, almost identical in size, shape and design to
the Constanza gem, indicates that although remarkable, the
design was not a ‘one-off’.
Included by Raffaele Garrucci in his comprehensive Storia
dell’ arte cristiana nei primi otto secoli della chiesa, this second
example was purchased in Rome by the English collector, the
Rev. George Frederick Nott (1767–1841).8 Although its
whereabouts are presently unknown, a plaster impression
made in the 19th century (Pl. 2) testifies that the gem was
fractionally larger and more elongated in shape than the
Constanza gem, and that the pattern of a crucified Jesus amid
twelve Apostles appeared with only minor variations. In
Plate 2 Nott Gem. Plaster cast, original: carnelian, slightly convex, c. 19 x
14mm. Syria (?), mid-4th century AD
examining the cast, these differences are most notable with
respect to the figure of Jesus, who at the central axis of the
composition is depicted on a similar scale to the Apostles but is
shown on a column, physically elevated above them, and
crowned with a nimbus. A further difference is the depiction of
the two Apostles to either side of Jesus, who are shown raising
their hands to touch the base of his cross. These pictorial
variations aside, as in the case of the Constanza gem, the
design is accompanied by a Greek version of Jesus’ name: ehco
x-pect-oc, with the final two letters of Christos split either side
of a lamb, placed strategically below the cross.9 As will be
discussed below, an early to mid-4th century ad date is
probable for this gem; and specifically, the nimbus indicates a
date no earlier than the Constantinian period, before which
time nimbi are unlikely to have been used in a Christian
context.10
The inscriptions on both the Constanza and Nott gems are
positive (being engraved directly onto the stone so that they
were intended to be read on the face of the gem by the wearer)
rather than negative (intended to be read in impression). In
this they follow what appears to be a characteristic of gems
engraved in Late Antiquity and as broadly symptomatic of the
general decline in skill that is witnessed prior to this period.11
The conjunction of the name of Jesus with Christian
iconography is, as noted above, a further feature common
amongst Early Christian gems. To better understand the
significance of the union of Jesus’ name with an image of his
Crucifixion in what must be seen as an avowedly Christian
context, it is necessary to examine both components of the
design in turn.
The invocation of the name of Jesus from the New
Testament period through to the time of Origen has been well
documented;12 yet in the context of this discussion it is
interesting to note that by the 2nd century ad, the name was
expanded by explicit references to Christ’s death. Justin Martyr
writes that many demoniacs were exorcised by Christians ‘in
the name of Jesus Christ, who was crucified under Pontius
Pilate’ (2 Apology 6.6); and his almost formulaic use of this
phrase, ‘crucified under Pontius Pilate’ emerges with particular
prominence in his Dialogue with Trypho (Dial. 30.3; 76.6; 85.2).
The phrase was employed by Irenaeus (Adversus haereses
2.49.3, and Epideixis, 97), and was possibly known by Origen
(Contra Celsum 1.6), suggesting a strong persistence of the
belief amongst Christians that such references enhanced the
perceived power resident in the name of Jesus itself. This belief
was already current in the Apostolic era, as witnessed in such
acts as Peter’s healing of a cripple in the name of Jesus Christ of
Nazareth, ‘whom you crucified, whom God raised from the
dead …’ (Acts 4.9–10). Aune suggested that underlying this
particular expansion of the name with the formulaic reference
to the Crucifixion is the notion that Jesus’ death on and
ensuing victory over the cross, spelled the destruction of
demonic powers.13 Viewed in this light, both gems can be seen
to follow and so provide just such an expansion of the name,
but doing so through image rather than word; they very clearly
reflect both the interpretation of the Crucifixion as a triumph
over death in the early Church, and the use of the name of
‘Christ crucified’ as a means of protection by Christians. In so
doing, they stand as key witnesses to the Early Christian use of
engraved gems as objects with a distinct devotional and
‘Gems of Heaven’ | 215
Harley-McGowan
religious function. It can be argued that this is confirmed in the
iconography. The depiction of Jesus crucified in the presence of
the Apostles is extremely unusual by comparison to
developments in the representation of the subject in the 5th
and 6th centuries ad, where Jesus is customarily depicted
hanging on the cross in the presence of his mother and John
the Evangelist and often the two thieves. Nevertheless, when
seen within the broader context of Christian iconography at
this period, it can be linked directly with contemporary
pictorial trends for the representation of Jesus among his
Apostles.
In the course of the 4th and early 5th centuries ad, a range
of pictorial formats for the depiction of Jesus among his
Apostles was developed and popularised across a variety of
media in Christian art. Invariably explored within and
understood to be set in a celestial context, this theme placed
especial emphasis on Jesus’ authority as well as his victory over
death in the resurrection (Pl. 3). Hence Jesus could be
portrayed in one of a number of guises: teacher, thaumaturge,
heavenly King, philosopher or giver of the new law.14 Yet
regardless of the role he assumed, and whether he was
presented standing, seated or enthroned as he fulfilled that
role, Jesus was always shown at the centre of the composition,
presiding over the assembly of his Apostles who flanked him in
strictly symmetrical and hieratic compositions. The Apostles
themselves could be shown seated in discussion, or standing
and processing ceremonially towards Jesus – sometimes with
one arm raised in a gesture of acclamation. In certain
iconographic formats, the iconic figure of Jesus was
complemented by the figure of a lamb beneath his feet (as
witnessed on the Nott gem). In other formats he was
substituted by an aniconic symbol: a throne (as on the front
panel of the Pola ivory casket, where it is combined with the
lamb as well as the four Rivers of Paradise);15 or the symbolic
monogram of the cross-trophy, containing the triumphant
cross of the Crucifixion surmounted by a victory wreath (Pl. 4).
Plate 3 Christ Teaching the Apostles / Giving the New Law, mosaic, probably late 4th century AD, Chapel of S. Aquilino, Basilica of S. Lorenzo, Milan
216 | ‘Gems of Heaven’
A subsequent variant on the theme evolved: the apostolic
veneration of the cross, an iconographic type that appeared in
the late 4th century ad on the sculpted reliefs of a small group
of Roman sarcophagi known as ‘star-and-wreath’. The entire
front of sarcophagi in this group is devoted to the single
composition, with twelve Apostles, shown wearing pallia as on
the Constanza gem, processing slowly and simultaneously
towards the central cross-trophy, the crux invicta, which they
venerate and with which Jesus is associated and identified.
Stars appear between the heads of each Apostle, and each is
crowned with a wreath representing the Crown of Life (Rev.
2.10). In certain examples, such as that now in Arles (Pl. 5),16
the Apostles extend their arms to touch the wreathed cross.17
Hence as Bianca Kühnel has argued, on the sarcophagi the
symbol of the cross and the figure of Jesus became
interchangeable, the substitution of the figure making no
significant change in the general meaning of the scene.18
As with contemporary manifestations of the theme of
‘Christ amongst the Apostles’ in other media, these ‘veneration’
scenes can be regarded as directly related to the iconographic
variant preserved on the gems, where the aniconic symbol of
victory, the wreathed cross, is replaced with the figure of
Christ, crucified yet simultaneously resurrected in triumph.
Jesus, in figural or aniconic form, was understood to be the
conqueror of death and thus shown receiving the victorious
crown of martyrdom, or gestures of veneration, from the
Apostles in heaven. In a rare version seen on a sarcophagus
previously in the Vatican, both the resurrected Christ
(appearing to the two women in the garden) and the victorious
cross are shown (Pl. 4). In the composition found on the Nott
gem, the figure of Jesus evokes the shape of the victorious cross
with his body, raised above the Apostles – two of whom touch
the base of the cross just as they touch the arms of the victory
wreath in other versions (Pl. 5). In the Constanza design,
although the cross itself is shown, and ties at Jesus’ wrists
indicate he is actually attached to it, scale rather than posture
 The Constanza Carnelian and the Development of Crucifixion Iconography in Late Antiquity
alone is used to convey the dominance of the cross-form of
Christ’s own body. Consequently, if this imagery is regarded as
being part of the broader search for and development of an
iconography that symbolically expressed the celestial
veneration of Jesus as a figure of triumph and of authority by
the Apostles, the iconography borne by the Constanza and Nott
gems (the latter moreover including the symbol of the lamb)
emerges as an entirely logical, if perhaps less popular variant
on a prevalent and powerful theme in 4th-century ad Christian
art.
While the two gems thus provide critical evidence for the
representation of the Crucifixion by Christians prior to the 5th
century ad, there is a third gem that predates both examples,
and which points to an even earlier manifestation of a pictorial
experiment with the subject. A large bloodstone intaglio (Pls
6a–b), acquired by the British Museum in 1986, preserves a
unique representation of Jesus who is again named and
depicted hanging from a tau cross.19 This gem (see also
Engemann in this volume) is larger than the two gems already
considered; and as has been argued elsewhere, the size, with
the style of the carving, material and inscription show it to be
typical of a large group of Graeco-Roman magical amulets
originating in Egypt and Syria and used widely in the Roman
Empire during the 2nd and 3rd centuries ad.20 Sharing
important iconographic features with later representations of
the subject in Christian art, this third gem can be seen to
present a distinct, amuletic, usage of emergent pictorial
practices in which the name of Jesus and an image of his death
are combined.
Like the group of amulets to which it belongs, the magical
Crucifixion gem is covered with a densely carved and largely
incomprehensible inscription that includes some Greek letters.
Plate 4 ‘Acclamation’ sarcophagus,
4th century AD. Palazzo del Duc di
Ceri in Borgo Vecchio. Previously in
the Vatican, now lost
Plate 5 ‘Star-and-Wreath’
sarcophagus, c. AD 375–400. Musée
de l’Arles Antique, Inv. no. FAN
92.002483 (casket), FAN 92.002484
(lid)
On the obverse (Pl. 6a) a nine-line inscription begins with the
invocation, ‘Son, Father, Jesus Christ’, followed by uncertain
magical names (including soam noam), vowels, and possibly, as
Derchain suggested in his original publication of the gem, the
word ‘hung up’ – which would of course correlate nicely with
the iconography. On the reverse (Pl. 6b), another nine-line
inscription, written in another hand, contains amongst its
strings of letters two names familiar from magical papyri and
other magical gems. Also present is the name Emmanuel from
Isaiah 7:14, taken by Christians to be a prophetic reference to
Jesus. So the crucified figure is possibly named twice (on the
front and back of the stone) as well as depicted.
With the exception of the image, the gem fits well
stylistically into the large body of magical gems of the 2nd and
3rd centuries ad. The iconography of the Crucifixion is
unusual. Jesus is portrayed as a nude, bearded and long-haired
man, hanging from a tau cross in a fashion – with legs split
apart – not seen again in Christian art for the representation of
Plates 6a–b Bloodstone, eastern Mediterranean (Syria?), late 2nd–3rd
century AD, 30 x 25 x 5.8mm. London, British Museum, PE 1986,0501.1; from
the collection of Roger Periere, Paris
‘Gems of Heaven’ | 217
Harley-McGowan
Plate 7 Graffito with parody of the Crucifixion, scratched into plaster wall,
Imperial Palace, Palatine Hill, Rome (excavated 1856), early 3rd century ad,
H. 380mm, W. 330mm. Museo Palatino, Inv. no. 381403
the Saviour. His arms are stretched out beneath the horizontal
bar of the cross and attached to it by two short strips at the
wrists. His elbows and hands fall loosely as a result, and the
iconography in this detail recalls the Constanza gem, as does
the turning of the head sharply to the left, and the use of the
tau cross. Jesus’ upper body is also rigidly upright. His legs are
carefully carved and shown in profile, bent at the knee, and
hanging open loosely, as though he is seated on a bar or peg,
although none is shown. The starkness of the position,
emphasising Jesus’ nudity, is wholly antithetical to the
triumphal symbolism of the crucified Christ seen in the
Constanza gem and in subsequent Early Byzantine
representations of the subject. Nevertheless, from literary
accounts as well as archaeological evidence we know that
executioners placed victims in different positions on the cross,
including having the legs ‘open’ rather than side by side.21 The
image does however share important, rudimentary visual
elements with another image of the Crucifixion in Late
Antiquity.
The flat, strictly frontal presentation of Jesus on the
magical gem, with erect carriage of the head and torso, is seen
in the more well-known crucified figure of the so-called
‘Alexamenos’ graffito (Pl. 7). The graffito, scratched into the
plastered wall of servants’ quarters in the Imperial Palace on
the Palatine Hill, Rome, and now in the antiquariam of the
Palatine, is dated to the early 3rd century ad.22 Customarily
interpreted as a satire of Christian belief in a crucified deity, it
shows the figure of a young man standing in the foreground,
saluting a second figure, tied to a cross and having the head of
a donkey. Unlike the magical gem, where the accompanying
218 | ‘Gems of Heaven’
inscription names the crucified figure, here the Greek
‘Alexamenos sebete theon’ (generally translated ‘Alexamenos,
Worship God’) is believed to refer to the rumour that Christians
worshipped a donkey-headed god – an accusation known to
Minucius Felix (Octavius 9.3, 28.7) and Tertullian (Apologeticus
16.12).
While much could be written about its iconography, in this
context it should be noted that there is a remarkable parity
between this representation of a crucified figure, and those
representations of Jesus crucified that emerge in the 4th and
5th centuries ad, and subsequently across the 6th and 7th
centuries ad. This includes the use of a tau cross (which
appears on the magical and Constanza gems); the upright
stance of the body against the cross; the tapering of the arms
upwards to suggest that they are attached to the crossbar; the
profile view of the victim’s head; the representation of the
victim clothed and not naked; and the more controversial
depiction of a foot support. With all of these features, this
visual conception of a crucifixion appears strikingly
sophisticated for its early date and suggests a ‘pagan’, or non-
Christian’s awareness of two things: firstly, the significance of
Jesus’ Crucifixion (at least in terms of it being a powerful and
efficacious symbol); and secondly, a consciousness of the
existence of Christian representations of the Crucifixion by the
early 3rd century ad. This awareness is supported by the
magical gem (Pl. 6), carved possibly in the eastern part of the
Empire at a similar point in time. On the gem a representation
of Jesus’ Crucifixion is attempted, not to deride, but to invoke
his power – and since there is a syncretism of Greek text,
invocation of the name of Jesus on the amulet (which has some
semblance, in its form, with liturgical language) and
iconography, it is not clear whether the owner would have been
a Christian or not.23
Although Church authorities strongly disapproved of
magical amulets, which were all pervasive in the Graeco-
Roman world, Christians were not averse to using them. In
addition, we know that the practice of accompanying the use of
amulets with magical incantations was widespread.24 This
included the appropriation of Christian liturgy and belief by
practitioners of Graeco-Roman magic. Thus we find the
3rd-century ad Christian theologian Origen (Clement’s near
contemporary and fellow Alexandrian) writing:
The name of Jesus is so powerful against the demons that
sometimes it is effective even when pronounced by bad men
(Contra Celsum 1.6).
The magical gem should therefore be viewed as illustrative of
the eclectic pagan use of various (including ‘Christian’) images
and words on gems engraved for magical purposes as distinct
from the specifically Christian invocation of Jesus’ name in
conjunction with devotional images on gems worn in finger-
rings by Christians.25
Conclusion
According to the surviving evidence, the earliest images of the
Crucifixion have generally been thought to have been
produced in the West in the 5th century ad, relatively late in
the broad development of Christian iconography. The first
dates to ad 420–30 and appears in the pictorial narrative of the
Passion that is arranged across the series of ivory panels known
as the Maskell Passion ivories (Pl. 8).26 The second, slightly
 The Constanza Carnelian and the Development of Crucifixion Iconography in Late Antiquity
Plate 8 Ivory plaque with the Crucifixion, Rome, AD 420–30. 75 x 98mm. London,
British Museum, PE 1856,0623.5
later, representation is found amongst the cycle of episodes
from the Old and New Testament illustrated on the wooden
doors of the church of Sta Sabina in Rome, carved in the ad
430s (Pl. 9). The evidence yielded from the study of engraved
gemstones, which furnish even earlier representations of the
subject, suggests that experimentation with pictorial
representations of Jesus affixed to the cross had begun by the
3rd century ad. While the experimentation might not have
been extensive, or the results popular (given the paucity of
evidence), the fact of it is very clearly if unexpectedly attested
by the magical Pereire gem in the first instance, and the
Constanza and Nott gems in the second.
The magical gem (Pl. 6), when placed alongside and with
the evidence of the near contemporary Palatine graffito
indicates several things. Firstly, that representations of a
crucified figure were being attempted in the eastern as well as
the western parts of the Roman Empire by the 3rd century ad.
The very portability of the gem, even if confined to private use
by an individual, would also suggest that designs for the
Crucifixion were in circulation (however limited) amongst
makers as well as customers at an early date. Secondly, that
pagans used such representations for their own purposes,
Plate 9 Detail of wooden door panel with Crucifixion, Sta Sabina , Rome, c. AD 430
including magical and satirical; and therefore thirdly, that the
Crucifixion was understood by Christians and non-Christians
alike as a powerful and efficacious symbol. The iconographic
evidence links directly with literary evidence, only briefly
discussed here, for the efficacy of verbal proclamations of
‘Christ crucified’ as a means of warding off evil amongst Jesus’
followers and non-followers alike.
Alongside this visual evidence, the Constanza and Nott
gems attest to the fact that in the following century, when
further experimentation with Crucifixion iconography was
clearly taking place, the interest in the subject among Christian
communities remained firmly on the death as a demonstration
of Jesus’ triumph and authority. Furthermore, that while the
magical and indeed Palatine representations seem to illustrate
pictorial practices of evoking some sense of ‘story’ or narrative
and practical detail in terms of documenting aspects of the
‘how’ Jesus was crucified (such as clearly rendering the cross,
the victim’s means of affixation to it, even posture and the
position of the legs), the Constanza and Nott gems appear to
have somewhat broader, symbolic concerns. These pertain
directly to the manifestation of Jesus’ power in defeating
death, and so to the depiction of his crucified yet victoriously
resurrected body symbolically adored in heaven by the
Apostles. The presence of the twelve Apostles beside the cross
is not in accord with accounts of the Crucifixion in the
canonical Gospels, which specifically note their abandonment
of Jesus (Matthew 26:56, Mark 14:50, Luke 22:54, John 18:15).
Indeed, from surviving evidence it appears that no such
depiction of the subject occurs again until the early 8th century
ad, in Rome on the triumphal arch of the church of Sta Maria
Antiqua (ad 705–7), where rows of angels and peoples replace
the Apostles in processing towards and adoring the crucified
but triumphant Christ.27 Hence rather than purporting to fulfil
a narrative function in the ways developed in other media for
the representation of Jesus’ death after the 5th century ad, the
Nott and Constanza gems seem to anticipate later symbolic and
even apocalyptic representations of the scene, such as occurs at
Sta Maria Antiqua.
The Constanza and Nott iconography suggests that before
arriving at a visual depiction of the Crucifixion in its historic/
narrative guise for public art (which extant evidence
documents as emerging in the 5th century ad),
experimentation with the depiction of the event occurred in
the miniature arts, utilising those triumphal artistic formats
developed with increasing success in other media for the
expression of the veneration of Jesus (and also, the veneration
of the cross). Given the dating of contemporary sarcophagi, it
can be argued that the remarkable survival of these two gems
provides rare evidence for the existence and circulation of such
a representation of the crucified Jesus amongst Christian
communities, probably by the mid-4th century ad. From the
rarity and lack of other evidence at this time it might also be
concluded that in competition with other highly successful
types for the representation of Jesus triumphant, the crucified/
resurrected variant on that overwhelmingly popular
iconographic pattern was not transferred to, or tested in, other
media for further development. Nevertheless, it did survive in
some form, as the re-emergence in such Early Byzantine
apocalyptic scenes as that preserved at Sta Maria Antiqua
seems to suggest.
‘Gems of Heaven’ | 219
Harley-McGowan
The question of the rarity of images of the Crucifixion in
Late Antiquity is by no means solved by this brief assessment of
the evidence provided by engraved gems. Nevertheless, an
attempt has been made here to raise awareness of the evidence,
pointing out not only the 3rd century ad existence of
Crucifixion iconography, but manifestations of that existence
in art used in magical and Christian contexts. As Harcourt-
Smith presciently observed, the iconography preserved by the
Constanza and Nott gems contributes key evidence in the
broader history of the representation of the subject in Christian
art, and so deserves to be fully incorporated into a detailed
account of that iconographic development. And as Derchain
further demonstrated, in the process of charting this
development, the testimony of magical gems should be allowed
to sit alongside that of Christian gems. As a result, when
considering the rarity of Crucifixion images in Late Antiquity, a
wider range of evidence can be taken into account. The
connection between the visual and literary sources for the
reference to and invocation of Christ crucified as a source of
power, is just one step in directly challenging that persistent
belief that persecuted Christians were too scared or too
ashamed to name and depict the subject explicitly. Certainly,
the rarity of representations remains something of a mystery,
given the evident ability of artisans to depict the subject. Yet
alternative explanations beyond claims of avoidance of or
refusal to depict the image are now able to undergo further
consideration.
Acknowledgements
I acknowledge with gratitude the support and encouragement of
Jeffrey Spier and thank him for critical observations and suggestions.
My thanks are also due to: Christopher Entwistle; Robin Jensen; and
Andrew McGowan, for his comments.
Notes
1 The gem was first published by C. Harcourt-Smith, ‘The
Crucifixion on a Greek Gem’, The Annual of the British School at
Athens 3 (1896/7), 201–6.
2 Franks gave at least 1,500 objects to the British Museum in 1895,
one year before he retired from the Keepership: M. Caygill, ‘Franks
and the British Museum – the Cuckoo in the Nest’, in M. Caygill and
J. Cherry (eds), A.W. Franks: Nineteenth-Century Collecting and the
British Museum, London, 1997, 96–7.
3 For the formula, its appearance on Christian gems and frequent
use in conjunction with symbols as well as narrative scenes, see
now J. Spier, Late Antique and Early Christian Gems, Wiesbaden,
2007, 34–5 and 35, n. 24.
4 The expansion of knowledge was only recently provided by Spier
in his comprehensive catalogue, cited above, and on this gem and
its date, no. 444, 73–4.
5 The contents of this paper present a brief synthesis of detailed
arguments contained in F. Harley, Images of the Crucifixion in Late
Antiquity: The Testimony of Engraved Gems, Unpublished PhD
dissertation, University of Adelaide, 2001, and in a forthcoming
monograph.
6 References to Clement here are from the Ante-Nicene Fathers. For a
close discussion of the passage see P.C. Finney, ‘Images on Finger-
rings and Early Christian Art’, Dumbarton Oaks Papers 41 (1987),
181–6.
7 See Spier (n. 3), chapters 3 (inscriptions and monograms), 4
(symbols), 5 (Good Shepherd) and 6 (narrative scenes, including
220 | ‘Gems of Heaven’
some 3rd century ad examples).
8 R. Garrucci, Storia dell’ arte cristiana nei primi otto secoli della
chiesa, 6 vols, Prato, 1872–81, vol. 6 (1880), 124, pl. 479, no. 15.
9 Variations on the spelling of Jesus’ name appear on other Early
Christian gems: Spier (n. 3), nos 205, 206 and 446.
10 Spier argues that the nimbus indicates a date no earlier than the
Constantinian period, ‘before which time nimbi are unlikely to
have been used in a Christian context’: Spier (n. 3), 74.
11 Spier (n. 3), 11.
12 W. Heitmüller, Im Namen Jesu: Eine sprach- und religions-
geschichtliche Untersuchung zum Neuen Testament, speziell zur
altchristlichen Taufe, Göttingen, 1903. The contents are
recapitulated in the discussion by D. Aune, ‘Magic in Early
Christianity’, in H. Temporini and W. Haase (eds), Aufstieg und
Niedergang der römischen Welt, Part II, 23/2, Berlin and New York,
1980, cols 1507–57, especially 1546–9.
13 Aune (n. 12), 1547–8.
14 Both depictions of Christ among the Apostles on the two long sides
of the Sarcophagus of Stilicho, sculpted in or near Milan in the late
4th century ad (now preserved beneath the pulpit, Basilica
Sant’Ambrogio, Milan), include a depiction of a lamb beneath his
feet, as on the Nott gem. J. Dresken-Weiland, Repertorium der
christlich-antiken Sarkophage, vol. 2, Mainz, 1998, 56ff, no. 150
with bibliography, and pl. 60.1.
15 Ivory ‘Samagher’ Casket, Rome (?), c. ad 450, 185 x 205 x 161mm.
Museo Archaeologico Nazionale di Venezia. D. Longhi, La Capsella
Eburnea di Samagher: Iconografia e Committenza, Ravenna, 2006, pl.
VIIa.
16 B. Christern-Briesnick (with H. Brandenburg and G. Bovini),
Repertorium der christlich-antiken Sarkophage, vol. 3, Mainz,
2003, 35–6, no. 49.
17 M. Lawrence, ‘Columnar Sarcophagi in the Latin West’, Art
Bulletin 14 (1932), 112–15 and 173, nos 90–5; G. Koch, Früh-
christliche Sarkophage, Munich, 2000, 315–16.
18 On this point: B. Kühnel, From the Earthly to the Heavenly Jerusalem,
Rome, 1987, 69, as part of her broader discussion of the association
and identification of Jesus with the cross in monumental artistic
contexts at this period. According to Lawrence, the sarcophagus in
Arles preserves the ‘fullest and best’ example of the star-and-wreath
type: Lawrence (n. 17), 173, no. 90, with early bibliography; and now
Koch (n. 17), 316, no. 50 with further bibliography.
19 First published by Philippe Derchain in 1964: ‘Die älteste Darstellung
des Gekreuzigten auf einer magischen Gemme des 3. (?) Jahr-
hunderts’, in K. Wessel (ed.), Christentum am Nil, Recklinghausen,
1964, 109–13.
20 J. Spier and F. Harley, ‘Magical Amulet with Crucifixion’, in J. Spier
(ed.), Picturing the Bible: The Earliest Christian Art (exh. cat., Fort
Worth), Fort Worth, 2007, 228–9, including bibliography.
21 J. Zias and E. Sekeles, ‘The Crucified Man From Giv’at ha-Mivtar: A
Reappraisal’, Israel Exploration Journal 35 (1985), 22–7.
22 M. Itkonen-Kaila and H. Solin in V. Väänänen (ed.), Graffiti del
Palatino vol. 1: Paedagogia (Acta Instituti Romani Finlandiae 3),
Helsinki, 1966, 40–1, fig. 35 and 209–12, no. 246; I. Di Stefano
Manzella, Le iscrizioni dei cristiani in Vaticano, Vatican City, 1997,
192–4; Harley in Spier (n. 20), 227–8.
23 Harley (n. 5), 133, 135–6; and Spier (n. 3), 75.
24 On Christian magic, see Aune (n. 12), cols 1507–57.
25 Spier (n. 3), passim, and specifically 82.
26 For a brief discussion, see Harley in Spier (n. 20), 229–32.
27 U. Nilgen, ‘The Adoration of the Crucified Christ at Santa Maria
Antiqua and the Tradition of Triumphal Arch Decoration in Rome’,
in J. Osborne et al. (eds), Santa Maria Antiqua al Foro Romano cento
anni dopo (confer. proc., Rome, 2000), Rome, 2004, 128–35. For a
recent (2008) reconstruction of the triumphal arch by Per Jonas
Nordhagen and Per Olav Folgerø see: P.O. Folgerø, ‘The Lowest,
Lost Zone in The Adoration of the Crucified Scene in S. Maria
Antiqua in Rome: A new conjecture’, Journal of the Warburg and
Courtauld Institutes 72 (2009), 215, fig. 9.
Seals in Transition
Their Change of Function and Value in Late Antiquity
Gertrud Platz-Horster
The question as to why gemstones almost disappeared in Late
Antiquity has concerned many scholars of ancient glyptics.
This conference at the British Museum has given me a welcome
opportunity to review it again.
In his ‘unhistorical historic comedy’ Romulus der Große, the
Swiss author Friedrich Dürrenmatt ridiculed the prevalent
opinion that culture ended with the decline of the Roman
Empire.1 Basic research in Late Antiquity during the last
decades has mainly taken place in Anglophone scholarship –
from Peter Brown2 to Bryan Ward-Perkins3 – discussing its
definition with quite different results. In 2005, László Török in
Transfigurations of Hellenism explored continuity and change
in the arts in Ptolemaic to Early Byzantine Egypt.4 Aziz
Al-Azmeh in his lecture ‘Rome, New Rome and Baghdad:
Pathways of Late Antiquity’ given at the Wissenschaftskolleg zu
Berlin and the Berlin-Brandenburgische Akademie der
Wissenschaften in March 2008, argued for the continuation of
Graeco-Roman culture in Early Islam: he focused on the
importance of Arabic historians in translating Greek literature
and on their influence on the transformation of Hellenic
culture post-antiquity.
With regard to engraved gems, Adolf Furtwängler in his
magnum opus, Die antiken Gemmen, published in 1900,
expressed his vehement disgust at the loss of sensibility for the
medium, the poor quality, and the small range of devices,
starting in the late 2nd century ad and degenerating more and
more towards Late Antiquity. This verdict – in the tradition of
the 19th century – deeply influenced glyptic studies for more
than 100 years. Only in 2007 was Jeffrey Spier the first to
dedicate an entire publication to ‘Late Antique and Early
Christian Gems’.5 His catalogue comprises over 1,000 items,
including lost objects known only from the literature. That is a
very small number compared, for example, to the up to 10,000
Roman engraved gemstones found in Aquileia alone. Both
Spier and Erika Zwierlein-Diehl6 have noted that the technical
ability to engrave gemstones did not vanish, as excellent
imperial portrait gems, elaborate vessels in hard stone or glass
and the unchanged high quality of the coinage testify.
Furtwängler wrote:
.... die Kunst der Glyptik tritt zurück; sie verliert ihre Stellung als
eine bevorzugte Modekunst, die sie in voriger Periode hatte, und
sie kommt herab, schon im zweiten Jahrhundert n. Chr.7
But was it only a fashion to use engraved gems set in finger-
rings for sealing? Zwierlein-Diehl has suggested that primarily
a change in sealing practices may have been responsible. When
and why did this change from sealing with engraved gems to
the use of crude metal finger-rings or lead bullae occur, and
what was its cause? My paper will focus on these questions. I
will approach them by giving a selected overview of the –
admittedly not very many – archaeological contexts with
engraved gems in hoards or burials which attest to their
change in function and value.
Dated contexts with engraved gems in the West
In all times of war, people have concealed their personal goods
of value. The treasure from Regensburg-Kumpfmühl8 in south-
eastern Germany was found in 1989 inside the wall of the
Roman castle. The hoard, hidden in a large bronze pot,
contained 25 gold, 650 silver and three bronze coins as well as
female jewellery. The coins date the deposit of the hoard to
after ad 166. At that time, the Romans defeated the constant
attacks by the Germanic tribe of the Marcomanni from the
eastern bank of the river Danube. Three gold finger-rings of
identical form with broad shoulders bear engraved nicoli of a
similar shape and style, whereas the tiny fourth one (inner
diam. 16.8 x 11.6mm) with a banded bezel holds a glass gem
imitating nicolo. All of them may be dated as contemporary
with the deposit. From their interior measurements (max. c. 14
x 17mm), these gold finger-rings must have belonged to a
woman together with the rest of the jewellery. Their settings
indicate that all four gems could have been used for sealing.
A hundred years later, in the mid-3rd century ad, the
situation for the Roman Empire was even more precarious,
both along its borders as well as inside the Empire. The large
number of deposits attests to the fears of the population.9 The
treasure of Eauze in south-west France included around 28,000
coins with a total weight of 120kg, minted between the reigns
of Commodus (ad 177–92) and Postumus (ad 260–8), the last
one issued in ad 261.10 As most of the coins are of silver
(denarii), the hoard had an enormous value and must once
have belonged to a military or political Roman commander.
Postumus, being one of the commanders against the Franks
and the Alamanni, proclaimed himself Caesar of the Imperium
Gallicum in ad 260 in opposition to the legal Caesar Gallienus,
who was defending the borders against Sasanian forces in the
East. In the ‘Eauze treasure’ among the gold jewellery of seven
necklaces, five pairs of earrings, three bracelets, pendants etc.,
there are four gold finger-rings, the two most elaborately
worked ones set with cameos. Six engraved gems, three nicoli,
two glass gems imitating nicolo and one carnelian of poor
quality, are unset – as is a fine carnelian cameo with a female
bust (see the paper by Guiraud this volume, Pls 4–7). The two
glass gems may be contemporary with the deposit, whereas the
others originate in the 1st and 2nd centuries ad. The unset
gemstones had obviously not been worn as jewellery before
being hidden in the treasure for their value per se.
The same phenomenon can be observed in a male grave of
the late 3rd century ad in Xanten on the lower Rhine (Pl. 1).11
Within a vast necropolis of more than 500 already excavated
graves, along a street connecting the veteran’s city Colonia
Ulpia Traiana to the former castle for two Roman legions Vetera
Castra, in 2000 the skeleton of an adult man was found buried
‘Gems of Heaven’ | 221
Platz-Horster
Plate 1 Xanten, Viktorstrasse, male grave no. 10
with rich grave goods which are unique to Xanten. Among local
ceramics and glass were the remains of a textile bag containing
a jet necklace, a gold pendant, a bronze finger-ring, a glass disk
and six unset gems; all but one of the gems are much older than
the burial. The most precious stone is an amethyst engraved
with the unfinished capita opposita of Nero and his mother
Agrippina the Younger, which can be dated by coins to ad 55/6.
Although the amethyst was damaged and unfinished, it was
kept for more than 200 years as an item of high value and
buried as such with the other gems – ultimately not used either
for sealing or as an inlay for jewellery.
This is one of the last dated contexts with engraved gems in
Xanten. Beneath the cathedral of St Viktor, in graves of the late
4th century ad, only two gems set in finger-rings have been
found: a silver ring from a man’s burial holds a 1st-century ad
carnelian depicting the goddess Nemesis;12 the second, found
in a boy’s burial, was a bronze ring with a glass gem imitating
nicolo, showing Orpheus among his herd.13 This is a well-
known scene on numerous glass gems found throughout the
Roman Rhineland, Trier, Gaul and Carnuntum. In 1902, during
the construction of the University Hospital in Bonn in the
canabae of the Roman legionary castle, a large complex of
cheap jewellery was discovered, including objects in pseudo-jet
(from the local brown coal, lignite), bronze and glass, dating to
the mid-3rd century ad.14 It preserved the contents of a
jeweller’s shop for the soldiers who lived nearby. Whereas the
bronze items seem to have been manufactured in Mainz, the
pseudo-jet and glass objects were presumably made in
Cologne, a centre of glass-making in the West – especially the
mass-produced glass gems imitating nicolo which were spread
all over the northern provinces of the Roman Empire from the
Rhineland to Gaul as far as Noricum in the East. This complex
in Bonn also signifies a trend of the 3rd century ad: the
222 | ‘Gems of Heaven’
replacement of precious gemstones by cheaper glass gems.
In 1988 Hélène Guiraud, on the basis of around 1,000 gems
found in Gaul from the period of the Roman occupation,
published a graph which clearly shows the increase of glass
gems imitating nicolo during the 2nd century ad with a peak at
around ad 200. At the same time the relationship between the
gems in glass and stone changes, mainly towards carnelian,
nicolo and jasper. And at the right end of both graphs, the
disappearance of gems in both glass and stone in the 4th
century ad is labelled ‘la quasi disparition de la glyptique au
IVe siècle’.15 Guiraud calculates their percentage up to 3% of all
gem finds in Gaul and the Rhineland.16
On the basis of the 637 gems and cameos found by 2009 in
and around Xanten, I have evaluated the 341 items from the
‘Fürstenberg’ site and the 125 items from the Colonia Ulpia
Traiana and the cemeteries in between (Pl. 2).17 The latest
gems in precious stone or glass from these sites can be dated by
style, setting and/or by context to the first half of the 3rd
century ad. The latest archaeological context including gem
stones is that mentioned above, the cathedral of St Viktor,
dating to the late 4th century ad, but the gems dating
respectively to the 1st and early 3rd centuries ad. Among all the
gems from Xanten, there is not a single one dating to the 4th
century ad. Only one glass cameo comes from the nearby
burgus in Goch-Asperden, established by Valentinian I after
ad 369: this delicate glass cameo shows the portrait bust of a
lady with a hairstyle comparable to coins of Aelia Flacilla
(ad 379–83).18
The disappearance of glyptics in Xanten during the late 3rd
century ad is even more noteworthy as none of the continuous
attacks by the Franks between ad 260 and 276 seems to have
caused the end of civil life in the Colonia Ulpia Traiana. But as
the population had diminished significantly, by the early 4th
 Sardony

Glas, un
Lagenac

gebände

„Nicolo
Karneol

Onyx s

K-onyx

Gesamt
früh/sp

rot/grü

Am/Be
Nicolo

Achat/
Jaspis,
Band-
achat

Glas,
Sard
Datierung
Seals in Transition
Die antiken Gemmen aus Xanten. Teil III: Neufunde, Neuerwerbungen, Nachträge und Auswertung 131
150–50 vC 1 3 10 2 5 4 4/- - - - - 1/-/- 30

50 vC–0 46 10 5 2 7 5 18 (1 K)/- - - 1/- 4 - 98

Achat/Chalc.

