AMERICAN JOURNAL

OF ARCHAEOLOGY
THE JOURNAL OF THE ARCHAEOLOGICAL INSTITUTE OF AMERICA

Volume 109 • No. 4 October 2005

THE AMERICAN JOURNAL OF ARCHAEOLOGY, the journal of the Archaeological Institute of
America, was founded in 1885; the second series was begun in 1897. Indices have been published
for volumes 1–11 (1885–1896), for the second series, volumes 1–10 (1897–1906) and volumes 11–
70 (1907–1966). The Journal is indexed in the Humanities Index, the ABS International Guide to Classical
Studies, Current Contents, the Book Review Index, the Avery Index to Architectural Periodicals, Anthropological
Literature: An Index to Periodical Articles and Essays, and the Art Index.

MANUSCRIPTS and all communications for the editors should be addressed to Professor Naomi J.
Norman, Editor-in-Chief, AJA, Department of Classics, Park Hall, University of Georgia, Athens,
Georgia 30602-6203, fax 706-542-8503, email nnorman@aia.bu.edu. The American Journal of Archaeology
is devoted to the art and archaeology of ancient Europe and the Mediterranean world, including the
Near East and Egypt, from prehistoric to late antique times. The attention of contributors is directed
to “Editorial Policy, Instructions for Contributors, and Abbreviations,” AJA 104 (2000) 3–24. Guidelines
for AJA authors can also be found on the World Wide Web at www.ajaonline.org. Contributors are
requested to include abstracts summarizing the main points and principal conclusions of their articles.
Manuscripts, including photocopies of illustrations, should be submitted in triplicate; original
photographs, drawings, and plans should not be sent unless requested by the editors. In order to
facilitate the peer-review process, all submissions should be prepared in such a way as to maintain
anonymity of the author. As the official journal of the Archaeological Institute of America, AJA will
not serve for the announcement or initial scholarly presentation of any object in a private or public
collection acquired after 30 December 1973, unless its existence was documented before that date or
it was legally exported from the country of origin. An exception may be made if, in the view of the
Editor-in-Chief, the aim of the publication is to emphasize the loss of archaeological context. Reviews
of exhibitions, catalogues, or publications that do not follow these guidelines should state that the
exhibition or publication in question includes material without known archaeological findspot.

BOOKS FOR REVIEW should be sent to Professor John G. Younger, Editor, AJA Book Reviews, Classics
Department, Wescoe Hall, 1445 Jayhawk Boulevard, University of Kansas, Lawrence, Kansas 66045-
2139, tel. 785-864-3153, fax 785-864-5566, email jyounger@ukans.edu. The following are excluded
from review and should not be sent: offprints; reeditions, except those with great and significant
changes; journal volumes, except the first in a new series; monographs of very small size and scope;
and books dealing with the archaeology of the New World.

THE AMERICAN JOURNAL OF ARCHAEOLOGY (ISSN 0002-9114) is published four times a year in
January, April, July, and October by the Archaeological Institute of America, located at Boston Uni-
versity, 656 Beacon Street, Boston, Massachusetts 02215-2006, tel. 617-353-9361, fax 617-353-6550,
email aia@aia.bu.edu. Subscriptions to the American Journal of Archaeology may be addressed to the
Institute headquarters. A subscription request form can also be downloaded from the AJA Web site
(www.ajaonline.org). An annual subscription is $75 (international, $95); the institutional rate is $250
(international, $290). Membership in the AIA, including a subscription to AJA, is $125 per year
(C$192). Student membership is $73 (C$118.50); proof of full-time status required. International
subscriptions and memberships must be paid in U.S. dollars, by a check drawn on a bank in the U.S.,
by money order, or by credit card. Subscriptions due 30 days prior to issue date. No replacement for
nonreceipt of any issue of AJA will be honored after 90 days (180 days for international subscriptions)
from the date of issuance of the fascicle in question. When corresponding about memberships or
subscriptions, always give the account number, as shown on the mailing label or invoice. A microfilm
edition of the AJA, beginning with volume 53 (1949), is issued after the completion of each printed
volume. Subscriptions to the microfilm edition, which are available only to subscribers to the printed
AJA, should be sent to ProQuest Information and Learning (formerly Bell & Howell Information and
Learning), 300 North Zeeb Road, Ann Arbor, Michigan 48106. Back issues of AJA and the Index 1907–
1966 may be ordered from the Archaeological Institute of America. Back issues of the AJA from 2000–
present are available in nonprintable pdf format on the AJA Web site. Exchanged periodicals and
correspondence relating to exchanges should be directed to the Archaeological Institute of America
in Boston. Periodicals postage paid at Boston, Massachusetts, and additional mailing offices. Postmaster:
send address changes to the American Journal of Archaeology, Archaeological Institute of America,
located at Boston University, 656 Beacon Street, Boston, Massachusetts 02215-2006.

The opinions expressed in the articles and book reviews published in the American Journal of Archaeology
are those of the authors and not of the editors or of the Archaeological Institute of America.

Copyright © 2005 by the Archaeological Institute of America

The American Journal of Archaeology is composed in ITC New Baskerville

at the offices of the Archaeological Institute of America, located at Boston University.

The paper in this journal is acid-free and meets the guidelines for permanence and durability of the
Committee on Production Guidelines for Book Longevity of the Council on Library Resources.
 New Archaic Coin Finds at Sardis
NICHOLAS CAHILL AND JOHN H. KROLL
Abstract
Two Archaic coins were discovered in excavations of
the western fortification of Lydian Sardis in the summer
of 2002. The coins, one gold and one silver, are Lydian
12th staters with the confronting foreparts of a lion and
bull on the obverse and a square incuse punch on the
reverse, the type of coins that are commonly referred to
today as “croeseids.” They come from under and around a
cobbled pavement belonging to the fortification. The
pavement and the coins under it were sealed by destruc-
tion debris when the fortification was demolished, an
event that can be confidently assigned to the capture of
Sardis by Cyrus the Great of Persia in the 540s B.C. A
third coin, although not recognized as such at the time,
was found in 1988 in the destruction debris itself with
the skeleton of a young man, probably a casualty of the
battle. Upon cleaning, this coin proved to be a croeseid
silver 24th stater. The discovery of these coins from a
secure context offers a new and much-needed fixed point
in the chronology of the Archaic coinage of Asia Minor.*
In reconstructing the chronology of early Greek
coinage, the bimetallic coinage conventionally at-
tributed to Lydian King Croesus has long served as
a crucial point of reference. The gold and silver
staters and fractions of this coinage display on their
obverses the confronting foreparts of a lion and a
bull; reverses are punched with one or two typeless
incuse squares. For more than a century and a half
it has been customary to refer to such coins as
“croeseids” on the grounds that the gold staters are
to be identified with the gold Kroiseioi stateres men-
tioned in ancient sources; accordingly, numisma-
tists have dated the introduction of this coinage to
the reign of Croesus, around the mid sixth century
B.C. Since the 1980s, however, several prominent
* N. Cahill, the excavator of the coins, is responsible for
writing pp. 589–609; J.H. Kroll for pp. 609–13. The authors are
grateful to Isabelle Pafford, numismatist during the 2002 sea-
son, when the coins were found, who did much of the prelimi-
nary work on the coins; conservator Kent Severson, who
cleaned the two impossible silver croeseids; Andrew Ramage
for his expertise on Lydian coinage and precious metals; Rich-
ard Stone, conservator at the Metropolitan Museum of Art; two
anonymous reviewers for their helpful suggestions; and par-
ticularly Crawford H. Greenewalt, Jr., field director of the Sardis
Expedition, for his permission to publish and for his constant
encouragement and wise advice. The debt to many other spe-
cialists, particularly on ceramics and chronology, is acknowl-
edged throughout the text.
1
Inv. C02.1/2002.2, 0.88 g, diam. 7.2 mm.
589
American Journal of Archaeology 109 (2005) 589–617
scholars have challenged the identification and/
or its chronological relevance, arguing that the coin-
age was inaugurated at some later time in the sixth
century by one of the Persian kings who ruled Lydia
after Croesus was overthrown. Against this back-
ground of speculation and chronological uncer-
tainty, new and decisive evidence for the chronology
unexpectedly emerged in the recovery of three of
these coins during recent excavations of the great
Lydian fortification at Sardis.
discovery of the coins
Two of the coins were found during excavation
of the fortification in the summer of 2002. These
are Lydian 12th staters (hemihektai) of the croeseid
type: one gold,1 the other silver (figs. 1–3).2 The third
coin is a silver 24th stater, also of the croeseid type
(figs. 1, 4). 3
The first two coins were found associated with a
recess on the west side of the fortification, origi-
nally uncovered between 1984 and 1989. 4 Three
sondages were subsequently dug to explore ear-
lier levels of the recess (figs. 5a–c, 6). One sondage
had been backfilled. A balk remained between the
other two sondages; excavation of this balk was a
goal of the 2002 season. 5 The gold croeseid was
found in the balk, sealed in a layer of undisturbed,
laminated sandy earth about 10 cm under the floor
of the recess (figs. 2, 5[1], 7[1]).6 The context of
the coin was, therefore, completely secure; there is
no realistic possibility of contamination.
The discovery of this coin prompted full excava-
tion of the remainder of the cobbled floor and asso-
2
Inv. C02.3/2002.8, 0.72 g after cleaning (1.06 g before
cleaning), diam. 7.9 mm.
3
Inv. M88.12:9609/C04.1/2004.11, 0.35 g after cleaning
(0.48 g before cleaning), diam. 6.3 mm. Published as “uncer-
tain metal object” (Greenewalt et al. 1993, 20–1, figs. 14, 15,
and 40 n. 19).
4
Greenewalt et al. 1987a, 20–1; 1987b, 71–2; 1990, 141–3;
1993, 14–23; Greenewalt 1990, 10–4.
5
Greenewalt 1990, 13–4; Greenewalt and Rautman 1998,
494–5; 2000, 663–4. The excavation was prompted by the col-
lapse of a roof covering the sondage in the southeast corner,
under the weight of an unusually heavy winter snowfall.
6
Bas. 6, E 107.4 / S 95.2, *100.27 (* elev.). On the system
of grids and levels at Sardis, see Hanfmann and Waldbaum
(1975, 7–11).
590 NICHOLAS CAHILL AND JOHN H. KROLL [AJA 109

Fig. 2. Gold croeseid 12th stater. (© Archaeological

Exploration of Sardis/Harvard University)

Fig. 1. Gold and silver croeseid 12th staters and silver

croeseid 24th stater. (© Archaeological Exploration of Sardis/
Harvard University)

ciated strata, both to learn more about the stratigra-

phy of this area and to forestall illicit excavation.
Parts of the floor had become eroded in the 14 years Fig. 3. Silver croeseid 12th stater. (© Archaeological
Exploration of Sardis/Harvard University)
since it had been exposed. The area was cleaned of
inwash and backfill before proper excavation began.
Figure 5 shows the preserved extent of the cobble
surface at the beginning of the 2002 season.
The silver 12th stater was found 2.3 m from the
west scarp of the trench, in an area where the pave-
ment of the recess had eroded (figs. 3, 5[2], 7[2]).7
This area had been cleaned of all later material,
and although the coin was not sealed by the cobble
pavement at the time of excavation, it must belong
with the stratum that predates the surface. When
found, it was almost unrecognizable as a coin but was Fig. 4. Silver croeseid 24th stater. (© Archaeological
enveloped in a hard, brown mantle of corrosion that Exploration of Sardis/Harvard University)
had to be painstakingly chipped away with a blade to
reveal the lion-and-bull motif and incuse punch. to recognize it as a coin, and the object was invento-
The cleaning of this initially unrecognizable sil- ried as a metal disk of uncertain function. Careful
ver coin prompted the re-examination of a tiny, manual removal of its hard, mineralized coating re-
heavily corroded metal disk excavated in 1988. This vealed that this was another Lydian croeseid coin,
disk was found next to the skull of a man in his early a 24th stater, with lion-and-bull protomes on the
20s, probably a fallen soldier, discarded in the “brick obverse and a deep punch on the reverse (fig. 4).
fall” destruction debris that filled the recess (figs. Another Archaic coin was turned in to the expe-
5[3], 8). The skeleton had been dumped outside dition two days after the discovery of the gold
the fortification shortly after the destruction and was croeseid: a silver 12th stater of Miletus with the
lying upside down, with his torso above his skull and forepart of a lion reclining to the left, its head turned
limbs splayed. At the time, a surface test for silver back to the right on the obverse, and a star com-
proved inconclusive; the surface was too corroded posed of floral elements on the reverse. 8 It was said

7
Bas. 11, E 102.95 / S 92.50, *99.95.
8
C02.2/2002.3, 1.08 g, diam. 10 mm. For the type, SNG
Danish National Museum, Copenhagen: Ionia, nos. 944–950,
952, 953; Becker 1988, 14–6, 32 (Early Series). For a tabula-
tion of hoards from Asia Minor containing this relatively com-
mon type of coin, which is typologically more advanced than
the thicker, unifacial croeseids, see Becker (1988, 28–9). The
authors wish to thank Richard Stone, Donna Strahan, Kent
2005] NEW ARCHAIC COIN FINDS AT SARDIS
to have been found in the sieving of earth cleaned
from the recess before proper excavation outside
the balk began. At first, the expedition believed
that, despite the mixed context, this coin was also
associated with the cobble surface, like the gold
and silver croeseids. In contrast to the two silver
croeseids, however, the Milesian coin was hardly
corroded and was legible with only light brushing.
This led Richard Stone, senior museum conserva-
tor at the Metropolitan Museum of Art, to conclude
that the coin had been deliberately cleaned before
it came into our hands, and other conservators have
agreed with his conclusions. We, therefore, now be-
lieve that this coin must be intrusive (see appx.).
The discovery of three Archaic coins in this small
area remains unprecedented and difficult to ex-
plain. They are among a very small number of Lydian
coins uncovered by the current expedition at
Sardis.9 No other Archaic coins have been found in
the decades of excavation of equivalent strata in
this sector, despite careful sifting that would have
caught such finds. Discovering three coins in this
one area, two from the same stratum, might just be
a peculiar and providential fluke, but there may
also be some specific reason why they were found
in this part of the fortification. The circumstances
of discovery of the two coins sealed beneath the
floor do not suggest that they were deliberately
placed in this fill, in the manner of a foundation
deposit or other intentional placement. Rather, they
seem to have been lost accidentally during the con-
struction of the wall. The third coin belonged to a
soldier who was killed in the capture of Sardis. The
findspot of the coin, next to the head of the skel-
eton, suggests that he was either carrying the coin
in a pouch hung from his neck or in his mouth,
like characters in Aristophanes (Vesp. 609, 791; Av.
503; Eccl. 818) who carried small change in their
mouths as a matter of course.
the fortification and recess
The recess that contained the coins is part of the
fortification of Sardis from the Lydian period, the
seventh and sixth centuries B.C. (fig. 9[64a]).
Twenty meters thick at the base and still standing
Severson, and Dylan Smith for their expert advice on the coin’s
condition. The coin is published in Greenewalt 2003a, 483,
fig. 7; 2003b, 65–6; 2005, 82.
9 11
A croeseid silver 12th stater discovered in sector PN, on
the floor of an apsidal building probably dating to the second
half of the sixth century B.C., is the only other croeseid found
in a stratified context (Buttrey et al. 1981, no. 133). A gold
sixth stater turned in to the expedition in 1963 was reportedly
found in the mud baths in the foothills of the Tmolus behind
591
up to 10 m high, this well-preserved structure was
one of the most massive fortifications of its era, ri-
valed only by the walls of the great cities of
Mesopotamia, and has been an important focus of
excavation at Sardis since its discovery in 1976, in
sectors MMS, MMS/N, and MMS/S (figs. 10, 11).
Recent fieldwork has uncovered remains prob-
ably belonging to the continuation of the fortifica-
tion around the Lydian city. To the south of sector
MMS, a stretch of wall follows a natural ridge toward
the acropolis, anticipating the line of the later Ro-
man fortification (fig. 9[64b]). To the north and
east, the Lydian fortification probably forms the core
of a series of mounds just inside the Roman city
wall (fig. 9[64c]). At the east edge of the city the
fortification again follows a natural ridge and un-
derlies the Roman city wall (fig. 9[64d]). 10 The
stretch of fortification in sector MMS was, therefore,
part of the western circuit of the Lydian city, and
the coins were found outside this line of defense.
Despite more than 25 years of research, impor-
tant questions remain about the design and phas-
ing of the fortification, even in its best-understood
stretch, sector MMS. The wall underwent consider-
able modifications in its relatively short history, and
even within one phase its construction is often het-
erogeneous and complex. The following summary
is simplified to give a general idea of the architec-
tural context of the coins.
In an early phase, some or all of the wall was built
of mudbrick set on a stepped stone socle (fig. 7).
This structure was about 20 m thick at the base, and
its core still stands at least 8 m high. The fortifica-
tion was broken by a gate just north of the modern
road. The way through the gate was paved with a
series of pebbled and cobbled surfaces, some of
which bear wheel ruts.11
The mudbrick structure may not be the earliest
fortification here. Monumental remains, perhaps
predecessors of the fortification, have been exposed
in sondages near the modern highway and in the
recess where the coins were found. A goal of the
excavation in 2002 was to clarify one of these early
features, a sandstone wall face partly exposed in
1995.12
Sardis (Buttrey et al. 1981, no. 132).
10
Greenewalt et al. 1987b, 80–4 (mound 2); Greenewalt
and Rautman 2000, 672–3 (mound 2); Greenewalt n.d.
Greenewalt et al. 1995, 12; Greenewalt and Rautman 2000,
658–60.
12
Greenewalt 1979, 24–5; Greenewalt et al. 1995, 12;
Greenewalt and Rautman 1998, 493, figs. 17, 18, 494–5, fig.
20.
592 NICHOLAS CAHILL AND JOHN H. KROLL [AJA 109

Fig. 5. Plan of recess. (© Archaeological Exploration of Sardis/Harvard University)

The original wall seems to have become weath-
ered or damaged, necessitating extensive repairs
and modifications.13 The relative and absolute chro-
nology of these changes is uncertain; they may be-
long broadly to one campaign of repair and
redesign, or they may have taken place piecemeal
over a longer period of time. The surface with the
Archaic coins was associated with one of these
phases of modification, late in the building history
of the fortification.
The gate and much of the eastern, inner face of
the fortification was resurfaced in stone masonry.
The west side of the fortification, in contrast, was
radically transformed. The eroded mudbrick face