Gesamt Dat.
„Nicolo-P.“,

Onyx schw.
Material

Lagenachat/

gebändert
Sardonyx

früh/spät
Glas, uni

K-onyx/
rot/grün
Karneol
augusteisch 41 5 - 2 5 2 19/- 3 - 2/1 - - 80

Am/Be
Nicolo

Jaspis,
Band-
achat

Glas,
Sard
Datierung
20–70 nC 57 2 1(?) 4 2 - 6/- 12 1/- 2/1 - -/1/1 90

150–50
70–130 vC
nC 1
16 3- 10- 2- 5- 4- 4/-
- -- -
1/- -- -- 1/-/-
-/1/- 30
18

50 vC–0
2./3. Jh. nC 46
11 10- 5- 2- 7- 5- 18 (1
-/3K)/- -- -
4/2 1/-
1/- 4- -- 98
21

augusteisch 41 5 - 2 5 2 19/- 3 - 2/1 - - 80

Gesamt Mat. 172 20 16 10 19 11 47/3 15 6/2 6/2 4 1/2/1 337
20–70 nC 57 2 1(?) 4 2 - 6/- 12 1/- 2/1 - -/1/1 90
[K = Kameo. Letzte Spalte: Karneolonyx, Amethyst, Bergkristall. –
70–130 nC 16 - -Dazu: 4 ungravierte
- - Gemmen:
- -
AGXanten - 1/-
II Nr.197–200] - - -/1/- 18

2./3. Jh. nC
Tabelle 1 11Fundorte:
- Fürstenberg/Vetera
- - -
I (13/12 v.- Chr.–69/70
-/3 n. Chr.)
- und Vetera
4/2 II1/- -
(71/72–275 -
n. Chr.) – 21
Material : Datierung : Anzahl.
Gesamt Mat. 172 20 16 10 19 11 47/3 15 6/2 6/2 4 1/2/1 337

[K = Kameo. Letzte Spalte: Karneolonyx, Amethyst, Bergkristall. –

Dazu: 4 ungravierte Gemmen: AGXanten II Nr.197–200]

Tabelle 1 Karneol
Material Fundorte: Fürstenberg/Vetera
Sardonyx/ Glas,I (13/12 v. Chr.–69/70Nicolo
„Nicolo-P.“, n. Chr.) und Vetera IIPlasma
Jaspis, (71/72–275 n. Chr.) – Gesamt
K-onyx/
Lagenachat Material : früh/spät
uni Datierung : Anzahl. rot/grün Am/Be/ Dat.
Datierung Cha

100 vC–20 nC 8 1 6 -/- - -/- - -/1/-/- 16

1. Jh. nC 8 1 2 3/2 (1 K) 5 1/- 3 -/2/2/- 29

2. Jh. nC
Material 17
Karneol -
Sardonyx/ -
Glas, -/1
„Nicolo-P.“, 16
Nicolo 10/2
Jaspis, -
Plasma 1/-/2/1
K-onyx/ 50
Gesamt
Ende 2./3. Jh. Lagenachat uni früh/spät rot/grün Am/Be/ Dat.
Datierung 3 1 (K) 2 -/17 2 2/- 3/-/-/-
Cha 30
nC
100 vC–20
Gesamt nC
Mat. 368 31 106 -/-
3/20 23- -/-
13/2 3- -/1/-/-
4/3/4/1 16
125

1. Jh. nC 8 1 2 3/2 (1 K) 5 1/- 3 -/2/2/- 29

[K = Kameo. Letzte Spalte: Karneolonyx, Amethyst, Bergkristall, Chalcedon]
2. Jh. nC 17 - - -/1 16 10/2 - 1/-/2/1 50
Tabelle 2
Ende2 2./3.
Fundorte: Colonia Ulpia Traiana (98/99–275/276 n. Chr.) und Tricensimae(?) (Ende 3. Jh.–Anfang 5. Jh.);
Jh. gems from Vetera Castra and Colonia Ulpia Traiana. Tables: find spot / material / dating / number
Plate Xanten,
3
Gräber 1 (K)
Viktorstrasse 2
und Dom (3.–4. -/17 2
Jh. n. Chr.) – Material 2/-
: Datierung : Anzahl. 3/-/-/- 30
nC

Gesamt Mat. 36 3 10 3/20 23 13/2 3 4/3/4/1 125

century ad the city was reduced to nine central insulae. The century ad the Sasanians attacked and destroyed the most
smaller city, called Tricensimae (?), was strongly fortified by important Roman cities on the eastern borders.22 The many
[K = Kameo. Letzte Spalte: Karneolonyx,
large walls of 4m thickness and 48 towers. Finally, Roman life Amethyst, Bergkristall,
clay Chalcedon]
seals from Kommagene-Doliche, destroyed in ad 256 by
in Xanten ended in the mid-5th century ad. 19
Shapur I (ad 240–70), 23
are now spread all over the world. But
Tabelle 2 Fundorte: Colonia Ulpia Traiana (98/99–275/276 n. Chr.) und Tricensimae(?) (Ende 3. Jh.–Anfang 5. Jh.);
The relationshipGräber
between the precarious
Viktorstrasse und Dom political situation,
(3.–4. Jh. at Zeugma
n. Chr.) – Material on the
: Datierung Euphrates in 1999–2000 the city archive of
: Anzahl.
the hiding of precious objects and the value of gems may be more than 102,000 items was excavated. The texture of their
attested by the treasure from Isny, found in 1968 in the Late reverses indicates that papyrus and leather were the materials
Roman castle of Vemania in southern Germany (Pl. 3).20 The on which the burnt documents were written. Seventy percent
193 silver and copper coins lay partly rolled in a linen sack on of the impressions reproduce small private seals, very few are
top of a wooden casket which contained female jewellery. The double seals. Among the 297 items published so far by Mehmet
latest coin was issued in ad 305. The various gold jewels are Önal, impressions from gems of the 2nd century bc to the first
combined with beads in coral, jet, glass and bone; also, three of half of the 3rd century ad are represented, including Roman
the five finger-rings are made of pure gold, but all of them are Republican devices (for example portraits, the Rape of the
set with simple glass gems – one in blue and the other four Palladion by Diomedes and Ulysses), gems with inscriptions
imitating nicolo (inner diameters: 15 x 18mm–18 x 22mm). The and those with anguipedes. Seleukia/Zeugma in the south-
jewellery is dated to the 3rd century ad, probably the second east of Turkey, from its foundation in around 300 bc, was a
half, and might have belonged to the wife of the praefectus flourishing city of military and administrative importance
based at the castle which was overrun by the Alamanni in situated at an easy crossing-point of the river Euphrates.24 The
ad 305. archive preserved documents for customs, trade and
transportation, the communication network, etc. After Shapur I
Seals and sealings in the West had burned and razed the city in ad 252, a violent earthquake
A look at the East of the Roman Empire finds a corresponding somewhat later buried everything beneath rubble.
political situation affirmed by excavated archives with clay Also in this region of the Empire, the ongoing battles
bullae. Among the many find-spots of clay seal impressions in followed by the migration of peoples during the second half of
both the eastern Mediterranean and the Near East,21 the latest the 3rd century ad caused a deep transformation of all social
in date are also the most recently found. During the mid-3rd structures. Although there is no archaeological evidence for

‘Gems of Heaven’ | 223

Platz-Horster
the use of Roman gemstones so late in this region, the art of
gem engraving was still cultivated by the Sasanians (ad 224–
642/51) after Shapur I – sometimes in very precious stone. Also
the custom of sealing is attested by clay impressions until the
Muslim Arabs’ conquest of the late 7th century ad. A large clay
bulla in Munich from Iran uniting 16 seal impressions (67 x 66 x
17mm) may once have served as the official certificate of a
document or of a valuable product. The larger seal in the centre
with an inscription in Pahlavi belonged to the governor of
Georgia.25 An earlier Sasanian bulla in a private collection,
together with five seal impressions, shows the portrait of a
Constantinian prince.26
Roman seal-boxes
Having looked at gemstones as original seals in the West and at
clay bullae as seals in the East of the Roman Empire, both of
which ended in the course of the political commotion during
the second half of the 3rd century ad, I will focus on another
sealing medium: Roman sealing capsules. Recently Alex Furger
224 | ‘Gems of Heaven’
Plate 3 The Isny Treasure, Stuttgart
published the entire finds from the excavations at Augusta
Raurica/Augst in northern Switzerland.27 On the basis of 138
bronze capsules – mostly excavated in dated contexts – and of
related material, he gives the first typology for this rather rare
species which first appeared around 100 bc and vanished
around ad 280, astonishingly at the same time as both
gemstones and clay impressions. These capsules served for
sealing the knot of the cord which tied up a document or a
parcel. The string was led through the lateral cavities and the
knot placed in the box; the capsule was then filled with heated
wax which could extend through the holes in the bottom of the
box and a hole in the closed lid. The device of a gemstone could
be pressed into the warm wax before closing the lid, but this
did not necessarily have to happen.
These tiny capsules – chiefly in bronze, but also in bone or
wood – occur almost exclusively in the north of the Roman
Empire. This confirms a statement by Cicero in 63 bc in his
oration ‘Pro Flacco’ (37–8) in which he speaks of a testimony
said to have been ‘sealed with that Asiatic clay’, but obviously

Plate 4 Twenty-two gold finger-rings from the Thetford Treasure. London, British Museum, PE 1981,0201.2–23
‘the seal was of wax’.28 Wax impressions protected in seal-
boxes replaced sealing in clay due to the climate: this cera
sealing in easily melted beeswax was unsuitable for the hotter
zones of the Empire, where sealing in quick-drying creta (clay)
was an ancient tradition. A look at the distribution of
Nomophylakia – archives with clay seal impressions preserved
by the accident of fire – confirms this.
Roman lead seals
Around ad 100 another method of sealing appeared: struck in
lead, these seals as well fixed the knot of a string whose hole is
usually preserved. The earlier lead seals are impressed on one
side only, the reverse being bulbous or flat, although some-
times with the impression of the material to which they were
affixed. From the Constantinian era onwards, however,
double-sided lead seals, made by a pincer-shaped bulloterion,29
predominated.
A high percentage of the lead seals from Trier (2,500 items)
were found in the river Mosel near the Roman bridge and the
harbour, but also near the Late Antique palace. Most of them
served for sealing goods from all over the Empire which had
Trier as their destination. After authenticating and opening the
cargo, these lead seals were thrown into the river to avoid their
second usage. Trier at that time was not only the military and
administrative capital of the western Empire, but also an
economic and trading centre, especially for shipping. The lead
seals found in the river indicate the customs and taxes the
sender had to pay. The close correspondence of the finer seals
to coinage has led to the suggestion that their engravers
belonged to the state mint. These lead seals are spread all over
the Late Roman world (from Istanbul to Lyon) and continued in
use throughout the Byzantine era.
The re-use of Roman gems
The Thetford treasure, discovered in 1979 in East Anglia, is
presented here as one example of quite a number of very rich
4th-century ad hoards. It contained 83 items of precious
materials, among which were 22 gold finger-rings (Pl. 4).30 Five
of these rings are set with re-used earlier engraved gems of
Seals in Transition
amethyst, chalcedony, jasper and nicolo; others are decorated
with garnets, emeralds, amethysts or coloured glass; two of the
settings are empty. The hoops, bezels and settings are
elaborately highlighted with delicate filigree and even relief
decoration. While the Roman intaglios, dating to the 1st to 3rd
century ad, are all re-used, the gold finger-rings and the other
jewellery look almost entirely brand new. They seem to be the
products of one workshop, probably situated in Gaul. The
re-used Roman engraved gems were chosen for their colour,
not being valued more highly than the beautiful plain
gemstones. One unmounted carnelian, cut down to a
rectangular form, a chalcedony intaglio and a sardonyx cameo
set in gold pendants, complement this part of the treasure,
presumably a jeweller’s or a merchant’s stock-in-trade.
During the 4th century ad the numerous examples of
elaborate gold work contrasts with the Roman gemstones, now
being prized and re-used for their precious and colourful
material, not to fulfil their former function as seals.31 The
‘Getty hoard’ of 15 pieces of jewellery, buried around ad 400
presumably in the east of the Roman Empire, contains six gold
finger-rings of similar shape and decoration, three of them set
with re-used Roman engraved carnelians, which could not
have functioned as seals due to the delicacy of their settings
(Pl. 5).32 The finger-rings might have come from one workshop,
as did the rings in the Thetford Treasure. As at Thetford, the
‘Getty hoard’ also included a re-used 3rd-century ad sardonyx
cameo representing the busts of a couple framed by a gold
brooch which later on was mounted with coarse loops for use
as a pendant.33
Already during the 3rd century ad ‘barbarian’ veterans
were settled by the Roman military within their territory, and
because of this, a gradual acculturation is evident in burials –
like Roman gems set in Germanic finger-rings within Germania
inferior and superior. On the other hand, the ‘barbarians’ by
attacking the Roman Empire carried off precious objects and
inserted them into their own personal goods, as attested, for
example, in Germanic graves in Germania libera like Leuna,
Haßleben, and Zethlingen. The high estimation in which
colourful Roman gemstones were held was continued by the
‘Gems of Heaven’ | 225
Platz-Horster
Plate 5 Finger-rings in the Getty Hoard. Malibu, The J. Paul Getty Museum
Merovingian and Frankish tribes who re-used them in
elaborately worked gold finger-rings and fibulae, often
neglecting the orientation of their devices: they were no longer
intended for sealing.34
This evidence attests to the inter-cultural influence
between the Romans and migratory tribes from the 3rd
century ad onwards and the decline of gemstones in their
function as personal seals. Ammianus Marcellinus (ad 330–c.
395) in his Res gestae described the kind of peace agreement
between Constantius and the Alamannic Kings Gundomad and
Vadomar in ad 354 near Augusta Raurica as gentium ritu, i.e.
according to their customs.35 This loss of power led the Roman
emperors to accept the conventions of foreign aggressors; and
this probably also negated the necessity of sealing contracts or
other documents, as was customary under Roman law.
Different Germanic tribes like the Franks and the
Alamanni, for example, not only settled more and more in
Roman territories, but in the end conquered them. Still, the
Plate 6 Gold ring from Krefeld-Gellep, male grave 1782
226 | ‘Gems of Heaven’
most precious of their grave goods were often Roman gems set
into delicate gold finger-rings. In the necropolis of Krefeld-
Gellep (grave no. 1782), the ring finger of the left hand of a
male skeleton, attributed by the excavators to the Frankish
duke Arpvar, buried around ad 520–30, was adorned with a
fine gold ring set with an Augustan-period nicolo gem
depicting two satyrs, one playing the aulos (Pl. 6).36 This
phenomenon can be observed in quite a number of burials of
the 6th and 7th century ad from Anglo-Saxon England to
northern Italy. In any case ancient gems were highly prized as
settings for elaborate finger-rings.37
From the time of Justinian I in the mid- 6th century ad to
Justinian II in the second half of the 7th century ad, even in
Rome, engraved gemstones were re-used for contemporary
jewellery. Beneath the ‘Crypta Balbi’, amongst the material of a
jeweller’s workshop of that period, were found nine Roman
gems, partly broken, as well as two Etruscan scarabs,
fragments of old glass, bone and ivory, and precious stones
such as garnet, lapis lazuli and sapphire, coral and mother of
pearl.38
The ‘crisis of the Imperium’: a change in sealing custom
To sum up: already during the attacks of different Germanic
tribes in the mid-2nd century ad some valuable hoards were
being buried. The ‘crisis of the Imperium’ in the second half of
the 3rd century ad resulted in the collapse of the Roman
borders and intensified the invasions of hostile troops in both
the west and the east of the Roman Empire. This upheaval
caused a concentration of Roman forces in a few locations, a
rapid inflation of the Roman currency through the constant
fresh minting of money for new troops, a reduction of the
alienated Roman population and their turn to real assets such
as gold and precious stones.

The following migration of peoples and the amalgamation
of the Roman population with the ‘barbarians’ – many of whom
became Roman citizens and climbed up the social ladder – also
caused ‘a drastic shift’39 in the custom of sealing: in the western
Roman Empire, engraved gems and bronze seal capsules
disappeared at the same time around ad 280, as did clay seal
impressions in the East. Although the art of gem engraving did
not vanish – and was continued also for daily life sealing, for
example, by the Sasanians at a high level – the Romans in the
future only used lead seals struck by or for the administration
to raise taxes on trade goods.
Augustine in a letter from ad 402 described the device of
the seal he had used to be sure that it arrived undamaged;40 and
the pagan Roman senator Q. Aurelius Symmachus, who
probably died in the same year, asked his friend Virius
Nicomachus Flavianus to confirm the receipt of his letter by
describing its seal.41 Other than written testimonies like these,
the archaeological evidence contradicts the use of a personal
engraved gemstone for sealing letters or documents at that
time. But actually we do not know which kind of annulus
Augustine or Symmachus really used: an engraved gem – or
perhaps a metal finger-ring, the latter surviving antiquity up to
the Middle Ages.42
The availability of precious stones
One additional reason for the decreasing numbers of
gemstones may have been the accessibility of the material
itself. It is obvious from excavated contexts that the superior
precious stones like emerald, garnet, amethyst, aquamarine or
even clear carnelian already during the 2nd century ad
diminish in burials and even in rich hoards.43 Less expensive
materials like impure carnelian or rock crystal, red and yellow
jasper, haematite or nicolo available from the Alps, the Balkans
or the low mountain ranges in Central Europe increasingly
dominate the engraved gems. This goes along with the
continued inflation and the unstable political situation which
constricted the trade in gemstones from distant countries, and
with the restriction of valuable materials like marble, gold,
silver or precious stones, which became accessible only to the
court and upper levels of the administration – enforceable by
imperial decree from the 4th century ad at the latest.44 Finely
engraved gems with royal portraits such as Constantine I (r. ad
306–37) in large intense amethyst, the Armenian princess
Warazadukta (c. ad 330) in garnet, the Visigoth king Alaric II
(r. ad 484–507) in sapphire or his Ostrogothic antagonist
Theodoric (r. ad 493–526) again in amethyst, confirm this
verdict but also the continuing ability of the craftsmanship.45
Returning to the introduction: at the end of the Roman
Empire, the use of gemstones for everyday sealing had already
ended almost 200 years before Romulus Augustulus was
displaced by Odoaker in August ad 476. Wearing a finger-ring
with an engraved seal stone in former times had been a normal
custom, as Cicero commented in the above cited oration ‘Pro
Flacco’ in 63 bc: ‘Everybody uses the impression on both public
and private letters’. And in ad 69/70 during the battle of Castra
Vetera/Xanten, one in every seven of the Roman soldiers who
died against the Batavian forces, also lost his engraved gem.46
For Roman citizens sealing with gemstones was not only a
fashion, but rather a common right and custom. It became
redundant in the course of acculturation and amalgamation
Seals in Transition
with their neighbours. The accession to power of Diocletian in
ad 284 marks the start of the transformation of the ancient
legacy. The ‘emperor on the move’ and the new elites, recruited
mainly from the migratory tribes, did not for the most part
uphold the old Roman custom of sealing with engraved gems.
In this respect, the 1,000 items Jeffrey Spier has brought
together in his book on ‘Late Antique and Early Christian Gems’
appear in a new light: they are numerous compared to the
otherwise then vanishing species.
Notes
1 F. Dürrenmatt, Romulus der Große. Eine ungeschichtliche
historische Komödie in vier Akten. Neufassung 1980, Zurich, 1998,
40: ‘Wo die Hose anfängt, hört die Kultur auf’. Dürrenmatt wrote
the comedy in the winter of 1948/49.
2 P.R.L. Brown, The World of Late Antiquity: from Marcus Aurelius to
Muhammad ad 150–750, London/New York, 1989 (2nd edn); A.
Cameron et al. (eds), The Cambridge Ancient History, vols 12, 13 and
14.2, Cambridge, 1998 (with recent literature).
3 B. Ward-Perkins, The Fall of Rome and the End of Civilization,
Oxford, 2005; idem, Der Untergang des Römischen Reiches und das
Ende der Zivilisation, Stuttgart, 2007.
4 L. Török, Transfigurations of Hellenism. Aspects of Late Antique Art
in Egypt ad 250–700, Leiden/Boston, 2005.
5 J. Spier, Late Antique and Early Christian Gems, Wiesbaden, 2007.
6 E. Zwierlein-Diehl, Antike Gemmen und ihr Nachleben, Berlin,
2007, 185, 232–7; eadem, Die antiken Gemmen des Kunsthistorischen
Museums in Wien, vol. II, Munich, 1979, 11; eadem, Die antiken
Gemmen des Kunsthistorischen Museums in Wien, vol. III, Munich,
1991, 13.
7 A. Furtwängler, Die antiken Gemmen, vol. III, Leipzig/Berlin, 1900,
359 (translated: ‘The art of glyptic steps back; it loses its privileged
position as an art in fashion which it had in the previous period’.
8 A. Boos, L.-M. Dallmaier and B. Overbeck, Der römische Schatz von
Regensburg-Kumpfmühl, Regensburg, 2000, 19–47.
9 See below. Additional: ‘Il Tesoretto di Zambana-Valle di Carpeni,
Adige’, post ad 272. L. Endrizzi and F. Marzatico (eds), Ori delle Alpi
(exh. cat., Castello del Buonconsiglio, Trento), Trento, 1997, 491–3,
nos 1167–80; nos 1171–77: seven silver rings with nicolo-glass gems;
no. 1178: gold ring with nicolo-glass gem. Total: 284 coins, three
necklaces, three lunular pendants, one silver needle, one round
silver pendant with chain, one blue glass tessera. With further
parallels from the Alps region.
10 F. Dieulafait, H. Guiraud, J.-M. Pailler and D. Schaad, Le trésor
d’Eauze (Gers), Toulouse, 1987, 16, fig. 11, 18–19, figs 19–20;
D. Schaad (ed.), Le Trésor d’Eauze. Bijoux et monnaies du IIIe siècle
après J.C., Toulouse, 1992, 46, passim.
11 G. Platz-Horster, ‘Neue Gemmen aus Xanten’, Bonner Jahrbücher
201 (2001), 53–68, fig. 2; Zwierlein-Diehl 2007 (n. 6), 324, 503, fig.
980a–b; C. Bridger, ‘Die Gräber der Spätantike (275-ca.430 n.
Chr.)’, in M. Müller, H.-J. Schalles and N. Zieling (eds), Colonia
Ulpia Traiana. Xanten und sein Umland in römischer Zeit (Xantener
Berichte, Sonderband), Mainz, 2008, 586, n. 2365–6, fig. 416;
G. Platz-Horster, ‘Die antiken Gemmen aus Xanten, Teil III:
Neufunde, Neuerwerbungen, Nachträge und Auswertung’,
Xantener Berichte 15 (2009), 137, 162–7, nos 45–6, 48–50, 52.
12 Platz-Horster 2009 (n. 11), 159, no. 40.
13 Ibid., 160–1, no. 42.
14 G. Platz-Horster, Die antiken Gemmen im Rheinischen
Landesmuseum Bonn, Bonn, 1984, 11–16, 38–48, nos 11–31.
15 H. Guiraud, Intailles et Camées de l’époque romaine en Gaule I, Paris,
1988, 74.
16 H. Guiraud, Intailles et Camées de l’époque romaine en Gaule II,
Paris, 2008, 84, n. 249; orally corrected from 1 % to 3 % on the basis
of recent finds, at the British Museum Conference in London, on
29.05.2009.
17 Platz-Horster 2009 (n. 11), 130–2, Tables 1–2, 136–7.
18 H.-J. Schalles in G. Platz-Horster, Die antiken Gemmen aus Xanten
II, Cologne, 1994, 192–3, no. 307.
19 T. Otten and S. Ristow, ‘Xanten in der Spätantike. Topografie und
Geschichte’, in Müller, Schalles and Zieling (n. 11), 549–82.
20 Stuttgart, Württembergisches Landesmuseum: J. Garbsch and
P. Kos, ‘Das spätrömische Kastell Vemania bei Isny I. Zwei Schatz-
‘Gems of Heaven’ | 227
Platz-Horster
funde des frühen 4. Jahrhunderts’, Münchner Beiträge zur Vor- und
Frühgeschichte 44 (1988), 27–56, finger-rings, 35, nos 10–13, no. 17,
43, 50. L. Wamser (ed.), Die Römer zwischen Alpen und Nordmeer.
Zivilisatorisches Erbe einer europäischen Militärmacht (exh. cat.,
Rosenheim), Mainz, 2000, 384, no. 141.
21 K. Vandorpe, ‘Seals in and on the papyri of Greco-Roman and
Byzantine Egypt’, in M.-F. Boussac and A. Invernizzi (eds), Archives
et sceaux du monde hellénistique (BCH Suppl. 29), Paris, 1997, 231–
91; R. Haensch, ‘Die Verwendung von Siegeln bei Dokumenten der
kaiserzeitlichen Reichsadministration’, in ibidem, 449–96;
D. Salzmann, ‘Porträtsiegel aus dem Nomophylakeion in Kyrene’,
Bonner Jahrbücher 184 (1984), 141–66; D. Plantzos, Hellenistic
Engraved Gems, Oxford, 1999, 28–9; D. Berges, ‘Die Tonsiegel aus
dem punischen Tempelarchiv in Karthago’, Römische Mitteilungen
109 (2002), 177–223; Zwierlein-Diehl 2007 (n. 6), 8–9, 11–2, 335, s.v.
Tonsiegel (lit.), Index 563, s.v. Tonabdrücke.
22 See the rock-cut relief of Bishapur, Fars, southern Iran, with
Shapur I (ad 241–72) on horseback, worshipped by the Roman
emperors Gordian III, Philip the Arab and Valerian: W. Seipel (ed.),
7000 Jahre Persische Kunst. Meisterwerke aus dem Iranischen
Nationalmuseum in Teheran (exh. cat., Kunsthistorisches Museum
Wien), Vienna, 2000, 271, fig. 1.
23 P. Weiß, ‘Tonsiegel aus Kommagene (Doliche)’, in J. Wagner (ed.),
Gottkönige am Euphrat (Sonderheft Antike Welt), Mainz, 2000,
102–3.
24 R. Ergec, M. Önal and J. Wagner, ‘Seleukia am Euphrat/Zeugma.
Archäologische Forschungen in einer Garnisons- und
Handelsstadt am Euphrat’, in Wagner (n. 23), 105–13; M. Önal, Clay
Seal Impressions of Zeugma, Ministry of Tourism and Culture,
Gaziantep Museum, 2007; idem, Emperors, Empresses, Kings,
Philosophers and Symbols on Clay Impressions from Zeugma (Asia
Minor Studien), Munster (forthcoming).
25 L. Wamser (ed.), Die Welt von Byzanz. Europas östliches Erbe (exh.
cat., Munich), Munich, 2004, 348, fig. 762: München,
Prähistorische Staatssammlung, Inv. no. 1219.
26 Spier (n. 5), 20, pl. 137, fig. 2 (a–c).
27 A.R. Furger, M. Wartmann and E. Riha, Die römischen
Siegelkapseln aus Augusta Raurica (Forschungen in Augst, vol. 44),
Augst, 2009.
28 Cicero, Pro Flacco, 37–8, trans. by C. Macdonald, London, 1977,
482–4; RE II A 1 (1923), s.v. signum Sp. 2381 (Wenger).
29 Spier (n. 5), 189–90; R. Loscheider, ‘Römische Bleiplomben’, in
A. Demandt and J. Engemann (eds), Konstantin der Große (exh.
cat., Trier), Mainz, 2007, 368–74; R. Loscheider, Plomben als
Zeugnisse der spätantiken Verwaltung und Repräsentation, ibidem,
CDRom IV.1.17.
30 C. Johns and T. Potter, The Thetford Treasure. Roman Jewellery and
Silver, London, 1983, 20–9; finger-rings: 80–95, nos 2–23, pl. 1, col.
pls 3–4; M. Henig, ‘The Gemstones’, in ibid., 30–2: ‘apparently
drastic shift in fashion and custom’.
31 For the phenomenon of re-using gems as well as coins in elaborate
jewellery as a reflection of changes in the political and social
situation in Late Antiquity see: G. Platz-Horster, ‘The message of
re-using coins and gems in late antique jewellery’, in D. Plantzos
(ed.), Proceedings of the Conference ‘Coinage and Jewellery’, Ios/
Greece, 2009, Athens (in press).
32 B. Deppert-Lippitz, ‘A Group of Late Antique Jewelry in The Getty
Museum’, Studia Varia from the J.P. Getty Museum, vol. I, Malibu,
1993, 107–40; finger-rings (Inv. 83.AM.228.1–7), at 122–5, nos 7a–g,
figs 16–22.
33 Ibid., 111–13, figs 4a–c (Inv. 83.AM.225.2); Platz-Horster (n. 31), fig.
7.
34 Platz-Horster (n. 14), 77–83, nos 71–7 ( with lit.). Several
4th-century ad gold rings exemplify the gap between the cheap
‘nicolo glass’ gems set in elaborate and expensive gold work: R.
Laser, ‘Die römischen Fingerringe und Gemmen aus dem Gebiet
der DDR’, Arbeits- und Forschungsberichte zur sächsischen
Bodendenkmalpflege 29 (1985), 133–58; H. Ament, ‘Zur
Wertschätzung antiker Gemmen in der Merowingerzeit’,
Germania 69 (1991), 401–24, at 414, 419; S. Ristow and H. Roth,
Reallexikon der Germanischen Altertumskunde² IX (1995), 63, s.v.
Fingerring; E. Zwierlein-Diehl, Siegel und Abdruck. Antike
228 | ‘Gems of Heaven’
Gemmen in Bonn, Bonn, 2002, nos 78–9; eadem, ‘Antike Gemmen
im Mittelalter: Wiederverwendung, Umdeutung, Nachahmung’,
in D. Boschung and S. Wittekind (eds), Persistenz und Rezeption.
Weiterverwendung, Wiederverwendung und Neuinterpretation
antiker Werke im Mittelalter, Wiesbaden, 2008, 237–84, at 237, n. 2,
238, n. 4.
35 A. Demandt, Die Spätantike. Römische Geschichte von Diocletian bis
Justinian 284–565 n. Chr (Handbuch der Altertumswissenschaften
III.6), Munich, 2007, 108, n. 47; Julian 129. Ammianus XIV, 10.1.
Demandt describes Constantius’ crossing the River Rhine at
Augusta Raurica in ad 354 with a peace-offering to the Alamannic
royal brothers Gundomad and Vadomar and interprets ‘gentium
ritu’ as ‘nach ihrem Brauch’. In contrast: W. Seyfarth, Ammianus
Marcellinus, Römische Geschichte, Vol. 1, Berlin, 1968, XIV.10.16:
‘icto post haec foedere gentium ritu perfectaque sollemnitate
imperator Mediolanum ad hiberna discessit’: Kommentar 270, n.
238: ‘gentium ritu’ does not mean ‘nach dem Brauch jener Völker’,
but ‘Völkerrecht’ = ‘iure gentium’. However, M. Avenarius,
University of Köln, agrees with Demandt’s interpretation (e-mail
9 May 2009).
36 R. Pirling, ‘Ein fränkisches Fürstengrab aus Krefeld-Gellep’,
Germania 42 (1964), 188–216, no. 2, fig. 33; eadem, Römer und
Franken am Niederrhein. Burg Linn, Krefeld, Mainz, 1986, 157,
fig.133.
37 The closest parallel to the finger-ring from grave 1782 in Krefeld-
Gellep (see n. 36), a delicate gold ring set with a 2nd-century ad
nicolo gem showing the standing Bonus Eventus, was found in 1862
in a ship burial at Snape, Suffolk, grave 1: W. Filmer-Sankey and
T. Pestell, ‘Snape Anglo-Saxon Cemetery: Excavations and Surveys
1824–1992’, East Anglian Archaeology 95 (2001), 195–8, pls II–III,
fig. 78, with further literature (I am grateful to Noël Adams for the
reference).
38 M. Ricci, ‘Produzione di lusso a Roma da Giustiniano I (527–565) a
Giustiniano II (685–95): L’atelier della Crypta Balbi e i materiali
delle collezioni storiche’, in M.S. Arena et al., Roma dall’Antichità
al Medioevo nel Museo Nazionale Romano, Crypta Balbi, Rome,
2001, 331–5, fig. 129: Ricostruzione grafica dell’attività di un’ ideale
officina romana del VII secolo (disegno Inklink); 339, no. II.4.54–
64; G.B. Andreozzi, G. Graziani and L. Saguì, ‘Gems from
archaeological excavations at Rome (Crypta Balbi)’, Zeitschrift der
Deutschen Gemmologischen Gesellschaft 45/4 (1996), 175–88.
39 Henig (n. 30), 30.
40 See: Zwierlein-Diehl 2007 (n. 6), 232–3: Augustine, Epistles 59,2:
‘Hanc epistolam signatam misi annulo qui exprimit faciem
hominis attendentis in latus’.
41 See: Zwierlein-Diehl 2007 (n. 6), 7: Q. Aurelius Symmachus,
Epistulae 2,12, ed. O. Seeck, Berlin, 1876, 46.
42 For Late Antique finger-rings in metal see: L. Wamser and G.
Zahlhaas, Rom und Byzanz. Archäologische Kostbarkeiten aus
Bayern, Munich, 1998, 217–26, no. 318 (bronze box with seven
finger-rings) and nos 319–38; C. Schmidt in Wamser (n. 25), 328–
32, nos 634–84; Spier (n. 5), 183–8; see below n. 45.
43 For the accessibility of high value gemstones see: G. Platz-Horster,
‘Kleine Praser and Chromium-bearing Chalcedonies. About a
small group of engraved gems’, Pallas. Revue d’études antiques 83
(2010), 179–202.
44 Demandt (n. 35), 413, 417: marble, gold, silver, and precious stones
were government property and restricted to export. Note 89:
Codex Theodosianus X 19,1; 2; 8; 10; 13 with 2,500 imperial edicts
since ad 312. Commentary: J. Harries and I.N. Wood, The
Theodosian Code: Studies in the Imperial Law of Late Antiquity,
London, 1993.
45 See: Zwierlein-Diehl 2007 (n. 6), 183–4, 445, figs 671–3; G.
Kornbluth, ‘The seal of Alaric, rex Gothorum’, Early Medieval
Europe 16 (2008), 299–332, with references to contemporaneous
royal portraits in gemstones or metal; D. Willers and L. Raselli-
Nydegger (eds), Im Glanz der Götter und Heroen. Meisterwerke
antiker Glyptik aus der Stiftung Leo Merz (exh. cat. Bern), Mainz,
2003, 160, no. 166 (R. von Kaenel); J. Spier, Treasures of the Ferrell
Collection, Wiesbaden, 2010, 66, no. 47.
46 Platz-Horster 2009 (n. 11), 140.
Myth Revisited
The Re-use of Mythological Cameos and Intaglios in Late Antiquity and the Early
Middle Ages

Gemma Sena Chiesa

Many of the classical gems which have come down to us were
re-used on objects in Late Antiquity or in the Middle Ages, most
of them on liturgical objects. Antique engraved gems were set
in gold together with smooth gemstones, and they made
reliquaries, crosses and gospel book-covers gleam with pomp
and beauty. The most well-known example is the famous
‘Gemma Augustea’ now kept in the Kunsthistorisches Museum,
Vienna (Pl. 1). This magnificent cameo of the Tiberian age, one
of the most important works of classical glyptic, has come
down to us because it was re-used in the 11th-century reliquary
of St Sernin in Toulouse.1 The re-use of ancient gems was a
constant practice also after the Middle Ages. In
commemoration of the precious gifts ornamented with gems
donated by Queen Theodelinda (7th century ad) to Monza
Cathedral, near Milan, the 15th-century painter, Bottega degli
Zavattari, depicted jewellers employed by the Lombard queen
dismantling ancient gem-set objects in order to create
reliquaries and gifts for Christian churches (Pl. 2).2
There have been many recent studies concerning the re-use
of gems.3 But numerous questions still remain as to the reasons
for such frequent re-use of classical engraved gems in the
period from the 4th to the 12th centuries ad. Already in the
Classical period gemstones were part of precious objects; they
were of great artistic value and kept for a long time, as we know
from the discovery in tombs of gemstones which were more
than a few centuries old. In addition to the customary use (and
therefore re-use) of engraved gemstones in Late Antiquity,
there was another reason. The production of engraved gems
apparently declined at the end of the 3rd century ad, probably
because of the economic crisis experienced by the Empire due
to the first barbarian invasions (see the paper by Gertud Platz-
Horster, this volume). In this period there were also great
difficulties in finding the source materials with the result that
the traditional function of an engraved gem as a seal declined.
Due to the lesser availability of newly-worked gems, during
the 4th century ad the phenomenon of re-use gained more
importance. Unengraved gems or cameos became quite
widespread. Very often large imperial cameos with portraits
were reworked to adapt them to the images of new protagonists
at court.4 Many old engraved gems came from family treasures
and were re-used in personal jewellery. An interesting example
is the gold ring (Pl. 3) belonging to the Parma treasure buried
in the 5th century ad.5 Other gems were conserved in the
imperial treasuries, like the cameos re-used in precious objects
for members of the upper classes.
Plate 2 Detail of a 15th-century
fresco painting with the stories of
Theodelinda: goldsmiths taking off
gems from an ancient statue to
re-use them on liturgical objects.
Monza, Basilica di S. Giovanni
Battista

Plate 3 Gold finger-ring with a

nicolo depicting a warrior (1st
century AD), from the Parma
treasure (3rd century AD)

Plate 4 Vicenza. Lombardic gold

finger-ring (7th century AD) with a
Plate 1 ‘Gemma Augustea’. Vienna, Kunsthistorisches Museum, formerly in glass gem depicting two divinities
the treasury of St Sernin, Toulouse (1st century AD)

‘Gems of Heaven’ | 229

Sena Chiesa
Plate 5 Benevento. Lombardic gold Plate 6 The Majesty of Sainte Foy
fibula with a sardonyx cameo with a (7th-9th century AD). Amethyst
bust of Minerva (4th century AD) intaglio depicting Isis (2nd century
Oxford, Ashmolean Museum AD). Conques, Treasury of the Abbey
In the Lombardic period ancient engraved gems were used
as a symbol of nobility in men’s rings (Pl. 4) and in large round
women’s fibulae (Pl. 5).6 At the same time the practice of
ornamenting sacred precious objects with re-used gems of
great financial value and aesthetic importance began. In the
Early Middle Ages, in a society which underwent profound
changes, and in a new, strongly hierarchical social
organisation, the use or the donation of engraved gems
signified the cultural continuity between paganism and
Christianity and the association with an upper class which was
thought to be the heir of the Roman Empire. It should not be
forgotten that in Christian society pagan iconographies were
reinterpreted as fabulae and as a connection with the ancient
world and certainly not as a sign of belief in the ancient gods.7
The quantity of antique gems that was in circulation
between the 8th and the 12th centuries is quite astonishing.
Many engraved stones once again came from the imperial
glyptic treasury in Constantinople; those of particular value
came from there as gifts or purchases for the new rulers of the
West. It is interesting to recall that the sacred objects decorated
with the antique gems which have come down to us were
nearly all made in the West.
Gemstones with myths on liturgical objects
Giving a few examples, I will now deal with a special re-use of
classical gems depicting mythological subjects employed on
Christian liturgical objects of the Early Middle Ages. The
objects are often real and true ‘collections’ of gems (sacrae
dactyliothecae). Mythological gemstones would seem
unsuitable for the ornamentation of artefacts used in liturgical
ceremonies. Their re-use witnesses the ongoing symbolic
power and prestige of classical mythology in Christian society.
Already in the 4th century ad pagan and Christian depictions
were often used contemporaneously, apparently without any
distinction.
An interesting example are the ‘mythological’ gems placed in
spectacular reliquaries in the treasury of Conques Abbey (7th–
10th century ad). I cite, inter alia, the lovely amethyst engraved
with Isis on the right sleeve of the reliquary of Sainte Foy (Pl.
6); the carnelian with Apollo Citharoedus on the reliquary of
the Circumcision; the cameo with a faun’s head (which is part
of a larger depiction of the Dionysian thiasos) on the
pentagonal reliquary, together with many other classical
spolia.8 I will analyse in detail three intaglios re-used on two
Christian liturgical objects of different periods, which have
230 | ‘Gems of Heaven’
Plate 7 Cross of Desiderius (8th-9th century Plate 8 Sardonyx cameo:
AD). Reverse, lower part: sardonyx cameo portrait bust of Domitian as
depicting Minerva (3rd-4th century AD). Minerva (1st century AD).
Brescia, Santa Giulia. Museo della città Paris, Cabinet des médailles
been studied recently: the Cross of Desiderius in Brescia (8th–
9th century ad)9 and the Dreikönigenschrein in Cologne (12th
century),10 both of which have, in time, taken on a different
symbolic meaning.
1. Imperial ideology
About 50 gems of the Classical period are contained on the
so-called Cross of Desiderius (dated to the 9th century ad, the
beginning of the Carolingian period), in Brescia. It is probably
one of the largest gemmed crosses to have come down to us. It
is still kept in Brescia in the Monastery of Santa Giulia, founded
in the Lombardic period, to which it was donated.11 Among the
Roman gems re-used on the Cross is a three-layered sardonyx
cameo, fitted into the middle of the lower arm of the reverse of
the Cross, depicting a bust of Athena-Minerva.12 The goddess
wears an Attic helmet with lophos (Pl. 7). Two long curls of hair
escape from the side of the helmet. The goddess wears the aegis
shaped into a bodice composed of a check pattern; in the
middle is a schematic gorgoneion. The details are engraved in
sharp clean lines without any kind of shaping.
The cameo is broken but it has been placed on the Cross in a
highly visible position, meaning that it was considered to be
very important. The position of the cameo on the Cross is in
axis with that of another gem of the 1st century bc representing
Mars wearing an Attic helmet and holding a shield, to which I
will return below.13 This cameo is engraved with an
iconographic type which was quite widespread in Roman
glyptic, often indistinguishable from the figures of Rome and
Mars with Attic helmet. It employs a different iconography to
the one used for the elegant heads of Minerva with Corinthian
helmet carved in two-layered onyxes produced between the
3rd and 4th centuries ad only for jewellery (earrings and
pendants) and which we know from numerous replicas.14 Some
of these replicas also became part of the glyptic decoration of
the Cross of Desiderius, perhaps reinterpreted as angels or
saints, but most probably chosen for their beauty.15
The Brescia cameo depicts instead the solemn image of the
goddess of the late Hellenistic age documented by other
cameos of the Roman Empire.16 The same iconography was
used in the Julio-Claudian period for portraits of princesses. I
cite the well-known cameos in the British Museum17 and those
on the Cross of Desiderius.18 But it was also used for the
amazing glyptic portrait of Domitian as Minerva, now in the
Cabinet des médailles, Paris (Pl. 8),19 which was also re-used in
the Middle Ages on the reliquary of St Castor in Koblenz.20
 Myth Revisited

Plate 9 Sardonyx cameo with emerging from her peplum.

bust of Minerva (4th century AD).
Venice, Museo Archeologico
Nazionale
On the Late Antique cameo of the Cross the naturalistic
shape is replaced by the dry style and the sharp transitions
between the levels of relief which is characteristic of glyptic
works of the 3rd or 4th centuries ad.21 It is therefore the product
of a workshop of that time. Among the oldest examples of this
‘late’ series is the cameo with Minerva wearing the Corinthian
helmet, fitted in a pendant with a gold setting dated to the
second half of the 3rd century ad and found in a woman’s tomb
in Goito.22 From the same period of production as the Brescia
cameo, if not from the same workshop, there is the cameo in
Venice (Pl. 9), from the Grimani collection (end of the 15th
century),23 and probably acquired in Rome.
The considerable quantity of Late Antique cameos
portraying Minerva that have come down to us assures us that
this iconography, whose popularity derived from the positive
message it inspired, must have been reproduced in vast
numbers even if it had numerous variations. Many of the Late
Antique cameos with Minerva wearing the helmet were
re-used a few centuries later in the Middle Ages. In addition to
the examples fitted on the Cross of Desiderius and the cameo
portraying Domitian/Minerva on the reliquary of St Castor, a
two-layered onyx cameo was re-used on a 7th-century ad
Lombardic fibula from Benevento, now in the Ashmolean
Museum, Oxford,24 and the cameo set on the book cover of the
gospel of Theodelinda in Monza (this also of the 7th century
ad) (Pl. 10).25 On the Monza gem Athena does not wear the
aegis, which is, however, curiously recalled by the small snake
All of these examples appear reworked because of the sharp
undercutting of the faces made in the light layer of the stone,
and the grooved intensification of the features and the eyes,
which takes on the almond shape typical in the late Lombardic
and Carolingian periods.26 It is possible, due to the traces of
reworking, that the Grimani cameo was also re-used in the
Early Middle Ages: I would mention that in the Crypta Balbi in
Rome in the 8th century ad, there still existed goldsmiths’
workshops where ancient gemstones could have been re-used.27
The iconography of Minerva wearing the Attic helmet
should therefore have been a great success in Late Antiquity
and interpreted as the symbol of maiestas imperii. The
symbolic meaning of the classical goddess was in time
transformed into a general sign of sovereignty through the
even more frequent depictions on coins of emperors wearing
the aegis and Attic helmet. But the identification of the goddess
with wisdom and virtue always remained widespread.
Fulgentius, a Christian writer in the 6th century ad who
wrote a compilation of mythological stories, tells us that
Athena has the gorgoneion on her chest because with wisdom
she terrorises her enemies: ‘Gorgonam etiam hic addunt in
pectore quasi terroris imaginem, ut vir sapiens terrorem contra
aduersarios gestet in pectore’.28 Also the goddess wears the
helmet for the same reason: ‘cristam cum galea ponunt ut
cerebrum sapientis armatum sit et decorum’ (‘they put the
crested helmet on the head so that the brains of the wise are
armed and dignified’). As already mentioned, the armed
Athena-Minerva was in the Early Middle Ages considered a
symbol of imperial power and confused with the
representations of Rome and Mars wearing the Attic helmet.
They are all glyptic subjects with the same symbolism which is
why they were also often used between Late Antiquity and the
Early Middle Ages on liturgical objects.29
Among the representations of Mars close to glyptic
depictions of Minerva is a cameo on the Cross of Desiderius
fixed above the Minerva cameo on the lower arm of the Cross.
A similar image is also engraved on a cameo in the
Kunsthistorisches Museum in Vienna, formerly in the treasury
in the church of Sts Etienne and Pierre in Troyes.30 The Minerva
cameos and the ones with Mars could also have been
interpreted as a figuration of Alexander the Great, often
depicted with the Attic helmet as on the extraordinary
medallion on one of the large silver plates belonging to the
Mildenhall treasure (Pl. 11).31 These objects therefore exercise
a strong reminder of the veneranda antiquitas, to the imperial
world which was always present like a fil rouge from the 5th
century ad to the Middle Ages.