13
Eroded mudbrick faces have been identified on the west
face under the earthwork near the highway (Greenewalt n.d.);
just north of the recess in which the coins were found (fig. 5
[8]) (Greenewalt et al. 1993, 15); and, refaced with stone, on
the east face of the fortification in MMS/S (Greenewalt et al.
1994, 19–20; 1995, 20–1). The mudbrick east face, first discov-
ered in 1979 (Greenewalt et al. 1983, 2–6), stands on what
seems to be the original stone socle but shows very little weath-
ering; this stretch may have been subsequently refaced with
mudbrick. The feature originally interpreted as a platform
extending from the west face (Greenewalt et al. 1983, 3) sub-
sequently proved to be the top of the eroded, earlier phase of
the fortification, upon which an upper, narrower wall was built
at about the time the glacis was created (Greenewalt n.d.).
2005] NEW ARCHAIC COIN FINDS AT SARDIS
Fig. 6. Recess, after excavation of sondages (1998). (© Archaeological Exploration of Sardis/Harvard University)
was buried by a sloping mound of earth (figs. 10,
11).14 The mound has been exposed to a width of
23 m from the face of the wall, and it certainly con-
tinued farther west; it more than doubled the origi-
nal 20 m thickness of the fortification to at least 43
m. It rises to a height of at least 12 m, completely
burying the original mudbrick face. Rain had
eroded rills into the mound in the one small area
where the original surface has been preserved and
excavated. One such rill contained the partly ar-
ticulated skeleton of a dog; another was filled with
destruction debris and contained a partial human
skeleton.
This mound has been interpreted as a glacis, built
to prevent siege engines from approaching the face
of the wall and to prevent sappers from undermin-
ing or tunneling through or under the wall. An
alternative interpretation, that it represents a siege
mound laid against the fortification, must be re-
jected because the walls of the recess were clearly
built as part of the same program to retain the
earthwork; these are too carefully constructed to
belong to a siege mound. The mound is, however,
14
Greenewalt et al. 1993, 15–6, with previous references;
Greenewalt n.d.
593
no steeper than the siege ramp at Lachish, and, there-
fore, not too steep for an enemy to have dragged up
battering rams and other siege engines.15
The Recess and the Findspots of the Coins
Some 70 m south of the gate, the glacis was inter-
rupted by a recess, in which were found the Archaic
coins. The recess was formed by a vertical stone back
face and two sidewalls (figs. 5–7, 10[1], 11). Both
sidewalls have finished faces only on one side since
they retained a fill of earth and are narrower toward
their west ends, perhaps because their termini re-
tained a lesser weight of earth. Although built with
different construction techniques, the three walls
bond and belong to the same phase. The recess
walls are preserved to a maximum height of 5.5 m,
and the glacis retained behind the north wall rises
higher still.
The recess was floored with a surface of pebbles
and fist-sized cobbles set in earth. This floor was
originally thought to be roughly at the level of the
natural ground in front of the fortification. Further
excavation showed, however, that the glacis contin-
15
Eph’al 1984; Ussishkin 1990.
 594
NICHOLAS CAHILL AND JOHN H. KROLL

Fig. 7. Composite section through recess. (© Archaeological Exploration of Sardis/Harvard University)

[AJA 109
2005] NEW ARCHAIC COIN FINDS AT SARDIS 595

ues down at least another 5 m to the west, so the

recess is perched on a slope. The slope may well
have continued significantly farther down, perhaps
into a ditch in front of the wall; the lower part of the
glacis remains buried and inaccessible under more
than 8 m of post-Lydian deposits.
The purpose of this recess is not clear. Located
on the outside of the fortification, it might have
served as a kind of staging area or defensive plat-
form, where soldiers could have gathered and kept
defenders at a distance. However, it seems to have
been open on the west, leaving the defenders ex-
posed. Temporary or movable structures might have
helped protect combatants. Four grids of nails still
preserving wood pseudomorphs were found on the
recess floor, the only remains of wooden structures
here. These are of different sizes: 2.3 x 0.45 m; 0.8
x 0.95 m; 1.04 x 0.90 m; and, more irregularly
shaped, 1.3 x 1.0 m. These have been interpreted
as shutters or other features fallen from above and
originally belonging to the fortification; alterna-
tively, they might have been the remains of protec-
tive wooden shields or other devices, abandoned
here in the recess. However, there is a notable ab-
sence of other military equipment directly associ-
ated with the use of the recess; while arrowheads
and additional weapons are found in other parts of Fig. 8. Skeleton within brick fall, associated with silver
croeseid 24th stater. (© Archaeological Exploration of Sardis/
the fortification, the only armament discovered Harvard University)
within the recess was a helmet, dumped here after
the fortification had been destroyed.
Three other recesses have been excavated in this fortification were burned and nearby houses were
stretch of the fortification. On the west face, about looted and razed, their contents smashed and aban-
16.5 m south of the recess with the coins, is another doned on the floors. The buildings were then bur-
retaining wall, beyond which is a cobble surface ied under a thick layer of debris. This destruction
(fig. 10[2a]). This might be the north wall of a dif- and associated debris are among the most distinc-
ferent recess, although farther south the fortifica- tive contexts excavated at Sardis.
tion has undergone still more modification, and The destruction involved extensive burning.
the situation remains unclear. Part of another grid Parts of the east stone face of the fortification in
of nails was found on the surface here, similar to MMS/S are reddened and spalled by fire. 17 The
but larger than those of the northern recess.16 The debris contains charcoal, and burned and shattered
east, inner face of the fortification is articulated with stones, perhaps belonging to buildings on the su-
two recesses (fig. 10[2b, 2c]). Unlike the western perstructure of the fortification, were found at the
recess, these were framed by vertical or near-verti- bottom of the layer on the east face. Most surviving
cal walls rather than sloping earthworks, and their stretches of the fortification, however, show no traces
floors are at ground level rather than being set high of fire. Perhaps these areas did not burn, or only the
up on the fortification. The purposes of these re- upper parts of the fortification burned, leaving the
cesses are also unclear. lower sections undamaged. The fortification in other
parts of the city, on the east and north, were also
The Destruction and “Brick Fall” extensively burned, presumably at the same time.18
The Lydian fortification was destroyed close to Shortly after the fire, the upper part of the fortifi-
the middle of the sixth century B.C. Parts of the cation was deliberately demolished and its remains

16 17
Greenewalt et al. 1985, 78; Greenewalt and Rautman 2000, Greenewalt et al. 1994, 19.
18
665–9. Greenewalt et al. 1987b, 80–4; Greenewalt n.d.
 596
NICHOLAS CAHILL AND JOHN H. KROLL

Fig. 9. Plan of Sardis. (© Archaeological Exploration of Sardis/Harvard University)

[AJA 109
2005] NEW ARCHAIC COIN FINDS AT SARDIS
dumped over the side, burying the lower part of
the fortification and adjacent areas under a thick
layer of debris (figs. 7, 10, 12). This rubble consists
mostly of whole and broken bricks and is, there-
fore, referred to as “brick fall.” The dense layer of
brick sealed and protected the floors beneath it, as
well as preserved the standing remains of the forti-
fication. Because the brick fall is easily recogniz-
able in different parts of the sector and on both
sides of the fortification, and because the destruc-
tion deposits contain consistent types of vases and
other material, this stratum can be considered as a
single horizon, and dates and conclusions derived
from one area apply to all. The layer provides a criti-
cal terminus ante quem for any material found in
and under it, including the Archaic coins.19
Brick fall filled the recess, resting directly on the
cobble floor and running up to the three walls. At
the back of the recess it reached a thickness of about
4 m; debris sloped down to the western edge of the
recess, where it was only 0.1–0.3 m thick, but it com-
pletely covered and sealed the floor of the recess.
It spilled down the slope beyond the recess, cover-
ing the surface of the glacis and filling erosion rills.
Where the croeseid gold coin was found, brick fall
was preserved to a thickness of between 1.5 and 2
m. Over the findspot of the croeseid silver 12th
stater it was about 0.7 to 0.8 m thick. The skeleton
with the croeseid silver 24th stater lay within the
brick fall at the northeast corner of the recess, sealed
by about 1.2 m of destruction debris and about 2.3
m above the floor.
Brick fall has been found elsewhere around the
fortification, piled against the faces of the wall that
had not been already covered by the glacis; indeed,
burial under its own destruction debris is respon-
sible for the extraordinary preservation of the forti-
fication. On the east side of the wall, brick fall filled
the shallow recess south of the gate (fig. 10[2c]). 20
In this area, the destruction debris covered not only
the fortification but also buried houses inside the
wall (fig. 10[3a]). Well-preserved pots and other
material from these houses, which were burned and
destroyed at the same time as the fortification, offer
some of the clearest evidence for the date of the
destruction.
19
Greenewalt et al. 1987a, 22–4.
20
Greenewalt et al. 1987a, 22, fig. 7.
21
Greenewalt et al. 1985, 78; Greenewalt and Rautman 2000,
667.
22
Greenewalt et al. 1994, 19–21.
23
Greenewalt and Rautman 1998, 495.
24
Greenewalt n.d.
597
Farther to the south, brick fall on the west side of
the fortification was piled up to 4 m thick in the
possible recess there (fig. 10[2a]).21 On the east
face of the fortification, debris filled another re-
cess, reaching a thickness of 7 m (fig. 10[2b]).22
Just outside the recess, however, there was much
less debris, although artifacts on the destruction
floor were well preserved.23 Unlike the portions far-
ther north, this face of the Lydian fortification con-
tinued in use into the fourth century B.C. or later.
Remains of what seems to be another Lydian house
are located 18 m away, buried under about 0.5 m of
debris from both the fortification and its own
mudbrick walls (fig. 10[3b]).24
The destruction debris within the gate was dif-
ferent in character, consisting of a thin layer of
crushed red brick. Material resembling brick fall
was piled against the casemate walls that blocked
the gate in the Late Archaic phase of the fortifica-
tion and continued some meters east of the gate.25
The Late Archaic Fortification
A new fortification was constructed on the ruins
of the old. The foundations of a stone wall 5.6 to 7.1
m wide survive in places at the top of the ridge, dug
into the mudbrick of the earlier wall. The wall in-
corporated many reused blocks, and its foundation
trench was filled with destruction debris. This later
wall also had a glacis on its west side, piled over the
earlier glacis and brick fall and further filling the
recess with up to 3 m of clay fill (fig. 7). This layer of
dense clay additionally sealed lower strata from later
intrusion. The gate court was blocked at this time
by the construction of a casemate wall with a sand-
and-gravel fill. There is no trace of a gate passage in
this phase of the fortification.26
the date of the context of the coins
The brick fall forms a distinctive break in the
stratigraphy and in the use of the area of the Lydian
fortification. Sealing the pebble surface and the
two 12th staters beneath and the 24th within the
destruction debris, brick fall provides a stratigraphic
fixed point and a chronological terminus ante quem
for the coins. The security and unity of this deposit,
and its absolute dating, are critical. The date of the
25
Greenewalt et al. 1994, 17; 1995, 11–2; Greenewalt and
Rautman 1998, 493. Very little of the area outside the gate
passage has been excavated, and much of the gate was dug be-
fore the nature of brick fall was understood; for these reasons
the identification of brick fall here is not as secure as elsewhere.
26
Greenewalt et al. 1983, 6–8; 1986, 8; Greenewalt and
Rautman 2000, 667 n. 60.
598 NICHOLAS CAHILL AND JOHN H. KROLL [AJA 109