2. Myth with a moral meaning

Plate 10 Book-cover of
Theodolinda’s gospel: sardonyx
cameo with bust of Minerva
wearing Attic helmet and peplum
and snake before the breast (4th
century AD, reworked in the 7th
century AD). Monza, Basilica di
S. Giovanni Battista
Plate 11 Mildenhall Trasure, Alexander
plate: detail of the central medallion
(4th century AD). London, British
Museum, PE 1946,1007.7
A two-layered onyx gem is fitted on the upper arm of the
reverse side of the Cross of Desiderius in a very prominent
position (Pl. 12).32 It is curiously fitted in a vertical position
which makes the scene depicted unclear. The stocky figure of
Hercules, with a curiously styled beard, is recognisable. He
holds a club, a quiver, and three round objects, perhaps the
apples of the Hesperides, are lying on the ground. An
overturned cup is nearby. Hercules is attacking a naked
woman, perhaps Omphale, who has fallen to the ground under
the force of his onslaught. Her hair is worn loose and she wears

‘Gems of Heaven’ | 231

Sena Chiesa
Plate 12 Cross of Desiderius (8th-9th century AD). Reverse: Plate 13 Pompeii, Casa del Principe di
sardonyx cameo depicting the fight of Hercules and Omphale Montenegro. Drunk Hercules and Omphale.
(4th century AD). Brescia, Santa Giulia. Museo della città Naples, Museo Archeologico Nazionale
bracelets on her wrists, and perhaps on her ankles, as well as a
necklace. Beneath her one can see the paws of a lion-skin. A
young man (maybe a personification of Virtus) bursts on the
scene, wearing a mantle on his shoulder and holding an object
in his right hand (probably a torch). His left is about to put a
wreath on the head of Hercules, perhaps to celebrate the hero’s
final victory over the queen who had kept him as a slave. The
apples of the Hesperides, which were sacred to Aphrodite, and
the overturned cup highlight the erotic and convivial aspect of
the scene.33 It is therefore a curious depiction, probably with a
moralistic implication, of one of the most popular myths in the
ancient world and in modern art. It was often reproduced in
figurative art, as on the well-known 18th-century painting by
François LeMoine now kept in the Louvre.
In the Classical period the myth of Hercules, who became a
slave to queen Omphale who, in order to humiliate him, wore
the attributes of the hero and forced him to wear women’s
clothing, is well known in both literary and figurative works.
This is due not only to the amatory significance of the story, but
also because of the hero’s thrill in being part of Dionysos’
world. Only in the reign of Augustus was the figure of the
Lydian queen considered negative as her triumph over the hero
was equated with the seduction of Anthony-Hercules by
Cleopatra-Omphale.34
The symbolic message would have been ambiguous, in part
negative because it represents the enslavement of the hero to
wine and love and in part positive as an example of female
seduction stripping the man of brutality and bringing him to
peaceful deeds.35 The story can also be interpreted as an
example of a joyful amatory life without any restraints. It also
evokes the adventurous life of a world ‘upside down’.36
The Omphale myth was very popular in ancient Rome. A
sumptuous fresco from the Vesuvian area shows Hercules
drunk and lying down: he has lost his club, while Omphale is
looking down at him from above surrounded by her servants
(Pl. 13).37 A round fresco from Pompeii (Pl. 14) depicts
Omphale wearing the lion-skin embracing Hercules wearing
the myrtle crown.38 They are perhaps portraits of the master
and mistress of the house appearing as mythological figures.
On gems Omphale is depicted naked, clothed only with the
lion-skin and with the attributes of Hercules. Moreover she is
depicted with the same elegant iconography, certainly inspired
232 | ‘Gems of Heaven’
Plate 14 Pompeii, Casa di Ercole. Hercules
and Omphale. Naples, Museo Archeologico
Nazionale
by a famous original (about 30 replicas are known, perhaps
intended for female rings). The beautiful figure of the queen
wrapped in the lion-skin seems to refer to the prevalence of the
female gender and to her victory in the world of Eros.39 This is
what Propertius says when praising the triumphant beauty of
Omphale.40 But the poet concludes with the image of Augustus
who in the name of Rome is triumphant over Cleopatra-
Omphale: ‘quid nunc juvat … si mulier patienda fuit? Cape
Roma, triumphum et longum Augusto salva precare diem’.41
The fight scene between Hercules and Omphale which
appears on the gem in Brescia is a unicum as far as I know. Nor
can it be linked to images of Omphale on 3rd–4th- century ad
magical gems.42 The curious Brescia iconography may perhaps
derive from a representations of Hercules with one of his lovers
in the late Hellenistic period. A superb example is represented
by an amethyst in the Museo Archeologico Nazionale in
Florence attributed to Teukros by Marie-Louise Vollenweider
(Pl. 15). Hercules has the lion-skin, the club and a quiver
nearby. In front of him there is a young nude girl (Jole?) with a
cloak. In this case the depiction represents more of an amorous
encounter than a violent seduction.43 The scene is repeated on
an interesting cameo fragment signed by Dioskourides but now
lost (Pl. 16). On the gem is an erotic scene with two naked
figures, one of which is a girl with a lion-skin.44 The engraving
is of very high quality and confirms that this iconography was
well known in the Augustan age.
Plate 15 Amethyst intaglio signed Plate 16 Lost cameo signed by
by Teukros. Hercules with a Dioskourides. Hercules and Omphale (?)
nymph (1st century BC). Florence,
Museo Archeologico Nazionale

Plate 17 Dionysiac sarcophagus: Hercules, retained by satyrs, assaults a
nymph. Naples, Museo Archeologico Nazionale
The theme of Hercules seducing a nymph, appearing
between the 2nd and 3rd century ad,45 is often present on
sarcophagi with the procession of Dionysos and satyrs (Pl.
17).46 The often drunk Hercules is supported by two satyrs
whilst he tries to seduce a nude nymph standing in front of
him.
But the more interesting comparison, because Hercules is
depicted wearing female clothes (therefore referring directly to
the story with Omphale), is that with the Townley puteal in the
British Museum, on which there is also a boy with a torch in his
hands.47
In Late Antiquity, when with Achilles he became the hero
par excellence, Hercules often appears in the Dionysian cycle, in
particular on sumptuous silverware like the large plate from
the Mildenhall treasure in the British Museum (Pl. 18).48 In the
festive celebration of Dionysian thiasos, the theme of Hercules
drunk supported by two satyrs is taken up. The composition of
the group is extraordinarily similar to the depiction on the
Brescia intaglio, which was certainly executed in Late
Antiquity, either at the end of the 4th or perhaps during the 5th
century ad. Very characteristic are the sharp contours and the
lack of details, which isolate the scene in a neutral space. The
somewhat ‘abstract’ composition is also significant. These are
elements which are also found in part on a group of Late
Antique mythological cameos already highlighted by Martin
Henig and Jeffrey Spier.49 In these pieces, however, the relief is
softer and more crudely cut. The most interesting aspect is that
the scene is highly stylised and completely dislocated. There is
no ground line, the figures are in low relief and the attributes
of Hercules are dispersed into emptiness. It is evident in the
Plate 18 Silver plate from the Mildenhall Treasure. Drunken Hercules
supported by two satyrs. London, British Museum, PE 1946,1007.1
Myth Revisited
Plate 19 Ivory relief with Apollo
and Daphne (late 5th century AD).
Ravenna, Museo Archeologico
Nazionale
apparent indifference towards spatial arrangement that there
is too much emphasis on the poses of the figures, which does
not allow an an easy interpretation of the subject. Examples of
such free spatial arrangement are not rare in figurative
production in Late Antique glyptic as can be seen, for example,
on the lovely onyx cameo in the Content Collection (4th
century ad) with a priestess sacrificing a bull,50 where the
arrangement of the young girl and the animal placed vertically
is forced and unnatural. This sort of untidy composition also
characterises the ivory relief depicting Apollo and Daphne
(dated to the end of the 5th century ad) (Pl. 19),51 in which a
bird is unnaturally placed between the figures. The same
unusual style is also found on the hunting scene in the circus
represented on the consular diptych leaf of Areobindus
(beginning of the 6th century ad).52
The Brescia cameo has all the elements to create a symbolic
and ‘cultured’ interpretation of the subject but it is evident that
the artist chose a reversed interpretation of the myth. Hercules
wins over Omphale and frees himself of his servitude. The
erotic subject acquires nobility through the interesting, finely
carved figure of the young boy (perhaps Virtus or Amor). The
young man offers Hercules the laurel wreath, symbol of moral
victory over weakness and drunkenness. It recalls the imperial
depiction of the genius who reaches over the victorious
emperor crowning him or giving him the globe, as in two
cameos of the 4th and 5th century ad celebrating respectively
the glory of Licinius and the crowning of Valentinian III
(Pl. 20).53
The theme is however interpreted with a new perception,
the figures are placed in the space in an abstract way, the
Plate 20 Sardonyx cameo: crowning of Valentinian III (?) (5th century AD).
St Petersburg, State Hermitage Museum
‘Gems of Heaven’ | 233
Sena Chiesa
Plate 21 Cross of Desiderius. Sardonyx cameo: the Muses in Parnassos (4th
century AD). Brescia, Sta Giulia, Museo della città
upside-down categories no longer exist, the gestures appear
allusive and unconcerned with the narrative: everything is
symbolic.54 The Brescia cameo was certainly created between
the 4th and 5th century ad for a person belonging to an already
Christian cultured society but who nevertheless liked luxury
objects ornamented with pagan myths. We know that the
perennial use of pagan themes survived in both imperial
circles and amongst the political elite, even if Christian. A few
centuries later, at the beginning of the Carolingian Empire, the
cameo was taken from a treasure by somebody from the
aristocracy to be used in decorating the precious Cross, a royal
gift to the church of San Salvatore in Brescia. I do not think,
however, that they are any traces of reworking as on the
Minerva cameo.
The ancient engraved stone, even if the depiction was
explicitly amorous and mythological, was fitted upside down in
the middle of the upper arm of the obverse of the Cross, the one
which has the most antique cameos. It is evident that the
depiction was not of much interest. One can however suppose a
new interpretation of the image in a Christian sense. The
victory of virtue over vice seen, as depicted often in the Early
Middle Ages, in the female sense. The figure which is given the
most visibility is the noble youth who is crowning the hero, in
this way emphasising the moral interpretatio of the
representation.
The Hercules and Omphale cameo is connected on the
Cross to two other mythological carvings which share a moral
interpretation: the cameo of the Muses (Pl. 21) and the cameo
with the myth of Pegasus and the nymphs.55 Both refer to the
necessity of the soul to reach heaven through art and beauty.
Even the fight between Hercules and Omphale may have had
the same allegorical interpretation of Christian salvation for
the purchaser of the Cross.
3. Leda and the swan
In the Dreikönigenschrein in Cologne, that marvellous
medieval dactyliotheca which we know so well today from the
studies of Erika Zwierlein-Diehl, an agate gemstone is fitted
that she dates to the 3rd century ad, but is perhaps older. The
artefact is placed on the side ‘of David’ of the precious casket
just above the representation of the biblical king (Pl. 22).56 The
engraving is very refined but quite simple and realistic. It
depicts the overtly erotic scene of Leda lying down with a swan
(Pl. 23). It is one of the most popular myths in the ancient
234 | ‘Gems of Heaven’
Plate 22 Dreikönigenschrein (11th century): side with David. Cologne, Dom
world related to the loves of Jupiter in animal form.57
Many different versions of this story have been told.58 Here I
will mention only some of them. Leda, wife of Tyndareus the
Spartan king and daughter of a powerful nymph, is seduced by
Jupiter who transformed himself into a swan according to at
least two versions: one of the deceived young woman who
picks up the wounded bird and the certainly more erotic one of
the amorous embrace between the queen and Jupiter
transformed into a swan. Leda lays two eggs from which two
heroes (the Dioskuroi) and two princesses, Helen and
Clytemnestra, emerge. But another ancient version (it has
always been an intricate myth) describes Leda as the custodian
of eggs born from the divine union of Zeus-swan and Nemesis
transformed into a goose.
Leda, like Omphale, represents feminine mythical history
in the ancient world and, like her, is the expression of amorous
seduction. Euripides, making Helen tell the story of Leda, her
mother, described the queen as a woman deceived by the
transformation of the god.59 On the contrary Ovid, as we will
see, places Leda in the category of women that participate in
the indecent amorous games of the gods and hence is one of the
‘bad girls’ of the heroic Greek world.
In addition to the many paintings on the subject,60 the myth
of Leda was represented in sculpture in three principal types
that became very popular in copies of the Roman period. In
particular the one by Timotheus, of the middle of the 4th
century bc, represents a frightened Leda as she receives the
swan which swoops down from the sky to lie on her lap (Pl.
24).61 The Hellenistic version is more explicitly amorous with
Leda standing up embracing a large swan:62 of this there are
many copies. The most important of these was the copy
exhibited in the 16th century in Venice (Pl. 25),63 which had
inspired many artists in the Renaissance.64 Both of these scenes
are depicted on gemstones and were very popular due not only
Plate 23a-b. Intaglio (and cast) depicting Leda and the swan (3rd century AD).
Dreikönigenschrein, Cologne, Dom

Plate 24 Leda receives the swan Plate 25 Leda with the swan (Roman
(Roman copy of a 4th century BC copy of a Hellenistic sculpture). Venice,
sculpture). Rome, Musei Capitolini Museo Archeologico Nazionale
to their erotic significance but also because they were derived
from statues of great prestige.
The third version, in which Leda is shown lying down
receiving the swan which wraps its wings around her, has
become well known from a gem in particular. This version may Pan (Pl. 28),71 a desecration of the myth which is already
derive from an original picture known also to Ovid, who
appears to recall it in his description of the myth. He narrates
65
that Arachne, an Athenian girl envious of Athena’s ability in
weaving, had embroidered the scene of Leda in a tapestry
between the crimina Jovis (the amorous sins of Jupiter) and his
simulationes (‘deceits’):66 ‘fecit et olorinis Ledam recumbare
sub alis’ (‘depicted Leda lying down under the wings of the
swan’). This depiction appears repeatedly on gems throughout
the period of the Roman Empire, perhaps due to its closeness to that rings were still fitted with gems depicting the licentious
Ovid’s text, as well as for the amorous subject and its very
attractive appearance. Moreover, due to its compact
67
elongated composition, it adapts well to the diminutive size
and oblong shape of a gem.
I cite as one example the beautiful onyx in the Museo
Archeologico Nazionale in Naples (Pl. 26),68 a masterpiece in
miniature relief, in which Leda’s hair is curiously styled in
ringlets (‘Melonenfrisur’) recalling the Alexandrinian period:
the decorative invention of the curve of the swan’s neck can be
clearly seen. The cameo described as ‘cignus concubans cum
Leda’ belonged to Cardinal Barbo who fitted it into a silver-gilt
table in the centre of four other cameos.69 This is therefore a
15th-century re-use, which testifies to the humanistic
Plate 26 Onyx cameo: Leda reclining with the swan (1st century BC). Naples,
Museo Archeologico Nazionale, formerly Barbo and then Medici Collection
Myth Revisited
Plate 27 Agate cameo: Leda reclining Plate 28 Carnelian intaglio: Leda and
with the swan (Roman Imperial swan on a chariot drawn by Pan (end
period). Geneva, Merz Collection of the 1st century AD). The Hague,
Royal Coin Cabinet
atmosphere of the papal court of that period. The gem was then
acquired by Lorenzo ‘il Magnifico’. As we will see later this gem
became famous in the Renaissance period and was the source
of inspiration for many works of art.
A simplified copy is the two-layered agate cameo in the Leo
Merz collection (Pl. 27).70 The rendition of the feathers on the
wings and the style of Leda’s hair would date this work to
around the end of the 2nd or 3rd century ad. A curious
‘grotesque’ gem, which I believe to be a unicum, studied by
Marianne Maaskant-Kleibrink and now kept in The Hague,
depicts Leda and the swan on a cart being pulled by the god
evident in the Imperial period and documented by the literary
sources.
The gem on the Dreikönigenschrein is a fine example of
re-use in the Middle Ages of a composition linked to erotic
Greek mythology which was disapproved of by the Fathers of
the Church for the scandalous behaviour of the gods,
demonstrating the inconsistency of the Christian religion. In
the 3rd century ad Clement of Alexandria found it immoral
lovers as Jupiter and Leda.72 However, until at least the 5th
century ad, Coptic art uses the myth of Leda in its most sensual
form, reopening the problem of the interpretation of pagan
mythology in a Christian context. On a Coptic stone relief in
the Ashmolean Museum in Oxford dated to the beginning of
the 5th century ad (Pl. 29),73 the depiction of the naked women
in the centre, caressing a large bird which follows her, is
thought to be Leda. In addition in the Middle Ages the
archdeacon of Soissons had in his seal a gemstone depicting
Leda lying down with the swan.74
Plate 29 Coptic relief: Leda and the swan (6th century AD). Oxford, Ashmolean
Museum
‘Gems of Heaven’ | 235
Sena Chiesa
Plate 30 Giorgione (attributed to), Leda and the swan. Plate 31 Filarete, bronze doors of Old St Peter’s
Padua, Museo Civico Basilica (15th century). Detail: Leda and the swan.
Città del Vaticano
There have been many allegorical interpretations of the
myth of Leda which attempt to interpret this licentious scene in
a symbolic way. Athenagoras, a Christian apologist of the 2nd
century ad, interpreted the coupling of Leda with Jupiter as a
symbolic depiction of the conjunction of the air with the
earth.75 In the Gnostic gospels (for instance the Book of Baruch
by Justinus) the myth was interpreted as the symbol of the
male figure (Elohim/Jupiter) who is united to the female one
(Edem/Leda) to create angels in an allegoric cosmogony of the
myth.76 In the apocryphal gospels (the protoevangelium of
James considered a Greek oriental manuscript of the end of the
2nd century ad and therefore the inspiration for many of the
depictions of Leda in Coptic art) a sparrow flies from a tree to
the sleeping Anna (as the swan flies to Leda), and Anna like
this conceives Mary.77
For Martianus Capella (De Harmonia) at the beginning of
the 5th century ad, Leda is a prudens puella (the ‘bad girl’ of
Ovid is transformed), deceived by Jupiter who conquers her
with the sweet song of the swan, a symbol of music which
refines the world.78 In the Fabula de cigno et Leda of Fulgentius
(6th century ad) Jupiter is a symbol of power, whilst Leda,
again a negative example, is the source of the misfortunes
which corrupt the imperial power.79In the figurative ambit of
the Middle Ages, the iconography of Leda with the swan was
also interpreted as eternity which has as its attribute the
phoenix, the magical immortal bird.
It was only in the late Renaissance period that the fable of
Jupiter and Leda reacquired its original meaning, due in part to
the great popularity that it had in Renaissance culture from
Ovid’s description and also for the great fame of the Leda gem
in Cardinal Barbo’s collection which subsequently passed onto
the Medici collections. In the Renaissance the depiction of Leda
and the swan is often reproduced from antique monuments as
well as on works of contemporaneous artists. Likewise the
myth of Leda is depicted by Giorgione in a beautiful painting
which sets the myth in a wide landscape (Pl. 30).80 Filarete uses
the figures of Leda and the swan to decorate the bronze doors
of St Peter’s in Rome, although in this case Leda appears
chastely clothed (Pl. 31).81 In the cultural climate of the
beginning of the Renaissance period the iconography of Leda is
now an intentional return to the artistic skill of the classical
world.82 It is through the Renaissance re-appropriation that the
236 | ‘Gems of Heaven’
Plate 32 Henry Matisse, Leda and the swan
(1944–1946). Private collection
iconography of Leda lying down entwined with the swan
comes down to us. Among the many modern representations of
the ancient myth, I would mention the colourful interpretation
dating to 1944–6 by Matisse (Pl. 32).83
Some conclusions
The use in liturgical objects of gems with motifs linked to myth
and to the pagan pantheon, of which I have drawn attention to
just a few examples, is not easy for us to understand. It seems
difficult to believe that the imperial patrons who made the
glyptic treasures available to sumptuously adorn the gifts that
they offered to the church were not interested in making a
selection of the chosen objects. Perhaps the subjects engraved
on the gems were considered less important than the colour of
the gemstones or their market value. On this issue of the
complex interpretation of the relationship between the
craftsman, buyer, and receiver of the engraved gems, different
interpretations have been recently advanced. I will not go into
the magical meaning assumed by some kinds of gemstones in
the Middle Ages, a problem which has already been studied in
some detail by Erika Zwierlein-Diehl.84 Instead, I mention two
other interpretations.
Firstly, that the engraved gemstones of the Roman period
acquired a higher value simply because they evoked the
grandeur of the Caesars of which the Church and the new
powerful aristocracy considered themselves their heirs. This is
certain for some ‘symbolic’ themes, like the bust wearing the
helmet (Rome or Minerva) which we have seen was frequently
re-used on objects in the Early Middle Ages.
A second hypothesis, already proposed by Zwierlein-
Diehl,85 is that those gems which depicted ancient myths were
the subject matter of an interpretatio Christiana. This
interpretation, according to Christian allegorical
interpretations of pagan depictions, perhaps originated in the
‘moral’ interpretation of mythological themes which were
already present in pagan philosophical reflections of the 3rd
and 4th centuries ad. For example, the heroes of the numerous
and not always salubrious adventures of the Classical period,
like Achilles and Hercules, become symbols of heroic ‘virtue’
and of the man free from ties. I believe, however, that the
insertion of gems with mythological pagan figures onto
reliquaries must be interpreted more simply as a desire to offer

to Christ the celestial emperor and to his church the same
homage of precious gems which decorated the imperial
crowns. A particular emphasis, equally symbolic, could have
come from the use of gemstones engraved by artists in the
Classical era. They were made even more precious by their
rarity and by their matchless engraving, which combined
technical virtuosity and great value in a multi-coloured effect.
Already Ambrose of Milan in the De Fide86 talked about
pagan mythology as poeticae fabulae which could, in this way,
be continuously appreciated. Between the 4th and 5th century
ad many Christian emperors ordered the eradication of the cult
of images of pagan divinities, but allowed them to be enjoyed
as works of art to be safeguarded and admired. It was ordered
that the statues of the gods (which were appreciated for their
artistic value rather than for their sacred nature) were to be
taken from their temples and arranged to decorate the city. A
similar ‘artistic tolerance’ was also true for those pagan
depictions on luxury objects which could have (or should have)
therefore been re-used to glorify the Church and to celebrate
the sovereignty of Christ.
Notes
1 E. Zwierlein-Diehl, Magie der Steine. Die antike Prunkkameen im
Kunsthistorischen Museum, Vienna, 2008, 119ff.
2 Bottega degli Zavattari, Monza, Basilica di San Giovanni Battista.
See B. Morigia, Chronicon, ad annum 1359: ‘Theodelinda idolum
aureum, margaritis et lapidibus praetiosis, mirabili modo
ornatum, in vasi set aliis ornamentis multis pulcherrimis et magni
sauri purissimi fabricari et se esisdem margaritiis et lapidibus
praetisosi mirabiliter ornari fecit (‘She transformed the golden
idol, wonderfully decorated with pearls and precious stones, into
many other objects, very beautiful and large, all in the purest gold,
and made them adorned with the same pearls and precious
stones’). On the value of classical carved gems in the Middle Ages,
see E. Zwierlein-Diehl, Die Gemmen und Kameen des Dreikönigen-
schreines. Der Dreikönigenschrein im Kölner Dom, Vol. I.1
(Denkmäler Deutscher Kunst. Die großen Reliquienschreine des
Mittelalters, Studien zum Kölner Dom, 5), Cologne, 1998, 62–70.
3 M. Henig, A Corpus of Roman Engraved Gemstones from British
Sites, Oxford, 1978 (2nd edn); J. Spier, Ancient Gems and Finger
Rings. Catalogue of the Collection. The J. Paul Getty Museum,
Malibu, 1992; M. Henig et al., Classical Gems. Ancient and Modern
Intaglios and Cameos in the Fitzwilliam Museum, Cambridge, 1994;
Zwierlein-Diehl (n. 2); eadem (n. 1); see also Gesztelyi in this
volume.
4 I mention only the famous cameo depicting Constantine’s family,
inserted in the 12th century into the upper binding of the Ada
Gospels in the Schatzkammer der Stadtbibliotek, Trier.
5 M. Calvani Marini, ‘Tesoro di Parma’, in G. Sena Chiesa and E.A.
Arslan (eds), Milano capitale dell’Impero romano 286–402 d.c. (exh.
cat., Milan), Milan, 1990, 351–4.
6 On the finger-rings see, G. Sena Chiesa (ed.), Gemme. Dalla corte
imperiale alla corte celeste, Milan, 2002, 21 (M. Dolci); E. Gagetti,
‘Anello longobardo in oro con pasta vitrea figurata di età romana
da Dueville’, in B. Zanettin and L. Dolcini (eds), Cristalli e gemme.
Realtà fisica e immaginario, simbologia, tecniche e arte, Venice,
2003, 422–8 (with literature). For the fibula from Benevento in the
Ashmolean Museum see: D. Limonta, ‘Abbigliamento e incontro di
culture: fibule con spolia glittici’, in Sena Chiesa ibid., 27–40, at 33,
pl. 3.
7 Poeticae fabulae: Ambrosius, De fide ad Gratianum Augustum,
3.2.3, ed. J.P. Migne, Patrologia Latina, Paris, 1844–55, 1862–5, vol.
16. In Fulgentius’ Mithologiarum libri tres, ed. R. Helm, Stuttgart,
1970, the stories of gods and heroes are called fabulae. On the
ideological significance of mythological motifs in Late Antiquity
see also, R.E. Leader-Newby, Silver and Society in Late Antiquity:
Functions and Meanings of Silver Plate in the Fourth to the Seventh
Centuries, Aldershot and Burlington, 2004, esp. 141ff.
8 D. Gaborit-Chopin and E. Taburet-Delahaye (eds), Le trésor de
Conques (exh. cat., Paris), Paris, 2001, 22, 33, 42.
Myth Revisited
9 Sena Chiesa (n. 6).
10 Zwierlein-Diehl (n. 2).
11 G. Sena Chiesa, ‘Il prestigio dell’antico e il riuso glittico fra IV e X
secolo’ and ‘Sacra dactyliotheca: la croce di Desiderio a Brescia ed il
suo ornato glittico’, in eadem (n. 6), 1–16 and 154–63.
12 G. Sena Chiesa, ‘Nova gloria vetustatis. Intailles et camées dans la
Croix de Didier de Brescia’, in M. Avisseau-Broustet (ed.), La
glyptique des mondes classiques. Mélanges en hommage à Marie-
Louise Vollenweider, Paris, 1997, 97–117, at 110, pl. 21; Sena Chiesa
(n. 6), gem V/C 167, 208–9, no. 49 (M. Geroli).
13 Sena Chiesa (n. 6), gem V/C 177, 213–4, no. 55 (M. Cadario).
14 M. Gramatopol, Les pierres gravées du Cabinet Numismatique de
l’Académie Roumaine (Collection Latomus, 138), Brussels, 1974;
I. Popović, ‘Roman cameos with female busts from Middle and
Lower Danube’, Pallas 83 (2010), 203–24. On this kind of cameo on
the Cross of Desiderius see, Sena Chiesa (n. 6), gem V/D 140, 205–
6, no. 42 (M. Geroli) and gem V/B 199, 217, no. 59 (M. Geroli).
15 On the spread of the type see, H. Guiraud, Intailles et camées de
l’époque romaine en Gaule (Territoire français) (48e supplément à
Gallia), Paris, 1988, nos 977–8.
16 See the example in the Museo Archeologico Nazionale in Florence:
A. Giuliano, I cammei della collezione Medicea del Museo
Archeologico di Firenze, with M.E. Micheli, Storia delle collezioni e
regesto, Rome-Milan, 1989, no. 9. The type seems to remake the
5th-century bc Greek iconography of the helmeted goddess, as,
e.g., on the coin of Thuroi: Lexicon Iconographicum Mythologiae
Classicae, Athena/Minerva, no. 305 (hereafter LIMC). Also very
similar is a cameo in the Cabinet des médailles, Paris: M.-L.
Vollenweider and M. Avisseau-Broustet, Camées et intailles, Tome
II: Les portraits romains du Cabinet des médailles, 2 vols, Paris,
2003, no. 131.
17 H.B. Walters, Catalogue of the Engraved Gems and Cameos, Greek,
Etruscan and Roman in the British Museum, London, 1926;
W.-R. Megow, Kameen von Augustus bis Alexander Severus (Antike
Münzen und geschnittene Steine, 11), 2 vols, Berlin, 1987, 260, no.
B27, pl. 29.3.
18 Sena Chiesa (n. 6), gem V/D 187, 214, no. 56 (C. Lambrugo).
19 Vollenweider and Avisseau-Broustet (n. 16), no. 132.
20 Ibid., 117.
21 On the style of Late Antique cameos: M. Henig, The Content Family
Collection of Ancient Cameos, Ashmolean Museum, Oxford, Oxford–
Houlton, 1990; J. Spier, Late Antique and Early Christian Gems,
Wiesbaden, 2007.
22 Sena Chiesa and Arslan (n. 5), 282–3, no. 4c.2g.1 (G. Sena Chiesa);
Limonta (n. 6), 27.
23 B. Nardelli, I cammei del Museo Archeologico Nazionale di Venezia
(Collezioni e Musei Archeologici del Veneto, 43), Rome, 1999, 43,
no. 16.
24 Sena Chiesa (n. 6), 33.
25 Sena Chiesa (n. 6).
26 E. Gagetti, ‘“Ex Romano vitro splendentes lapilli”. Ricezione di
iconografie della glittica ellenistico-romana in cammei vitrei
altomedievali’, in S. Fortunelli (ed.), Sertun Perusinum Gemmae
oblatum. Docenti e allievi del Dottorato di Perugia in onore di
Gemma Sena Chiesa, (Quaderni di Ostraka, 13), Naples, 2007,
161–96, at 181, n. 67.
27 M.S. Arena et al. (eds), Roma dall’antichità al medioevo nel Museo
Nazionale Romano Crypta Balbi, Milan, 2001, 333 and 139. See also
Gagetti (n. 26), 183–5.
28 Fulgentius, Mithologiarum libri tres, II, De Minerva.
29 Sena Chiesa (n. 6), gem V/C 179, 213–4, no. 55 (M. Cadario).
30 Vienna, Kunsthistorisches Museum, Antikensammlung, IX a 43;
Zwierlein-Diehl (n. 1), 46, pl. 20.
31 K.S. Painter, The Mildenhall Treasure, London. 1977.
32 G. Sena Chiesa, ‘La “Croce di Desiderio” a Brescia ed il problema
del riuso glittico in età tardoantica ed altomedioevale’, in
Splendida civitas nostra. Studi in onore di Antonio Frova (Studi e
ricerche sulla Gallia Cisalpina, 8), Rome, 1995, 429–41, at 433–6;
eadem (n. 6), gem V/A, 118, no. 36 (M. Cadario).
33 On the erotic-Dionysiac meaning of the scene, see Sena Chiesa (n.
6), gem V/A 118, 202, no. 36 (M. Cadario).
34 Plutarch, Antonius, 90.4, ed. Rev. W.W. Skeat, London, 1875.
35 The representation of Omphale as a Roman matron who died in the
Severan age can have such a meaning: P. Zanker, ‘Eine römische
matrone aus Omphale’, Mitteilungen des Deutschen
Archäologischen Instituts. Römische Abteilung 106 (1999), 119–31.
‘Gems of Heaven’ | 237
Sena Chiesa
36 V. Dasen, ‘Le secret d’Omphale’, Revue Archéologique 46 (2008/2),
265–81.
37 Pompeii, Casa del Principe di Montenegro. Another fresco from
Pompeii (Naples, Museo Archeologico Nazionale, inv. no. 8992)
depicts Hercules, standing with the distaff, and Omphale, wearing
the lion-skin. On the subjects: A. Coralini, ‘Ercole e Onfale nella
pittura pompeiana. Problemi di iconografia’, Ocnus 8 (2000),
69–92.
38 Pompeii, Casa di Eracle (Naples, Museo Archeologico Nazionale,
inv. no. 9004): LIMC, VII, Omphale, no. 14.
39 M.-L. Vollenweider, Die Steinschneidekunst und ihre Künstler in
spätrepublikanischer und augusteischer Zeit, Baden-Baden, 1966,
pl. 32.3–5; LIMC (n. 38), nos 71–6 (J. Boardman); S. Toso, ‘Le cattive
ragazze: le Amazzoni, Onfale, Medea’, in I. Colpo, I. Favaretto and
F. Ghedini (eds), Iconografia 2001. Studi sull’immagine (Antenor
Quaderni, 1), Rome, 2002, 289–307, at 290.
40 Propertius, Elegiae, 3.11.17–20: Omphale in tantum formae processit
honorem, ed. L. Mueller, Sex. Propertii Elegiae, Leipzig, 1898.
41 Ibid., 3.11.49–50 (‘woe betide if we should lie under a woman. But
celebrate, Rome, your triumph now that you are safe and invoke
long life for Augustus’).
42 Dasen (n. 36).
43 Vollenweider (n. 39), pl. 37.4–5. Interesting is the gem with Apollo
and Marsia (C. Gasparri [ed.], Le Gemme Farnese, Naples, 1994, no.
89), in a Hellenistic taste for fables.
44 Vollenweider (n. 39), pl. 68.7; A. Furtwängler, Die Antiken
Gemmen, 3 vols, Leipzig-Berlin, 1900, III, pl. LVII.8.
45 See, for instance, a bronze mirror case with Hercules in New York
(Metropolitan Museum, inv. no. oo6.1228c): LIMC, IV, Heracles,
no. 1554.
46 E.g. on the Dionysian sarcophagus in Villa Albani (LIMC, V,
Heracles, no, 3262), on the sarcophagus at Bolsena (LIMC, V,
Heracles, no. 3264), on the small but interesting bronze vase in
Boston (LIMC, V, Heracles, no. 3267) and on a sarcophagus in the
Museo Archeologico Nazionale in Naples (P. Zanker and B.C.
Ewald, Mit Mythen leben. Die Bildenwelt der römischen
Sarkophagen, Munich, 2004, pl. 125).
47 LIMC, VII, Omphale, no. 33 (London, British Museum, inv. no. M
2541).
48 Painter (n. 31).
49 Henig (n. 21), nos 91, 100, 119, 123–5, 136; Spier (n. 21), 140.
50 Henig (n. 21), no. 149; Spier (n. 21), no. 767.
51 C. Rizzardi, L. Martini, C. Muscolino and E. Cristoferi (eds), Avori
bizantini e medioevali nel Museo Nazionale di Ravenna, Ravenna,
1990, no. 1, pl. I.
52 Paris, Musée National du Moyen Age, inv. no. Cl. 13135.
53 Paris, Bibliothèque Nationale, Cabinet des Médailles, ‘The
Triumph of Licinius’ (Spier [n. 21], no. 718); St. Petersburg, State
Hermitage, ‘Crowning of an Emperor’ (Spier [n. 21], no. 572).
54 On the meaning of ‘above’ and ‘under’ in Roman imperial art:
G.L. Grassigli, ‘Il sotto, il sopra. Per una semantica della
composizione nell’arte imperiale’, in I. Colpo, I. Favaretto and F.
Ghedini (eds), Iconografia 2005. Immagini e immaginari
dall’antichità classica al mondo moderno (Antenor Quaderni, 5),
Rome, 2005, 133–44.
55 Sena Chiesa (n. 32); Sena Chiesa (n. 12); M. Cadario, ‘La toilette di
Pegaso nella Croce di Desiderio a Brescia’, ACME: Annali della
Facoltà di Lettere e Filosofia dell’Università degli Studi di Milano
52/2 (1999), 201–18.
56 E. Zwierlein-Diehl, ‘Interpretatio Christiana: Gems on the Shrine
of the Three Kings in Cologne’, in C.M. Brown (ed.), Engraved
Gems: Survivals and Revivals, Washington DC, 1997, 63–84, at 74;
eadem (n. 2), 344, no. 250; A. Dierichs, ‘Leda Schwangruppen in
der Glyptik und ihre monumentalen Vorbilder’, Boreas 13 (1990),
37–50, pl. 6.
57 Propertius (n. 40), 1.13.30: ‘gods change into animals not to upset
men, because of their glittering look’.
58 Apollodorus, Bibliotheca, 1.7.10; 3.10.5–7, trans. J.G. Frazer,
Cambridge, MA, and London, 1921; Hyginus, Fabulae, 77, trans.
238 | ‘Gems of Heaven’
and ed. M. Grant, University of Kansas Press, 1960; Ovid, Heroides,
8.77, trans. G. Showerman, Cambridge, MA, and London, 1931, and
Metamorphoses, 6.10.9, trans. A.D. Melville, Oxford, 1986. On the
ancient sources of the myth and on its different versions: M.C.
Monaco, ‘L’eredità dell’antico’, in R. Nanni and M.C. Monaco, Leda,
storia di un mito dalle origini a Leonardo, Florence, 2007, 21–54.
59 Apollodorus (n. 58), 3.10.7.
60 Such as Euripides, Helen, 16–22 (Zeus, as a swan flying away from
an eagle to Leda and possessing her through deception), trans.
E.P. Coleridge, London, 1891..
61 I recall the beautiful copy in the Musei Capitolini, inv. no 302, and
the specimen in the Museo di Antichità at Parma (LIMC, VI, Leda,
nos 73–4: J. Boardman).
62 LIMC, VI, Leda, 240, no. 16; Monaco (n. 58), 40.
63 Venice, Museo Archeologico, inv. no. 30: C. Anti, Il regio Museo
archeologico nel palazzo reale di Venezia, Rome, 1930, 30, no. VIII.7;
LIMC, VI, Leda, no. 96; Monaco (n. 58), 38–9.
64 R. Nanni, ‘La formazione della recezione medioevale del mito’, in
Nanni and Monaco (n. 58), 55–92.
65 Her. (n. 58), 8.77; Met. (n. 58), 6.10.9
66 Recalled too by Statius, Silvae, 2.6.45, ed. and trans. D.R.
Shackleton Bailey, Cambridge, MA, and London, 2003.
67 E.g. Furtwängler (n. 44), pls XXVIII.17, XXLII.74; J. Boardman,
Engraved Gems. The Jonides Collection, London, 1968, no. 74; LIMC,
VI, Leda, no. 288. On Leda depicted on gems see: Dierichs (n. 56).
68 Naples, Museo Archeologico Nazionale, inv. no. 25967/134:
N. Dacos, A. Giuliano and U. Pannuti (eds), Tesoro di Lorenzo il
Magnifico, I, Le gemme, Firenze, 1973, no. 48; U. Pannuti, Museo
Archeologico Nazionale di Napoli. La collezione glittica, II, Rome,
1994, 36, no. 74.
69 Dacos et al. (n. 68); Inventario del Cardinal Pietro Barbo (1457), 87
ff.
70 M.-L. Vollenweider, Deliciae Leonis. Antike geschnittene Steine und
Ringe aus Privatsammlung, Mainz am Rhein, 1984, no. 457;
D. Willers and L. Raselli-Nydegger (eds), Im Glanz der Götter und
Heroen, Meisterwerke Antiker Glyptik aus der Stiftung Leo Merz,
Mainz am Rhein, 2003, no. 59.
71 M. Maaskant-Kleibrink, Catalogue of the Engraved Gems in the
Royal Coin Cabinet, The Hague, The Hague, 1978, no. 351.
72 Clement of Alexandria, Protrepticus, 60.1–2, ed. M. Marcovich,
Leiden, 1995, and Paedagogus, 3.59–60, ed. M. Marcovich, Leiden,
2002; Spier (n. 21), 15. In general: R. Nanni, ‘Sulle tracce di modelli
di età classica’, in Nanni and Monaco (n. 58), 60 ff.
73 Oxford, Ashmolean Museum, inv. no. 1970-403; LIMC, VI, Leda, no.
52 (L. Kahil et al.); Monaco (n. 58), 50–1.
74 Zwierlein-Diehl (n. 2), 99–100, pl. 58.
75 Athenagoras, A Plea for the Christians, 22, trans. B.P. Pratten (Ante-
Nicene Fathers, Vol. II), Edinburgh, 1867.
76 M. Simonetti, ‘Note sul “Libro di Baruch” dello gnostico Giustino’,
Vetera Christianorum 6 (1969), 71–89. On the medieval
interpretation of the myth of Leda: Nanni (n. 64), 61–2.
77 Nanni (n. 64), 63–4.
78 De Harmonia, in De nuptiis Philologiae et Mercurii, chapter IX, ed.
A. Dick, Leipzig, 1925.
79 Fulgentius, (n. 7), 2.13 (fabula de cygno et Leda).
80 School of Giorgione, Leda e il cigno (Padua, Museo Civico): Nanni
(n. 72), 119, pl. 53.
81 Antonio Averlino known as Filarete, Leda and the Swan, Città del
Vaticano, bronze doors of Old St. Peter’s Basilica. On the
allegorical meaning given by Filarete to the figure of Leda (the
worldly pleasures?): Nanni (n. 64), 89.
82 On the influence of classical Leda upon Renaissance artists: Nanni
(n. 64).
83 Henry Matisse, Leda and the Swan, 1944–1946 (part of a triptych):
private collection.
84 Zwierlein-Diehl (n. 2).
85 Zwierlein-Diehl (n. 56).
86 Ambrosius (n. 7), 3.2.3.
Inscriptions on Portrait Gems and Discs in
Late Antiquity (3rd–6th centuries ad)
Between Epigraphical Tradition and Numismatic Particularism

Sébastien Aubry
In this paper I will first present five different epigraphical historisches Museum, Vienna (Pl. 6),13 and is still used on the
configurations on gems and silver discs that depict only single Byzantine-influenced Lombard coinage in the 8th century ad.14
portraits, and also describe all the legends found on them.1 These legends can be contracted or complete, but also
I will start by examining the most legible inscriptions, like sometimes include a qualifying adjective expressing affection,
terms or formulas. Secondly I will touch on some related as on a silver disc and a plasma gem, both of the 4th century ad
problems such as monograms, Christian symbols and (Pl. 7);15 this indicates that they were probably used as gifts.
representations of couples and families, which will allow me to
demonstrate a sixth epigraphical type. This two-part
introduction aims at addressing the crux of the matter, that is
to say, engraved backgrounds with abbreviations, contractions,
uncertain directions of reading and a seventh epigraphical
configuration. The basis of my article comprises some 50
gemstones and discs dating from the 3rd to the 6th centuries
ad, and I will discuss these in relation with various –
sometimes earlier – archaeological parallels which shed light
upon these inscriptions.2

1. Epigraphical configurations (Pl. 1)

Plate 2 Grey mottled stone Plate 3 Brown carnelian intaglio with
(serpentine?) intaglio with draped male bust. Surrounding legend (Type
1.1. Type 1 bust of a young, bearded man. 1) ΠΟΛΕΜΩΝΟC; 12.4 x 9.8mm; 4th
A serpentine-like gemstone once on the London market (Pl. 2),3 Surrounding legend in two parts century AD
a brown carnelian in the Hermitage (Pl. 3),4 a silver disc in the (Type 1) CΙCΙΝ-ΝΟΥ; 14.4 x 12.5mm;
4th century AD
Numismatic Museum, Athens,5 a one-sided seal,6 all dating
from the 4th century ad, and a 5th-century ad phalera,7 serve
to illustrate the first type in which the inscription is written
clockwise on the upper part of the gem (or disc), sometimes
interrupted in the middle by the figure, with the tops of the
letters turned to the outside. These inscriptions may either be
in the nominative or genitive case as they indicate the name of
the owner or the represented person (or both). This
configuration, which is known in the Republic,8 the Imperial
period9 and in the Late Empire, probably derives directly from
coin legends (Pl. 4) as for example on a 5th-century ad garnet
depicting the Emperor Theodosius II (Pl. 5).10 The legend Plate 4 Solidus of the Emperor Plate 5a Garnet intaglio with a
surrounding the bust is probably a metaphor for the crown or Theodosius II in consular garb. facing bust of the Emperor
Obverse: surrounding legend in two Theodosius II. Surrounding legend
the halo as one can see on a 4th-century ad medallion of parts (Type 1): DN THEODO - SIVS PF (Type 1): DN THEODOSIVS AVG; 26 x
Constantius II,11 or a 6th-century ad imperial seal of Justinian AVG; D. 20.5mm; mid-5th century 22mm; mid-5th century AD
AD. WT 4.48g; AD 416–18
I.12 This hagiographic representation of Type 1 can also be seen
on a 5th-century ad sapphire of Alaric II in the Kunst-

Plate 5b Cast of Plate 5a Plate 6 Sapphire of Alaric II (cast).

Surrounding legend (Type 1): ALARICVS
Plate 1 Epigraphical configurations (Types 1 to 7) REX GOTHORVM; 20.6 x 16.7mm;
5th century AD
‘Gems of Heaven’ | 239
Aubry

probably a token of affection like ‘In my mind’.22

Plate 7 Silver disc (cast) with bust of Plate 8 Silver disc with a draped
a woman wearing an elaborate male bust (detail). Surrounding
headdress. Surrounding legend in legend (Type 1): SPESINIANE VIVAS;
two parts (Type 1): CORN D-VLCIS; 16.5 x 15mm; 4th century AD
11.9 x 10.5mm; 4th century AD
To correlate with this kind of addition, one often finds portraits
on discs of the 4th century ad with the word vivas ('May you
live'), either with a name in the vocative case or alone (Pl. 8).16
Shifting this theme to the divine, one finds on a 4th-century ad
disc a Type 1 prayer with vivas in deo (Pl. 9),17 and without this
last word, two 3rd-century ad chalcedony stamp seals with an
indication of Christian allegiance.18 Finally, the Greek counter-
part of vivas, namely ζήσαις, is found mainly on text-cameos,
sometimes in combination with εὐτυχι;19 as an example one can
point to a 3rd-century ad female portrait on a carnelian in Paris
which carries a similar (but incorrectly transcribed)
inscription.20
1.2. Type 2
The next epigraphical configuration is very similar to Type 1: in
this instance, the inscription appears all around the border,
with the letters still turned to the outside. With this new type
one finds nearly all the different kinds of legend noted above,
but because of the space left free for the surrounding
inscription, the latter is sometimes more developed, and has a
more complex structure. Three examples will illustrate the
point.
First is the famous 4th-century ad garnet engraved with the
portrait of an Armenian princess, now in Vienna (Pl. 10).21 Her
name, in the nominative, is preceded by her title and followed
by another name in the genitive case and a word presumably in
the dative. The one in the genitive indicates that the stone was
probably offered by the princess to a girl called Hermione: it is
an apposition implying possession. The last term ΨΥΧΗ is
distanced from the previous legend due to the inversion of the
direction of reading: it involves quite a particular thought,
On the second example, a portrait of the end of the 3rd
century ad 23 from Cadès’ Impronte Gemmarie,24 one finds again
a wish with the word vivas. Here the inscription, as on the
Vienna garnet, is inscribed in two lines but respecting the
direction of reading: the legend makes up a whole. It concerns
a marriage pledge in which the names of the husband25 and
wife26 are mentioned. In fact, as I will discuss below, the
interruption of a clockwise reading of the inscription
(especially for Type 2) tends to indicate a modification in the
meaning of the whole inscription.
Finally, a 3rd-century ad agate in Naples (Pl. 11)27 is very
interesting because the legend has never been correctly
transcribed: on it is the name of the represented person with a
wish including another appellation, ΕΥΤΥΧΕΙ ΑΝΠΕΛΙ. The
Greek-koiné phonetic particularism of the imperative28
suggests the probable origin of the gem to be Aphrodisias (in
Caria) where the two names are attested.29 As we are dealing
here with a wish in the imperative singular, this last one can
only apply to one of the subjects; the dots serve to confirm the
appellative status of the name’s specific part of the inscription,
and also allows separation from the wish formula.
Consequently the function of the gemstone is the same as the
one of the Armenian princess: a gift, with the name of the lady
in the nominative case and, in apposition, a wish towards the
man to whom she offered her portrait.
This specific grammatical structure is also attested on two
brown carnelians: in one instance in Greek, but where the
sequencing of the legend and the link between it and the figure
is exactly the same;30 in the second instance in Latin on a
2nd-century ad text-intaglio with vivas.31
1.3. Type 3
For the third epigraphical configuration, I will concentrate on
inscriptions written only in the lower part of the gem, and, for
the first time, with the letters facing inwards. For single
portraits, the best examples are cameos, but the legends only
function as complements to the figure or as wishes with vivas.
Other instances, dealt with below, have couples and family
types. For now one may cite a 3rd-century ad cameo from the
Content Collection, on which a slightly incurved inscription
defines the portrait of a lady and stresses her beauty (Pl. 12).32

Plate 11 Agate intaglio with bust of Plate 12 Onyx cameo (three

Plate 9 Amethyst intaglio (cast) with
draped bust of a young man.
Surrounding legend in two parts
(Type 1): VIVAS-S-IN DEO; 13.5 x
13mm; 4th century AD
Plate 10 Garnet with Armenian
princess. Surrounding legend in three
parts (Type 2): BACIΛICCA OYA-
PAZAΔOYKTA EPMIONHC ΨYX-H;
18.5 x 14.6mm; 4th century AD
Aurelia Papiana. Surrounding legend
with palm-branch as stop-mark
(Type 2): ΑΥΡ•ΠΑΠΙΑΝΗ•ΕΥΤΥΧΕΙ
ΑΝΠΕΛΙ (palm-branch); 26.7 x
20.3mm; 3rd century AD
layers) with bust of a woman.
Legend in exergue (Type 3): ΤΗ
ΚΑΛΗ; 24 x 15mm; 3rd century AD

240 | ‘Gems of Heaven’

 Inscriptions on Portrait Gems and Discs in Late Antiquity (3rd–6th centuries AD)
Plate 13 Silver disc with confronted
male and female busts. Surrounding
legend in two parts (Type 2):
LVCIANE AVGVST-A V; 11.5 x
11.7mm; 4th century AD
Plate 15 Brown chalcedony with a
male bust. Legend on each side of the
portrait (Type 5): ΘΕΟΤΕ/ΚΝΟΥ;
16.8 x 14.4mm; 4th century AD
In Geneva and in Paris, one finds similar portrait-cameos
with incurved inscriptions. The first, a portrait dating from the
Tetrarchy (ad 300) is inscribed [aΠ]a [g]regori vivas.33 It
cannot be transcribed as a marriage pledge as on a 4th-century
ad gold ring34 or the Cadès’ gem mentioned above;35 on the
contrary, the word γρηγορι, written in Latin,36 has to be
understood as an imperative,37 and the legend translated as
‘Ap(p)a,38 may you live and be wakeful!’.39
The second is a 5th-century ad cameo inscribed her vir v40
whose legend has been wrongly transcribed as Herrenia
virgo Vestalis according to a similar medallion of the 2nd
century ad.41 The latter is inscribed belliciae modeste v/v, the
last part of which – designated as Type 7 – I will deal with
below (Pl. 34). This was translated as the nonsensical plural
virgini Vestali – (there is only one Vestale represented) – instead
of virgo Vestalis. The problem is knowing whether modeste is a
name in the vocative case, or an adverb. Assuming the second
to be true, the whole legend should be translated as:
‘(Medallion of) Bellicia, modestly, virgin of the Vestales’, and in
this case, with reference to several other examples of portraits
dating from the 1st century bc,42 the portion ‘v/v’ (Type 7),
restored as virgo Vestalis, would be a qualificative or the name
of the person represented as duo nomina initials. But if we
consider the possibility of an appellation in the vocative case, it
might be correct to suppose that v/v really is a plural,43 but that
of the subjunctive vivas, that is vivatis; it would apply to both
Bellicia and Modestus. Whichever, the use of either vocative or
genitive terms, the Type 7 portion v/v,44 as well as the
masculine Modestus (incompatible with virgo), does not
validate this hypothesis, despite the fact that one finds a
similar grammatical configuration on a 3rd-century ad nicolo
from Berlin.45 So it is wiser to consider this inscription as a
legend that qualifies the medallion as the property of the
person depicted.
To return to the Paris cameo, if the legend is herennia with
virgo vestalis, how can it be explained that the first two
words are contracted in three letters and not the third, even
though it is abbreviated to v/v on the Vestale medallion? It
would have been more logical to inscribe either her vir ves, in
Plate 14 Sardonyx cameo (three layers) representing Septimius Severus,
holding spear and patera, between Caracalla and Geta, both crowned by a
Victory; in the foreground an altar. Legend in exergue (Type 4): [ΥΡΕΡ ΤΗΝ]
ΝΕΙΚΗΝ ΤΩΝ ΚΥΡΙΩΝ CΕΒ(ΑCΤΩΝ); 31 x 32mm; early 3rd century AD
which each word is as important as the others, or, in the case of
a lack of space in the exergue, her v v. As the Cadès’ cast46 has
shown that a single portrait could carry a marriage pledge
referring to both husband and wife, it seems to me that it could
be the same here, unless it is a joint pledge of duo nomina type,
or, as on the Geneva cameo,47 a legend with vir as an
acclamation or an ablative (for example herennia virtute
vivas, ‘Herennia, may you live with courage!’). But the V is
undoubtedly the beginning of vivas or vivatis as on a 4th-
century ad silver disc said to have come from Hungary (Pl. 13).48
1.4. Type 4
The inscriptions, portraits of couples or families have their own
keys to reading; therefore I will return to this in the next
section. However, there is a small number of cameos on which
the legends may be related to those of Type 3: the only
difference between them is the use of a ground line to separate
the field from the exergue. To exemplify Type 4, one can point
to two remarkable cameos: firstly, a 3rd-century ad cameo in
the Cabinet des médailles, Paris, bearing a family portrait of
Septimius Severus with a dedication to ‘the victory of the
Emperors’ (Pl. 14).49 Secondly, a 5th-century ad sardonyx
intaglio in the Hermitage depicts the investiture of Valentinian
III, with a Christogram, two Greek letters on each side of it (A
and W), and the signature of the engraver under the ground
line (see the paper by Sena Chiesa, this volume, Pl. 20).50
1.5. Type 5
Another marginal epigraphical configuration on portraits
creates a close link between the figure and the inscription and
is exemplified by a chalcedony of the 4th century ad (Pl. 15).51
In this case, the legend does not surround the figure as Types 1
and 2 do, but is separated into two parts, one with the letters
turned to the outside, the other with the letters to the inside
(like a new line); it is as if the inscription had the bust in a
stranglehold. The inscribed name is that of the figure, and in
the correct genitive case, the name of the owner as well. This
epigraphical configuration can also been found on a portrait
dating from the Imperial period.52
‘Gems of Heaven’ | 241
Aubry

Plate 16 Amethyst (cast) with bust
of Gothic king and Latin monogram;
21.5 x 16.6mm; 5th century AD
Plate 17 Sard intaglio with male
bust. Monograms on each side of the
portrait (Type 7): POB/AT; 15 x 11.5
mm; mid-1st century BC (40–30 BC)
Plate 20 Silver disc with male bust at
left, facing female bust wearing
necklace and earrings. Surrounding
legend with cross as pivot (Type 1):
ΘΕΟΠΟ (cross) ΝΤΟΥ; 11.8 x 11.5mm;
4th century AD
Plate 21 Garnet intaglio (cast) with
female bust, draped, wearing a
necklace. Legend on each side of the
figure with a cross as stop-mark
(Type 2): ΚΑΡΙ/ΑΝΗ (cross); 11.8 x
8.8mm; 5th century AD

2. Related questions an upper reference point or punctuation mark (Pl. 21).63

The second part of this paper touches on various related
questions. I shall introduce some more complex figure types,
namely portraits of couples and families.
2.1. Monograms
Close to Type 3, a Gothic monarch’s facing portrait on the
famous amethyst from the Merz Collection53 is identified by a
Latin monogram, whose form is typical of the late 5th century
ad (Pl. 16).54 Despite the fact that it has not yet been clearly
deciphered, various interpretations have been offered, one of
which is to read it as theodericvs for Theodoric the Great.55
However, the use of monograms to characterise a bust is
nothing new on Late Roman gems (Pl. 17).56
2.2. Christian symbols
Closely connected with monograms, we may address the use of
Christian symbols such as the cross,57 Christogram (Pl. 18),58
and staurogram59 on portraits. Used to indicate that the bearer
belongs to the Christian religion, they are also of a certain
importance in the configuration of the inscription. On a silver
disc of the 4th century ad featuring a couple,60 the meaning of
the Type 2 legend in the vocative case is subordinated to a
Christogram on each side of which there is an i and a n. It is a
wish: crescentine apule in ch(risto), implying in christo
vivatis. As on a 4th-century ad mosaic61 and other 4th- and
5th-century ad discs and gemstones, Christian symbols play a
part in the symmetrical arrangement and balance of the
epigraphical composition, either as a pivot (Pls 19–20),62 or as
Remember the garnet of Emperor Theodosius II (Pl. 5)64 or the
agate from Naples,65 for which, however, the symbol was a
palm-branch in the lower register. The survival of such usage
in the 9th century ad is demonstrated by a rock crystal intaglio
seal stamp of Lothar II of Lotharingia.66
2.3. Couples and families (Type 6)
These portrait types are interesting for several reasons. Firstly,
as on intaglios and rings with a single bust, one may find the
name of the owner or of one of the persons represented. It is
easier to differentiate these possibilities by analysing the whole
composition. On a 4th-century ad silver disc with two
confronted busts,67 is a name in the genitive case which
indicates the owner, but which could also refer to one of the
two portraits. Sometimes the choice is much easier when there
are busts of two recognisable persons, but a legend which does
not fit. One may take as an example a 4th-century ad jasper
with the draped busts of Peter and Paul with a Type 3 Egyptian
theophoric name inscribed below.68 The same kind of
inscription on a 2nd-century ad jasper,69 featuring a couple and
referring also to the owner, shows that such an heterogenous
association was not specific to the Late Empire.
Secondly, as gems and discs were often used as marriage
gifts, the legend can also confirm that tradition, as on a bronze
ring of the 4th century ad.70 Assuming that gems with couples
were also used as love tokens, one may find on them, as for
single portraits, the word vivas as a blessing formula invoking
the protection of God. But here the important point is that,
apart from vivas/vivatis which we have already met, we can
find on 4th-century ad couple portraits several types of
abbreviated forms or declensions for vivas.71

Plate 18 Silver disc (cast) with
confronted male and female busts;
above and between them, a chi-rho;
14 x 14mm; 4th century AD
Plate 19 Quartz crystal intaglio with Plate 22 Silver disk (cast) with busts
confronted male and female busts. of a woman, wearing a necklace,
Surrounding legend with chi-rho as facing a male; between them the bust
pivot (Type 2): ΜΑΤΡΩΝ (chi-rho) of a small boy. Surrounding legend in
ΑΠΡΟ[ΒΙ]ΑΝΟΥ; 14 x 11.3mm; 4th four parts (Type 2): DO-M VICT
century AD CAEN; 11 x 11mm; 4th century AD

242 | ‘Gems of Heaven’