Fig. 10. Schematic plan of Lydian levels at sector MMS. (© Archaeological Exploration
of Sardis/Harvard University)
2005] NEW ARCHAIC COIN FINDS AT SARDIS
Fig. 11. Reconstruction of Lydian fortification. (Drawing by P. Stinson) (© Archaeological Exploration of Sardis/Harvard
University)
destruction is based on imported Greek pottery
found in and under the brick fall, on a 14C date from
the destruction level, on a dendrochronological
date, and on the general nature of the destruction
deposits. All these methods agree in establishing a
date near the mid sixth century B.C. The destruc-
tion is, therefore, almost certainly associated with
the capture of Sardis by Cyrus the Great in the 540s
B.C. Although this identification has been argued
in print on a number of occasions, it is worth re-
examining the context and date.27
Ceramics
Relatively little pottery was found on the cobbled
surface of the recess itself, and all was of local manu-
facture. An almost-intact round-bottomed oinochoe
was discovered in the southeast corner of the recess
(fig. 5[5]). Large fragments of a wave-line hydria were
found scattered in the destruction debris and on
the floor of the recess.28 Both are similar to local vases
from other contexts that can be dated to the sixth
27
Greenewalt et al. 1983, 6; 1987a, 22–4, 29–31; 1995, 18;
Ramage 1986; Greenewalt 1992; Greenwalt and Heywood
1992, esp. 1–4 (the helmet was found within the destruction
debris in the recess where the coin was subsequently discovered).
28
Oinochoe P89.66:9709; Greenewalt et al. 1993, 40 n. 22.
Hydria P89.69:9716 (unpublished).
29
Greenewalt et al. 1987a, 22, 23, fig. 7, 24–31; 1987b, 62–
599
century B.C., particularly those from Lydian houses
near the fortification, but are not closely datable.
The largest number of datable ceramics found
so far in the destruction level comes from houses
on the east side of the fortification, which had been
burned and then buried under brick fall.29 Unlike
the relatively bare spaces immediately adjoining the
fortification, these houses were packed with arti-
facts, many still where they had been before the
destruction, with pots resting on their stands and
cooking vessels filled with foodstuffs. Most of the
290 or so complete pots from the floors of these
houses were local and not precisely datable. A num-
ber of imported vases and local figured wares can
be closely dated, however, including the following
catalogued objects:
1. Complete Attic komast cup, attributed to the
Vienna Komast Painter, among the latest cups
of this series, almost identical to the painter’s
name vase, which is dated by Brijder to ca. 560–
555 B.C. (fig. 13a).30
70; 1990, 143–56; 1995, 13–9; Cahill 2000. A monograph on
the Lydian houses and their contents, with fuller discussion
of the vases, is forthcoming.
30
P84.56:8909, found with Attic black-figure skyphos no.
2 (Ramage 1986; Greenewalt et al. 1987a, 28, fig. 13;
Schaeffer et al. 1997, Attic 46). Cf. Brijder 1983, 82–4, pl. 6d
no. K98.
600 NICHOLAS CAHILL AND JOHN H. KROLL
Fig. 12. Recess with brick fall partially excavated (1989). (© Archaeological Exploration of Sardis/Harvard University)
2. Almost-complete Attic Little Master band
skyphos, attributed to the Group of Louvre CP
10252, “just before the mid-sixth century,” by
J.T. Haldenstein (fig. 13b).31
3. Fragments of an Attic black-figure Little Mas-
ter band cup, preserving the profile and about
one-third of the lower body (fig. 14).32
4. Fragments of an Attic black-figure Little Mas-
ter cup, with feline in tondo (fig. 15d).33
5. Body fragments of an Attic black-figure closed
vessel, with greaved leg and shield of warrior
and woman in upper register, lion(?) and si-
ren in lower register; perhaps by the Camtar
Painter or the Painter of London B 76, 560s
B.C., according to Ann Brownlee (fig. 15b).34
6. Two sherds from outer wall of Lakonian cup,
showing legs of komast(?), attributed to the
Typhon Painter by C.M. Stibbe, “probably his
work,” ca. 560–555 B.C. (fig. 15a).35
31
P84.57:8910, found with no. 1 (Ramage 1986; Greenewalt
et al. 1987a, 28, fig. 13; Schaeffer et al. 1997, Attic 85).
32
P93.36:10092 (Greenewalt et al. 1995, 18, fig. 20).
33
P93.49:10117 (Greenewalt et al. 1995, 18, fig. 18).
34
P93.44:10109 (Greenewalt et al. 1995, 18, fig. 18).
35
P93.56:10124 (Schaeffer et al. 1997, Lakonian 5).
36
P94.12:10149 (unpublished).
[AJA 109
7. Fragment of wall of Attic black-figure cup, with
uncertain figural decoration, perhaps harpy
or bird (fig. 16a).36
8. Fragment of Attic black-figure vessel, possibly
an oinochoe neck, with bird (fig. 16b).37
9. Two Middle Corinthian aryballoi (one com-
plete, one fragmentary) with quatrefoil pat-
tern (fig. 16c).38
10. Fragments of a Corinthian black-figure lip
cup, with uncertain figural decoration (fig.
15c).39
11. Six fragments from the lower part of a Chiot
chalice.40
12. Complete lebes of local manufacture in
Orientalizing style, decorated with sea serpents
and fish (fig. 17). The style is similar to Early
Fikellura.41
13. Table amphora of local manufacture, deco-
rated in Orientalizing style with wild goats,
37
P86.100:9351 (Greenewalt et al. 1990, 153).
38
P85.21:9070, P85.32:9112 (Schaeffer et al. 1997,
Corinthian 123, 124; Greenewalt et al. 1987b, 68, figs. 13–15).
39
P93.45: 10111 (Greenewalt et al. 1995, 18, fig. 18).
40
P04.3:11780 (unpublished).
41
P93.25:10069 (Greenewalt 1994–1995; Greenwalt et al.
1995, 17–8, figs. 16, 17). Early Fikellura, see Greenewalt (1971).
2005] NEW ARCHAIC COIN FINDS AT SARDIS
Fig. 13. Pottery from destruction level in houses: a, Attic komast cup no. 1; b, band skyphos no. 2. (© Archaeological Exploration
of Sardis/Harvard University)
birds, and floral ornament, ca. 575–550 B.C.,
according to K. Iren (fig. 18).42
14. Fragments of an Ephesian Ware lid.43
15. Fragment of closed vessel decorated in
Orientalizing style, preserving hind leg and
haunch of goat (fig. 19a).44
Several other vessels with simple Orientalizing
decoration were found, including two lids, a num-
ber of fragmentary stemmed dishes, a fragmentary
stand, two Lydian “pyxides,” and two Myrina ampho-
ras.45 All are compatible with a mid sixth-century date,
although they are less closely datable than the fig-
ural sherds. An Orientalizing vase in the shape of a
boat, decorated in Early Fikellura style, was found
discarded in the house as well.46
A small excavated area of a second house pro-
duced mostly local pottery vessels of the same gen-
eral types as those from the other houses (fig. 10
[3b]). The only clearly datable sherd was a frag-
ment from the interior of a Lakonian black-figure
cup (no. 16), with hairy satyr or cyclops, attributed
to a follower of the Naucratis Painter by C.M. Stibbe,
ca. 550–540 B.C. (fig. 19b).47 Outside of the houses,
the brick fall contained very few artifacts, and the
surfaces it rested on were often almost bare. In the
brick fall and stone layer near the northeastern re-
42
P95.42:10256 (unpublished).
43
P93.41:10103 (Greenewalt et al. 1995, 17). Ephesian Ware,
see Greenewalt (1973).
44
P86.103:9355 (Greenewalt et al. 1990, 153).
45
Lids P85.73:9315, P93.46:10112 (Greenewalt et al. 1987b,
70 n. 19 and fig. 10; 1995, 17); stemmed dishes P93.26:10070,
P93.47a:10113, P93.47b:10038; stand P84.106:8978 (Greene-
walt et al. 1987a, 28 n. 15 and fig. 11); pyxides P86.47:9275
(Greenewalt et al. 1990, 154 n. 30 and fig. 20); P86.48:9276
(unpublished); Myrina amphora P94.41:10202.
46
P86.15:9223 (Greenewalt et al. 1990, 152, figs. 15, 16;
1993, 2, fig. 20).
47
P98.234:11083 (Greenewalt and Rautman 2000, 668 n.
62).
601
cess (fig. 10[2c]), fragments were found of an Attic
black-figure Little Master cup (no. 17) (fig. 19e)48
and of a Fikellura amphora with grazing goat (no.
18), attributed to R.M. Cook’s Lion Group BI.1 and
BI.2 by G. Schaus, ca. 550 B.C. (fig. 19d).49 Just above
the brick fall layer, in the southwestern recess (fig.
10[2a]), fragments were discovered of an Attic
black-figure vessel (no. 19) with decoration in two
registers, showing a panther and the legs of feline
and bird(?) in the manner of Sophilos (fig. 19c).50
Many local vessels were found in front of the for-
tification near the southeastern recess (fig. 10[2b]).
All are characteristic types of the mid sixth century
B.C., although none is more closely datable.
The destruction level inside the gate contained
30 to 40 local vessels, many of them at least partly
restorable, and about 135 arrowheads. Again, al-
though not precisely datable they are similar to those
found in the houses. The only imported vessel is
an Attic black-figure band cup (no. 20), with no
figural decoration preserved.51
All of the datable pottery in and under the brick
fall, therefore, dates to near the middle of the sixth
century or somewhat earlier. In 25 years of continu-
ous excavation in this sector, no material has been
found in or under the destruction layer that must
48
P80.21:8523 (Greenewalt et al. 1983, 6). A surface find
from this same area of the fortification, discovered at the top
of the hillock where brick fall and the mudbrick of the fortifi-
cation lie just below modern surface, is a fragment of a plate
attributed to the Polos Painter by G. Bakir, published (as an
Anatolian or East Greek black-figure fragment) in Greenewalt
et al. 1982, 20, and figs. 20, 21 (NoEx78.13). This may belong
with the brick fall.
49
P82.43:8665 (Greenewalt et al. 1985, 73, fig. 22).
50
P82.10A, B:8599–8600. The context is not necessarily
below the destruction debris (Greenewalt et al. 1985, 78;
Greenewalt and Rautman 2000, 667 n. 57).
51
P95.65:10294 (Greenewalt and Rautman 1998, 490–1 n.
35).
602 NICHOLAS CAHILL AND JOHN H. KROLL
Fig. 14. Pottery from destruction level in houses: Little Master band cup no. 3. (© Archaeological Exploration of Sardis/Harvard
University)
be dated after ca. 550 B.C. Red-figure vases and
other material datable to the second half of the sixth
century and later are found in other Sardian con-
texts, but none has been discovered in or under
the brick fall.
Equally striking are the changes in the ceramic
assemblage in strata immediately postdating the
destruction. Among the characteristic shapes found
in fills just above the brick fall are Achaemenid
bowls and open, thin-walled lamps of Broneer Type
1.52 Achaemenid bowls are easily recognizable and
are not uncommon in the first deposits above the
brick fall. The bowls were certainly in use at Sardis
by the early fifth century B.C. but may have been
introduced earlier.53 No Achaemenid bowls have
been found in or under the brick fall, which indi-
cates that this Near Eastern shape was introduced
shortly after the destruction, further associating it
with the capture of Sardis by a Near Easterner; its
sudden appearance would be hard to account for
in other historical circumstances.
Radiocarbon and Dendrochronology
Radiocarbon analysis of carbonized seeds from the
destruction layer in the Lydian house (fig. 10[3a])
52
Dusinberre 1999; 2003, 172–95. Both illustrated in
Greenewalt et al. (1987a, 26, fig. 10) and Dusinberre (2003,
173, fig. 62). Achaemenid bowls from Early Hellenistic con-
texts in Rotroff and Oliver (2003, 61–2, nos. 215–218).
53
Deposit 1 in Dusinberre (1999; 2003, 172, fig. 62, 179).
In the recess, brick fall is covered with a thin layer of rain-
washed earth, and then by a dense layer of stones, the detritus
from robbing the recess walls (Greenewalt 1990, 10, fig. 12).
Both of these strata, as well as the Late Archaic glacis above,
contained Achaemenid bowl fragments. They did not, how-
ever, include imported pottery that could be more closely dat-
able. These strata might date as late as the end of the sixth
century, but the absence of any intervening strata without
Achaemenid bowls between the destruction debris and these
[AJA 109
yielded a date of 2520 ±50 b.p. (uncalibrated). Be-
cause the radiocarbon curve during this period is
so flat, the calibrated calendar date could range
from 800 B.C. to 480 B.C. (table 1). Further testing
is unlikely to provide more precise results.54
Although large quantities of charcoal have been
found in the destruction debris, the fragments dis-
appointingly preserve too few rings to allow den-
drochronological dating. The one piece that
preserved a sufficient number of rings (94) was dis-
covered in the destruction level just inside the gate
passage and has been dated to 585 B.C. The char-
coal fragment does not preserve the bark, however,
and so only offers a terminus post quem for the
cutting of the wood and the subsequent ruin of the
building to which it belonged.55
While neither of these methods offers a precise
date for the destruction, they do provide indepen-
dent evidence that the date is very unlikely to be as
late as the end of the sixth century B.C. Vickers and
Gill’s radical down-dating of not only croeseid and
other early coins but also of Archaic pottery and
sculpture would reject the traditional ceramic chro-
nology adopted here.56 Under this low chronology
one might hypothesize, for instance, that the de-
layers suggests that the bowls were introduced to the ceramic
repertoire at Sardis shortly after the destruction.
54
Greenewalt et al. 1987a, 29 n. 20. The date reported there
of 570 ±50 b.c. was based on the calibration curves in use in
1984 but must be recalibrated according to more recent cali-
bration curves. Similar results were obtained using the programs
Calib 4.4 (Stuiver et al. at http://depts.washington.edu/qil/
calib/) and the Oxford Radiocarbon Calibration Program
“Oxcal” (http://www.rlaha.ox.ac.uk/orau/). Many thanks to
Peter Kuniholm for his help with understanding 14C calibra-
tion curves.
55
Greenewalt and Rautman 1998, 492 n. 38.
56
Francis and Vickers 1983; Vickers 1985.
2005] NEW ARCHAIC COIN FINDS AT SARDIS 603