 Inscriptions on Portrait Gems and Discs in Late Antiquity (3rd–6th centuries AD)
Plate 23 Banded agate intaglio with busts of a man, woman and child. Legend
(Type 6): ΕΥΤΥΧΙ ΠΑΝΧΑΡΙ ΜΕΤΑ ΤΗC ΚΥΡΙΑC ΒΑCΙΛΙCCΗC ΚΑΙ
ΠΑΥΛΙΝΑC / ΙC ΘΕ-ΟC | (Type 1) ΖΟΗ; 31 x 24.5mm; 3rd century AD
Finally, one of the most interesting aspects of Late Antique
gems is the family portrait on which the legend complementing
the image is divided between Type 2 and a sixth epigraphical
configuration (Type 6) which is composed of a combination of
Types 1 and 3. The main difference between Types 2 and 6 is
the inversion of the direction of reading between the upper and
lower register. Type 2, inscribed on the border, surrounds the
image with letters turned to the outside in a single direction
(Pl. 22).72 Type 6 – a horizontal variant of Type 573 – remains
parallel to the border but still in the same plane as the ground
line,74 the upper part of the legend with letters to the outside,
the lower with letters to the inside.75 On a 4th-century ad silver
disc, each person is named by an abbreviated word76 with a
Type 2 configuration of the inscription; but on a famous
banded agate of the 3rd century ad (Pl. 23),77 the upper part of
the legend is a wish,78 and the lower another wish and an
acclamation perhaps denoting a Christian origin.79 The
inversion of the direction of reading underlines the end of the
first wish and introduces two other elements, each in its own
line. Like epigraphical Type 5, Type 6 also closely links the
inscription and the design: the legend taken as a whole is a
wish for those depicted on the gem. On another Type 6 gem, a
4th-century ad carnelian from the eastern Mediterranean,80
4he upper part of the inscription identifies the three engraved
busts with their so-called ‘diacritic names’,81 and the lower part,
with a break in the direction of reading as a new line, mentions
the signum82 common to the three persons.
Plate 24 Carnelian intaglio (drawing) Plate 25 Glass paste with three
with draped bust of Septimius busts, two with laurel wreaths. By
Severus, laureate and crowned with the figures the letters: AVB (reversed)
the modius, facing right between N; 18.6 x 16.3mm; 4th century AD.
Caracalla and Geta (?); below the Paris, Cabinet des médailles
figures, the letters (Type 3): ΕΖ
(reversed) C; present location
unknown; 3rd century AD
3. Abbreviations and contractions
3.1. Initials and reversed letters
As seen above, the limitations of space on the engraved surface
area sometimes impose the need for abbreviations;83 therefore I
will now deal with these and with uncertain directions of
reading. The first matter addressed will be an attempt to
explain the meaning and specific arrangement of some initials
on gems with group portraits, using the cases already met with
as well as various archaeological parallels.
Firstly, on a 3rd-century ad carnelian intaglio (Pl. 24),
whose present location is unknown,84 is engraved a bust of
Septimius Severus between his two sons. Below, the inscription
– ΕΖ (reversed) C – has been interpreted as the acclamation Εις
Ζεύς Σάραπις because of the fact that on the one hand Severus
is depicted with the modius of Sarapis, and on the other hand
this acclamation is common on representations of the god and
on amulets.85 But, if the inscription is not difficult to restore,
the interesting point is the particular reversal of the Z; one
could detect in that back-to-front letter an engraver’s error, a
regional particularism or a passing fashion as in the 2nd
century bc.86 But in fact this mistake was intentionally made to
underline the Greek nature of the abbreviated legend. As an
indication, one could mention the famous 4th-century ad Wint
Hill Bowl on the inner border of which an inscription runs
around the design in easily legible serifed capitals.87 If the first
three words are in Latin,88 the next four letters are in fact the
Greek formula ΠΙΕ Ζ[ΗCΗC], πιε ζήσαις,89 written in Roman
characters. In this case, too, the reversed Z underlines the
Greek nature of the last part of the Latin wish. The palm-
branch at the end represents a punctuation mark, as on the
Naples agate (Pl. 9).90
On another gem, a glass paste from Würzburg91 dating from
the end of the 4th century ad (Pl. 25), are represented three
busts, two of them with laurel wreaths: probably an emperor,
his co-regent, the crown prince as Caesar, or noblemen dressed
as the imperial family of the time. Indeed the letters above, N B
V A, don’t belong to a recognisable group.92 And, on the
impression as on the gem, one of the letters is back to front, B
and N respectively. There are three defects: this inversion, the
terminal strokes of the V, which are not parallel to the bottom,
and the break in reading direction between A and the other
letters. These anomalies could give us the impression that the
intaglio had been manufactured by an engraver who was
ignorant of the rules of Latin epigraphy. However, the letter
forms are typical of the 4th century ad,93 and the reversed N is
not exceptional: the legend could be legible in the positive
(impression) as well as the negative (stamp), as with the stamp
on the pan of a Roman bronze steelyard.94
These epigraphical considerations prevent us from
interpreting the inscription either as a formula addressed to
the group – for example N(o)b(iles) V(ires) A(ugusti)95 – or as an
appellation in tria nomina with two cognomina. It is more likely
the specific form of the tria nomina with a separate agnomen, 96
a usage attested during the 4th century ad;97 the break in the
direction of reading serves to reinforce this possibility by
implying the formula ‘also known as’ between the V and A.98
Otherwise the letters N B V could possibly be the initials of the
busts represented, with the addition of the A. Because of its
position, which is not on the same scale as the other letters, it
‘Gems of Heaven’ | 243
Aubry
Plate 26 Silver disc with a female Plate 27 Brown carnelian intaglio
bust, wearing a hair band, facing a with bust of a young man facing
male bust; between them, the bust of Nemesis. Legend behind the two
a small child (detail); above, three figures, on each side of the gem
letters (Type 1): ΥΦΕ; 11 x 11mm; 4th (Type 1): Ν / ΦΛΠ; 14 x 10.7mm; 2nd
century AD century AD
could be acting as a common qualifying adjective or signum,
like the Hermitage carnelian,99 which also has a break in the
direction of reading (Type 6).100
One last example, which demonstrates the use of
abbreviations to complement the figure: on a 4th-century ad
silver disc with a couple and their child are written three letters
ΥΦΕ, which were thought to be part of a peripheral legend,
apparently supported by fragments of letters behind the nape
of the woman’s neck (Pl. 26).101 However it is more probably a
defect in the metal itself. In any case, the suggested likely
restoration as ‘Euphemios’102 may be countered by one objection
and two alternative hypotheses. First of all, Euphemios and all
its inflections103 are spelt with an eta and not with an epsilon
after the phi. So the letters are likely to be part either of
another nomen, or the abbreviated tria nomina of the owner;
the latter possibility is not so plausible because we know that
nearly all of the denominations on gems and discs in Late
Antiquity are in the form of signa or so-called ‘diacritic names’.
Following the example of the glass paste in Würzburg,104 the
most logical interpretation is to regard each letter as the initial
of the name of each figure on the gem. But on some earlier
items – for example on a 2nd-century ad carnelian from Vienna
(Pl. 27)105 – the overall structure of the field indicates clearly
which interpretation to apply to the abbreviated Type 1 legend:
the N, not set vertically due to the desire for left-right
symmetry, applies to the figure of Nemesis, and the letters ΦΛΠ
to the tria nomina initials of the owner.106
3.2. Breaks in the direction of reading
Having dealt with the various types of abbreviations on group-
figure gems, I will now briefly examine the question of breaks
in the direction of reading on single portrait intaglios which, as
we have seen, implies a connection between the inscription
and the figure. It is exemplified by a 4th-century ad nicolo from
Plate 28 Nicolo intaglio with male Plate 29 Nicolo intaglio (cast) with
bust. Legend in three parts (Type 3): bust of a Roman lady. In the field, the
P-E-R-E; 14.6 x 11mm; 4th century AD letter P; 12.5 x 10mm; 2nd century AD
244 | ‘Gems of Heaven’
the Hermitage (Pl. 28),107 on which the Type 3 contracted
legend is displayed in the lower part of the bust, like on some
other examples during the Republic and the Empire,108 and
names the owner. But, the first P, set vertically, breaks the
reading direction and indicates that the bust too is that of the
owner.
Likewise, a single letter inscribed in the same way can be
directly linked to the figure it complements, as on a 4th-
century ad rock crystal featuring Paulus,109 or as on an earlier
example, a 2nd-century ad portrait from Paris (Pl. 29),110
whose legend presents Matidia111 as the personification of
Pietas, as on a silver denarius celebrating her as Augusta.112
Besides, the use of an initial to express a virtue that the design
implies is already known on gems.113
3.3. Type 7
Sometimes, as we have seen before with the 'Vestale'
medallion,114 the initials of the legend can be set parallel to the
ground line115 on both sides of the figure, as if gripping it. This
constitutes the seventh epigraphical configuration on gems
and discs in Late Antiquity (Type 7). Indeed, on a 4th-century
ad glass paste from Würzburg (Pl. 30),116 the two letters H and
F define the bust as that of Horatius Flaccus, of whom another
portrait also exists on a cameo in Geneva with a Type 1
nominative inscription.117 Other earlier examples of such a
configuration can be mentioned, in particular a 3rd-century ad
nicolo in Vienna: the inscription on it marks it as the portrait of
Fulvia Plautilla, rendered as f(ulvia/ae) c(ai filia), daughter of
Caius Fulvius Plautianus and wife of Caracalla (Pl. 31).118 So, if
two Type 7 letters can constitute a formula like the senatus
consulto S C on coins with a portrait bust,119 they could also be
either the initials of the owner – but the Type 7 configuration
implies that the portrait, too, is his – or the first two letters of a
word, as with the mint on Roman provincial coinage.120 Indeed,
on a 5th-century ad jasper featuring a male bust a letter is
inscribed on each side of the figure, X and P.121 It is either the
name of the owner – as duo nomina initials or signum
abbreviation122 – or a metaphor implying Christian allegiance.
There would be no sense, however, in translating the X/P as a
Christogram, for two reasons. On the one hand, there is no link
with the figure, contrary to a 3rd-century ad carnelian123 or a
late 3rd-century ad bronze ring,124 both with I and X,
respectively for Ἰχθῦς and Ἰησοῦς χριστός, in a similar
epigraphic configuration (Types 6 and 7). Furthermore it
would have been much more appropriate to engrave a cross or
a staurogram to imply a Christian symbol.125 It is, however,
necessary to concede that the Type 7 initials are sometimes
intended to replace, even to complete, certain meaningful
Plate 30 Glass paste of Horatius Plate 31 Nicolo intaglio of Plautilla.
Flaccus. Legend on each side (Type 7): Legend on each side (Type 7): F / C;
H / F; 18 x 15.2mm; 4th century AD 13.5 x 9.5mm; 3rd century AD
 Inscriptions on Portrait Gems and Discs in Late Antiquity (3rd–6th centuries AD)
Plate 32 Sardonyx intaglio with a bust Plate 33 Solidus of the Emperor Honorius standing to right, holding a
of Poppaea. Legend on each side and standard and Victory on a globe, and trampling on captive. Reverse:
below the portrait (Type 7): O / P / surrounding legend in two parts (Type 1): VICTORI-A-AVGGG | (Type
(cornucopia); 7 x 14.5mm; AD 60–5 7) M / D | (Type 4) COMOB; D. 19mm; WT 4.46g; AD 395–400
symbols.126 A 1st-century ad sardonyx in the Cabinet des
médailles, Paris, representing a crowned bust of Poppaea, the
wife of Nero (Pl. 32),127 shows the use of initials and symbols of
the same size which are put together in order to specify the
status of the engraved figure. The two letters O P apply to Ollia
Poppeia,128 and the miniature cornucopia to the symbol of
Fecunditas, as on some imperial silver coins.129 Indeed,
according to Tacitus (Annales, XV, 23), after Poppaea had given
a daughter to Nero in ad 63, the latter had a temple built to
Fecunditas, dedicated to the fertility of the empress.
4. Conclusion
During the 4th and 5th centuries ad, while legends retain the
traditions inherited from previous periods of referring to the
owner or the design, they also innovate in the quantity of
unique appellations, that is signa and so-called ‘diacritic
names’, by comparison with compound nouns. They sometimes
accompany devices, wishes or Christian professions of faith,
and the majority of their epigraphical configurations (Types 1
to 4 and 7) derive specifically from coin legends. Regarding
Type 7, in the case of portraits, the inscriptions are only used to
complete the image or to qualify its function, and can on no
account have an exogenous value regarding the figurative area;
therefore they apply to the name of the gemstone’s owner only
if the latter is himself depicted on it. With regard to Types 5 and
6, even though they are quite rare, they possess the same
function as Type 7, which is to envelop the image on the gem
and to use it as a pivot or as a complementary insertion. The
connection between the two elements is based on the
hypothesis that the text defines or clarifies the image, and vice
versa.130 It is doubtless the stylistic interpenetration between
coins and gemstones that is behind the survival of this reading
system until the end of the Late Empire; we have evidence of it
in the marginal combination of various types (1, 4 and 7) which
are found both in numismatics and in glyptics from the end of
the Republic (Pls 33–34),131 and which testify to the mainly
descriptive function of inscriptions on the stones and discs of
Late Antiquity.
Notes
1 Such as a name, wish, belief, symbol.
2 Or underlining the tradition, spread or continuity of each
epigraphical configuration.
3 J. Spier, Late Antique and Early Christian Gems, Wiesbaden, 2007,
no. 20 (CΙCΙΝ-ΝΟΥ).
4 Ibid., no. 18 (ΠΟΛΕΜΩΝΟC [3rd declension]).
5 Ibid., no. 58 (ΓΑΥ-ΔΕΝΤ).
Plate 34 Sard intaglio with bust of a beared
man. Surrounding legend (Type 1): EPI-CRATES
| (Type 7) Q / Q; 14 x 12mm; mid-1st century BC
6 R. Cormack and A. Eastmond (eds), The Road to Byzantium.
Luxury Arts of Antiquity, London, 2006, no. 147 (ΗΓΕΜ-ΟΝΙΟΥ).
7 M.-L. Vollenweider and M. Avisseau-Broustet, Intailles et camées II.
Les portraits romains du Cabinet des Médailles, Paris, 2003, no. 265
(fragmentary inscription [–] antoniniae).
8 M. Maaskant-Kleibrink, Royal Coin Cabinet the Hague. Catalogue of
the Engraved Gems in the Royal Coin Cabinet, the Hague: the Greek,
Etruscan and Roman Collections, Wiesbaden, 1978, no. 310
(ΕΠΙΚΟΥΡΟΥ, 1st century bc).
9 E. Zwierlein-Diehl, Antike Gemmen des Kunsthistorischen
Museums in Wien, III, Vienna, 1991, no. 1750 (ΑΛΕΞΑΝΔ-ΡΑ, late 2nd
century ad).
10 Spier (n. 3), no. 76 (legend: dn theodosivs avg). The letters PF do
not appear on the gem as on the legend of a solidus with the same
facing bust (Numismatica Genevensis, Auction 2, November 18th
2002, no. 142).
11 Cormack and Eastmond (n. 6), no. 63.
12 Ibid., no. 151, with the legend: [d n ivstini]anvs pp[avg],
D(ominus) N(oster) Iustinianus p(er) p(etuus) Aug(ustus).
13 Spier (n. 3), no. 83.
14 E. Spagnoli and M.C. Molinari, ‘Le monete’, in A. Salvioni (ed.), Il
tesoro di Via Alessandrina, Rome, 1990, 77–107, no. 18: solidus of
Grimoald III, Duke of the Lombards (obverse legend: grimvald).
The mint technique betrays a Byzantine influence.
15 E. Zwierlein-Diehl, Antike Gemmen in Deutschen Sammlungen.
Staatliche Museen Preussischer Kulturbesitz. Antikenabteilung
Berlin, Munich, 1969, no. 516b, double-sided plasma gem with the
inscription procvla rarissima (the face no. 516a shows a
2nd-century ad figure of Eros: the intaglio was undoubtedly a love
gift). Spier (n. 3), no. 60, for a silver disc with the legend corn
dvlcis. One might think that dvlcis was the surname of the
aforementioned Cornelia (corn) if it did not already appear on a
jasper from Vienna as the qualifying adjective ‘sweet’ (celade/
vicas/n – for vincas – helpi/dia dv/lcis, ‘Celadus, may you win,
(wish your) sweet Helpidia!’): see  Zwierlein-Diehl (n. 9), no. 2154.
For vincas on another gem, see Spier (n. 3), no. 85 (ricime-r
vincas, ‘Ricimer, may you win!’).
16 Spier (n. 3), nos 52 (spesiniane vivas) or 54 (vi-vas).
17 Ibid., no. 23 (viva-s in deo).
18 Ibid., nos 2 (IHCOY XPICTOY) or 3 (XPICTOY).
19 Ibid., no. 737 (HPAKΛITOC ZHCЄC ЄN ΘЄW); also, M. Henig, The
Content Family Collection of Ancient Roman Cameos, Oxford, 1990,
no. 38 (ЄYTYXI ZHCAIC [O Φ]OPWN); see also, ibid., 6–7, nos 28–37.
20 Vollenweider and Avisseau-Broustet (n. 7), no. 239 (EYTYX[I O]
ΦOPWN [‘Good luck to the bearer!’]). Vollenweider transcribed
EYTYX ΨOYWN. See also M. Henig, A Corpus of Engraved
Gemstones from British Sites (BAR, British Series 8), Oxford, 2007
(3rd edn), no. 743.
21 Spier (n. 3), no. 70 (BACIΛICCA OYAPAZAΔOYKTA ЄPMIONHC/
ΨYX-H). For a complete description and a restitution of the legend,
see: Zwierlein-Diehl (n. 9), no. 1731, 72–3. For ΨYXH, see also M.
Schlüter, G. Platz-Horster and P. Zazoff, Antike Gemmen in
Deutschen Sammlungen. Hannover und Hamburg, Wiesbaden,
1975, Hannover, no. 1667 (ΨYXH/KAΛH).
22 See E. Zwierlein-Diehl, Antike Gemmen und ihr Nachleben, Berlin,
2007, 183–4.
23 The hairstyle (like the Empress Severina, wife of Aurelian), the
epigraphy, as well as the etymology of the nickname (Fructosa) are
‘Gems of Heaven’ | 245
Aubry
characteristic of the 3rd century ad. See also CIL VIII, no. 2756
(Ennia Fructosa).
24 T. Cadès, Impronte gemmarie dell’Istituto, Rome, 1836, Collezione
Nott, no. 67 (silvane vivas cvm frvcco-sa). On CVM, see n. 87.
25 Silvanus: vocative in -E implies nominative in -VS (and vocative in
-I = nominative in -ivs).
26 Fructosa (cum + ablative).
27 U. Pannuti, Museo Archeologico Nazionale di Napoli. Catalogo della
collezione glittica II, Rome, 1994, no. 216 (with the legend:
AYP•ΠAΠIANH [‘Aurelia Papiana’]•ЄYTYXЄI AN (sic) ΠЄΛI
[‘Ampelius’]).
28 Compared to the epigraphical tradition of the Roman West, the
imperative case of the verb εὐτυχέω is not contracted here; the
verb in Latin is declined in the standard vocative case. See also
E. Zwierlein-Diehl, Antike Gemmen des Kunsthistorischen
Museums in Wien, II, Vienna, 1979, no. 1200.
29 J. Reynolds, C. Roueché and G. Bodard, Inscriptions of Aphrodisias
(IAph 2007), available <http://insaph.kcl.ac.uk/iaph2007>,
12.631 and 12.101. The use of the letter N instead of M is an epi-
graphical subtlety already attested on a stone dating from the
same period (Pannuti [n. 27], no. 198). See also,  IAph 2007, 2.19,
4.202 and 8.609.
30 M. Gramatopol, Les pierres gravées du Cabinet numismatique de
l’Académie Roumaine (Latomus 138), Brussels, 1974, no. 845 ([E]
IBIC EYTYXEI ΠOLONA). The first word, ‘eibis’, relates to the
portrait, offered as a gift to Polona. There is a possibility, however,
that this gem is post-antique.
31 Zwierlein-Diehl (n. 9), no. 2159 (gelasivs/zosime vi/vas, ‘A
laugher (wish): Zosimus, may you live (happy)!’). E. Zwierlein-
Diehl had already demonstrated the etymology of Gelasius,
derived from γελάω, ‘to laugh’, and brought together other signa
on seals, such as Hilarius and Gaudentius, often associated with
vivas, and which represented wishes for a happy life. See, Henig (n.
19), no. 35 (ЄYTYXI ΓЄΛACI, ‘May you have luck and laughter!’ or
‘Good luck to you, Gelasios!’). For the arbitrary hyphenation, see
also Zwierlein-Diehl (n. 9), no. 2154 – see n. 15 above.
32 Henig (n. 19), no. 45 (TH KAΛH, ‘To the beautiful girl’). This legend
perhaps bears out the suggestion that partial nudity is meant to
indicate a compliment to the wearer as being like the goddess
Aphrodite (see ibid., 25, notes); consequently the cameo was
probably worn by a woman.
33 M.-L. Vollenweider, Musée d’Art et d’Histoire de Genève. Catalogue
raisonné des sceaux, cylindres, intailles et camées II. Les portraits, les
masques de théâtre, les symboles politiques, Mainz am Rhein, 1979,
no. 272bis.
34 Spier (n. 3), no. 35 (septimi elia vivatis).
35 See n. 24 above.
36 It would be nonsense to transcribe the first missing letter as a Γ – as
Vollenweider thought – instead of a G when the rest of the name is
in Latin.
37 Henig (n. 19), no. 47 (AKAKIN ΓPHΓOPI [‘Innocent one, be
wakeful!’]); see also, idem, no. 48.
38 Lexicon Iconographicum Mythologiae Classicae I, 63; 67 and CIL VI,
23997; 13406. Vollenweider, following M. Guarducci, transcribed
the name in Greek letters: a few other bilingual inscriptions on
gems and glass are attested (see n. 87 below – or H.B. Walters,
Catalogue of the Engraved Gems, Greek, Etruscan and Roman, in the
British Museum, London, 1926, no. 2689 (ΛYCIMAX/
OC•L•EPHORI).
39 The gesture of the left hand seems to indicate a Christian
influence: vivas could thus mean vivas in deo and be a metaphor of
aspirations to eternal life. See Vollenweider (n. 33), 262.
40 E. Babelon, Catalogue des camées antiques et modernes de la
Bibliothèque nationale, Paris, 1897, no. 328; F. Buonarroti,
Osservazioni istoriche sopra alcuni medaglioni antichi all’Altezza
Serenissima di Cosimo III, Granduca di Toscana, Rome, 1698, 406–
11, pl. XXXVI/3, had read ner vir v (ner for ‘neratia’).
41 Buonarroti (n. 40), 406–11, pl. XXXVI/1.
42 See L. Pirzio Biroli Stefanelli, ‘I gioielli’, in A. Salvioni (ed.), Il
tesoro di Via Alessandrina, Rome, 1990, 41–75, at no. 5, epicrates
q/q (probably a tria nomina or a qualificative). See  also, M.-L.
Vollenweider, Die Porträtgemmen der römischen Republik, Mainz
am Rhein, 1972–4, no. 113/1 ; IGI IV, no. C178, L/L (portrait of
Licinius Lucullus; the dolphin holding an olive branch is perhaps
an allusion to the naval victory of Lucullus in Lemnos); A.
Furtwängler, Beschreibung der Geschnittenen Steine im
246 | ‘Gems of Heaven’
Antiquarium, Berlin, 1896, no. 6536, P/P (portrait of Pompeius
Magnus, qualified as Pompeius patronus).
43 For the doubling of the consonant to emphasise the plural, as
attested in epigraphy and in numismatics, see: M.C.J. Miller,
Abbreviations in Latin, Chicago, 1998, xi–xii.
44 See Section 3.3., Type 7, paragraph below. The legend has the
figure as if in a grip: it creates a direct link between the two
elements. Therefore it cannot be a verb like vivatis, but a qualifying
word.
45 C. Weiss, Die antiken Gemmen der Sammlung Heinrich Dressel in
Antikensammlung Berlin, Berlin, 2007, no. 652 (avsenti/vivas/
ivcvndae).
46 See n. 24 above.
47 See n. 33 above.
48 See Spier (n. 3), no. 43 (lvciane avgvst-a v: 4th-century ad silver
disc). See also n. 71 below.
49 Vollenweider and Avisseau-Broustet (n. 7), no. 228 ([YΠEP THN]
NЄIKHN TWN KYPIWN CЄB(ACTWN). Probably related with the
peace of ad 209, when Geta received the title of Augustus (see ibid.,
179). Translation of the dedication from: A.M. McCann, The
Portraits of Septimius Severus (MAAR XXX), Rome, 1968, 182.
50 Cormack and Eastmond (n. 6), no. 65 (see also, Spier [n. 3], no.
572, or Zwierlein-Diehl [n. 22], no. 822), •fl•romvl•vest•fecit•
(‘Flavius Romulus Vest(…) made this’). D. Prozorovskii (see ibid.,
149) suggests that vest means vestiarius (keeper of the imperial
clothing), and consequently that Flavius Romulus was not the
creator but the commissioner of the stone; but the addition of the
word fecit presents evidence to the contrary.
51 Spier (n. 3), no. 19 (ΘEOTE/KNOY).
52 Vollenweider and Avisseau-Broustet (n. 7), no. 179 (ΠOΘOYM/
ЄNH; 2nd century ad).
53 M.-L. Vollenweider, Deliciae Leonis. Antike geschnittene Steine und
Ringe aus einer Privatsammlung, Mainz am Rhein, 1984, no. 317.
See also Spier (n. 3), no. 84.
54 See Spier (n. 3), no. 27. Compare to Vollenweider (n. 53), no. 523
(monogram D(ominus) N(oster) Theodoricus; late 5th century ad).
55 To identify the character, the authors of the proposition, P.E.
Schramm and J.D. Breckenridge (see Vollenweider [n. 53], no. 317,
bibliographical notes), point out the abnormality of the upper lip,
which gives him a particular grimace, and which we also find on a
gold medallion of Theodoric in Rome (see C. Barsanti, A. Paribeni
and S. Pedone, Rex Theodericus. Il medaglione d’oro di Morro
d’Alba, Rome, 2008).
56 Vollenweider and Avisseau-Broustet (n. 7), no. 106/8 (pob(licius)
at(ilianus).
57 Spier (n. 3), no. 79 (5th-century ad garnet).
58 Chi-rho symbol: see ibid., no. 42 (4th-century ad silver disc).
59 Staurogram: rho-cross symbol: see ibid., no. 26 (4th-century ad
nicolo).
60 Ibid., no. 40 (crescentine apvle 1 (christogram) n, inscription
naming both the busts as ‘Crescentina’ and ‘Apulus’.
61 See N. Zaïd, ‘Augustin, l’enfant de Thagaste’, Archéologia 406
(2003), 34 (4th century ad).
62 Spier (n. 3), no. 25, 4th-century ad quartz crystal with inscription
MATPWN (chi-rho) A ΠPO[BI]ANOY, or ibid., no. 45, 4th-century ad
silver disc with ΘЄOΠO (cross) NTOY (the inscription does not refer
to the figure: it must be the name of the ring’s owner).
63 Ibid., no. 74, 5th-century ad garnet with KAPI-ANH (cross).
64 See n. 10 above.
65 See n. 27 above.
66 Zwierlein-Diehl (n. 22), no. 897 (xpe adivva hlotharivm agv,
Ch(rist)e adiuva Hlotharium Aug(ustum), ‘Christ, help Emperor
Lothar!’); see also, eadem, 288.
67 Spier (n. 3), no. 46 (EYΦHMIOY); see also idem, nos 45 (ΘЄOΠ/
ONTOY – see n. 62 above) and 47 (E[Λ]-EYΘEP-IO-Y).
68 Ibid., no. 451 (ANOYBIWN (for ‘anubis’). There is another theophoric
Egyptian name on an 2nd-century ad carnelian in The Hague,
representing a prize table and a capricorn (Maaskant-Kleibrink [n.
8], no. 696: EICIWN, for ‘isis’).
69 Zwierlein-Diehl (n. 22), no. 699 (ΦOPT-OYNAT-OY).
70 Spier (n. 3), no. 36, inscribed concordia which, as with rings
inscribed in Greek OMONOIA (see  Zwierlein-Diehl [n. 22], nos 208
and 769; Henig [n. 20], nos 50–1), signifies ‘marriage, concord’.
71 For example, Spier (n. 3), nos 43 (lvciane avgvst-a v - see n. 48
above), 41 (pr-imvl-e v-iv), 59 (viva in/xp), 48 (vi-vas in d-eo),
and 35 (septimi elia vivatis). On other 4th-century ad discs and
 Inscriptions on Portrait Gems and Discs in Late Antiquity (3rd–6th centuries AD)
rings without couples, we can also find gvrda v1 (ibid., no. 55), and
vivai/in deo (M. Henig, The Lewis Collection of Engraved
Gemstones in Corpus Christi College (BAR Supplementary Series 1),
Cambridge, 1975, no. 294).
72 Spier (n. 3), no. 39 (do-m vict caen).
73 See n. 51 above.
74 Even if this last one is imaginary.
75 See Cormack and Eastmond (n. 6), no. 96 (with the inscription
OYHC ANHC NINAC / OYΛPHOYHC).
76 See n. 72 above. For the name of the boy, there is the problem of the
toga virilis (a child usually got it at the age of 15 to 17): see C.B. Horn
and J. Martens, ‘Let the little children come to me’. Childhood and
Children in Early Christianity, Washington DC, 2009, 17.
77 Spier (n. 3), no. 1 (EYTYXI ΠANXAPI META THC KYPIAC
BACIΛICCHC KAI ΠAYΛINAC/ZOH/IC ΘE-OC). See also, Zwierlein-
Diehl (n. 22), no. 183.
78 ‘Good luck to Pancharios with the lady Basilissa and (daughter)
Paulina’.
79 ‘Life’ and ‘There is one god!’.
80 See n. 75 above. The origin of the appellations is probably Sebastia
in Palestine: see, Cormack and Eastmond (n. 6), 163.
81 OYHC for vis, ANHC for anis and NINAC. For ‘diacritic name’: J.-M.
Lassère, Manuel d’épigraphie romaine, Paris, 2005, 102 , 110–13.
82 OYΛPHOYHC for Ulrivis. For signum, see ibid., 102 (Late Antique
onomastics) and 110–13 (signum); it can be worn by several
members of a family, and it appears mostly isolated, unlike the
agnomen.
83 See Section 2.3, Type 6, paragraph above. Indeed the silver disc is
11 x 11mm (Spier [n. 3], no. 39), the cameo in Vienna, 31 x 24.5mm
(ibid., no. 1), and the carnelian from St Petersburg is 51.5 x 39mm
(Cormack and Eastmond [n. 6], no. 96).
84 McCann (n. 49), 183, pl. XCII, with the inscription EZ (reversed) C.
85 G.M.A. Richter, The Metropolitan Museum of Art. Catalogue of
Engraved Gems: Greek, Etruscan and Roman, Rome, 2006 (2nd
edn), no. 253 (E-IC ZEYC•C/AP/AΠIC).
86 See E. Brandt, E. Schmidt, A. Krug and W. Gercke, Antike Gemmen
in Deutschen Sammlungen. Staatliche Münzsammlung München,
Munich, 1968–72, no. 909, z (reversed) alviuz (reversed). See
also idem, nos 630, 832, 934 and 943 (2nd–1st century bc).
87 D.B, Harden, ‘The Wint Hill hunting bowl and related glasses’,
Journal of Glass Studies 2 (1960), 45–82, at 48–51, figs 1–2 and 4–7,
hunting scene with an inscription running around the design,
vivas cvm tvis pie z (reversed); see also, Henig (n. 19), 9.
88 It may be translated as ‘Life to you and yours’: see, Henig (n. 19), 9.
89 Translatable as ‘Drink and good health to you’: Henig (n. 19), 9, and
C. Carletti, Epigrafia dei Cristiani in Occidente dal III al VII secolo,
Bari, 2008, 195, no. 82 (pie zeses).
90 See n. 27 above.
91 E. Zwierlein-Diehl, Glaspasten im Martin-von-Wagner-Museum der
Universität Würzburg I, Munich, 1986, no. 801.
92 Like Pupienus, Balbinus and Gordianus III, or Severus, Caracalla
and Geta: O. Neverov, Antique Intaglios in the Hermitage Collection,
Leningrad, 1976, no. 139. For a description, see Zwierlein-Diehl (n.
91), 265 (late 2nd– early 3rd century ad); the lengthening of the
faces on the Würzburg gem (eadem, no. 801), compared to that of St
Petersburg, is a sign of late engraving (late 4th–early 5th century
ad). See also, Schlüter et al. (n. 21), Hamburg, no. 67.
93 See Schlüter et al. (n. 21), Hanover, no. 1628, for the letters V and N.
94 M. Hainzmann and R. Wedenig (eds), Instrumenta Inscripta Latina
II. Akten des 2. Internationalen Kolloquiums. Klagenfurt , Mai 2005,
Klagenfurt, 2008, 195–7, figs 4–5 (1st century ad): b (reversed) ann
(reversed) af, Banna f(ecit); see also, CIL XIII, 10027, 139.
95 It should be a legend in the nominative case, or possibly a
dedication in the dative: these last ones, usually of Types 3 and 4
(Vollenweider and Avisseau-Broustet [n. 7], no. 228 – see n. 49
above) are also of Types 1 or 2 in relation to portraits of emperors
(Richter [n. 85], no. 657 (divo clavdio imperato[ri]).
96 Roman agnomina were placed after the tria nomina and were
sometimes introduced with the formula qui et (vocatur), sive or
idem, ‘also known as’, which differentiates them from signa.
97 See Lassère (n. 81), 102 and 113.
98 See B.H. McLean, An Introduction to Greek Epigraphy of the
Hellenistic and Roman Periods (323 bc– ad 337), Ann Arbor, 2002, 124.
99 See n. 75 above.
100 In any case, the A is common to all three persons, and creates a
link, a complementary rapport between the inscription and the
design, like vertical Type 5 and horizontal Type 6 legends.
101 Spier (n. 3), no. 44.
102 No doubt by drawing a parallel between the inscription and
another silver disc carrying that name (see ibid., no. 66).
103 See H. Solin, Griechische Personennamen in Rome, Berlin, 2003, 66
(‘Eupheros’), 520 (‘Euphemos’) and 1395 (‘Euphemia’).
104 See n. 91 above.
105 Zwierlein-Diehl (n. 9), no. 1741.
106 The Roman discontinuous method of contraction for Greek and
Latin words – first, middle and last letter preserved – means that
ΦΛΠ could not be a single appellation (nomen or signum);
consequently it must be the initials of a tria nomina: see
A.N. Oikonomides, Abbreviations in Greek Inscriptions, Papyri,
Manuscripts and Early Printed Books, Chicago, 1974, 21–9.
107 Spier (n. 3), no. 17 (p-e-re, for ‘Perennius’?).
108 See Vollenweider and Avisseau-Broustet (n. 7), no. 182 (Π-PO-KΛ,
‘Procles’; 2nd-century ad jasper), and Vollenweider (n. 42), no.
33/4 (pla-c-[–], ‘Placid(i)us’)
109 Zwierlein-Diehl (n. 9), no. 2538.
110 Vollenweider and Avisseau-Broustet (n. 7), no. 162.
111 The hypothesis that P implied ‘Plotina’ has been advanced but, as
M.-L. Vollenweider demonstrated, the portrait is much like the one
of ‘Matidia’ (see RIC III, Hadrian, no. 34).
112 RIC II, Trajan, no. 759, a denarius which celebrates Matidia Augusta.
On the reverse, Salonina Matidia embodies the goddess Pietas, and
holds hands with her daughters Sabina and Matidia Minor.
113 Brandt et al. (n. 86), no. 924, the letter F with the cornucopia
implies Fortuna (prosperity).
114 See n. 41 above.
115 See n. 74 above.
116 Zwierlein-Diehl (n. 91), no. 819.
117 Vollenweider (n. 33), no. 32 (orat[i]vs).
118 Zwierlein-Diehl (n. 9), no. 1729.
119 For example, see RIC I, Tiberius, no. 56, divvs avgvstvs s/c (obv.).
120 See RPC I, no. 1756, Π/E for Perinthos (Thrace). Regarding the
mint, the same principle rules the well-known Hellenistic sapphire
signed by Pyrgotheles (Π/Y): it is not the mint, but the workshop.
See J. Boardman, Engraved Gems. The Ionides Collection, London,
1968, no. 15, or J. Boardman, Greek Gems and Finger Rings, London,
2000 (2nd edn), no. 997.
121 Schlüter et al. (n. 21), Hamburg, no. 67.
122 It could be a signum like Χρῆσιμος, Χρῆσανθος or a theophoric
Christian name like Χριστόδουλος ('servant of Christ' – see Spier
(n. 3), no. 341) or Χριστόφορος ('bearer of Christ'). See Lexicon of
Greek Personal Names I, 486; II, 479–80; IIIa, 478–9; IIIb, 445; IV
357–8 and F. Preisigke, Namenbuch: enthaltend alle griechischen,
lateinischen, ägyptischen, hebräischen, arabischen und sonstigen
semitischen und nichtsemitischen Menschennamen, soweit sie in
griechischen Urkunden... , Amsterdam, 1967 (2nd edn), 479.
123 Spier (n. 3), no. 193 (fish).
124 Ibid., no. 319 (Good Shepherd).
125 See ns 57 to 59 above.
126 Zwierlein-Diehl (n. 91), no. 537, bust of Marcus Iunius Brutus, one
of the murderers of Julius Caesar, with the dagger, instrument of
his crime; an image of the dagger can also be found on the EID
MAR silver denarius (RRC, no. 508/3) issued by Brutus.
127 Vollenweider and Avisseau-Broustet (n. 7), no. 135.
128 Titus Ollius was the name of her father. As on Zwierlein-Diehl (n.
9), no. 1729 (Fulvia Cai – see n. 118 above), the name of the father is
used to characterise the figure.
129 Coinage of Iulia Maesa, Orbiana, Iulia Mamae, Etruscilla, Salonina
and Severina (2nd–3rd centuries ad).
130 If the text is the name of a slave, the image shows how he got the
stone: as a gift after having won a competition: see Walters (n. 38),
no. 2666, a hand holding a palm-branch (symbol of victory);
around the figure, Type 5 legend stefan-o/t hatili ti s, Stefano, T.
Hatili(us) Ti() s(ervo).
131 For example, the combination of Type 1 with Type 4 can be found
on gems (Pirzio Biroli Stefanelli [n. 42], no. 5, epicrates q/q, on
coins (RIC 1, no. 56, divvs avgvstvs s/c – see n. 119 above) and on
medallions (Buonarroti [n. 40], 406–11, pl. XXXVI/1, belliciae
modeste v/v). On the reverse of a solidus of Honorius (Spagnoli
and Molinari [n. 14], no. 17, with the legend victori avgg, mint
(Mediolanum) and metal (Comitatus Obryzum, ‘pure gold’), we
find the association of Types 1, 4 and 7, as on the reverse type of a
solidus of Grimoald III with the legend dom(inv)s car(olvs) r(e)x
g/r vic (ibid., no. 18).
‘Gems of Heaven’ | 247
Roman Intaglios Oddly Set
The Transformative Power of the Metalwork Mount
Genevra Kornbluth
Several of the essays in this collection deal with Roman seal
stones, intaglios that were set in rings worn all over the Empire.
Such gems were often re-used in medieval art. Many are
prominently set on Early Medieval fibulae, reliquaries and
other metalwork.1 It has been argued that their flawless,
enduring luxury epitomised beauty for medieval viewers. But if
simple admiration is all that was going on, why are so many
stones set with apparent disregard for their proper orientation?
I will focus here on engraved gems in metalwork mounts of the
6th to early 9th centuries ad, arguing that the figures on such
stones were neither ignored nor misunderstood, but rather
intentionally subverted.
The imagery and inscriptions on magical gems, discussed
many times in this volume, demonstrate that Roman
gemstones were thought to have power, and that the power of
the materials was meant to be enhanced by engraving with
appropriate figures. The Latin version of the Orphic Lapidary
associated with the legendary magus Damigeron and king
Evax2 proves that such ideas continued at least into the 5th–6th
century ad in the West. But increasingly from the 6th century
ad, Christian belief was also a factor in the way gems
functioned. On a mature Early Byzantine work like the
Dumbarton Oaks amethyst from the 6th–7th century ad, for
example, Christian content is clear and overt (Pl. 1).3 At the
same time, the engraver was plainly still working within the
tradition of magical amulets. In Christ’s left hand is a scroll
with the protective inscription, ‘In the beginning was the
Word,’ and below his raised right hand is a list of powerful
archangelic names (Pl. 2). This stone is linked to both biblical
texts and works like the Orphic Lapidary.
Gems that are closely associated with texts have complex
meanings and function on several levels at once. I argue that
the same is true of stones that cannot be so neatly linked with
textual evidence. The concept of lithic power, dependent on
Plate 1 Amethyst intaglio with Plate 2 Detail of Pl. 1:
standing Christ, 6th-7th century ad; inscription
Washington DC, Dumbarton Oaks
Byzantine Collection, BZ.1953.7
248 | ‘Gems from Heaven’
both the inherent potential of the material and its physical
form, continued to be a part of Merovingian, Anglo-Saxon and
early Carolingian culture. Apotropaic stones have been found
in graves from the mid-5th to the late 7th century ad, in church
burials as well as pagan and Christian cemeteries. Rock crystal,
long associated with natural energy, was painstakingly shaped
into spheres that were wrapped in metal bands and worn as
‘bound pendants’, markers of identity and social status (Pl. 3).
Fossil sea urchins and iron ore nodules, naturally spheroidal
and also thought to hold power, were used in much the same
way. Both crystal and cobalt blue glass were shaped into
elaborate polyhedra that reinforced the power of materials
with the strength of forms.4
In this context, how were Roman figured intaglios
perceived? Some were quite straightforwardly set into rings,
attesting to formal continuity. But outside the literary culture
of the Mediterranean, and in partially or fully Christian
communities, the intellectual framework underpinning gem
usage cannot have remained unaltered. In order to perceive
whatever changes there might have been, we need to turn
away from rings.
In a composition that may follow Merovingian precedent,
the Augustan cameo famously adorning the Cross of Lothar
stood in for Christ and/or the current emperor, and perhaps for
God the Father as well (Pl. 4).5 On an Ottonian cross the image
has acquired new layers of meaning, and we would be missing
a great deal of what the object has to say if we were to think of
it as merely a Roman emperor. The more distant from its
primary usage a gem’s secondary mounting becomes, the less
difficult it is for us to discern transformation. On the Cross of
Herimann in Cologne, changes of both religion and gender
make it easy to see the mutation of a lapis lazuli female portrait
into the head of Christ.6
Plate 3 Rock crystal, silver-bound pendant Plate 4 Lothar Cross, gold and gems on
from Giberville (Calvados) ‘Le Martray’ wood core, c. ad1000; Aachen
grave 27, second half of the 6th century ad; cathedral treasury
Musée de Normandie de Caen, 79.4.80

Plate 5 Teuderigus Reliquary (front), gold, gems, and glass on wood core,
second half of the 7th century AD; St-Maurice d’Agaune abbey treasury
Far distant indeed from Roman rings are Merovingian and
early Carolingian reliquaries. The most popular western forms
of 6th–7th century ad personal reliquaries, strap-ends and
buckles, were not normally set with gemstones. With the
emergence of the burse reliquary as a major form, however, the
situation changed.7 Although this type was often just as small
as its predecessors, it could be much more elaborate. Two
extremely rich early examples are set with large numbers of
figural gems, and unlike others, have not suffered extensive
later modifications: the Teuderigus reliquary in St-Maurice
d’Agaune and the Enger reliquary in Berlin.8 Like later
medieval reliquaries, these display intaglios with a range of
subjects and materials greater than those used for seals.9 The
Teuderigus Reliquary (Pl. 5) takes its name from the
Plate 6 Teuderigus Reliquary (right end)
Roman Intaglios Oddly Set
Plate 7 Teuderigus Reliquary (front), Plate 8 Teuderigus Reliquary (front),
detail: intaglio with lion detail: intaglio with Mars
inscription written diagonally across its back, naming the
priest Teuderic as originator of the project to honour St
Maurice, as well as two secular patrons and two artists.10 The
inscription is linguistically and palaeographically pre-
Carolingian,11 and the reliquary is now generally dated to the
second half of the 7th century ad and localised in the region of
south Germany/Switzerland. Three sides prominently display
engraved gems, along with a contemporary glass cameo, inlaid
blue and green glass, and literally hundreds of cut garnets. The
intaglios on both short sides are set so as to be easily legible. On
the right end (Pl. 6), all three show standing figures facing
toward their proper left, extending the left arm out or
downward. Like reset Roman inscriptions advertising the
literacy and heritage of medieval monks,12 these prominently
and carefully displayed gems manifest pride of ownership,
reflect and enhance the prestige of the saint housed within,
and perhaps declare a victory over paganism.
Compositional focus on all the intaglios and the careful
orientation of figures on those of the right end make it all the
more noticeable that most of the gems on the large front
surface are not vertically set. On the viewer’s left, a lion is
mounted with its head facing downward, its legs extended out
toward the side (Pl. 7). Above that stone a red one shows two
facing figures standing upright. On the other side of the cameo,
a nicolo Janus head is upside down and on a diagonal, the neck
pointing up toward the right. Beyond that, near the right edge
of the reliquary, an armed Mars stands on his head (Pl. 8).
These inversions were probably not caused by botched
restoration. Only the upright stone with two figures shows
evidence of tampering with the setting. Others, including the
upside-down Mars, are remarkably untouched. Are the odd
positions the result of artistic negligence? The regularity of
poses on the side panels suggests not. The stones’ imagery was
probably considered when they were set.
The overall composition of the front panel is based on
diagonal crosses made by lines of pearls and gems, flanking an
orthogonal cross (as on the side panels) formed by a central
cameo with enlarged blue and green teardrop cells at its top,
bottom, and sides. The salvific power of the cross, central to
the function of any Christian reliquary, is thus foregrounded.
Antje Krug has identified the large cameo figure as Dea Roma,13
and argued that the image was purposely chosen. If her theory
is correct, perhaps it was not only crosses that alluded to
‘Gems of Heaven’ | 249
Kornbluth
Plate 9 Sant’Agata de’ Goti, cathedral Plate 10 Sant’Agata de’ Goti,
portico cathedral portico, detail: spolia
Christian power, but also the inclusion of Rome herself within
the Christian orbit. In any case, the potency of the reliquary
and the relics contained within it appear to be of paramount
importance to its program, as one might expect. I suggest that
intaglios set out of true were also part of that program.
‘Abrasax’ gems were not the only stones of power in the
Late Antique world. Far more visible, and famously problematic
to Christian viewers, were the cult figures of pagan gods. Such
sculpture could be positively dangerous, and had to be
destroyed.14 A temple in Uley, Gloucestershire, was still in use
in the late 4th century ad even after part of it collapsed or was
demolished, but in the 5th century ad its large Mercury was
mutilated and its defaced altars were used in the construction
of a timber basilica.15 Such thoroughness bespeaks not just a
desire to clear land for new construction, but active hostility.
Gregory of Tours, in The History of the Franks completed before
ad 594, related a story told him by St Vulfolaic, who with divine
aid annihilated a figure of Diana. When a crowd of people was
unable to pull down the heavy sculpture, Vulfolaic
hurried off to the church and lay prostrate and weeping on the
ground, praying to God for help, that in his divine power he would
destroy what human strength was powerless to overturn. [After
that] at the very first heave which we gave the idol crashed to the
ground. I had it broken to pieces with iron hammers and then
reduced to dust.16
Simple destruction (including disposal in lime kilns, the most
complete negation of all) must have been the fate of many
statues classified as idols by Christians. But in some cases
destruction was not simple. Many earlier figures were buried
after mutilation, as at Uley. In other medieval contexts earlier
material was made visible but not normally legible. On the
Romanesque cathedral portico of Sant’Agata de’ Goti, for
example, a relief of three standing figures was incorporated
into the wall on its side (Pls 9 and 10).17 If the only thought was
to re-use the stone for the new structure, why were its figures
made visible at all? If the figures were meant to be given a
Christian meaning and proudly exhibited, why were they
placed on their sides? The combination of visibility and
abnormal orientation suggests that a statement was being
made about the sculpture itself. Pure triumphalism could be
involved, but with a large variety of Roman stone probably
250 | ‘Gems of Heaven’
available for re-use,18 why use such anonymous figures? The
answer could lie in the perception of power. If the relief (like
threatening idols) was thought to have some potency that had
to be negated, it could have been destroyed or mutilated and
buried. But if there was a desire to retain but make safe, to
harness or bind rather than nullify that energy, an ‘upset’
mounting could have been ideal. Inscriptions on 2nd-century
ad funerary cippi were similarly mounted visible but upside
down as capitals in the 8th- to 12th-century ad church of S.
Giorgio di Valpolicella.19 In the 8th-century ad church of St
John at Escomb another inscription was placed on its side as
part of an interior window jamb (obscured?).20 Non-Christian
altars were frequently used in Britain as medieval fonts, stoups,
and cross bases. In many cases the original inscriptions were
left truncated but visible even when the rest of the stone was
thoroughly recut,21 suggesting the possibility of simultaneous
triumphalism and the channeling of power. While most datable
examples are later than our period, the crypt of St Andrew’s,
Hexham, (ad 674–8) includes a Roman altar re-used as a lintel
with part of its inscription intact and visible, set on its side but
not effaced.22 There must have been a full range of approaches
available to owners of spolia, from ostentatious and
unproblematic exhibition to complete destruction, with the
diversion or harnessing of power by means of upset display
somewhere in the middle.
Like relief sculpture mounted sideways in a wall, engraved
gems on metalwork could also have been meant to deal with
power. If people who no longer venerated Mars wanted to
preserve lithic strength enhanced by imagery but channel it in
a new direction (Pl. 8), the odd mounting on the Teuderigus
Reliquary could have provided them with a means to do so. Old
power could be harnessed in the service of new ideas.
A similar phenomenon occurs on the Enger Reliquary in
Berlin (Pl. 11).23 In this case one major side is gemmed, the
other three covered with repoussé figures (Christ with angels,
the Virgin, and saints). On the basis of the figural imagery and
inlay technique, the reliquary is thought to have been made in
the third quarter of the 8th century ad or the early 9th, north of
Plate 11 Enger Reliquary (front), gold, gems, and enamel on wood core, third
quarter of the 8th to early 9th century AD; Berlin, Kunstgewerbemuseum, 88.632

Plate 12 Enger Reliquary (front), detail: intaglio with satyr
the Alps.24 The front side of this work displays 13 large
gemstones, of which four, dating from the 1st century bc/ad to
the 2nd–3rd century ad, are engraved.25 Again, some of the
settings are unexpected. On the upper left a nude satyr on a
nicolo agate is mounted upright, but opposite it a similar stone
with Oedipus and the Sphinx is placed upside down (Pls 12 and
13). In the centre of the upper left quadrant a pale stone is set
more than 90° off axis, putting its cock, goat, and cornucopia
on their backs and heads. And in the centre of the lower right
quadrant a chalcedony is mounted with Jupiter Capitolinus
standing on his head (Pl. 14). Care has been taken to maintain
symmetry and colour differentiation in the overall design, with
an orthogonal cross established by the large unengraved nicolo
in the centre and dark gemstones at the ends, and a diagonal
cross marked out by that same centre, pale stones in the middle
of each quadrant, and blue or purple stones at the ends.
Between the gemstones, enamel animals represent (in the
reading of Victor Elbern)26 the fish, birds, and land dwellers of
Genesis 1: 26–8, surrounding the double cross of the new
creation. By this reasonable interpretation the Berlin reliquary,
like the one in St-Maurice d’Agaune, deals with salvific power.
In this case that strength is understood as the power of (and
over) creation. As ‘creatures,’ gemstones are of course part of
the same overall scheme. It has been argued that upset
Plate 13 Enger Reliquary (front), detail: intaglio with Oedipus and the Sphinx
Roman Intaglios Oddly Set
Plate 14 Enger Reliquary (front), detail: intaglio with Jupiter Capitolinus
mountings in fact nullify the engraved figures here: that the
makers of such reliquaries did understand earlier intaglios, but
chose to ignore their imagery in favour of compositional
balance (and so placed an unengraved stone in the centre).27
This theory could well explain the pale stone in the upper left
quadrant, whose angled disposition lets it roughly match the
vertical orientation of the other three pale stones near the
centre. And in terms of formal pattern, variant placement
results in all three of the gems with human figures gesturing
inward, toward the centre nicolo. Perhaps this was a means of
reinforcing symmetry and a central focus. But the evidently
purposeful use of colour in all the stones’ placements suggests
that more is happening here. The metalworkers could easily
have turned the engraved sides invisibly inward. The
chalcedony Jupiter, at least, would not have altered its
observable colour (Pl. 14). (While it is usually impossible to see
whether engraved gems have been turned inward by medieval
goldsmiths, translucent intaglios, which do leave figures visible
to close examination, were set in this way in the 11th century.)28
The fact that this procedure was not used on the Enger
Reliquary suggests that glyptic imagery had some interest for
the object’s makers after all.
The one upright intaglio, the nicolo satyr, may have
appealed to Early Medieval metalworkers just as it was. It could
easily have lent itself to interpretatio Christiana. Lacking
documentation, we cannot know precisely what new meaning
the figure might have gained. But we have ample evidence that
the practice of creating new meanings was widespread. It has
long been recognised that non-Christian gems in medieval
settings have in many cases undergone a kind of ‘baptism’.29
The transformation is most obvious when stones have been
given new inscriptions, as for example an Imperial amethyst
later engraved with a cross and the Greek ‘O Petros’ (Pl. 15).30
In another case a nearly contemporary description is very
suggestive. Albertus Magnus, writing in the mid-13th century
about a cameo set on the Shrine of the Three Kings at the end
of the 12th century, noted the presence of three heads, one
‘Ethiopian’.31 The stone in fact depicts two people, one wearing
a helmet decorated with a dark, barely visible profile figure.
Albertus probably described the third small face because he
thought of the cameo as an image of the three Magi, one
black.32 The intaglio of Venus and Mars placed above the
‘Gems of Heaven’ | 251
Kornbluth
Plate 15 Amethyst intaglio of Caracalla or Geta with added cross and
inscription; Paris, Cabinet des médailles, Chabouillet 2101
metalwork Adoration on the same reliquary should perhaps be
understood as an abbreviated repetition of that scene, while
the nearby cameo of Nero and Agrippina could be a Christ of
the Last Judgment adored by a standing angel, echoing the
Christ in judgment above.33 The Shrine of the Three Kings is of
course much later than the material under discussion here. The
inscription added to the ‘Petros’ intaglio can be dated only
generally to the Byzantine era. But both objects correspond to a
broad range of Christian interpretations documented by 12th-
and 13th-century seal impressions. Those imprints were made
using matrices that encircled early ring stones with new
metalwork inscriptions. One combination, for example,
surrounded a winged Victory with the inscription ‘ecce mitto
angelum meum’, as on the 1282 seal of Abbot Nicolas of
St-Etienne, Caen.34 While we cannot be sure that this type of
reinterpretation was current in the late 8th or early 9th century
ad, royal and imperial seals suggest that it was. Charlemagne
(d. ad 814) reigned as an emphatically Christian king and
emperor, as did his successors. Yet he and many other
Carolingian rulers sealed their official charters with ‘pagan’
ring stones surrounded by new metalwork inscriptions (Pl.
16).35 While the words do not necessarily identify the lapidary
images, they do imply that the gems had acquired new
meanings either religiously neutral or Christian. The dancing
satyr set upright on the Enger Reliquary (Pl. 12) is not one of
the motifs commonly ‘baptised.’ Nonetheless, it could have
been transformed in much the same way that a three-headed
creature became the Trinity (‘+caput nostru trinitas est’) for
Archbishop Roger de Pont L’Eveque of York (1154–81, document
of 1154–66),36 and perhaps for the Carolingians as well.37
Thoughtful theology might have found the ad hoc possibility
for an interesting new meaning.38 Perhaps the bunch of grapes
in the satyr’s hand suggested a eucharistic significance.
What then of the Berlin reliquary intaglios that were set
upside down, and presumably not given such an additional
layer of Christian signification? Their inherent power as gem
252 | ‘Gems of Heaven’
Plate 16 Wax impression of a seal of Charlemagne, on a charter of 7 August,
ad 807; Munich, Bayerisches Hauptstaatsarchiv, KS3
stones, reinforced by their imagery, was not denied or
destroyed by their settings, but rather channeled. Like the
Teuderigus Reliquary, this object appears to focus on the
salvific power of Christ and the relics of his saints, here with
further references to the recreation of the world. Perhaps lithic
power was also being created anew, channeled for a fresh use.
Reliquaries offer the clearest examples of stones in variant
alignments, simply because they have obvious tops and
bottoms. The orientation of fibulae is more problematic. Most
surviving brooches with engraved stones are late 6th- to late
7th-century ad disc and quatrefoil filigree fibulae from the
Rhineland and around Württemberg.39 Their designs are
symmetrical, often using polished stones along the edges to
form diagonal and orthogonal crosses (whether for religious or
decorative purposes, or both). In one case, a fibula with only a
single stone, we can be quite certain of the intended
orientation. An Italian brooch in the Ashmolean Museum has
preserved three gemstone beads hanging from its lower rim
(Pl. 17).40 If the fibula were worn in any but one direction, the
beads would fall awry and hit the metalwork disc. The cross
that organises the composition must therefore have been worn
vertically.
The front and backplates of these brooches were produced
separately and then joined together. The pin that allowed each
jewel to be fastened to a garment was attached to the back-
plate, but was usually aligned with a front axis.41 Although few
excavations have been sufficiently well documented to allow
certainty, it appears that these fibulae were normally fastened
to clothing with the pin in a horizontal position. Three disc
brooches found in undisturbed 7th-century ad graves at
Iversheim, for example, were oriented in that way.42
When intaglios are used on fibulae, they occupy the
centres. All examples from Italy except the one in the
Ashmolean have unfortunately lost their backplates. Among
the northern brooches, the ambitious survey of Bettina Thieme
has uncovered seven or eight (out of the 220 studied) with

Plate 17 Disc fibula with cameo of helmeted Minerva; Oxford, Ashmolean
Museum, AN1909.816
figural gems, one a cameo and the rest intaglios.43 Of the five
with discernable original figures and surviving backplates,
three are set as one would expect. A quatrefoil fibula from
Heilbronn, for example, presents a carnelian intaglio with an
armed Mars striding to the left.44 Although the fastening pin is
now missing, clear marks on the backplate indicate that it was
placed at a 90° angle from the vertical axis of the gem. If the
brooch was indeed worn with the pin horizontal, then the
figure was displayed with its feet towards the ground. But if a
horizontal pin allowed this gem to be easily read, it set others
at distinctly odd angles. A quatrefoil brooch from Kobern has in
its centre another Mars (Pl. 18).45 The vertical axis of this figure
lies along the same line as the pin, so the fibula was most
probably worn with the figure lying on its side, the face looking
up or down. Another example probably from Bad Hönningen
Plate 18 Quatrefoil fibula from Kobern with intaglio of Mars; Bonn,
Rheinisches Landesmuseum, 3068
Roman Intaglios Oddly Set
holds a gem set roughly parallel to the back pin, suggesting that
the brooch was worn with the engraved standing goddess lying
nearly flat (Pl. 19).46 Even then, though, the figure was not
quite aligned: it does not exactly follow the axis of the fibula,
but has its head angled to the left of one large border setting
and its feet tilting to the right of the opposite (former) stone.
Even if the wearer of this fibula chose to insert the pin
vertically, and oriented the brooch according to its other major
axis, she still necessarily wore the gem at an angle.
Other fibulae not directly related to the Goldscheibenfibeln
also set intaglios in decidedly odd ways. A brooch possibly from
a grave in or near the church of St Severin in Cologne displays a
carnelian intaglio of Apollo. Remarkably, the almost fully
transparent stone has been set with its image turned inward to
the gold mount.47 Given the unusual clarity of the material, the
setting both reveals and conceals. It should probably be classed
with the fibulae under discussion here, preserving but veiling
its figure. Its image may also have been worn off-axis. The
hinge for the brooch’s pin is located at the figure’s feet, so if the
pin were inserted into cloth from the top down (a sensible
arrangement to avoid loss), the figure would have been
displayed standing on its head. If the fibula were worn
horizontally, its figure would of course have lain on its side.
Another fibula found in Kerć (Crimea), outside our immediate
geographical area but probably from the first half of the 5th
century ad, likewise set its translucent intaglio with the image
turned inward, as does a medallion from Georgia, and (later)
an Ottonian fibula.48
A single early 6th-century ad great square-headed brooch
from south-eastern England is set with a carnelian with Cupid
milking a goat (Alveston Manor grave 5).49 This bow fibula is
the only one known to me that includes an engraved gem, and
could belong to the group discussed here. Its intaglio is set
perpendicular to the long axis and attachment pin, and it may
well have been worn with the gem on its side, perceptibly
tilted, or upside down.50 It appears that few Early Medieval
objects from England incorporated earlier engraved gems, but
a 5th- or 6th-century ad belt buckle attachment-plate from
Lyminge51 presents an interesting parallel to the brooches. In
its centre it displays a jasper intaglio with a standing Ceres. In
Plate 19 Disc fibula from Bad Hönningen with glass intaglio of Victory; Bonn,
Rheinisches Landesmuseum, 35.54
‘Gems of Heaven’ | 253
Kornbluth
Plate 20 Pendant with nicolo agate intaglio from Kirchheim am Ries grave 326,
mid-7th century AD; Stuttgart, Württembergisches Landesmuseum, F 70/269
326 (22)
order to function properly the plate had to be oriented with the
stone’s figure lying on its side, and so it was found when
excavated at the waist of the figure in grave 32.
Pendants can also demonstrate that not all medieval users
were either simply struck with admiration for the beautiful
images on ring stones or oblivious to them. When Roman
artisans set cameos and intaglios as pendants, they normally
mounted them to be easily legible. Early Medieval
metalworkers in the north, on the other hand, appear to have
been happy to upset or partially hide intaglio figures. Two
engraved gems found in a mid-7th-century burial in Kirchheim
am Ries were set in pendants that made the figures lie
sideways, one on its face and one on its back (Pls 20 and 21).52
An intaglio lion from Anglo-Saxon Sibertswold, from a grave
coin-dated after c. ad 650, was mounted in its pendant head
down and tail up.53 A white glass intaglio found in a 5th- to
7th-century ad cemetery in Marchélepot was likewise
suspended from one side (Pl. 22).54 The gem appears to portray
Anubis behind a lion that tramples a body, facing another
figure in profile, a scene not uncommon on ‘Abrasax’ gems.55 It
must always have been somewhat puzzling, and perhaps the
more powerful for that obscurity. Turned on its side, its
strength may have been tamed.
Plate 22 Pendant with intaglio from Marchélepot; formerly Berlin, Museum für
Vor- und Frühgeschichte, Va 5139
254 | ‘Gems of Heaven’
Plate 21 Pendant with carnelian intaglio from Kirchheim am Ries grave 326,
mid-7th century AD; Stuttgart, Württembergisches Landesmuseum, F70/269
326 (4)
Perhaps the most remarkable object, however, is one whose
image is not in itself at all obscure. A pendant found in an
unstratified context in the Merovingian cemetery in Dettingen
an der Erms contains a large rock crystal intaglio of an eagle
with spread wings (Pls 23 and 24, cf. Pl. 3).56 The stone has
suffered major fire damage since its first publication, but
fortunately many details were already recorded. Its image was
similar to eagles on numerous Roman objects, from small
intaglios to the column base of Antoninus in Rome itself, and
can be related to eagle fibulae from Spain and Italy. This highly
recognisable image was turned on its head, mounted upside
down in four encircling bronze straps that were bound together
at the top (the eagle’s feet). One might argue that it was meant
to be seen by a wearer who turned the image right-side-up
when she lifted it on its suspension strap or chain; but even
flipped over, the supporting metal bands partially obscured the
bird. Can we assume, then, that the figure was meaningless to
the stone’s Early Medieval user? This seems unlikely in any
society that had contact with the Roman Empire. The official
symbol of the Roman legions57 might well have been regarded
as an image powerful in its own right. It would have been
reasonable for a later owner to retain that power, neither
destroying nor completely hiding the image, but rather taming
it.
Intaglios are rare in bound pendants. One was buried in
Frénouville,58 but its original orientation can no longer be
determined. Another was recently discovered in an early- to
mid-6th century ad burial in Kent.59 The stone retains its
Plate 23 Bound pendant from Dettingen an der Erms, crystal intaglio in silver,
drawing made before damage; object Stuttgart, Württembergisches
Landesmuseum, Dettingen/E 29/154

Plate 24 Plaster impression of eagle intaglio in bound pendant from Dettingen
an der Erms, made before damage; Stuttgart, Württembergisches Landes-
museum
supporting bands, which partially hide the engraved figure of
Omphale in the same manner as the Dettingen pendant.
Perhaps here too, the energy of a figural gemstone was being
harnessed for use.
The redirection of power that I propose lay behind the
upset mounting of Roman engraved gems was certainly not the
norm– there are many more intaglios set upright than on their
heads or sides– but it does appear to have been a real
phenomenon. It may explain some puzzling features of
medieval gem usage. It would also be profoundly compatible
with concepts and practices that we know characterised the
Late Antique/Early Medieval period. Political and religious
potency were harnessed when Roman and Islamic gemstone
vessels, transformed into huge cabochons on the oversized
crux gemmata of the Ambo of Henry II in Aachen, lent their
semiotic power to the Ottonian patron as well as to the
Christian liturgy.60 The most familiar examples of transformed
and channeled power are indeed from thoroughly Christian
contexts, though we cannot assume that the phenomenon was
limited to that environment. The Gospels of John buried with
St Cuthbert (d. ad 685) and later recovered from his
sarcophagus61 did not cease to be a biblical manuscript, but
gained new importance as a relic of the saint. And of course
Cuthbert’s and any saint’s primary relics, his or her bones, had
undergone a transmutation even more radical. When a person
achieved sanctity those bones ceased to be merely shaped clay.
They became vessels able to carry divine energy, both as they
continued to support life and long after they ceased to do so.
Bones as relics are perhaps the most apposite comparison here:
taken from their original context within the body, they
continued to be recognisably bones, but their new sanctified
functions became paramount. The living power of the body
became spiritual power. Just so gem spolia may still have been
recognisable, and recognised, as earlier engravings, but the
strength of their materials and imagery was turned in new
directions.
Notes
1 W.S. Heckscher, ‘Relics of pagan antiquity in medieval settings’,
Journal of the Warburg Institute 1 (1938), 204–20; A. Krug, ‘Antike
Gemmen und das Zeitalter Bernwards’, in M. Brandt and A.
Eggebrecht (eds), Bernward von Hildesheim und das Zeitalter der
Ottonen (exh. cat., Hildesheim), Hildesheim, 1993, vol. 1, 161–72.
2 R. Halleux and J. Schamp (eds and trans.), Les Lapidaires Grecs,
Paris, 1985.
Roman Intaglios Oddly Set
3 G. Kornbluth, ‘No. 1: Intaglio with standing Christ’, in A. Kirin
(ed.), Sacred art, secular context: objects of art from the Byzantine
Collection of Dumbarton Oaks, Washington, D.C. (exh. cat., Athens,
Ga.), Athens, Ga., 2005, 57–8.
4 G. Kornbluth, Amulets, Power, and Identity in Early Medieval
Europe, forthcoming.
5 N. Wibiral, ‘Augustus patrem figurat: Zu den Betrachtungsweisen
des Zentralsteines am Lotharkreuz im Domschatz zu Aachen’,
Aachener Kunstblätter 60 (1994), 105–30.
6 R. Wesenberg, ‘Das Herimannkreuz’, in A. Legner (ed.), Rhein und
Maas: Kunst und Kultur 800–1400 (exh. cat., Cologne), Cologne,
1973, vol. 2, 167–76; U. Bracker-Wester, ‘Christuskopf vom
Herimannkreuz ein Bildnis der Kaiserin Livia’, ibid., 177–80.
7 D.v. Reitzenstein, Privatreliquiare des frühen Mittelalters (Kleine
Schriften aus dem Vorgeschichtlichen Seminar der Philipps-
Universität Marburg 35), Marburg, 1991; J. Braun, Die Reliquiare
des christlichen Kultes und ihre Entwicklung, Freiburg im Breisgau,
1940, 198–205.
8 Others: M.-M. Gauthier, ‘Le Trésor de Conques’, in Rouergue
Roman (la nuit des temps 17), La Pierre-qui-Vire, 1968, 98–145, at
138–9, 143–4, pls 44–8, 60; E. Poletti Ecclesia, ‘“L’incanto delle
pietre multicolouri”: gemme antiche sui reliquiari altomedievali’,
in G.S. Chiesa et al. (ed.), Gemme dalla corte imperiale alla corte
celeste, Milan, 2002, 55–74.
9 Krug (n. 1), 165–6.
10 G. Haseloff, ‘Der Abtsstab des heiligen Germanus zu Delsberg
(Delémont)’, Germania 33 (1955), 210–35, at 227–35; D. Thurre, ‘Les
trésors ecclésiastiques du haut Moyen Âge et leur constitution.
Éclairage à travers deux exemples helvétiques: Saint-Maurice
d’Agaune et Sion’, in J.-P. Caillet (ed.), Les Trésors de sanctuaires, de
l’Antiquité à l’époque romane (Centre de recherches sur l’Antiquité
tardive et le haut Moyen Age 7), Nanterre, 1996, 43–81, at 52–4.
11 C. Jörg, Die Inschriften des Kantons Wallis bis 1300 (Corpus
Inscriptionum Medii Aevi Helvetiae), Freiburg, 1977, 89–91.
12 J. Mitchell, ‘Literacy displayed: the use of inscriptions at the
monastery of San Vincenzo al Volturno in the early 9th century’, in
R. McKitterick (ed.), The Uses of Literacy in Early Medieval Europe,
Cambridge, 1990, 186–225, at 198–9.
13 A. Krug, ‘Antike Gemmen am Armreliquiar des hl. Blasius in
Braunschweig’, in J. Ehlers and D. Kötzsche (eds), Der Welfenschatz
und sein Umkreis, Mainz, 1998, 93–109, at 107–9.
14 E. Sauer, The Archaeology of Religious Hatred in the Roman and
Early Medieval World, Charleston, SC, 2003.
15 A. Woodward and P. Leach, The Uley Shrines. Excavations of a
Ritual Complex on West Hill, Uley, Gloucestershire: 1977–9 (English
Heritage Archaeological Report 17), London, 1993, 316–21.
16 Gregory of Tours, Libri Historiarum X (eds B. Krusch and W.
Levison, Monumenta Germaniae historica, Scriptorum rerum
Merovingicarum 1 [1]), Hanover, 1951, Bk 8, Ch.15, 382; Gregory of
Tours, The History of the Franks, trans. L. Thorpe, Harmonds-
worth, 1974, 446.
17 L.R. Cielo, Monumenti Romanici a S. Agata dei Goti: il Duomo e la
Chiesa di San Menna, Rome, 1980.
18 M. Greenhalgh, The Survival of Roman Antiquities in the Middle
Ages, London, 1989.
19 G. De Angelis d'Ossat, ‘L'architettura del S. Giorgio di Valpolicella:
una chiesa castrense’, in Verona in Età Gotica e Longobarda,
Convegno del 67 dicembre 1980, Atti, Verona, 1982, 149–84, at 165,
fig. 13.
20 R.P. Wright, ‘A fourth-century Roman inscription from the Saxon
church at Escomb, County Durham’, Archaeologia aeliana, ser. 4,
48 (1970), 45–9.
21 T. Eaton, Plundering the Past: Roman Stonework in Medieval
Britain, Stroud, 2000, 67–75, 90–1.
22 R.G. Collingwood and R.P. Wright, The Roman Inscriptions of
Britain 1: Inscriptions on Stone, Oxford, 1965, no. 1122.
23 P. Lasko, ‘The Enger Cross’, in M. Gosebruch and F.N. Steigerwald
(eds), Helmarshausen und das Evangeliar Heinrichs des Löwen,
Göttingen, 1992, 79–108.
24 G. Haseloff, Email im Frühen Mittelalter: frühchristliche Kunst von
der Spätantike bis zu den Karolingern, Marburg, 1990, 85–91, but
see: D. Buckton, ‘Byzantine enamel and the West’, Byzantinische
Forschungen 13 (1988), 235–44; E. Wamers, ‘VII/17 Engerer
Bursenreliquiar’, in J. Fried et al. (ed.), 794 Karl der Grosse in
Frankfurt am Main: Ein König bei der Arbeit (exh. cat., Frankfurt),
Sigmaringen, 1994, 155–7.
‘Gems of Heaven’ | 255
Kornbluth
25 A. Krug, ‘Antike Gemmen an mittelalterlichen Goldschmiede-
arbeiten im Kunstgewerbemuseum Berlin’, Jahrbuch der Berliner
Museen NS 37 (1995), 103–19.
26 V.H. Elbern, ‘Ein fränkisches Reliquiarfragment in Oviedo, die
Engerer Burse in Berlin und ihr Umkreis’, Madrider Mitteilungen 2
(1961), 183–204; idem, ‘Das Engerer Bursenreliquiar und die
Zierkunst des frühen Mittlealters’, Niederdeutsche Beiträge zur
Kunstgeschichte 10 (1971), 41–102.
27 Krug (n. 25), 115.
28 Krug (n. 1), 168, fig. 53; Krug (n. 25), 114, fig. 15; D. Kötzsche, Der
Welfenschatz im Berliner Kunstgewerbemuseum, Berlin, 1973, no. 9.
29 E.g. G.A.S. Snijder, ‘Antique and mediaeval gems on bookcovers at
Utrecht’, The Art Bulletin 14 (1932), 5–52.
30 J. Durand and M.P. Laffitte (eds), Le trésor de la Sainte-Chapelle
(exh. cat., Paris), Paris, 2001, no. 36.
31 Mineralia 2,3: Albertus Magnus: Book of Minerals (trans.
D. Wyckoff), Oxford, 1967, 131.
32 J. Hoster, ‘Der Wiener Ptolemäer-kameo einst am Kölner
Dreikönigenschrein’, in F. Dettweiler, H. Köllner and P.A. Riedl
(eds), Studien zur Buchmalerei und Goldschmiedekunst des
Mittelalters: Festschrift für Karl Hermann Usener zum 60.
Geburtstag am 19. August 1965, Marburg a.d. Lahn, 1967, 55–64.
33 E. Zwierlein-Diehl, ‘Interpretatio christiana’: gems on the Shrine
of the Three Kings in Cologne’, in C.M. Brown (ed.) Engraved Gems:
Survivals and Revivals (Studies in the History of Art 54),
Washington DC, 1997, 62–83.
34 J. Cherry, ‘Antiquity misunderstood’, in M. Henig and D. Plantzos
(eds), Classicism to Neo-classicism: essays dedicated to Gertrud
Seidmann (BAR International Series 793), Oxford, 1999, 143–7.
35 M. Dalas, Corpus des Sceaux français du Moyen Âge, 2: Les Sceaux
des Rois et de Régence, Paris, 1991.
36 J.P. Dalton, The Archiepiscopal and Deputed Seals of York 1114–1500,
York, 1992, 48.
37 E. Zwierlein-Diehl, Antike Gemmen und ihr Nachleben, Berlin,
2007, 288–9.
38 Cf. Krug (n. 1), 167; D. Kinney, ‘The horse, the king and the cuckoo:
medieval narrations of the statue of Marcus Aurelius’, Word and
Image 18 (2002), 372–98.
39 B. Thieme, ‘Filigranscheibenfibeln der Merowingerzeit aus
Deutschland’, Bericht der Römisch-Germanischen Kommission 59
(1978), 381–500 and plates; F. Rademacher, Fränkische
Goldscheibenfibeln aus dem Rheinischen Landesmuseum in Bonn,
Munich, 1940.
40 A. MacGregor et al., Ashmolean Museum Oxford. A Summary
Catalogue of the Continental Archaeological Collections (Roman
Iron Age, Migration Period, Early Medieval) (BAR International
Series 674), Oxford, 1997, no. 97.1.
41 E.g. W. Janssen, ‘Die Goldblechscheibenfibel aus Grab 42 des
älteren fränkischen Gräberfeldes unter dem Dom St Viktor zu
Xanten’, Studien zur Sachsenforschung, Herrn Dr. Albert Genrich
zum 65. Geburtstag gewidmet, Hildesheim, 1977, 235–71, fig. 7.
42 C. Neuffer-Müller, Das fränkische Gräberfeld von Iversheim Kreis
Euskirchen (Germanische Denkmäler der Völkerwanderungszeit
Serie B, Die fränkischen Altertümer des Rheinlandes 6), Berlin,
1972, 14–7, graves 53, 94, 142; Thieme (n. 39), nos 78, 79, 81.
256 | ‘Gems of Heaven’
43 Thieme (n. 39), 397: nos 68, 70, 88, 92, 120, 155, 188, 192 [gem
probably added later]; also G. Graenert, Merowingerzeitliche
Filigranscheibenfibeln westlich des Rheins, Montagnac, 2007, 237–
8, no. III/31d.
44 Thieme (n. 39), no. 68; Graenert (n. 43), 291, no. V/27b.
45 Thieme (n. 39), no. 92; Graenert (n. 43), 199–200, no. II/41a.
46 Thieme (n. 39), no. 70; Graenert (n. 43), 194, no. II/32.
47 A. Krug, Antike Gemmen im Römisch-Germanischen Museum Köln
(Wissenschaftliche Kataloge des Römisch-Germanischen
Museums Köln 4), Frankfurt am Main, 1981, no. 82.
48 I.G. Damm, ‘Goldschmiedearbeiten der Völkerwanderungszeit aus
dem nördlichen Schwarzmeergebiet: Katalog der Sammlung
Diergardt 2’, Kölner Jahrbuch für Vor und Frühgeschichte 21 (1988),
no. 18; H. Westermann-Angerhausen, ‘Eine unbekannte Fibel aus
dem ottonischen Kaiserinnenschmuck?’, Mainzer Zeitschrift 70
(1975), 67–71, pl. 20a.
49 J. Hines, A New Corpus of Anglo-Saxon Great Square-Headed
Brooches, London, 1997, 319; M. Henig, A Corpus of Roman
Engraved Gemstones from British Sites (BAR British Series 8, 2nd
edn), Oxford, 1978, no. 140.
50 Cf. Hines (n. 49), figs 121–9, burial positions of fibulae.
51 A. Warhurst, ‘The Jutish cemetery at Lyminge’, Archaeologia
Cantiana 69 (1955), 1–40, at 24, 33–5, pl. 9.3; Henig (n. 49), no. 264.
52 Grave 326: C. Neuffer-Müller, Der alamannische Adelsbestattungs-
platz und die Reihengräberfriedhöfe von Kirchheim am Ries
(Ostalbkreis) (Forschungen und Berichte zur Vor- und
Frühgeschichte in Baden-Württemberg 15), Stuttgart, 1983, 172–4.
53 Grave 172: A. Richardson, The Anglo-Saxon Cemeteries of Kent
(BAR British series 391), Oxford, 2005, vol. 2, 71, 343, fig. 5.
54 C. Boulanger, Le Cimetière Franco-Mérovingien et Carolingien de
Marchélepot (Somme): Étude sur l’Origine de l’Art Barbare, Paris,
1909, 104.
55 Cf. A. Delatte and Ph. Derchain, Les Intailles Magiques Gréco-
Égyptiennes, Paris, 1964, nos 120–4.
56 W. Veeck, Die Alamannen in Württemberg (Germanische Denk-
mäler der Völkerwanderungszeit 1), Berlin, 1931, 52, pl. G8a–c.
57 A. von Domaszewski, ‘Der Legionsadler’, in Paulys Real-
Encyclopädie der Classischen Altertumswissenschaft, ed.
G. Wissowa, Stuttgart, 2 (1896), 317–8.
58 C. Pilet, La Nécropole de Frénouville: Étude d’une population de la fin
du IIIe à la fin du VIIe siècle (BAR International Series 83), I–III,
Oxford, 1980, vol. II, 301–3, pl. 157 (8, 9).
59 N. Adams, ‘The rock crystal pendant from Grave C’, in K. Parfitt
and T. Anderson et al., Buckland Anglo-Saxon Cemetery, Dover:
Excavations 1994 (Canterbury Archaeological Trust, British
Museum), in press. I am grateful to Barrie Cook for showing me an
early version of this paper.
60 I.H. Forsyth, ‘Art with History: The Role of Spolia in the
Cumulative Work of Art’, in C. Moss and K. Kiefer (eds), Byzantine
East, Latin West: art-historical studies in honor of Kurt Weitzmann,
Princeton, 1995, 153–62, at 156–7.
61 R.A.B. Mynors and R. Powell, ‘The Stonyhurst Gospel’, in C.F.
Battiscombe (ed.), The Relics of Saint Cuthbert, Oxford, 1956, 356–
74.
The Re-use and Re-interpretation of Gemstones in
Medieval Hungary
Tamás Gesztelyi
Re-use
By the year 1000 the confederation of Magyar tribes which had
arrived in the Carpathian basin at the end of the 9th century ad
had transformed itself into a kingdom along European lines
under the leadership of King St Stephen. It is possible to trace
the re-use of antique gems at that time in different parts of the
country. In east Hungary, south of the city of Debrecen, a silver
pendant, similar in shape to a Hercules club, with rich
Byzantine-style decoration, was found in a 10th-century
(Conquest Period in Hungary) cemetery.1 Thus the re-use of the
jasper gem, placed in the bottom of the pendant, could have
happened on Byzantine territory and passed into Magyar
ownership as a gift or booty. The scene on the obverse of the
gem is well known and belongs to the mask/animal type
combination popular from the time of the early Empire
onwards. The most likely interpretation of the richly detailed
iconography is that its function was to bring good luck, wealth
and security with the help of magical imagery. The function of
the re-used gem as an amulet is also supported by the fact that
it was placed on the breast of a one-year old child.
In the western part of the country, east of Bratislava (now
Slovakia), and also from a 10th-century Magyar cemetery, a
gem depicting Faustina minor, the wife of Marcus Aurelius, set
in an adult sized silver alloy (billon) ring came to light (Pl. 1).2
The grave itself had been destroyed during sand mining and
hence it is impossible to tell anything for sure about the origin
of this extraordinary (in size and workmanship) gem.
However, it is possible to state that its find spot may reflect its
original place of loss or manufacture in Pannonia, or even
modern day Slovakia, because during the Marcomannic Wars
Marcus Aurelius spent considerable time in the region. The
original owner of the gem could have been the emperor
himself, or somebody from his close circle. The high quality
workmanship could be that of an artisan close to the imperial
court, or even working in the court. On the basis of the form of
Plate 1 Billon ring set with a gem depicting Faustina minor, 10th century, from
Szered (modern south-west Slovakia)
the ring, one could look for the re-user of the gem amongst the
Magyars.
After that period there is no information regarding the
re-use of antique gems for nearly 200 years. This can probably
be explained by the lack of written and archaeological
evidence, rather than by the disappearance of the gems
themselves. There is no data about liturgical objects decorated
with gems, about jewellery set with gems or about the usage of
gem seals in this period.
From the end of the 12th century, a little bit later than in
western Europe, the practice of placing antique gems into rings
and their use as seals started to spread in Hungary as well. In
the second half of the 12th century and the first half of the 13th
century, during the reigns of certain kings (notably István III,
Béla III and Imre) the seal rings of the rulers were impressed
onto the front side of the document seal as a control seal, not as
a counter sealing on the reverse side of the seal of the
document. In the first half of the 13th century the practice of
sealing started to spread amongst the laity as well, but this
has hardly left any traces in the archives and written sources,
because these seals (made with seal rings) were less formal, and
hence less authentic. Thus they were only suitable for the sealing
of private documents and letters, which disappeared without any
traces together with the family archives of the lower rank nobility.3
From the 13th and 14th centuries we have some rings which
have antique gems in their mounts. From their inscriptions one
can guess their function. We only know from descriptions and
drawings about a gold ring from the former Andrássy
collection,4 which, according to its inscription, was the
property of Sándor, bishop of Várad (1219–30; Pl. 2). The ring
held a very good quality sapphire gem engraved with a portrait
which was probably that of the wife or daughter of Constantine
the Great (r. ad 306–37). The bishop accompanied András II,
King of Hungary, to the Holy Land, so it is possible that he got
the gem as a present or purchased it during that trip.
Plate 2 Gold ring of Sándor, bishop of Varad (1219–30)
‘Gems of Heaven’ | 257
Gesztelyi
In the medieval collection of the Hungarian National
Museum are six more rings with gems and inscriptions:
1. A gem depicting a jumping lion, inscribed: s(igi) ll(um)
secreti.5 These smaller, private seals were used on documents
of no great importance, and also in letter-writing as
independent seals or as counter seals on the reverse of one-
sided seals. The ‘secret seal’ was probably used for the closing
of charters; later its function was probably expanded and it was
used for general sealing.
2. A gem depicting a combination of Pan and a satyr mask,
inscribed: ave maria vhis.6 A possible reading of the last word
is: v(irgo) h(umil)is. The humilitas of Mary is known from the
gospel of Luke (1:48), in the Visitation scene, when Mary
answers to Elisabeth’s greeting. The phrase humilis often
occurs in medieval hymns to Mary; moreover the humilitas of
Mary has its own liturgical feast on the 17th of July: Humilitas
Beatae Mariae Virginis. Another possible reading occurs if we
read n instead of the letter h (which is not very clear) – v(e)
n(erabil)is. The word venerabilis occurs several times in the
liturgy, in the Graduale, in the following form: benedicta et
venerabilis es.
3. An Etruscan scarab with a depiction of a centaur raising
one of its legs.7 There are inscriptions on the both sides of the
mount: on the reverse: ave maria gra; and on the front
defining its function: + s(igillum) ianee co(mi)t(i)s. The
inscription on the back was not used for sealing, because of the
bulging back of the scarab; its role was probably as a prayer.
There is a similar inscription on a seal of Thomas Lovel of
Chesterton, on which the text is given in full: ave maria gratia
plena.8
4. A gem depicting Sol in a quadriga, inscribed: + sigillum
comitis.9 Its worn surface suggests long usage. One cannot
determine the name of the person who used the seal; however,
from his title, comes, it can be deduced that he was a member of
the aristocracy of the Arpadian period (ad 900–1301).
5. A gem depicting Triton, inscribed: s.d.ladislavi.f.d.b.
bosinas (Sigillum Domini Ladislavi Filii Domini (?) Bani
Bosinae).10 Thus this could be the seal of Ladislaus, the ban of
Bosnia, although we do not have any other data about him.
6. A gem depicting a hound chasing a deer, inscribed:
s. nicolai.f.ionni (Sigillum Nicolai Filii Johannis), i.e. of
Nicolaus, son of Johannes.11 The people behind the names are
unknown.
From the 15th and 16th centuries we have several antique
gems re-used in rings, but they are uninscribed. Hence we
know nothing about their owners or their function. However,
from the medieval imprints of the seals, it is clear that the
majority of gem seals were without inscriptions.
In medieval Hungary the most important sources for the
re-use of antique gems are charters. Here gems were used as
seals, in accordance with their original function. It is obvious
that this practice was widespread in every region of western
Europe, yet little research has been done in this field. In recent
years German and Italian scholarship has mostly dealt with the
analysis of antique gems on liturgical objects, royal insignia
and jewellery. This is understandable as these objects are
representative and are the classical objects of contemporary
applied arts; moreover they are the precious historical
mementos of a country or a region. There could also be another
reason behind the lack of scholarly interest, mentioned by
258 | ‘Gems of Heaven’
Martin Henig in his catalogue of gems found in Britain, which
includes a whole chapter on the re-use of Roman gems in the
Middle Ages. In this chapter there are 34 gems; out of these 31
were in rings and only three were from the seal of a charter.
Henig gave the following explanation: ‘a complete corpus
would entail long research in medieval archives’.12
Thus the research of this group of objects is challenging.
The research of charters is controlled by strict regulations and
to find more gem sealings it is necessary to check hundreds of
charters continuously. This is not possible – quite
understandably – in any of the archives. A long term solution
could be the digitalising of the medieval material of the
archives. During that process it would be important to make
good quality photos of the sealings as well. This work has been
started in the State Archives of Hungary, but I have not been
able to use its results as yet. My research was only possible
because there are two collections of seal copies which were
produced in the 1960s in order to help the historical and artistic
research on medieval seals. On the basis of these collections,
I have managed to collect 114 antique gem seals. The upper
time limit was the year 1526 because in Hungary charters
produced before that date are considered medieval. This
cut-off date corresponds to the battle of Mohács, when the
country suffered a catastrophic defeat at the hands of the
Ottoman Empire. As a result of this, the medieval kingdom of
Hungary fell apart. After that time in the remaining area of the
kingdom documents took the place of charters.
A Hungarian scholar (Dezső Csánki) at the end of the 19th
century thought that the extraordinary popularity of gem seals
was a Hungarian speciality.13 This implied that antique
depictions were used as family coat-of-arms. This presumption
could only be justified if we had a clear picture about the usage
and spatial distribution of gem seals in other European
countries as well. Csánki’s other presumption that ‘a particular
lord or nobleman, of course used the same gem every time’ is
not correct. I have found two exceptions. The first is Lőrinc
Újlaki, who in the 1490s used a Bonus Eventus representation,14
but in 1517 used a gem depicting a lion.15 The other was István
Werbőczi, who in 1517 used a full-length figure of Hercules (Pl.
3), but in 1524 only used a bust of Hercules as a seal (Pl. 4).16
The first aspect of my evaluation of the collected gem seals
was the usual approach taken to glyptic research: the grouping
of the gems according to their date and subject matter. The
result of that undertaking can be summarised as follows:
between the 4th and 2nd centuries bc, five gems; between the
1st century bc and the middle of the 1st century ad, c. 30 gems;
between the middle of the 1st century ad and the beginning of
the 3rd century ad, c. 76 gems; from the 4th and 5th centuries
Plate 3 Sealing of Werbőczi, 1517 Plate 4 Sealingof Werbőczi, 1524
(impression)
 The Re-use and Re-interpretation of Gemstones in Medieval Hungary
ad (Sasanid), three gems could be identified.17 Gems belonging
to the first and last groups are not known from Roman
Pannonia, but represent types which probably occurred in
Dalmatia. The second group is also not representative of
Roman Pannonia. The third group is where most of the
Pannonian and Dacian gem finds belong. Thus in the case of
medieval gem seals, we are referring to finds which did not
come from the territory of medieval Hungary.
If we summarise the spatial distribution of the usage of
medieval gem seals, we can conclude that after the capital
Buda they were used in the largest numbers in the south-
western regions of the country (e.g. Pl. 5).18 This is understand-
able for several reasons. Firstly, – surely as a result of the
impact of Italy – this is the area where the medieval usage of
seals spread for the first time. Secondly, the south-western
regions of the country, i.e. Dalmatia, was the richest area in
gem finds. In ancient towns along the coast it was even possible
to find Greek gems, but Italian ones also commonly reached
the area. In addition, in the Imperial period, we should account
for important production and trading centres here. The
subjects also suggest the Dalmatian origin of several gems.
While in Pannonia the most popular representations were
Jupiter, Mercury, Victory and the eagle,19 in the medieval
material Amor, satyrs and Minerva are the most common.20
The Amor and satyr representations are more typical of Roman
Dalmatia than Pannonia.21
Apart from the south-western areas, in Transdanubia
(Pannonia) and in Transylvania (Dacia) the use of gem seals
was not widespread. However local finds contribute evidence
for the medieval usage of gems. In those regions of the country
where there was no Roman presence, gem seals are rare, apart
from the north-eastern counties: Sáros, Ung and Bereg.22 This
latter phenomenon could be explained by the fact that from the
Adriatic coast through Buda towards Poland an important
trade route (the so-called ‘Wine Road’) passed through this
area. Supposedly, thanks to active trade, antique gems reached
this region in larger numbers than other regions.23
If we investigate the chronological sequence of the usage of
medieval gem seals in medieval Hungary, then the following
picture is revealed.24 At the end of the 12th century one seal is
recorded in royal use. In the 13th century eight gems occurred,
of which four were at the disposal of the king, one was in use by
Plate 5 Charter of 1475 from Slavonia with
six medieval gem seals (south-west
Hungary, modern Croatia)
a lord, one was in the use of a lord (who in a later sense was a
familiaris), one belonged to a yeoman and one to an
ecclesiastical institution (loca credibilia). From the 14th century
we know of six gem seals: one belonged to the office-holder of a
noble hall, two to county office-holders, one to an ecclesiastical
institution, one to a town citizen and one to a foreign church
dignitary. Within these finds nearly all social strata are
represented, all of which were involved in the production of
charters. In the first three quarters of the 15th century, there
are a similar amount of gem seals in the collection, but the
ecclesiastical middle class becomes dominant. This is the
period when private individuals appear amongst the users of
seals: a town citizen and a nobleman. In the last quarter of the
15th century and the first quarter of the 16th there is a steep
growth in the number of gem seals compared with previous
periods. In both quarters the number of gem seals exceeded 40
gems. Amongst the owners, noblemen and lords became
dominant: the rest of the users were churchmen, mostly
members of the middle class, like prebends, but there are some
simple parish priests and bishops as well.
Re-interpretation
For the medieval re-interpretation of antique depictions
expressis verbis we have only one piece of data. The explanation
for this is that we hardly have any descriptions from that period
regarding the imagery of seal rings. In the inscriptions of the
rings there is not a single reference concerning the depicted
scene. The only example is from the turn of the 15th and 16th
century. In the coat-of-arms of the Hunyadi family the raven,
which is holding a ring in its beak (Pl. 6), is very similar to the
eagle, which holds the wreath of victory in its beak.25 The latter
Plate 6 Coat-of-arms of the Hunyadi
family (alias Corvinus)
‘Gems of Heaven’ | 259
Gesztelyi
Plate 7 Sealing of John Corvinus, Plate 8 Sealing of bishop Demeter,
1503 1374
was common on both the imperial coinage and gems. Antonio
Bonfini, who worked in the court of King Matthias Corvinus
(born Hunyadi, r. 1458–90), used these depictions as material
proof for the king of the Roman ancestry of the Hunyadi
family.26 Matthias’ affection for Roman antiquities is indicated
by the fact that Beatrix, his wife, wanted to buy him the coin
and cameo collection of Fr. Gonzaga, the cardinal of Mantua,
which was pawned by Lorenzo Medici.27 I have not been able to
find any proof for the use of gem seals by Matthias, but in the
case of his son, John Corvinus, I have. In the City Archives of
Debrecen there are three charters in which he confirmed the
privileges of the city. On these charters there are three
different versions of his ring seals. On the first the eagle was
depicted in profile, turning its head backwards, with a wreath
in its beak and a palm branch in front of it (Pl. 7. 28 This is surely
an antique gem. The second version is similar, but the eagle
seems to stand on a column and its figure is slimmer than the
previous example.29 This piece could be made after an antique
model, but the bird is closer to a raven. The third is the well-
known coat-of-arms of the Hunyadi family, a raven holding a
ring in its beak, looking forward and with a leafy branch at its
feet (Pl. 6).30 The branch likely derives from an antique
exemplar as well. The date of the charters does not suggest that
the seals were successively used, but rather that they were in
use at the same time.
Even if it not expressed directly, the meaning assigned by
medieval users to certain antique depictions can be determined
indirectly. As an example I can cite Demeter, Bishop of
Transylvania, who later became the Archbishop of Esztergom,
and a cardinal. On his seal an armed man steps on a figure,
who is lying on the ground (Pl. 8).31 On the basis of the spear in
his hand, it is possible to deduce that the soldier wants to stab
the unarmed man. In the modern literature we can find
descriptions which match that interpretation of the scene on
the seal: ‘it is depicting a figure, who has been floored by an
armed man, with an arrow and a spear in his hand...’; and also:
a winged figure (St Michael), stepping to the right and trampling
a blurred figure and at the same time stabbing him with a spear.32
It is likely that the scene was interpreted similarly by the bishop
himself, because on another seal of his there really was a St
Michael, but in a different composition: the archangel is
standing on the back of a dragon, girdled with his wings, and
the archangel has a falchion in his right hand.33 But the
dynamic scene of the antique gem depicts a totally different
story. We know this scene from several depictions which are
preserved in good condition. The figure on the ground is Rhea
260 | ‘Gems of Heaven’
Plate 9 Sealing with Mars and Rhea Plate 10 Administrative seal of the
Silvia chapter of Eger, 1256
Silvia, the naked man who hovers above her, is the armed Mars
(Pl. 9).34 Behind him we can see his cloak, not his wings.
Similarly it is obvious that the figure of the winged Victory
for a Christian was nothing other than an angel. We have a
direct proof for this from the end of the 13th century from
France. On a gem of the abbot of Caen which depicts Victoria,
there is an inscription: ecce mitto angelvm mevm.35 The
winged Nemesis could also be identified with an angel. This
occurs on the earliest ecclesiastical gem sealing from Hungary,
the administrative seal of the chapter of Eger.36 Its description
in the literature is as follows:
On the seal there is an eagle, extending his wings at both sides, it is
a symbol of John, the evangelist … On the imprint from 1256
between the wing and body of the eagle there is an administrative
seal, which as far as it can be judged by the imprint, was an antique
gem.37
The cautious author did not identify the scene. But the
19th-century drawing of the seal in the same book makes it
clear that it was interpreted as an angel, who is standing
frontally, with wings on both sides (Pl. 10). In reality, on the
basis of its pose, we can deduce that the figure appears in
profile in a long robe and is stepping to the left. On the right
side we can see a wing, while at the left side there is a raised
arm. The details and attributes of the figure cannot be
identified, but from the gesture it is likely that it is Nemesis,
who with one hand raises the edge of her peplos and with the
other hand holds an olive branch in front of her (Pl. 11).
The masks and mask-animal combinations of antique gems
carried a more complicated system of symbols. Even today we
have to rely on mere guesswork regarding the meaning of these
fantastic combinations. However, they were not just re-used,
but in the Carolingian period, imitations were made. One
example of this is the sapphire gem on the reliquary of Abbot
Plate 11 Intaglio with Nemesis Plate 12 Sealing of the archbishop
of York, 1154
 The Re-use and Re-interpretation of Gemstones in Medieval Hungary
Plate 13 Sealing of a Hungarian Plate 14 Sealing of Béla (Adalbertus) IV,
bishop, 1523 1266
Anno, on which there is a depiction of the combination of three
masks.38 The arguments of Professor Zwierlein-Diehl regarding
this gem are totally convincing. According to her, these
depictions were interpreted as representations of the Holy
Trinity. A couple of centuries later, from the 12th century, we
have definite proof of this thesis. On the sealing of the
archbishop of York, around an antique gem depicting a
combination of three heads, there is an inscription which
interprets its meaning: + capvt nostrvm trinitas est (Pl.
12).39
We encounter similar examples in the usage of seals in
Hungary, although the interpreting inscriptions are missing. In
one instance a bishop used a seal which combined the depiction
of the head of an eagle, Silenos and Pan (Pl. 13).40 It is likely
that he interpreted this as the representation of the Holy
Trinity, because it was obvious to identify the head of the bird
with the Holy Spirit, the older man with the Father and the
younger with the Son. There is also a preserved ring seal,
whose owner is unknown, but from the inscription (ave maria)
we can assume that he was a devotional Christian.41 It is well
known that between the 14th and 16th century there were
attempts to depict the Holy Trinity in the form of three faces.
This was the tricephalus or trifrons.42 We can assume that these
were wide-spread because Pope Urbanus VIII in 1628 banned
and declared these depictions as heretical and in 1745 Pope
Benedict XIV took a stand against them again in his encyclical
letter Sollicitudine nostrae... . 43.
Finally I present an example which demonstrates that, even
if modern scholars – who obviously aspire for objective and
professional definition – presume Christian themes on antique
seals, then it was quite understandably typical in the Middle
Ages, when users were obviously prejudiced on a religious
basis. There is a sealing, which has been assigned to King Béla
IV (1235–70), on which there is visible a hand which holds some
kind of object, and around it there are the remains of some
letters (Pl. 14).44 According to a Slovakian historian this is a
representation of the right hand of St Stephen, the Holy Right
(Pl. 15), which was the most important relic in medieval
Hungary; it probably originally held flowers.45 From the
inscription he tried to reconstruct the following words: domine
(?) yolanty (?) (the second word would refer to the wife of
Béla). In reality the image is well known from antique gems: a
hand holding an ear (Pl. 16).46 Its interpretation is clear from
the Greek inscription: mnhmoneye, which means ‘remember!’.
In antiquity such gems were likely used as a present, or a
warning between lovers. Some gems of this type have been
found in Pannonia. The sealing is beyond doubt from an
Plate 15 Holy Right of St Stephen. Plate 16 Cameo with hand holding
Budapest, St Stephen’s Basilica an ear
antique gem and the interpretation of the inscription as a
13th-century majuscule is without any justification. However I
cannot reject the possibility, that just like the 20th-century
scholar, the 13th-century user of the gem associated it with the
holy right hand of St Stephen, which was already adored as a
relic at that time.47 Another interesting aspect of this gem is
that the depiction is in the cameo technique, i.e. in positive,
which in antiquity was not used for sealing.
In my presentation I exceeded the time range of the
conference – of course with the consent of the organisers. This
look into the more recent period of the Middle Ages probably
will inspire some of us to draw more and more medieval
artefacts into the research on antique gems and with that it will
be possible to widen our knowledge of their re-use and
re-interpretation.
Acknowledgements
This work was supported by the TÁMOP 4.2.1./B-09/1/KONV-2010-
0007 project.The project is implemented through the New Hungary
Development Plan, co-financed by the European Social Fund and the
European Regional Development Fund.
Notes
1 T. Gesztelyi, ‘Antik gemmák honfoglalás kori sírokban (Antike
Gemmen in landnahmezeitlichen Gräbern)’, Déri Múzeum
Évkönyve (2002–3), 65–8, at 65.
2 Ibid., 65–7.
3 P. Lővei, ‘Pecsétek, pecsétnyomók, sokpecsétes oklevelek (‘Seals,
stamps, charters with many seals’)’, in T. Wehli (ed.), Középkori
magyar uralkodói pecsétek, MTA Művészettörténeti Kutatóintézet,
Budapest, in press.
4 M. Hlatky, A magyar gyűrű (The Hungarian Ring), Budapest, 1938,
48.
5 T. Gesztelyi, Antike Gemmen im Ungarischen Nationalmuseum,
Catalogi Musei Nationalis Hungarici, Series Archaeologica III,
Budapest, 2000, no. 204.
6 Ibid., no. 249.
7 Ibid., no. 5.
8 M. Henig, A Corpus of Roman Engraved Gemstones from British
Sites (BAR British Series 8), Oxford, 1978 (2nd edn), 285, M 11.
9 Gesztelyi (n. 5), no. 147. The inscription Ave Maria gratia is used
very widely across Europe on seals, brooches and other items.
10 Ibid., no. 37.
11 Ibid., no. 57.
12 M. Henig, A Corpus of Engraved Gemstones from British Sites, Part
II, (BAR 8), Oxford, 1974, 106. See also, M. Henig, ‘The Re-use and
Copying of Ancient Intaglios set in Medieval Personal Seals,
mainly found in England: An Aspect of the Renaissance of the
‘Gems of Heaven’ | 261
Gesztelyi