Fig. 15. Pottery from destruction level in houses: a, Lakonian cup no. 6; b, Attic black-figure closed vessel no. 5; c, Corinthian
lip cup no. 10; d, Attic Little Master cup no. 4. (© Archaeological Exploration of Sardis/Harvard University)

struction was that of the Ionian revolt in 499 B.C.
However, the radiocarbon and dendrochronologi-
cal dates make such a late date highly unlikely and
strongly support the traditional chronology.
Weapons and Casualties of Battle
In addition to pottery and other material offer-
ing an absolute date in the mid sixth century B.C.,
the destruction contained evocative remains of a
battle: weapons, armor, and the skeletons of casual-
ties.57 Two human skeletons were found in the re-
cess (figs. 5[3, 6], 8). They were not formally buried,
nor were they lying where they fell in battle, but
had been ignominiously dumped with the brick
fall, presumably as part of the cleanup and demoli-
tion of the fortification. The skeleton in the corner
of the recess, high up in the destruction debris,
has already been mentioned as the owner of the
silver croeseid 24th stater. The other body lay in an
erosion gully on the slope below the recess. Both
were males in their early to mid 20s, and both dis-
played physiognomic traits of soldiers, such as dif-
ferential development of the arms and shoulders
consistent with carrying a shield on the left and
wielding a weapon with the right arm. The neck
vertebrae of the skeleton in the corner of the re-
cess were compressed, probably by wearing a heavy
object, such as a helmet, on his head. Both had sur-
vived injuries during their lives, including wounds
to the head, arms, and hands; the casualty in the
erosion gully had suffered a broken rib three to
four weeks before death, perhaps at Pteria, as de-
scribed by Herodotus (1.76). And both died vio-
lently, with broken forearms, probably from warding

57
In general, Greenewalt 1992, 1997.
604 NICHOLAS CAHILL AND JOHN H. KROLL [AJA 109
the fortification (fig. 10[2b]). Just inside the gate
were about 135 iron and bronze arrowheads of dif-
ferent shapes and sizes.61
The superstructure of the fortification must have
been demolished very shortly after the battle, fill-
ing the recess and sealing the coins. The skeleton
in the corner still clutched a small stone in his right
hand, showing that rigor mortis had not yet released
when he was dumped out with the destruction de-
bris. The houses covered by rubble show no signs
of weathering or disturbance other than some loot-
ing before they were burned. Pots were smashed
but still articulated on the floor, many broken in
situ on stands or hearths. Fragile organic material
such as carbonized grain was well preserved. The
brick fall must have been deposited soon after the
houses were burned and the soldiers killed—on
the order of days rather than weeks or months.
Fig. 16. Pottery from destruction level in houses: a, Attic
In short, the destruction assemblages offer an
black-figure fragment no. 7; b, Attic black-figure fragment
no. 8; c, Corinthian aryballos no. 9. (© Archaeological unusually complete picture of a military conquest,
Exploration of Sardis/Harvard University) with casualties of soldiers and civilians, weapons,
and armor. Houses inside the fortification were
burned with their contents still intact. This was
off blows (“parry fractures”), and wounds in the back shortly followed by the deliberate destruction of
inflicted at the time of death. The casualty in the the fortification. All chronological indications
gully had a broken neck and bore three head point to a date in the mid sixth century B.C.; noth-
wounds from at least two different weapons, two from
a sharp-edged weapon such as a sword and one from
a narrower, blunter one, perhaps a battle axe.58 A
third skeleton, belonging to an older man in his
late 30s, lay in one of the houses.59
Pieces of armor and military equipment were
found in the destruction level as well. An iron-and-
bronze helmet, crushed and shattered but largely
complete, lay within the brick fall between the two
skeletons in the recess (fig. 5[4]). This may have
belonged to one of the casualties; it rested on the
same slope as the skeleton in the corner of the re-
cess, whose neck had been compressed by wearing
a heavy object—perhaps this very helmet.60 Remains
of a long wooden object, possibly the shaft of a
weapon like a spear or javelin, were found along
the same slope. Other military equipment found
in the destruction level elsewhere in this sector
includes an iron saber, a sickle-like implement,
perhaps a war sickle, and clusters of iron hardware, Fig. 17. Pottery from destruction level in houses:
possibly belonging to a wheeled vehicle such as a Orientalizing lebes no. 12. (© Archaeological Exploration
chariot, found near the recess on the east side of of Sardis/Harvard University)