Twelfth Century’, in N. Adams, J. Cherry and J. Robinson (eds), 32 Sz. Süttő, Anjou-Magyarország alkonya I–II, Szeged, 2003, vol. I,
Good Impressions: Image and Authority in Medieval Seals, London, 93, and vol. II, 299.
2008, 25–34; also: http://www.britishmuseum.org/pdf/1b%20 33 Vienna, Staatsarchiv, AUR 1385. VII. 28.
rev%20order.pdf. 34 E. Brandt and E. Schmidt, Antike Gemmen in Deutschen
13 D. Csánki, ‘Harminczhat-pecsétes oklevél 1511-ből’, Turul 5 (1887), Sammlungen I. Staatliche Münzsammlung München 2, Munich,
1–11, 49–59 and 130–41, at 55. 1970, nos 1467–8.
14 T. Gesztelyi and Gy. Rácz, Antik gemmapecsétek a középkori 35 E. Zwierlein-Diehl, ‘Antike Gemmen im Mittelalter: Wieder-
Magyarországon – Antike Gemmensiegel im mittelalterlichen verwendung, Umdeutung, Nachahmung’, in D. Boschung and S.
Ungarn, Debrecen, 2006, no. 47. Wittekind (eds), Persistenz und Rezeption, Wiesbaden, 2008, 243–
15 Ibid., no. 96. 4.
16 Ibid., nos 95, 110. 36 Gesztelyi and Rácz (n. 14), no. 7.
17 Ibid., 22–38. 37 J. Jerney, Magyar Történelmi Tár 2, Budapest, 1855, 155, fig. 17.
18 Ibid., 39–44. 38 E. Zwierlein-Diehl, Antike Gemmen und ihr Nachleben, Berlin–New
19 Gesztelyi (n. 5), 12. York, 2007, 487, pl. 898. See also the paper by Lapatin, this volume.
20 Gesztelyi and Rácz (n. 14), 24, 29, 32. 39 Ibid., 488, pl. 899.
21 S.H. Middleton, Engraved Gems from Dalmatia, Oxford, 1991, 40 Gesztelyi and Rácz (n. 14), no. 108.
passim. 41 Gesztelyi (n. 5), no. 249.
22 Gesztelyi and Rácz (n. 14), 40–2. 42 W. Braunfels, ‘Dreifaltigkeit’, in E. Kirschbaum (ed.), Lexikon der
23 A. Kubinyi, ‘Die Märkte Ungarns im Spätmittelalter’, in F. Irsigler Christlichen Ikonographie I, Rome-Freiburg-Basel-Vienna, 1968,
and M. Pauly (eds), Messen, Jahrmärkte und Stadtentwicklung in 528.
Europa, Trier, 2007, 253–62. 43 Z. Kovács, ‘Sollicitudine nostrae...’ XIV. Benedek pápa a
24 Gesztelyi and Rácz (n. 14), 42–3. Szentháromság ábrázolásáról (Papst Benedikt XIV. über die
25 Ibid., 57. Darstellung der Trinität)’, Művészettörténeti Értesítő 50 (2001),
26 A. De Bonfinis, Rerum Ungaricarum decades, I–IV/1, (I. Fógel, B. 65–76.
Iványi and L. Juhász [eds]), Leipzig–Budapest, 1936–1941, libri 3, 44 Gesztelyi and Rácz (n. 14), no. 8.
9, 302. 45 A. Húscava, ‘Turcianske a liptovské lisziny Belu IV. z r. 1266’,
27 J. Balogh, A művészet Mátyás király udvarában I–II, Budapest, Sbornik Filozofickej Fakulty University Komenskeho 9 (1958), 193–4.
1966, 302–3. 46 Gesztelyi (n. 5), no. 284. See also the paper by Molesworth and
28 Gesztelyi and Rácz (n. 14), no. 68. Henig, this volume, Pl. 9.
29 Ibid., no. 58. 47 A. Farkas, Nemzeti ereklyénk-Szent István király jobbja. Az ereklye
30 Ibid., no. 140. története, Budapest, 1991.
31 Ibid., no. 17.

262 | ‘Gems of Heaven’

Reflections on Gems Depicting the
Contest of Athena and Poseidon1
Hadrien Rambach
Greek and Roman mythology tell us how the patron deity of the
newly founded city of Athens was chosen in a competition
between Athena and Poseidon.The gods raced to the Acropolis,
where they offered gifts to Athens.2 Athena caused an olive tree
to spring up, whereas Poseidon struck the ground with his
trident, prompting a spring of water to gush forth. Although
Poseidon had reached the Acropolis first, Cecrops – the
legendary king of Athens, shown in art as half-human and half-
snake – favoured Athena. This displeased Poseidon, who
cursed the city with a flood. Zeus intervened and allowed the
Athenians to choose the goddess. This story has inspired
several works of art of different periods, sizes and materials.
This article examines the depiction on engraved gems – with
essential detours to marble sculpture and numismatics – of the
‘dispute of Athena and Poseidon’, or rather the presentation of
the gifts, with the gods facing each other.
The most famous is a cameo in Naples, formerly owned by
Lorenzo de’ Medici (Pl. 1).3 One of the best known items in this
celebrated collection of engraved gems, this large cameo
measures over 50mm in height. Believed to date from the late
1st century bc, it has been variously and controversially
attributed, for example to Aspasios (although this has never
been thoroughly argued), and even to Pyrgoteles (despite this
evidently being impossible chronologically). This dating seems
very likely, and stylistically the cameo can be compared with
signed Augustan gems and contemporary coins. Part of the
attraction – and mystery – of the cameo is the unexplained
series of engravings in the exergue: two palm trees, two shells,
two wheels (?), the ΠΥ monogram,4 and another, unidentified,
Plate 1 The Medici cameo, 52 x 43mm. Naples, Museo Archeologico Nazionale
Plate 2 Detail of Medici cameo: the exergue
object (Pl. 2). The cameo is very well composed and rich in
ancillary details, such as the snake that aggressively confronts
Poseidon. The earliest known reference to the cameo is the
1465 Medici inventory, which suggests that it was purchased by
Piero ‘il Gottoso’ de’ Medici (1416–69). It remained in the
family’s possession in Florence until the death of Alessandro ‘il
Moro’ de’ Medici (1510–37), when it was taken by his widow,
Margaret of Austria (1522–86) to first Rome, then the
Netherlands (1559–67), and finally Ortona. Her son Alessandro
Farnese (1545–92) brought it back to Rome, where it remained
until 1735, when it was taken to Naples.
The possible origins of the iconography
To what degree was this representation of the myth the
personal creation of a gifted engraver as opposed to an
adaptation from pre-existing iconography? If it was not
original, did it derive from a single source, such as a Hellenistic
relief, or does it combine influences from a variety of works
of art?
The depiction of Athena is not unique: she is shown in a
similar pose in other works, such as the 1st-century bc silver
Coppa Corsini,5 a marble statuette in Athens,6 and a statue in
the Musée Rolin in Autun.7 The figure of Poseidon is not
original either: he is the so-called ‘Lateran Poseidon’, resting
his foot on a rock or on a prow, holding a trident and sometimes
a dolphin, as shown in a marble copy of a lost 4th-century bc
bronze sculpture by Lysippos (Pl. 3).8 Many slight variations
exist,9 including a number of stone copies from the Roman
period, such as a 2nd-century ad marble statuette from
Plate 3 The Lateran Poseidon, marble, H. 2m. Vatican City, Museo Lateranense
‘Gems of Heaven’ | 263
Rambach

Plate 4 Silver denarius of Sextus Pompey, Sicily, Plate 5 Silver denarius of Sextus Pompey, Sicily, Plate 6 Silver drachm, Bruttium, 216–214 BC
40–39 BC 38–37 BC

Plate 7 Silver tetradrachm of Demetrios Plate 8 Impression of garnet intaglio with Poseidon, Plate 9 Impression of nicolo intaglio with
Poliorketes, mint of Amphipolis c. 290–289 BC 20 x 14mm, ex-Marlborough Collection Poseidon, 10 x 8mm, ex-Marlborough Collection

Plate 10 The Portland Vase, H. 245mm. London, British Museum, GR Plate 11 Detail of the Portland Vase
1945,0927.1

Eleusis10 that is almost identical to the figure on the cameo,
apart from the spear which is inclined and not vertical. The
coinage of the period in which the Medici cameo was carved
includes two denarii of Sextus Pompey with a comparable type.
He issued a coin from Sicily in 40–39 bc with a complex scene
on the reverse depicting Neptune (holding an aplustre and
resting his right foot on a prow, naked but for a chlamys on his
left arm) standing facing left between the Catanaean brothers
Anapias and Amphinomus (Pl. 4).11 In 38–37 bc Sextus Pompey
struck a denarius depicting on the obverse a galley in front of
the Pharos of Messana, decorated with a statue of Neptune (Pl.
5).12 But I do not think that this is a reason to associate the
Medici cameo with Pompey, as – even in numismatics – the use
of the ‘Lateran Poseidon’ was nothing novel: the pose had
already appeared on coins of the Brettii in southern Italy
(c. 216–14 bc) that show Zeus standing facing left with his right
foot on an Ionic capital (Pl. 6),13 and even earlier (soon after
300 bc) on Macedonian silver tetradrachms struck for
Demetrios Poliorketes (Pl. 7).14 The type is commonly found on
1st-century bc/ ad gems, and I shall not attempt to make a list
here of known specimens. Two examples that illustrate the
point were both formerly in the collection of the fourth Duke of
Marlborough (1739–1817): one in garnet (Pl. 8),15 and one in
nicolo (Pl. 9).16 The ‘Lateran Poseidon’, with a snake, also
appears on the magnificent Portland Vase, which was certainly
engraved by a gem-carver (Pls 10–11).17

264 | ‘Gems of Heaven’

 Reflections on Gems Depicting the Contest of Athena and Poseidon
Plate 12 Glass paste with Poseidon and Amymone, Plate 13 Glass paste with Poseidon and Amymone,
24 x 19mm. Hanover, Kestner-Museum signed by Aulos. D. 31mm. London, British Museum,
GR 1923,0401.978
Three interesting and unusually large glass gems dating
from the same period as the Medici cameo show Poseidon and
Amymone.18 One, in Hanover, is of pale grey glass (Pl. 12);19
another, in the British Museum, is in white paste and is signed
by Aulos (Pl. 13);20 and a third, in Paris, was probably cast from
the same mould as the latter (Pl. 14).21 These illustrate a well-
known mythological story, in which the ‘blameless’ daughter of
Danaos is saved by Poseidon from a threatening satyr at
Argos.22 Poseidon is standing, his right foot raised on a rock, his
right hand holding the trident; his left arm, with drapery
wrapped round it, is behind his back. Amymone’s hydria lies
overturned at Poseidon’s feet. She is standing on the left,
wearing a long chiton and himation, raising her veil with her
left hand. The prototype for these gems was probably a Late
Classical or an Early Hellenistic relief. But, as Gertrud Platz-
Horster wrote, ‘you cannot argue that an image on a gem was
inspired by a sculpture if that sculpture is unknown’.23 It is in
fact not possible to say if the inspiration was a statue, a relief or
a painting. Could it simply be a coincidence that the
compositions of the Athena and Poseidon Medici cameo and
the Poseidon and Amymone pastes are so close? They are of the
same period. Could they be from the same workshop, or could
they have been engraved for the same patron? What are the
connections between these two types? They were both old
motifs, and we must ask why they were chosen by the
1st-century bc engravers.
Plate 15 Drawing of the Parthenon west pediment by Ludolf Stefani (1875)
Plate 14 Glass paste probably cast from the
same mould as Pl. 13. 19 x 21mm. Paris,
Cabinet des médailles
The Parthenon pediment
The 5th century bc is very significant for the Athena and
Poseidon myth, as that was when the Parthenon was built.
Between 438 and 432 bc a relief was carved in Pentelic marble
for the west pediment. Pausanias saw it around ad 160, and he
tells us that:
[on the Acropolis is a] group [of statues] dedicated by Alkamenes.
Athena is represented displaying the olive plant, and Poseidon the
wave
and that:
As you enter the temple [of Athena on the Acropolis at Athens] that
they name the Parthenon, all the sculptures you see on what is
called [...] the rear pediment represent the contest for the land
between Athena and Poseidon.24
This makes it clear that there were two depictions of Athena
and Poseidon on the Acropolis.
Fragments survive from the ‘violent’ depiction on the
Parthenon’s west pediment – i.e. that in which the gods are
confronting each other and preparing to strike the ground with
their spears and tridents. Depicting both Poseidon’s and
Athena’s torsos and the back of her head, these sculptures are
divided between the British Museum and the Acropolis
Museum. However, they are so fragmentary that we need to
rely on modern reconstructions of the Acropolis pediment: a
drawing by Ludolf Stefani (Pl. 15),25 and a full-scale replica of
‘Gems of Heaven’ | 265
Rambach
the Parthenon in Nashville (Pl. 16).26 More authentic are
drawings made in 1674 by Jacques Carrey (Pl. 17).27 Erika
Simon has written that:
the pediment-compositions must have been so impressive that
they had repercussions for various works of art.28
Nothing survives from the ‘peaceful’ group, which could have
been by Phidias’ younger contemporary Alkamenes. It depicted
Poseidon resting and the gods presenting their gifts to the
newborn city, but Pausanias’ passing reference gives no useful
information about its appearance. Simon has suggested that
this might be the composition depicted on the Medici cameo29 –
Plate 17 Drawing of the Acropolis pediment by Jacques Carrey (1674)
266 | ‘Gems of Heaven’
Plate 16 Detail of the Nashville
Parthenon by Dinsmoor and Hart
(Nashville, Tennessee)
an idea with which I agree. However, Luigi Beschi doubts this
because he regards the cameo as being too late in style to relate
to the Parthenon.30 In such cases, ancient engraved gems and
coins are essential iconographical records.31 When – as is
probably the case with the contest of Athena and Poseidon –
the original sculpted group has disappeared, contemporary
copies are major sources, as Francesca Ghedini describes:
It seems we can conclude, after comparison of the representations,
that they correspond iconographically to the monument described
by Pausanias, at least allusively. The Athenian coins [...], those of
Hadrian’s time (better readable in the Restitutio of Marcus
Aurelius), the neo-Attic relief, and the many gems of various
provenances, allow us without any doubt to recover the original
iconography of the Athenian ex-voto. At that time, it included the
two affronted gods, with the twisted form of an olive tree between
them. Poseidon rested his left foot on a rock from which flowed, or
should flow, the sea; his torso was erect, and his right hand tightly
gripped a trident. Athena stood unarmed, but not defenceless,
since her martial attributes lay beside her: her feared spear was
propped against an olive tree, and her shield lay on the ground,
supported invisibly. And the snake lazily uncoiled, perhaps
alluding to the myth remarked by Lucian as he described the holy
contest.32
This motif can be found on a variety of materials and in various
sizes. The fact that these depictions are of different periods
raises new questions: why does this subject so often appear on
gems, and what were the engravers’ sources?
The major differences between these two types, the quarrel
and its aftermath, lead me to propose two distinct iconographic
types: a ‘Parthenon’ type and a ‘Medici’ type. On the Parthenon
type, the gods are attacking each other, with crossed weapons
and feet and aggressive looks, as described by Ovid
(Metamorphoses 6):
the Sea-God standing, striking the rough rock with his tall trident.
She herself gives a shield, she gives a sharp-tipped spear, she gives
a helmet for her head; the aegis guards her breast, and from the
earth struck by her spear, she reveals an olive tree.33
 Reflections on Gems Depicting the Contest of Athena and Poseidon
Plate 18 Cameo fragment with Athena and Poseidon. Plate 19 Glass paste, impression, 19 x
L. 33mm. Vienna, Kunsthistorisches Museum 18mm. Vienna, Kunsthistorisches Museum
Plate 21 Red jasper intaglio with Athena and Poseidon. 14 x 11mm. Utrecht,
Geldmuseum
The events on the ‘Medici’ type take place afterwards, when
Athena has won. The Parthenon scene is therefore the more
dramatic. The Medici scene is all the more peaceful in that we
cannot see the pointed end of the trident with which Poseidon
created the Erekhtheis sea;34 only what seems to be a spear, or
long pole, is visible.
The Medici type on Roman gems
Although exceptional for its size and quality, the Medici gem is
not unique in depicting Athena and Poseidon. Other Roman
gems of the 1st century bc, and of the following two centuries,
depict the Medici type. The Kunsthistorisches Museum in
Vienna owns a cameo (Pl. 18), which may originally have been
even larger than the Medici one (the surviving fragment is up
to 33mm high, whereas the Medici cameo measures 52mm).35
Equally fine in style, it is a late Republican or early Imperial
piece, like the Medici cameo.36 In fact, this fragment is so close
in composition to the Medici cameo (for example, the position
of Athena’s shield and the way that Poseidon’s hand is almost
inside the bent trunk of the tree) that the two must be related.
Although they are not necessarily by the same engraver, one
certainly copies the other.
In the same period, an intaglio was carved with exactly the
same type, as is evident in particular in the shape of the snake
and the presence and position of the dolphin. On the basis of a
comparison with the paste of a dancing Bacchus in Berlin,
Erika Zwierlein-Diehl has attributed it to Dioskurides.37
Unfortunately, the gem seems not to have survived, and is
known only from a damaged glass copy in Vienna (Pl. 19).38
Although the figure of Athena has disappeared, her presence
can be discerned.
Plate 20 Carnelian intaglio with Athena and
Poseidon. 18.5 x 16mm. Utrecht, Geldmuseum
Plate 22 Silver belt buckle from Pompeii or Herculaneum. Naples, Museo
Archeologico Nazionale
In Utrecht there is a carnelian intaglio carved during the
same decades (Pl. 20).39 Although not particularly large (18.5 x
16mm), it is well detailed, harmoniously composed, and
undoubtedly the work of a master-engraver. On the basis of a
comparison with a gem in Vienna, Maaskant-Kleibrink has
proposed that it might by Heius, but it does not seem
sufficiently Hellenistic. As she wrote, it is a ‘fairly exact replica’
of the Medici cameo – in fact, it is identical, with the exception
of Athena’s spear and the dolphin in Neptune’s hand, neither of
which appears in the intaglio. The museum in Utrecht owns
another intaglio of Athena and Poseidon, in red jasper, 14mm
in length, poorly engraved in the ‘Small Grooves Style’ of the
1st/2nd century ad (Pl. 21).40 The depiction of the scene is very
crude, but the engraver undoubtedly understood who the two
figures were, as he has included Poseidon’s attribute (a
dolphin) in his extended hand. I shall group it with the Medici
type, although it is strictly speaking a variant with significant
differences: Poseidon does not rest on a rock, and is not leaning
towards Athena.
The existence of varied versions of this type confirms
Gisela Richter’s comment of 1956:
The gem engravers of the Roman age only differ from the sculptors
of that time in that they did not reproduce the works that they
copied mechanically. They had to copy freehand, like the painters,
or like the sculptors who adapted earlier works in new
compositions. And their reproductions have likewise become a
valuable source for our knowledge of many lost Greek
masterpieces.41
Any iconographical study of engraved gems should start with
this differentiation between mechanical copies, reproductions
and adaptations.
‘Gems of Heaven’ | 267
Rambach
Plate 23 Bronze medallion of Hadrian. D. 41mm. London, British
Museum, CM 1857,0812.2
The myth in metal – silversmiths and numismatics
Small works depicting the Medici type were created by Roman
silversmiths during the late 1st century bc and early 1st century
ad, such as the elegant silver belt buckle (cingulum) discovered
in Pompeii or Herculaneum (Pl. 22).42 Writing in 1970, Gertrud
Platz-Horster remarked that, ‘sometimes gems and coins show
strikingly similar images’.43 This is the case with the Athena
and Poseidon myth, both types of which appear on a number of
coin reverses during the first three centuries of the Roman
Empire. The Medici type can be found on the reverse of some
rare bronze medallions of Hadrian. Gnecchi in 1912 listed four
specimens: in Venice (D. 42mm, 54g), London (Pl. 23; D. 41mm,
42.26g), Rome (D. 38mm, 42.6g) and Vienna (D. 33mm,
23.88g).44
A well-preserved silver medallion depicts the Medici type
on its reverse (Pls 24a–b).45 The inscription states that it was
struck during the third consulate (ad 128/29–38) of Hadrian,
but some numismatists consider it to be a 16th-century
Renaissance work: the presence of a sceptre next to Hadrian’s
draped and bare-headed portrait, reminiscent of a Cretan
tridrachm of Caligula,46 would be unique for a 2nd-century ad
coin, and the reverse follows very closely the Medici type: the
dolphin has been split into two and becomes meaningless, as if
the engraver had copied the type mechanically, without
understanding what was depicted. This would be surprising for
the so-called ‘Alphaeus Master’ to whom this medallion has
been ascribed. The same reverse appears on a medallion of
Marcus Aurelius (Pl. 25),47 but this too may in fact be a
Renaissance medal (of a type derogatorily called ‘Paduans’).
However, the bronze coins of Hadrian are undoubtedly
genuine, and the use of the type under that emperor is likely to
be significant. Nicknamed Graeculus in his youth, he had by the
time the coins were struck become Hadrianos Sebastos Zeus
Plate 25 Bronze medallion of Marcus Plate 26 Bronze drachm, mint of
Aurelius. D. 39mm. Bonn, Rheinische Athens, AD 120–40
Landesmuseum
268 | ‘Gems of Heaven’
Plate 24 a–b Silver medallion with Hadrian (obverse) and Athena and Poseidon (reverse).
D. 33.5mm. Private collection
Olympios.48 He was now portraying himself as the new patron
of Athens. Also during the 2nd century ad, bronze drachms
were struck in Athens with the Parthenon type on the reverse.
The earlier issue, c.ad 120s–40s, displays the type with an owl
and a dolphin (Pl. 26);49 the later issue of the ad 140s/50s–170s
lacks the owl and dolphin (Pl. 27).50 A third Athenian issue,
around ad 264–7, under the reign of Gallienus (ad 260–8), uses
very similar dies.51 The type appeared for the last time on a
bronze coin struck for Philip the Arab (ad 244–9) in Phokaia in
Ionia (Pl. 28),52 but the tree has become tiny, and it is not clear
whether Poseidon is resting his foot on a rock or on the snake;
the engraving is poor, and the figures, which have been
romanised, have become very rigid. It is not surprising that
Athenian bronze coins used the Parthenon type: artists there
undoubtedly knew the friezes on the Acropolis. These local
engravers could also have seen the group of the Medici type,
and been inspired by it for their coins of the ad 120s–140s,
before it disappeared. But the choice of the type in Rome for
the Hadrianic medallions, and in Ionia for the coins of Philip, is
more intriguing. Why illustrate the contest of Athena and
Poseidon? How did they know the iconography?
Whoever he was, and there is no need to attempt to name
him, the carver of the Medici cameo is likely to have been a
late-Hellenistic artist, taught in Greece but active in Rome. The
gem might have entered the imperial collections, and therefore
have been available to court artists, as the choice of its type for
the medallions of Hadrian suggests. But, more simply, the motif
may have been circulating widely throughout the Roman
territories, by means of glass gems in daktyliothecæ, seal
impressions, plaster casts (of gems, stone sculptures or
silverware) or pattern books, etc. It is highly implausible that
artists could have created these various works independently,
without a common source – which they adapted to their own
Plate 27 Bronze drachm, mint of Plate 28 Bronze drachm, mint of
Athens, AD 140s/150s–170s Phokaia in Ionia, AD 244–9
 Reflections on Gems Depicting the Contest of Athena and Poseidon
Plate 29 Attic red figure hydria, mid-4th century BC, H. 51cm. St Petersburg,
Hermitage
style and time, to the scale desired, and to the material they
were using. The real question, therefore, is the identity of the
original Classical model: a mid-4th-century bc red-figure
terracotta, such as an Attic hydria in the Hermitage depicting
the Parthenon type (Pl. 29);53 or a Campanian vase in Madrid
with a variation of the Medici type;54 a stone sculpture, such as
a damaged cylindrical neo-Attic (2nd century ad) relief in
Cordova depicting the Medici type but showing Poseidon
resting on a prow rather than a rock,55 or a bronze sculpture? In
any case, the type was known and copied not only in the early
Roman Empire, but also for centuries later: in the early 5th
century ad the Medici type was faithfully reproduced on a
silver vase found near Oradea in Romania.56
The Late and Post-Antique gems
The theme of this symposium was Late Antiquity, and the
physical re-use of ancient gems during the Medieval period is
discussed elsewhere in this volume. Although so far I have
discussed only Classical and Early Roman gems, I am
interested in the re-use of ancient iconography: the dispute of
Athena and Poseidon provides an opportunity to study the
imitation of Classical gems in later periods. Indeed, as far as
gems are concerned, this iconographical type did not die out in
the 2nd century ad. The Medici type undoubtedly inspired an
elaborate sardonyx cameo probably engraved during the 6th
century ad in the ‘Mythological Workshop’, to use Jeffrey
Spier’s term (Pl. 30).57 This cameo was in a Russian collection a
century ago, but both its earlier provenance and present
Plate 30 Sardonyx cameo, 6th century AD. Formerly in a Russian collection
location are unknown. According to Spier, it is part of a small
group of cameos depicting mythological scenes [which] is very
difficult to categorise and date. The works are highly stylised, [...]
characterised by the stocky figures and the distinctive treatment
of the musculature, which is outlined by shallow cutting.
There are recurring subjects in this group, such as Dionysos
and Apollo. On this cameo, although the centre of the
composition is the traditional Medici type, it has become part
of a larger and more complex scene, in which Dionysos stands
behind Athena, holding a thyrsos and with a panther at his feet,
and Apollo stands behind Poseidon, holding a lyre and with a
swan at his feet. This poses a real mystery.58
First, would the Medici type have been identified correctly
at this date, and, second, why were Apollo and Dionysos
introduced? It is interesting to note that on the Hermitage
hydria and the Madrid crater mentioned earlier, Athena is
supported by Dionysos. Although Late Antique engravers could
have added figures because they did not understand what the
original group depicted, it is also possible that the Medici type
is a simplified version of a more complex story in which
Dionysos (and possibly also Apollo) had a role.59 Moreover,
where was this important piece engraved? We have seen that
the Medici type was well known in a variety of media:
silverware, gems, casts and probably drawings. But we have no
clue as to where the Medici cameo was in the Late Roman and
Early Byzantine periods.60 It has no provenance before the
Renaissance, and we cannot locate it before its appearance in
the inventory in Florence in 1465. It could, like most important
Classical artefacts, have been lost in the first centuries ad and
excavated during the Renaissance. However, it is worth
considering that it may never have been buried and instead was
a prized possession, passing from one court to another as a
gift.61 In that case, official engravers could have admired it and
used it in their own compositions, which would explain how,
half a millennium later, copies were made again, as soon as the
art of glyptics was reborn, during the Hohenstaufen period. In
southern Italy during the early 13th century at least two
cameos depicting the fight of Athena and Poseidon were
carved: a very large one with three layers, now in Paris (Pl.
31),62 and a smaller one now in Vienna (Pl. 32).63 The Paris
cameo was modified at a later date, with the addition of a
Hebrew inscription identifying the figures as Adam and Eve.64
The Viennese gem is faithful to the traditional Medici type, but
the Parisian one has a complex exergue (Pl. 33): is this a
coincidence – albeit a surprising one – or was this inspired by
the strange exergue of the Medici cameo (Pl. 2)?
‘Gems of Heaven’ | 269
Rambach
Plate 31 Cameo with Athena and Plate 32 Cameo with Athena and
Poseidon, early 13th century. 95 x Poseidon, early 13th century. 35 x
78mm including mount. Paris, 36mm. Vienna, Kunsthistorisches
Bibliothèque nationale Museum
Plate 33 Detail of the exergue of Plate 31
In the late 15th century, a medallion depicting the Medici
type was carved in Florence for a wall in the Palazzo Medici
(Pl. 34).65 There can be no argument about its source: all
Florentine artists of the time would have had access to the
Medici art collections, and this medallion forms part of a series
designed to celebrate some of the most valuable gems in that
collection, so this relief can therefore be described as ‘a copy’ of
the Medici cameo. In the Cabinet des médailles in Paris there is
a very large (46 x 38mm) 16th-century cameo that is quite
faithful to the Medici type, but its engraving is of very poor
quality.65 Various details, such as Athena’s hand, suggest that
the engraver neither had a good model to copy, nor understood
the subject. What then was his inspiration? Late Renaissance
and post-Renaissance gem engravers have occasionally used
ancient coins as models, as in the case of a 16th- or 17th-century
intaglio of Tellus and the Seasons that was ‘evidently derived
from a Roman coin type’, according to Erika Zwierlein-Diehl.66
As for the modern Athena and Poseidon gems, the engravers
are more likely to have been directly inspired by the Medici
cameo rather than by coins, either directly or via casts or
engravings.
Plate 34 Marble medallion, late 15th century. Florence, Palazzo Medici
270 | ‘Gems of Heaven’
Plate 35 Sard intaglio with Athena and Poseidon. L. 11mm. London, British
Museum, GR 1913,0307.28
The modern gems
The British Museum owns a small sard intaglio (11mm long) of
good quality that has been published as a 16th-century work
(Pl. 35).67 Although not an exact copy, it is close to the
iconography of the Medici type, even incorporating the dolphin
next to Poseidon: the engraver most probably knew the Medici
cameo. But the date given to this intaglio needs to be
reconsidered, as it closely resembles another gem considered to
be a much later work.
By the mid-18th century, large daktyliothecæ (collections of
gem-impressions in sulphur, plaster or glass) had been formed,
for instance by Stosch and (later) Tassie.68 Neo-classical artists,
especially those living in Italy, therefore had easy access to the
best ancient gems. This was probably the iconographical source
of a splendid intaglio carved by Giuseppe Cerbara (1770–1856).
This gem is undoubtedly inspired by the Medici cameo, as is
shown by the dolphin on the left and the position of the snake. I
have failed to locate Cerbara’s intaglio, which is known from a
Paoletti cast and an image in Lippold, who does not record its
material.69 It is not the example in the British Museum, which
has a few minor differences, such as the length and waviness of
the branch ‘cut’ by Athena’s spear. In addition, I discovered in a
London private collection an unpublished agate intaglio (Pl.
36)70 which is virtually identical to the Cerbara intaglio, and
can with reasonable certainty be ascribed to the same
engraver. The most significant difference is the branch above
the spear, which divides into two in the unpublished gem and
into three in the Cerbara intaglio. Some of the proportions
differ also. A rather charming detail that the engraver has
added to the Medici type is the dolphin biting a snake (the
original dolphin peacefully rests on a rock).71
Plate 36 Agate intaglio. 24.6 x 19.4mm. London, private collection
 Reflections on Gems Depicting the Contest of Athena and Poseidon
Plate 37 Sard intaglio by Louis Siriès. Vienna, Kunsthistorisches Museum
Before concluding, comprehensiveness requires that I refer
to a sard intaglio carved in 1755 by Louis Siriès (c. 1686 – c. 1757)
(Pl. 37).72 Although he was then in Florence, he did not copy, or
even really draw inspiration from, the Medici cameo. He
created a new iconographical type, in which the two central
figures are surrounded by a very large and complex scene. We
must hope that one day archives will reveal why Siriès decided
to illustrate this theme.
Prince Stanislas Poniatowski (1754–1833) had gems carved
with the myth of Athena and Poseidon, one with the fighting
type, the other with the conversing type. The former, a
carnelian which is now lost (number 143 in the 1841 catalogue
of Tyrrell’s gems), is signed by Pyrgoteles (ΠYPΓOTEΛEΣ) (Pl.
38).73 The latter, a carnelian set in a necklace with 11 other
intaglios, no. 146 in the 1839 Christie’s sale catalogue, and later
in the Wellington collection, is signed by Philemon
(ΦIΛEMONOΣ EΠOIEI) (Pl. 39).74 Both are true creations by so
far unidentified artists, undoubtedly inspired by a knowledge
of ancient gems.
In conclusion, I wish to emphasise the fact that the copies of
the ‘Medici’ cameo of Athena and Poseidon prove that this type
has enjoyed an unusually large audience. This ancient stone,
and a number of other Roman gems, such as the so-called Seal
of Nero, have had a lasting influence on art, and they can be
traced in Renaissance glyptics, medals and paintings. I hope to
have shown that such ‘reproductions’ also existed much earlier.
I have raised questions, but not provided many answers: it is to
be hoped that future research will help us to understand better
the manner in which ancient artists drew on pre-existing
compositions.
Plate 38 Carnelian intaglio signed Pyrgoteles. Ex- Poniatowski collection)
Plate 39 Carnelian intaglio signed Philemon. Ex- Poniatowski collection
Notes
1 This is the text of a lecture given at the British Museum on 31 May
2009, at a symposium devoted to recent research on engraved
gemstones in Late Antiquity. I am very grateful to Chris Entwistle
for inviting me to give this paper. I have benefitted from many
critical comments and suggestions by Gertrud Platz-Horster, to
whom I offer my most sincere thanks. Similarly, Olga Palagia has
been of invaluable help. Arianna D’Ottone, Francesca Ghedini,
Erkinger Schwarzenberg, Jeffrey Spier and Erika Zwierlein-Diehl
have also very kindly helped me at various stages of my research.
Finally, Edward Bigden and Michael Hall have generously tried to
improve my written English. Six months after my lecture at the
British Museum, Angela Gallottini published her Studi di Glittica
(Rome, 2009) with eight illustrations of Athena and Poseidon
gems. She also used two pictures of the ‘seal of Nero’ (on which I
gave a lecture on 18th May 2010 to be published in the Jahrbuch für
Numismatik und Geldgeschichte 61 (2011). I take this as promising
auspices, and as a sign of renewed interest in these gems and
iconographic researches in glyptics.
2 Pseudo-Hyginus, Fabulae (trans. M. Grant), Lawrence KS,
University of Kansas Press, 1960, 164.
3 Naples, Museo Archeologico, Medici-Farnese collection, inv. no.
25837/5, onyx, 52.1 x 43mm; see: N. Dacos, Il tesoro di Lorenzo il
Magnifico. Le gemme, Florence, 1973, cat. no. 6; U. Pannuti, Museo
Archeologico Nazionale di Napoli. La Collezione Glittica, vol. II,
Rome, 1994, cat. no. 82; L. Fusco and G. Corti, Lorenzo de’ Medici,
Collector and Antiquarian, Cambridge, 2006, 244, n. 40. In the 1465
inventory of Piero de’ Medici it was probably the ‘uno cameo legato
in oro con 2 figure et uno albero in mezzo di rilievo’ and valued at
180 florins. In 1492, in Lorenzo’s inventory, it was valued at 800
florins and described as ‘uno chammeo grande leghato in oro chon
dua figure intagliate di mezzo rilievo, 1° maschio e una femina,
chon un albero in mezzo che hanno a pie 2 serpe, champo nero’.
The fame of this piece was always so great that, in 1787, Angelica
Kauffmann (1741–1807) decided to represent it as a centre-
decoration of her belt in a self portrait (Florence, Galleria degli
Uffizi): see the description by Marta Bezzini in R. Gennaioli, Pregio
e bellezza, cammei e intagli dei Medici, Florence, 2010, cat. no. 172.
4 This monogram is not yet fully understood. I do not think that the
series of engravings on the exergue is a later addition. I favour the
hypothesis that it is a pi-upsilon monogram, rather than an upsilon
in a three-sided frame, but any link with Alexander the Great’s
engraver Pyrgoteles seems improbable (as well as it being the
‘Pythagorean letter’ – see, T. Titti, in Studi di glittica, Rome, 2009,
88). It is most likely the signature of an unidentified engraver, in
the style of the monograms found on a number of ancient Greek
silver tetradrachms (e.g. Seleucus I, Susa, after 301 bc:
A. Houghton and C. Lorber, Seleucid Coins – A comprehensive
catalogue, New York, 2002, no. 173, and E.T. Newell, The coinage of
the Eastern Seleucid Mints.  From Seleucus I to Antiochus III, New
York, 1938, no. 426; Lysimachos, Amphipolis, 288/282 bc:
M. Thompson, ‘The Mints of Lysimachus’, C. Kraay and G. Jenkins
(eds), Essays in Greek Coinage Presented to Stanley Robinson,
Oxford, 1968, 163–82, no. 199.
5 A kantharos in the Palazzo Corsini in Rome.
6 See: F. Ghedini, ‘Il Gruppo di Atene e Poseidon sull’ Acropoli di
Atene’, Rivista di archeologia 7 (1983), ill. 9; W. Gauer, ‘Eine
‘Gems of Heaven’ | 271
Rambach
Athenastatuette des Athener Nationalmuseums: Zum Iudicium
Orestis’, Archäologischer Anzeiger (1969), figs 1–3.
7 See: Ghedini (n. 6), n. 133; E. Espérandieu, Recueil général des bas-
reliefs, statues et bustes de la Gaule romaine, III, Paris, 1910, 73, no.
1861.
8 Vatican City, Lateran Museum, inv. no. 10315. 2.01m high, it was
found in 1824 in the port of Ostia: see W. Helbig and H. Speier,
Führer durch die öffentlichen Sammlungen klassischer Altertümer in
Rom, Tübingen, 1963 (4th enlarged edn), vol. I, 798–9, no. 1118.
9 Deciding which is the original composition (with the rock or with
the prow, with or without the dolphin, etc.) is not easy. Paolo
Moreno has listed a number of versions, and suggested a tentative
chronology (‘Una cretula di Cirene ed il Posidone del Laterano’,
Quaderni di Archeologia della Libia 8 (1976), 81–98). According to
him, the depiction on the Medici cameo would be the earliest,
whilst the Lateran sculpture would be a free (later) depiction. The
obvious differences between the two are the torso (straight on the
Medici figure), the trident-holding arm (raised high on the Medici
figure), and the other hand (resting on the thigh on the Lateran
figure).
10 Eleusis Museum, inv. no. 5087; H. 54cm. The product of a
Classicising Attic workshop, probably c. ad 140–60. Published:
Ghedini (n. 6), ill. 8; K. Kourouniotes, Eleusis. A Guide to the
Excavations and Museum, Athens, 1936, 89–90, ill. 33; E. Bartman,
Ancient Sculptural Copies in Miniature, Leiden/New York/
Cologne, 1992, 132–3.
11 Denarius, 3.91g, sold Gemini (auction II, 11 January 2006, lot 278)
and Stack’s (ex Knobloch collection, May 1978, lot 739):
M.H. Crawford, Roman Republican Coinage, London, 1974, no.
511/3a; E.A. Sydenham, The Coinage of the Roman Republic,
London, 1952, no. 1344; D.R. Sear, The History and Coinage of the
Roman Imperators 49–27 bc, London, 1998, no. 334.
12 Classical Numismatic Group, auction Triton V, 15 January 2002, lot
1847, denarius, 3.50g: Crawford (n. 11), no. 511/4a; Sydenham (n.
11), no. 1348; Sear (n. 11), no. 335; H.A. Seaby, Roman Silver Coins,
vol. II, London, 1979 (3rd edn), no. 2.
13 Classical Numismatic Group, mail-bid auction 75, 23 May 2007, lot
30, silver drachm, 4.55g (Attic standard), Second Punic War issue:
M. Arslan, Monetazione aurea ed argentea dei Bretti, Glaux 4,
Milan, 1989, dies 14/22; B. Head, Historia Nummorum – Italy,
Oxford, 1911, no 1969; L. Naville, Monnaies grecques antiques
provenant de la collection de feu le Prof. S. Pozzi, Geneva, 1920, no.
266 ; See F. Lenzi, ‘Ripostiglio di monete d’argento dei Bruttii’,
Rassegna Numismatica 11 (1914), 1–14.
14 Demetrios Poliorketes (305–284 bc), silver tetradrachm,
Amphipolis mint, c. 290–289 bc, 16.98g. Bank Leu auction 83, 6 May
2002, lot 202; E.T. Newell, The Coinage of Demetrius Poliorcetes,
London, 1927, no. 116 (obverse cxi).
15 Foiled almandine garnet intaglio, set in gold as a seal, 20 x 14mm,
from the Marlborough collection. See: J. Boardman (with D.
Scarisbrick, C. Wagner and E. Zwierlein-Diehl), The Marlborough
Gems: Formerly at Blenheim Palace, Oxfordshire, Oxford, 2009, cat.
no. 402.
16 Worn nicolo intaglio, 10 x 8mm, from the Marlborough collection:
Boardman (n. 15), cat. no. 599.
17 Glass cameo: H. 245mm, Max. diam. 177mm (93mm at mouth);
handle: ht. 96mm, W. 18mm. Possibly from Rome, early Imperial
(c. ad 5–25) or Augustan (27 bc–ad 14). Supposedly found in 1582
on the Monte del Grano, a property owned by Fabrizio Lazzaro,
who claimed that it came from a sarcophagus containing the body
of Severus Alexander (ad 222–35). Recorded in 1601 in the
collection of Cardinal Francesco Maria Bourbon del Monte (1549–
1626). Bought after del Monte’s death by Cardinal Antonio
Barberini (1607–71) for 500 scudi (or 200 only?), and first
published in 1642 by Girolamo Teti in his Aedes Barberinæ. Kept in
the Barberini family until its purchase in 1778 by Sir William
Hamilton, who sold it for about £2,000 to Margaret, dowager
Duchess of Portland in 1784, in whose family it remained until its
purchase for £5,000 by the British Museum in 1945 (inv. no. GR
1945.0927.1: H.B. Walters, Catalogue of the Engraved Gems and
Cameos, Greek, Etruscan and Roman, in the British Museum,
London, 1926, no. 4036). It has been most recently published by:
K. Painter and D. Whitehouse, ‘The History of the Portland Vase’,
Journal of Glass Studies 32 (1990), 24–84; S. Walker, The Portland
Vase, London, 2004; H.-C. von Mosch, ‘Outdoorsex’ unter dem
Pfirsichbaum? Die Portlandvase im Lichte einer sensationellen
272 | ‘Gems of Heaven’
Neuentdeckung’, Quaderni Ticinesi. Numismatica e Antichità
Classiche 39 (2010), 195–223. I am grateful to Simone Michel for
bringing the figure of Poseidon on the vase to my attention, and to
Susan Walker for her help in writing this note. All interpretations
of the scene have been contested; the figures have no attributes,
and it is proving a challenge to identify the subject of the carving:
Greek mythology, Roman historical allegory, universal theme?
Most theories identify the ‘Lateran’ figure as Poseidon, but it has
also been suggested that this bearded figure could be Augustus’
mythical ancestor Anton (son of Herakles). Also, the snake has
been believed to be Cleopatra’s asp, and not a mythological figure
such as Cecrops. Very interestingly, it has been (unconvincingly)
argued that the Portland vase is in fact a forgery, dating back to the
Renaissance (J. Eisenberg, ‘The Portland Vase: a glass masterwork
of the later Renaissance?’, Minerva, 14.5, Sept./Oct. (2003), 37–41)
– a thesis similar to that held for some of the medallions presented
below in this paper.
18 E. Simon, ‘Amymone und Poseidon’, Lexicon Iconographicum
Mythologiae Classicae, vol. I.1 (1981), 745–7 (hereafter LIMC).
19 It measures 24 x 18.9 x 4.8mm: P. Zazoff et al., Antike Gemmen in
Deutschen Sammlungen, Band IV: Hanover, Kestner-Museum,
Wiesbaden, 1975, cat. no. 244.
20 Diam. 31mm; inscribed AYΛOC\ΛEΞAEΠOIEI. From the
Barberini collection. See: A. Furtwängler, Jahrbuch des Deutsches
Archaologisches Institut IV (1899), pl. 2.3; Walters (n. 17), no. 3723;
M.-L. Vollenweider, Die Steinschneidekunst und ihre Künstler in
spätrepublikanischer und augusteischer Zeit, Baden-Baden, 1966,
pl. 30, ill. 1–2. Son of Alexas (AYΛOC AΛEΞA EΠOIEI), Aulos was
the brother of Quintus ([KO]INTOC AΛEΞA EΠOIEI). It is
remarkable that Francesco Vettori (1693–1770) owned gems by
both Aulos and Quintus: he owned not only the Venus and Cupid
intaglio by Aulos, found in a sarcophagus in 1735 and purchased by
Charles Townley (1737–1805) for £50 in 1773 (since 1815 in the
British Museum [O.M. Dalton, A Catalogue of the Engraved Gems of
the Post-Classical Periods in the British Museum, London, 1915, cat.
no. 643, where it is erroneously described as a neo-classical
forgery]), but also a magnificent carnelian engraved with a
dancing figure of Mars by Quintus (Christie’s South Kensington.
Antiquities. Including an English Private Collection of Ancient Gems,
Part II, London, 29 October 2003, lot 299 [Todhunter Collection])
probably wrongly believed to have been sold by Vettori to Gian
Gastone de’ Medici (1671–1737). No provenance is known for the
Quintus, but it does not show the same fire damage as the Aulos
intaglio, and it is unlikely that Vettori would not have given the
place of discovery had he known it, so it must be ruled out that they
were found together.
21 A blue and white paste fragment, 19 x 21mm, found in Rome. See:
E. Babelon, Catalogue de la collection Pauvert de La Chapelle, Paris,
1899, cat. no. 162; G. Richter, Engraved Gems of the Romans. A
Supplement to the History of Roman Art, London, 1971, ill. 649.
22 Amymone, the ‘blameless’ one, was one of the 50 daughters of
Danaus, the brother of Aegyptus. Danaus married his 50 daughters
to his brother’s 50 sons and instructed them to kill their husbands
on their wedding night. All complied but Amymone, who refused
because her husband Lynceus honoured her wish to remain a
virgin; thus she received the epithet ‘blameless’. Amymone and
Lynceus went on to found a dynasty of Argive kings that led to
Danaë, the mother of Perseus. While at Argos she went to collect
water and was rescued from a threatening satyr by Poseidon.
23 G. Horster, Statuen auf Gemmen, Bonn, 1970, 46.
24 Pausanias, Description of Greece (trans. W.H.S. Jones and H.A.
Omerod), London, 1918, I.24.2–5.
25 L. Stefani, ‘Erklärung einiger im Jahre 1871 im Südlichen Russland
gefundenen Kunstwerke’, Compte-rendu de la Commission
Impériale Archéologique pour l’année 1872, St Petersbourg, 1875,
142.
26 This reconstruction was started in Tennessee in the 1890s, by
William Dinsmoor and Russell Hart. The current building dates
from the 1920s.
27 The marble statues were already fragmentary when seen by
Jacques Carrey (1649–1726), even though he drew them before the
1687 Venetian bombardment. In 1749 Richard Dalton (c. 1715–91)
could still see 12 figures on the west pediment when preparing his
A Series of engravings, representing views of places, buildings,
antiquities, etc. (London 1752). In any case, in the first years of the
19th century, when Giovanni Battista Lusieri (c. 1755–c. 1821) came
 Reflections on Gems Depicting the Contest of Athena and Poseidon
to draw the Parthenon at Lord Elgin’s request, he found only four
figures left. See: O. Palagia, The Pediments of the Parthenon,
Leiden, 1993, pl. 3; eadem, ‘Fire from heaven: pediments and
akroteria from the Parthenon’, in J. Neils (ed.), The Parthenon:
from Antiquity to the Present, Cambridge, 2005, 228.
28 E. Simon, ‘Poseidon’, LIMC, vol. VII.1 (1994), 473.
29 Ibid.
30 L. Beschi, Pausania, Guida della Grecia Libro I, L’Attica, Milan,
1982, 351: ‘esso è una tarda realizzazione di tipo classicistico sia per
la presenza di un albero naturalistico di ulivo, sia per il reimpiego
di precedenti e noti tipi statuari’.
31 The groundwork for this research has been laid by Ludolf Stefani
(n. 25), 5–142 (the section on gems is at 136–42), when he found,
published and commented on the ancient literary sources. His
work was continued by Carl Robert (‘Das Schiedsgericht über
Athena und Poseidon’, Mitteilungen des Deutschen Archäologischen
Institutes in Athen VII (1882), 48–58. I am grateful to Elisabeth
Furtwängler for finding a copy of this publication for me). Both
Stefani and Robert also studied the iconography, but more
examples have since been found, and the standard work on the
west pediment of the Parthenon is now that by Olga Palagia (n. 27),
whilst that on the other group is the article by Francesca Ghedini
(n. 6), 12–36, pls 1–9).
32 Ghedini (n. 6), 17. I owe the translation into English, and many
suggestions, to Massimiliano Tursi.
33 Ovid, Metamorphoses (trans. A.D. Melville), Oxford, 1986, 6, 70ff.
34 Pseudo-Apollodorus, Bibliotheca (trans. J.G. Frazer), Cambridge
(MA), 1921, 3. 14.1.
35 Vienna, Kunsthistorisches Museum, Collection of Greek and
Roman Antiquities, Inv. No. IXb 8. Chalcedony, opaque white over
uncoloured background, max. length 33mm, the break possibly
reworked. See: F. Eichler and E. Kris, Die Kameen im Kunst-
historischen Museum, Vienna, 1927, cat. no. 37.
36 We should not be misled by its probably having been restored in the
Renaissance, when it was given its perfectly smooth shape.
37 E. Zwierlein-Diehl, Antike Gemmen in deutschen Sammlungen,
Band II: Berlin, Munich, 1969, no. 445 (white paste; 35.2 x 26.5 x
4.9mm).
38 Light yellow-brown paste, fragment; 19.0 x 18.2 x 3.8mm: E.
Zwierlein-Diehl, Die antiken Gemmen des Kunsthistorischen
Museums in Wien, vol. II, Munich, 1979, cat. no. 565 (inv. no.
1821.169 Nr25).
39 Utrecht, Geldmuseum, RCC inv. no. 129, bright red carnelian, 18.5
x 16mm. See: M. Maaskant-Kleibrink, Catalogue of the Engraved
Gems in the Royal Coin Cabinet, The Hague, 1978, no. 1156; this
intaglio, mistakenly said to be in The Hague, is nicely illustrated in
Gallottini (n. 1), 65. Maaskant-Kleibrink compares it to a carnelian
intaglio of c. 60–50 bc, depicting Asklepios and Hygieia, signed
HEIOY, and measuring 19.0 x 13.8 x 6.1mm in Vienna (E. Zwierlein-
Diehl, Die antiken Gemmen des Kunsthistorischen Museums in
Wien, vol. I, Munich, 1973, cat. no. 205 [inv. IX.B.1550]).
40 Utrecht, Geldmuseum, RCC inv. no. 130, pale orange-red jasper, 14
x 11 x 2.5mm: Maaskant-Kleibrink (n. 39), cat. no. 749.
41 G. Richter, Metropolitan Museum of Art, New York. Catalogue of
Engraved Gems, Rome, 1956, XXXIII.
42 Naples, Museo Archeologico Nazionale, inv. no. 5689 (inv.
Sangiorgio 117): 48 x 48mm, diam. of medallion 25mm, W. 81mm.
See: E. Künzl, ‘Cingula di Ercolano e Pompei’, Cronache Pompeiane
3 (1977), 180, pl. 3. The drawing from G. Finati, Real Museo
Borbonico, Naples, 1831, vol. VII, pl. 48 (Ferd. Mori del. et sculp.).
Künzl’s article was kindly drawn to my attention by Jeffrey Spier.
43 Horster (n. 23), 4.
44 The Venetian specimen is bi-metallic, as is the Viennese one
(Certosini collection). The Roman specimen, from the Vitali
collection, is in the Vatican. The London specimen is in the British
Museum (Inv. no. 1857.0812.2). See: F. Gnecchi, I medaglioni
romani, descritti ed illustrati, Milan, 1912, vol. III, 20, cat. no. 100,
pl. 146.8–9.
45 It is a magnificent, and unique, silver piece. It measures 33.5mm in
diam. and weighs 24.11g, which is roughly the weight of 8 denarii.
The coin is believed to have been struck in Rome. The dies have
been attributed to the Alphaeus Master (i.e. possibly Antoninianus
of Aphrodisias?) – notably on the basis of the very characteristic
letter shapes. From the G. Mazzini collection (Monete Imperiali
Romane, Milan, 1957, vol. 2, 92, pl. XXXII d. 475). After his death, it
was acquired by Ratto in Milan and sold to Leo Biaggi (photo plate
no. 2398) in the late 1950s. After his death, when his collection
was dispersed, it returned to Ratto, and later appeared as
Numismatica Ars Classica (Auction 18, Zurich, 29 March 2000), lot
519; most recently, on 4 January 2009, it reappeared in Heritage
(Auction 3004, New York, 4 January 2009), lot 20075. I am grateful
to Alan Walker for helping me to trace the coin’s provenance, and
for his advice on the Athenian coinage.
46 Caligula (ad 37–41), silver tridrachm, Crete. The British Museum
owns a specimen: CM 1842,0726.4: W. Wroth, British Museum.
Catalogue of the Greek Coins of Crete and the Aegean Islands,
London, 1886, no. 2 (23mm, 7.55g); this provincial issue was kindly
brought to my attention by Richard Abdy. On this coin, the image
of Caligula with a sceptre likens him to the local ‘Dictaean Zeus’
(Zeus of Mount Dicte). Hadrian is occasionally depicted as Zeus,
with an aegis cloak, but never with a sceptre. The bust of the
emperor holding a sceptre would become common under Probus.
47 Bronze medallion of Marcus Aurelius, Bonn, Rheinische
Landesmuseum, inv. no. RLMB 3327, 39mm, 41.48g. See:
F. Imhoof-Blumer and P. Gardner, ‘A Numismatic Commentary on
Pausanias’, Journal of Hellenistic Studies (1885–7), 130, pl. Z.15;
Ghedini (n. 6), ill. 2; C. Klages, ‘Athena gegen Neptun. Ein
Medaillon des Antoninus Pius’, Das Rheinische Landesmuseum
Bonn. Berichte aus der Arbeit des Museums 4 (1990), 55–7.
48 A. Birley, Hadrian the Restless Emperor, London, 2001, 220.
Interestingly, the silver medallion does indeed portray Hadrian as
Zeus.
49 Classical Numismatic Group, auction ‘Triton V’, 15 January 2002, lot
363, 12.19g. Cf. Imhoof-Blumer and Gardner (n. 47), 130, pl.
Z.11/12/14; J. Svoronos, Trésor des monnaies d’Athènes, Munich,
1923–26, pl. 89.6–10; J. Shear, ‘The coins of Athens’, Hesperia II.2
(1933), 276, cat. no. C.1; J. Kroll and A. Walker, The Athenian Agora.
XXVI. The Greek Coins, Princeton, 1993, 174; H.-C. von Mosch,
Bilder zum Ruhme Athens, Milan, 1999, 16. Based on a mistake by
Josephine Shear (‘Athenian imperial coinage’, Hesperia V (1936),
297, pl. 8.1), Francesca Ghedini erroneously considered some of the
Athenian coins to predate the gems, and therefore concludes that
the gems were copied from the coins. Nonetheless, she knew that
Shear was contradicting other numismatists (Ghedini [n. 6], n.
46), and Shear’s theory is now rejected.
50 Künker (Osnabrück), auction 124, 16 March 2007, lot 7961, 7.61g. Cf.
Imhoof-Blumer and Gardner (n. 47), pl. Z.16; Svoronos (n. 49), pl.
89.11; Kroll and Walker (n. 49), 261.
51 Cf. Shear 1936 (n. 49), 297, pls 8.9–8.11; Kroll and Walker (n. 49),
355.
52 Gorny & Mosch (Munich), auction 159, 8 October 2007, lot 300,
25.73g.
53 From the Panticapaeum Necropolis. Last quarter of the 4th century
bc, 51cm high. Excavations of A.E. Lyutsenko, 1872.
54 Madrid, Museo Arqueológico Nacional, inv. no. 11095. See: Ghedini
(n. 6), ill. 4.
55 Cordova, Museo Arqueológico. See: Ghedini (n. 6), ill. 3; C. Picard,
Manuel d’archéologie grecque. La sculpture, vol. IV.2, Paris, 1963,
495, pl. 205; A. García y Bellido, Esculturas romanas de España y
Portugal, Madrid, 1949, pl. 409. Picard also refers to a relief from
Smyrna, but gives no details.
56 See, S. Dumitrascu, Tezaurul de la Tauteni-Bihor, Oradea, 1973, pls
34–42.
57 Sold at the Parke-Bernet Galleries in New York (Part II of the Notable
Art Collection belonging to the Estate of the Late Joseph Brummer,
11–14 May 1949, lot 227), it was formerly in the collection of Prince
Nicholas Gagarine (1822–1905). Mistakenly said to be at the Walters
Art Museum, where exhibited in 1947, by Gallottini (n. 1), 70, pl. 3.
See: Early Christian and Byzantine Art. An Exhibition held at the
Baltimore Museum of Art April 25-June 22 (exh. cat., Walters Art
Gallery), Baltimore, 1947, 113, no. 551, pl. 67; J. Spier, ‘Late Antique
cameos’, in M. Henig and M. Vickers (eds), Cameos in Context,
Oxford and Houlton, pl. 3.7; idem, Late Antique and Early Christian
Gems, Wiesbaden, 2007, no. 759.
58 Ghedini (n. 6), n. 64, believes that these two figures could be the
Dioscuri; she noticed that Poseidon was associated with the
Dioscuri on the lost base of Poseidon’s statue at the Isthmus of
Corinth (Pausanias II.1.7–9: see, C. Picard, Revue d’Etudes Latines
35 (1957), 299).
59 A similar hypothesis (that the Medici type is a simplified one) was
made by Goffredo Bendinelli (Sulle tracce di opere fidiache andate
perdute, Turin, 1954), who saw the Madrid crater as being more
‘Gems of Heaven’ | 273
Rambach
complex and the most original, since it is the oldest version of the
composition.
60 As Antje Krug kindly remarked after my paper at the British
Museum, the Byzantine court liked relics, but not especially
ancient jewels, and we have no reason to think that the cameo ever
travelled to the Eastern Empire.
61 On the subject of gift giving and gem replicas in the Renaissance,
see: L. Clark, ‘Transient possessions: circulation, replication, and
transmission of gems and jewels in Quattrocento Italy’, Journal of
Early Modern History 15.3 (2011), 185–221.
62 Paris, Bibliothèque nationale, Cabinet des médailles, 95 x 78mm
including gold mount, mid-13th century. It was in the royal
collection in 1379, from which it passed to ‘une des plus anciennes
églises de France’, from which it was acquired c. 1685 according to
Marc-Antoine Oudinet (Keeper, 1689–1712). As is well known,
Hohenstaufen gems have long been misidentified: for example,
the Paris cameo was considered by Gisella Richter (n. 21), no. 65, to
be Roman and to have simply ‘suffered some retouching’ in the
Renaissance, whilst Erika Zwierlein-Diehl (n. 38), cat. no. 565,
considered it to be possibly Claudian and heavily re-engraved after
antiquity. See: J. Labarte, Inventaire du mobilier de Charles V, roi de
France, Paris, 1879, 308: ‘Item, un cadran d’or, où il a ung grant
camahieu, ouquel il a ung homme, une femme et ung arbre ou
mylieu, et aux cins dudit cadran, a, par embas, ung saphir et ung
balay, chascun environné de trois perles, et deux perles à l’un des
costez, pesant quatre onces cinq estellins’; E. Babelon, Catalogue
des camées antiques et modernes de la Bibliothèque nationale, Paris,
1897, cat. no. 27; Dacos (n. 3), pl. 62; R. Haussherr et al., Die Zeit der
Staufer, Stuttgart, 1977, vol. 1, cat. no. 886, 693–4, vol. 2, pl. 660; R.
Distelberger, Die Kunst des Steinschnitts, Vienna, 2002, 60–2, cat.
no. 17).
63 Vienna, Kunsthistorisches Museum, Inv. no. XII 143; 35 x 26mm, 62
x 50mm including gold mount. The mount is typical of the courtly
jewellery made in Prague for the Habsburgs around 1600; it first
appears in a 1750 inventory. See: Haussherr (n. 62), vol. 1, cat. no.
887, 694–5, vol. 2, pl. 659, vol. 5, 497; Dacos (n. 3), pl. 63.
64 I am most grateful to Shua Amorai-Stark for examining the
Hebrew inscription engraved in Sephardic letters on this cameo. It
reads ‫ִּׁשה ּכִי טֹוב ָהעֵץ ְל ַמ ֲאכָל‬
ָ ‫הּוא ָלעֵינַיִם וְנֶ ְחמָד ָהעֵץ וַ ֵּתרֶא ָהא‬-‫ וְכִי ַת ֲאוָה‬and is the
first half of a sentence in Genesis 3:6. A study of the script reveals
spelling mistakes and miscomposed letters, due to the
incompetence of the engraver (and/or the incorrectness of his
model). Indeed, and I quote Dr Amorai-Stark, ‘the mistakes in the
writing of certain letters, for example of the ‫ ש‬ (sh) in ‫ִּׁשה‬
woman) or the ‫( ט‬t ) in ‫( טֹוב‬good); and some of the words are
divided into units all show that the engraver did not know
Hebrew’, which indicate that neither engraver nor patron were
Jews. The tradition of non-Jews introducing Hebrew words into
works of art intensified in the late Renaissance (Rembrandt’s
paintings are an example), and Dr Amorai-Stark considers that this
inscription is likely to have been added in the 16th or 17th century,
and is not contemporary with the engraving of the gem.
65 Dacos (n. 3), pl. 81. This relief was formerly considered to be by
Donatello (c. 1386–1466), because of Vasari having written that ‘in
the first court of the Casa Medici [i.e. the Palazzo Riccardi] there
are eight marble medallions containing representations of antique
cameos’ by him (G. Vasari, The Lives of the Painters, Sculptors and
Architects, London, 1970, vol. I, 306). Some scholars have believed
that it is in fact by Donatello’s studio (U. Wester and E. Simon, ‘Die
reliefmedaillons im Hofe des Palazzo Medici zu Florenz’, Jahrbuch
der Berliner Museen VII–1 (1965), 15–91), but it now seems that they
are not related to Donatello at all: indeed, on 2 July 1452 Maso di
Bartolomeo (1406–56) was paid for drawing these medallions
(I. Hyman, Fifteenth century Florentine Studies: the Palazzo Medici
and a Ledger for the Church of San Lorenzo, New York and London,
1977, 208–9).
274 | ‘Gems of Heaven’
66 A. Chabouillet, Catalogue général et raisonné des camées et pierres
gravées de la Bibliothèque impériale, Paris, 1858, cat. no. 425:
‘imitation grossière’; Babelon (n. 62), cat. no. 462: ‘travail
médiocre’.
67 E. Zwierlein-Diehl, ‘Tellus and the Seasons: Gem Copies of a
Roman Medallion type’, in M. Henig and D. Plantzos (eds),
Classicism to Neo-classicism: Essays dedicated to Gertrud Seidmann,
Oxford, 1999, 67–77.
68 Ex-Blacas collection, 1866. Cast Cades VI-78. See: Dalton (n. 20),
cat. no. 591. I should add to this list of 16th-century works of art and
gems that the reverse of a 1527 bronze medal by Giovanni Maria
Pomedelli (1478/9 – after 1537) for Giovanni Ermo (c. 1483–1542) is
inspired by the Athena and Poseidon scene (J. Klauß, Die
Medaillen-sammlung Goethes, Band I, Berlin, 2000, 33, no. 40;
J.G. Pollard, Renaissance Medals, Washington DC, 2007, vol. 1, 217,
cat. no. 200). I am not certain whether the reverse of a 1572 bronze
medal by Antonio Abondio (1538–91) for Jacopo Nizzola da Trezzo
(1514?–89) is also inspired by the same design (S. Scher, The
Currency of Fame: Portrait Medals of the Renaissance, New York,
1994, 170, ill. 60).
69 On the subject of daktyliothecæ, see the exhibition catalogue by
V. Kockel and D. Gäpler, Daktyliotheken, Götter und Caesaren aus
der Schublade: Antike Gemmen in Abdrucksammlungen des 18. und
19. Jahrhunderts, Munich, 2006.
70 No signature is visible on the cast, and the ascription to Cerbara is
based on the hand-written catalogue by Tommaso Cadès (book 70,
no. 659); I would like to thank Getrud Platz for confirming this.
According to the biographies by Forrer, the theme of this gem
seems closer to the type of subjects engraved by his brother Niccolo
Cerbara (1793–1869). See: G. Lippold, Gemmen und Kameen des
Altertums und der Neuzeit, Stuttgart, 1922, pl. CII-4; L. Pirzio Biroli
Stefanelli, ‘Fortuna delle gemme Farnese nel XVIII e XIX secolo’, in
C. Gasparri (ed.), Le Gemme Farnese, Naples, 1994, pl. 144; eadem,
La collezione Paoletti II, forthcoming, VI-232.
71 From the collection of Raphael Esmerian (1903–76), then S.J.
Phillips, then Richard Trescott, now in a private collection. The
intaglio measures 24.6 x 19.4 x 6.6mm; 3.54g.
72 This applies to the privately owned piece: the British Museum gem
is too worn to be certain, and the picture in Lippold is not detailed
enough.
73 Vienna ,Kunsthistorisches Museum, AS inv. no. XII 683, Catalogue
des pierres gravées par Louis Siriès orfèvre du roi de France,
Florence, 1757, p. 30, no. 38. See: A. Bernhard-Walcher,
ָ ‫ ָהא‬ (the ‘Geschnittene Steine des 18. und 19. Jahrhunderts in der Antiken-
sammlung des Kunsthistorischen Museums Wien’, Zeitschrift für
Kunstgeschichte 59.2 (1996), 163. Mention should also be made of a
neo-classical basso-rilievo with Athena and Poseidon, in the
fourth wall of the entrance hall of the Villa Borghese (Rome),
commissioned by Marcantonio IV Borghese (1730–1800), which
features ‘cameo’ stuccoes by Vincenzo Pacetti (1746–1820),
Francesco Carradori (1747–1825), Massimiliano Laboureur (1767–
1831), Tommaso Righi (1727–1802), Luigi Salimei (1736–1817) and
A. Brunetti.
74 See: Catalogue des Pierres Gravées Antiques de S.A. le Prince
Stanislas Poniatowski, Florence, 1830–2, cat. no. I.218; J.
Prendeville, Explanatory catalogue of the proof-impressions of the
antique gems possessed by the late Prince Poniatowski, and now in
the possession of John Tyrrell, ... an essay on ancient gems and gem-
engraving, London, 1841, cat. no. 143.
75 Poniatowski (n. 74), cat. no. I.217; Berlin Daktyliothek Poniatowski
(a set of 419 plaster impressions presented to the King of Prussia by
the prince himself in 1832), no. 32; Christie’s, London, sale of the
Poniatowski gems, 29 April–22 May 1839, cat. no. 146, bought by
Norton; S.J. Phillips. The Wellington Gems published on the occasion
of an exhibition, London, 1977, cat. no. 418.9.