58 61
Greenewalt and Heywood 1992; Greenewalt et al. 1993, Saber, sickle, and wheeled vehicle: Greenwalt 1997;
20–1; Greenewalt n.d. Greenewalt and Rautman 1998, 496–7 nn. 49, 50, fig. 22. Ar-
59
Greenewalt et al. 1990, 150. rowheads: Greenwalt 1997; Greenewalt and Rautman 1998,
60
M87.7:9444 (Greenewalt and Heywood 1992). 490–2.
2005] NEW ARCHAIC COIN FINDS AT SARDIS
ing later than the middle of the century has been
recovered. The archaeological evidence over-
whelmingly supports the identification of this de-
struction as the famous capture of Sardis by Cyrus
the Great of Persia.
The Date of the Capture of Sardis
Cyrus’ capture of Sardis was one of the critical
turning points in the history of Anatolia, indeed of
the ancient western world. Although the event can
be dated generally to the 540s B.C., the exact date,
usually placed in 547/546 or 546/545 B.C., is not
as secure as is sometimes believed.
The only detailed accounts of Cyrus’ capture of
the city come from classical Greek and Roman
sources; there are no contemporary Near Eastern
descriptions. Herodotus (1.46–89) offers the full-
est and most familiar version. Worried by the rising
power of Cyrus and encouraged by his misunder-
standing of the oracle from Delphi, Croesus set out
to invade Persian territory. Cyrus assembled his army
and met the Lydians in central Anatolia. After an
inconclusive battle at Pteria, Croesus returned to
Sardis to await reinforcements, as the campaigning
season was over. Cyrus, however, marched quickly
toward Croesus’ capital and forced a battle on the
Hermos Plain before Sardis. His camels terrified
the Lydian cavalry, and the Lydians retreated into
the city. While Cyrus set siege to Sardis, Croesus
dispatched messengers to his allies, urging them
to come to his aid immediately. But after a 14-day
siege, Cyrus’ soldiers scaled the acropolis at a spot
that had been considered too steep to need fortifi-
cation. Sardis was captured and Croesus was placed
on a funeral pyre, only to be miraculously rescued—
a tradition known in Greek art and literature be-
fore Herodotus, as early as Bacchylides and Myson.62
It is not clear from Herodotus’ account whether
the whole city was besieged, or whether the lower
city fell after the battle on the plain and only the
acropolis held out for two weeks.
Xenophon (Cyr. 7.2.1–7.3), Justin (Apol. 1.74.3–
4), Ktesias (FGrHist 688.F9), Polyaenus (Strat. 7.6.2),
and others give versions that differ from Herodotus’
in important respects.63 According to Xenophon,
62
Bacchylides 3.23–62; Myson ARV 2 238, no. 47. Most re-
cently, Burkert 1985; West 2003.
63
Pedley 1972, 37–42.
64
Parth. Amat. narr. 22; cf. Haslam 1986.
65
Briant 2002, 34. The only ancient source who places the
fall of Babylon before Sardis is Justin (Apol. 1.7.3).
66
How and Wells 1928 I.86; Wade-Gery 1951, 219 n. 38 (544
B.C.), followed by Miller 1963 (547/546 B.C.); Drews 1969;
605
for instance, Cyrus captured both the city and
acropolis on the day following the battle before the
city, omitting the two-week siege and the embassies
to Croesus’ allies. Polyaenus specifies that Cyrus
took the lower city the night after the battle on the
Hermos Plain, using ladders to climb the un-
guarded walls, and that the citadel was taken later,
by threatening to hang the captured relatives of
the besieged rather than by scaling the unfortified
cliffs of the acropolis. Ktesias describes dummy sol-
diers made of wood, which were hoisted onto the
walls and caused the Lydians to panic; in a roman-
tic version of the story, the citadel was betrayed by
Croesus’ daughter Nanis through her love for
Cyrus.64
The historical date of the capture has occupied
scholarly debate for many years. The lower chrono-
logical limit is the fall of Babylon in October 539
B.C., as nearly all ancient sources agree that Cyrus
captured Sardis before he took Babylon.65 This is
the most secure terminus ante quem for the sack
and, therefore, for the loss of the coins. The upper
limit for the fall of Sardis must be Cyrus’ defeat of
Astyages and capture of Ecbatana, probably in 550
B.C. A precise date within this span, however, re-
mains elusive.
Nothing in Herodotus’ account gives an exact
date for the capture. He offers a number of syn-
chronisms to derive dates for the capture, but these
seem to be internally inconsistent. Scholars have
used these synchronisms together with other, later
sources to argue for dates ranging from 547 to 544
B.C. However, as Evans argued, by Herodotus’ time,
memories of the fall of Sardis were probably incon-
sistent and difficult to disentangle, and Herodotus’
discrepancies probably reflect various strands in the
ancient narrative tradition.66
Hellenistic and Roman chronographic sources
offer precise but problematic dates. The text of the
Marmor Parium is broken at this point; Jacoby restores
a date of 541 B.C.67 But the restoration is purely
hypothetical. A fragment of the Chronicon Romanum
in the Getty Museum, composed in A.D. 15/16 but
probably derived from Apollodorus, preserves a
date of 546/545 B.C.68 This is one of our few secure
Sealey 1976, 171, 192; Evans 1978 (547/546 B.C.);
Mosshammer 1979, 259–60; Burkert 1985.
67
FGrHist 239.A41, A42.
68
Burstein 1984; Haslam (1986) on the restoration of the
text, which follows the romanticized story preserved also in
Parthenius. Cf. Balcer (1972, 108), written before the new frag-
ment was published.
606 NICHOLAS CAHILL AND JOHN H. KROLL
Fig. 18. Pottery from destruction level in houses: rendering of Orientalizing table amphora no. 13. (Drawing by C. Alexander)
(© Archaeological Exploration of Sardis/Harvard University)
dates outside the manuscript tradition. Other dates
in the Chronicon Romanum are problematic, however,
and we must be wary of placing too much faith in it.
Mosshammer argues that Eusebius dated the fall of
Sardis to 547/546 B.C.—in the same year as the
death of Thales, Anaximander’s 64th year, and
Pythagoras’ 25th year—and makes the case that the
date was derived from Apollodorus. Jacoby, on the
other hand, believed that Eusebius dated the fall
of Sardis to 546/545 B.C.69
If such late, Hellenistic and Roman chrono-
graphic sources are to be trusted, it should be be-
cause they drew on the more precise records
preserved in Near Eastern archival sources. Moss-
hammer suggests that a plausible intermediary is
Berossus, a Babylonian author writing in the Early
Hellenistic period.70 The only surviving Near East-
ern archival source that might mention the fall of
Sardis is the often discussed and problematic tab-
let, the Nabonidus Chronicle. This account of the
reign of the last king of Babylon was inscribed in
the Early Hellenistic period but was perhaps writ-
ten in the fifth century B.C. and based on earlier
archival sources; it should be more reliable than
any classical account. The passage in question oc-
69
Mosshammer 1979, 255–62; Jacoby, FGrHist 244. Miller
(1963) also argues for 547/546. Mansfield (1983) casts doubt
on the synchronism of Thales’ death with the fall of Sardis and
[AJA 109
curs in the chronicle of Nabonidus’ ninth year, 547/
546 B.C. It relates that in April, Cyrus mustered his
army and crossed the Tigris, and in May he marched
to a place whose name, unfortunately, is damaged
on the tablet. He either killed or defeated its king
(depending on the translation of the Akkadian
verb), took its possessions, and set up a garrison in
the city. Since the 1880s this passage has frequently
been understood as referring to the capture of
Sardis, offering a precise date of 547/546 B.C. and
an independent, Near Eastern account of this fa-
mous event.71
However, Cargill argued in 1977, and nearly all
scholars now agree, that the toponym for the land
against which Cyrus marched simply cannot be read
with confidence. The name has been variously read
and restored as LU[du] for Lydia, IS[parda] or
IŠ[parda] for Sardis, ZU[. . .], and other forms. But
the sign is broken, and the plethora of restorations
over the years have been based more on arguments
of historical probability—to fit in with the classical
story of the fall of Sardis—than on the actual read-
ing of the tablet. The restoration, however, informs
the accounts of even those historians who explic-
itly accept Cargill’s warning, so that many modern
suggests that the fall might have come in a subsequent year.
70
Burstein 1978; Mosshammer 1979, 262.
71
Smith 1924, 98; Grayson 1975, 107–8.
2005] NEW ARCHAIC COIN FINDS AT SARDIS
Fig. 19. Pottery from destruction level in houses and brick fall: a, Wild Goat–style fragment no. 15; b, Lakonian cup no. 16;
c, Attic fragment no. 19; d, Fikellura amphora fragment no. 18; e, Attic cup fragment no. 17. (© Archaeological Exploration of
Sardis/Harvard University)
discussions date the fall to 547/546 B.C. despite
the uncertainty of the reading.72
Although the Nabonidus Chronicle does not of-
fer an independent date for the fall of Sardis in
547/546 B.C., it does preserve other useful infor-
mation about the conquests of the first Persian king.
In an important article on the career of Cyrus,
Mallowan pointed out that in addition to year 9, the
chronicles for year 10 (546/545) and part of year
11 of Nabonidus’ reign are preserved and they do
not mention any campaign by Cyrus.73 This absence
is significant: the Lydian campaign was a turning
point in his rise to power, and the chronicler was
clearly interested in Cyrus’ campaigns beyond the
Tigris. It is very unlikely that the chronicler would
have described a campaign in year nine against an
uncertain enemy across the Tigris that was not
Croesus and then omit the Lydian campaign in a
subsequent year. Since he does not mention such a
campaign in year 10, Mallowan argued that this
ought to rule out 546/545 as a possible date for the
capture of Sardis.
72
Cargill 1977, with previous bibliography; Schaudig 2001.
Cf. Storck 1989; Briant 2002, 34–7, 882–3.
607
Mallowan further noted that, according to the
Nabonidus Chronicle, Cyrus crossed the Tigris in
April and marched against his enemy in May, while
according to Herodotus, Sardis did not fall until
about November, after the campaign season had
ended. Mallowan, therefore, argued that the cam-
paign in year nine could not have been against
Lydia, whatever the reading of the broken sign, and
proposed Lycia as a possible alternative. He con-
cluded that the campaign against Lydia did not take
place in either years nine or 10 but must be dated
to one of the years where the Nabonidus Chronicle
is broken, in 545 B.C. or shortly afterward. This dat-
ing is close to the year 544 B.C. proposed by Wade-
Gery based on his reading of synchronisms in
Herodotus. 74
The discrepancy in year nine between a May cam-
paign and a November capture, however, need not
rule out the possibility that the campaign in year
nine was against Lydia. Although the chronicle is
generally quite specific about chronology of events
it describes, it is also very terse, and neither the
73
Mallowan 1972, 7 n. 34; 1985, 404–5.
74
Wade-Gery 1951, 219 n. 38.
608 NICHOLAS CAHILL AND JOHN H. KROLL
Table 1. Calibrated Radiocarbon Dates
Calibration per CALIB 4.4
Confidence Level Date (B.C.) Probability (%)
68.3% (1 sigma) 791–757
693–541
95.4% (2 sigma) 799–502
490–484
465–449
440–427
423–413
Nabonidus Chronicle nor the other surviving ac-
counts are complete enough to reconstruct fully
Cyrus’ campaign into Anatolia. There may have
been other battles and other enemies for him to
face before the battle at Pteria and siege of Sardis.75
And while the toponym is broken, it does not fol-
low that it was not Lydia or Sardis; it simply cannot
be read. We need not exclude Sardis as the ulti-
mate target of Cyrus’ campaign in 547/546 B.C. An
exact date is probably impossible to reconstruct with
certainty, but of the many dates proposed, either
547/546 or around 545 or 544 B.C. seem the most
likely.
Pre-Destruction Levels in the Recess
Whatever the precise date, the destruction of
Sardis by Cyrus provides a secure terminus ante
quem for the croeseid coins of no later than 539
B.C. It is more difficult to determine a precise strati-
graphic date for the minting and loss of the coins.
The croeseid 24th stater was found in the destruc-
tion debris itself and so was in circulation at the
moment Sardis was captured. The context of the
other two coins is somewhat earlier; sealed under
the floor, they must have been lost sometime be-
fore the destruction. But how long before, and how
long had they been in use? Can the coins be dated
archaeologically to the reigns of Croesus or his fa-
75
Cf. Cargill 1977, 105–6.
76
Four distinct surfaces were identified in excavation, and
other, more ephemeral surfaces survived in patches. These ear-
lier surfaces, however, seem to be working surfaces associated
with the building of the recess walls and the landscaping of
the area, rather than separate phases of use. The uppermost
cobbled surface ran up to—and so belongs with—all three walls
of the recess. Below this was a second surface, partly cobbled
and partly pebbled, which extended to the north and east walls
of the recess but was cut by the south wall. The gold croeseid
and the silver 12th were found in fill between these two sur-
faces. A number of worked and semi-worked blocks were found
lying on this second surface; these may represent building de-
[AJA 109
Calibration per OXCAL
Date (B.C.) Probability (%)
19.3 800–700 12.9
80.7 700–540 55.3
94.3 800–480 90.2
0.8 470–410 5.2
2.0
1.8
1.1
ther, Alyattes? The stratigraphy of this area offers
some hints but not an absolute answer.
The cobble floor is the uppermost of at least four
cobbled and earthen surfaces of the recess (fig.
7).76 These surfaces, however, all belong to a single
phase, the construction of the recess; no appre-
ciable time elapsed between them. The two coins
were probably lost at the same time the recess was
being built.
Predating the recess are at least four earlier
phases of construction. As noted above, the recess
and glacis represent a repair of an earlier, mudbrick
phase of the fortification. This earlier phase was in
turn cut into a structure built from large purple
sandstone blocks, probably a still earlier fortifica-
tion (fig. 6, in the sondage at right). Under the
purple sandstone building was a pair of cobble sur-
faces, and below these was a domestic space con-
sisting of at least two subphases of small-scale walls.
This was the lowest level reached in excavation.77
These different phases, explored only in narrow
sondages on the east side of the recess, have pro-
duced relatively little datable pottery, and assign-
ing precise dates is difficult. The foundation
trenches of the recess walls yielded two datable
sherds: a Lakonian II cup sherd (no. 21), with a file
of birds on the exterior and the head and torso of
person and the head of a bird on the interior, iden-
bris. Below this, a third surface was preserved only in patches.
The fourth (lowest) surface contained numerous working chips
and other building debris, as well as more worked blocks, and
probably represents the level from which the recess was built.
As the three walls bond and thus belong to a single building
campaign, these lower surfaces probably represent working
levels used during the construction of the recess.
77
Greenewalt 1990, 13–4, fig. 19. Another, intermediate
phase of the fortification, between the mudbrick phase and
the recess walls, is hinted at by the presence of a wall or other
feature built from small sandstone blocks preserved under the
north wall of the recess, but the nature of this phase is un-
clear.
2005] NEW ARCHAIC COIN FINDS AT SARDIS
tified by C.M. Stibbe as a work of the Painter of the
Fishes of Taranto, ca. 620–580 B.C. (fig. 20a);78 and
a North Ionian skyphos fragment (no. 22) in Wild
Goat style of mixed technique, showing the wing of
a sphinx(?), dated to the first half of the sixth cen-
tury by M. Kerschner (fig. 20b).79 None of this mate-
rial can be dated precisely enough to determine
whether the recess was built in the reign of Alyattes
or Croesus and so does not resolve the critical nu-
mismatic issue of when the coins were lost.
Datable pottery from the construction of this sec-
tion of the earlier, mudbrick phase of the fortifica-
tion wall includes a Middle Wild Goat I oinochoe
fragment (no. 23) showing hatched chevrons and a
cable band, identified as South Ionian, probably
Milesian, and dated to the second third of the sev-
enth century B.C. by M. Kerschner, from the upper
fill of the foundation trench (fig. 20c),80 and an
Early Wild Goat–style oinochoe fragment (no. 24)
showing a frieze of goats, from the mudbrick detri-
tus above the surface from which the mudbrick
phase of the fortification was built (fig. 20d).81 Sev-
eral Ionian cup and Bird Bowl fragments dating to
the late seventh and first half of the sixth centuries
B.C. were found in these and earlier levels, as well
as local pottery.82
Most important, two figured sherds were found
sealed beneath the cobble surface that divides the
earliest, domestic phase of occupation from the
subsequent construction of the fortifications: a
North Ionian Late Wild Goat–style oinochoe(?) frag-
ment (no. 25), in mixed technique, showing a lion
walking to the left, ca. 610–580 B.C. (fig. 20e);83
and a Wild Goat–style plate fragment (no. 26) with
rosette(?) in the center and a frieze of overlapping
goats (fig. 20f).84 These probably date to the late
seventh century or perhaps even later, providing a
terminus post quem for the entire sequence. Other
78
P97.121:10708 (Greenewalt and Rautman 2000, 663–4
n. 47 and fig. 21 [top]).
79
P95.49:10273 (Greenewalt and Rautman 1998, 495 n. 44).
Excavation in 2002 and 2003 clarified the stratigraphic posi-
tion of this piece, which was previously thought to belong to
an earlier phase. Local pottery from the construction fills was
notably eroded and undiagnostic, and a number of prehistoric
sherds suggest that this fill was imported from elsewhere; its
ceramic assemblage may, therefore, be somewhat earlier than
the actual date of deposit. Fragments of local column kraters
and lydions suggest a date in the first half of the sixth century.
80
P95.50:10274 (Greenewalt and Rautman 1998, 495 n. 44).
Again, excavation in 2003 showed that the context in which
the sherd was found does not predate the fortification wall, as
originally thought, but was part of its foundation trench.
81
P03.302:11584 (Greenewalt n.d.)
609
pottery from this phase includes Bird Bowls and
Ionian cups, including a local(?) Ionian cup with
black-on-red decoration found broken in situ.85
As the domestic structures were destroyed no
earlier than ca. 610–580 B.C., all phases of the forti-
fication must be compressed to the 40 to 70 years
between this date and the destruction in the 540s.
This period saw the construction of the early, sand-
stone phase of the fortification, the erection of the
20-m thick mudbrick wall, which stood for long
enough to become very eroded, and the repair with
glacis and recess. Thus, even if we cannot assign an
exact date to the creation of the recess, the final
phase to which the coins belong ought to be placed
relatively late in that short sequence, most probably
in the reign of its last king, Croesus. There is no
indication that this phase dates any earlier than
Croesus’ reign.
the coins
The Lydian gold 12th, silver 12th, and silver 24th
(see nn. 1–3) are relatively early specimens of the
Lydian bimetallic coinage of gold and silver lion-
and-bull staters and fractions that numismatists have
conventionally associated with King Croesus. Inas-
much as all previous Lydian coinage had been made
of electrum—a variable and easily manipulated
gold-and-silver alloy—and was probably significantly
overvalued, the substitution of the more reliable,
intrinsically valued lion–bull coinage of pure gold
and silver marked a momentous departure from
earlier monetary practice.
As confirmed by the three major investigators of
the coinage—Paul Naster, Ian Carradice, and
Cynthia Nimchuck86—it is divided into two large sty-
listic periods, the earlier with a relatively naturalis-
tic rendering of the two obverse animals (Phase I)
and the later in which the animals are depicted in a
82
Cook and Dupont 1998, 26–8, 129–31. Local pottery asso-
ciated with this phase does not seem to include typical sixth-
century shapes such as column kraters, lydions, and lamps, but
this may be an accident of preservation or excavation. Such
sixth-century pieces were found in the sondage dug in 1987
in levels that seem to predate the recess, but the stratigraphic
relationship between those levels and the mudbrick phase of
the fortification is unclear (Greenewalt 1990, 13–4).
83
P98.212:11057 (Greenewalt and Rautman 2000, 663 n.
44, fig. 21 [center]).
84
P03.306:11756.
85
Greenewalt and Rautman 2000, 663 nn. 43, 44. The black-
on-red Ionian cup is P98.205:11040 (cf. Greenewalt 1979, 25).
86
Naster 1965, 31–2; Carradice 1987, 74; Nimchuck 2000,
6–12. For the weights, see Naster 1976.
610 NICHOLAS CAHILL AND JOHN H. KROLL
87
harder, stylized manner (Phase II). The naturalis-
tic Phase I involved two different weight standards
for the gold; nearly all scholars have assumed that
the coinage began with the 10.7 g weight for staters
and fractions in both metals (Subphase A) and that
the shift to the lighter gold stater of 8.06 g and its
fractions occurred subsequently (Subphase B). 88
Table 2 summarizes the full sequence of styles,
weights, and denominations in both metals.
At 0.88 g, the Sardis gold 12th belongs to the
initial or “heavy” phase of the gold coinage that was
minted with a stater weight of 10.7 g.89 Since this
same weight continued to be employed for the sil-
ver throughout the “naturalistic” phase of the coin-
age, there is no way of knowing whether the Sardis
silver 12th and 24th should be assigned to the ear-
lier (Subphase A) or later (Subphase B) part of
Phase I. As best we can tell, the obverse of the 12th
should be classified under Nimchuck’s Style B,
which was common to both subphases. The obverse
of the 24th is too damaged for judgment.
Because stylized, Phase II silver croeseid half-
staters occur in early fifth-century hoards alongside
specimens of early Achaemenid silver sigloi, schol-
ars since the 1930s have recognized that much of
the lion–bull coinage continued to be minted un-
der Achaemenid auspices after the fall of Croesus
and served as royal Persian money in Asia Minor
until late in the sixth century, when Darius I inau-
87
The naturalistic Phase I was divided by Naster into two
substyles (“massive” and “nervous” obverse styles) but by
Nimchuck into three substyles (Styles A–C). Similarly, Naster
recognized only two substyles in the stylized Phase II of the
coinage (transitional and fully stylized obverse styles), whereas
Nimchuck recognized three (Styles D–F).
88
Nimchuck (2000, 11, 12, 21), whose analysis of the coin-
age is by far the most detailed to date, disagrees with this se-
quence in one fundamental respect. On the basis of stylistic
grouping and transferred dies, she believes that the gold staters
and fractions on both weight standards were struck continu-
ously and concurrently over the entire earlier phase of the
coinage, rather than sequentially. She is, however, unable to
offer a convincing explanation of how simultaneous minting
on two standards could make any monetary sense. One cannot
argue that coins of the two weights were minted for two differ-
ent destinations or regions of use since gold coins of both
weights have been found at Sardis and attest that both the
heavy and the light gold circulated locally. It seems most likely,
therefore, that the dies of the different naturalistic styles were
all used for striking heavy gold coins up to the point that the
weight of the gold was reduced, at which time the dies contin-
ued in use with gold of reduced weight. As Nimchuck acknowl-
edges, she was unable to determine the striking sequence of
dies that were used for gold staters of both weights (and some-
times also for silver staters). We could find nothing in her die
ρες.” Texts and translations in Melville
data that would preclude the possibility that dies of Styles B
and C might have been put into service alongside those of the
[AJA 109
gurated a new imperial gold and silver coinage with
initially similar weights but his own Persian types.90
The conventional view that the coinage was inau-
gurated by Croesus was first proposed in the early
1830s by L.-E. Cousinery91 and argued in detail in
1840 by H.P. Borrell,92 on the grounds that the gold
lion–bull staters, some of which were known to have
been found in the region of Lydia, should be iden-
tified with the gold coins referred to in several an-
cient sources as Kroiseioi stateres. 93 So widely
accepted did this identification become that for
most scholars of the past century and a half it has
served as the primary fixed point in the numismatic
chronology of sixth-century western Asia Minor.
Nevertheless, other interpretations have been
advanced. In 1938 Walther Giesecke contended
that the term “Kroiseioi” did not pertain to the gold
lion–bull coinage as a whole but only to the later,
lighter gold staters, by far the most common gold
staters of this type because of their long period of
production.94 Accordingly, Giesecke assigned the
initial heavy gold and its accompanying silver to
Alyattes (ca. 610–560 B.C.). But this suggestion,
while noted by Naster in his influential 1965 paper
on the croeseid coinage, has been otherwise ig-
nored, as the evidence of dies and style convinced
Naster and others that the heavy and early light
phases of the gold comprised a single, relatively
compact whole.
original Style A before the first reduced-weight gold was minted.
Some heavy gold staters from dies of all three styles are, in
fact, linked through a common pair of reverse punches
(Nimchuck 2000, 14–5, punch pair 3/5); it is notable that within
this linked group and throughout the Phase I coinage gener-
ally, Style A dies were used for twice as many extant heavy gold
staters than light staters, and Style C dies for vastly more ex-
tant light staters than the heavy staters (only one in Nimchuck’s
study).
89
It happens to share the same reverse punch with the three
heavy 12ths known to and published by Walburg (1991, 16, pl.
2, nos. V.1–3) and probably shares with them the same obverse
die (we cannot be sure from photographs). We believe that
the obverse should be classified under Nimchuck’s Style A.
90
Tabulations of hoards: Carradice 1987, 79; Nimchuck 2000,
37. Attribution of the later croeseids to Achaemenid author-
ity: Naster 1965, 26 nn. 5, 6 (with early bibliography).
91
Quoted in Mionnet 1833, 405.
92
Borrell 1840. Price (1984, 211) notes the coins were ear-
lier attributed to Samos, Akanthos in Macedonia, or Salamis
on Cyprus.
93
Beginning with IG 13, 458, the financial accounts for the
chryselephantine statue of Athena Parthenos in Athens (440/
439 B.C.), lines 29, 30. The other testimonia are much later:
Plut. Mor. 822F–823A; Poll. Onom. 3.87 and 9.84; and Hesychius,
s.v. “Κροσειοι στατ
Jones (1993, nos. 155, 410, 650, 656, 760).
94
Giesecke 1938, 51–3.
2005] NEW ARCHAIC COIN FINDS AT SARDIS
Fig. 20. Pottery predating the destruction: a, Lakonian fragment no. 21; b, N. Ionian skyphos rim no. 22; c, Wild Goat sherd no.
23; d, Early Wild Goat–style sherd no. 24; e, North Ionian Late Wild Goat–style sherd no. 25; f, Wild Goat–style plate fragment
no. 26. (© Archaeological Exploration of Sardis/Harvard University)
More recently, several influential discussions of
the coinage have advanced a third, even more un-
orthodox, notion that the lion–bull coinage did not
commence until after the death of Croesus and
should in fact be identified in its totality as a coinage
of the early Achaemenid rulers of Sardis beginning
with Cyrus. Since Martin Price boldly proposed this
post-Croesus hypothesis in 1984,95 at least three com-
mentators have endorsed it as probable or virtual
fact, 96 while others have respectfully treated it as a
plausible, if unproved, alternative to the conven-
tional chronology. As a concerned member of the
latter group, the distinguished French numisma-
tist Georges Le Rider devoted two long essays to an
exhaustive and carefully balanced re-examination
of the chronological arguments for the entire
croeseid coinage. In his second survey, “Qui frappa
les premiers créséides: Crésus ou Cyrus?,”97 he was,
however, no more able to reach a satisfactory answer
than in the first, in which he resignedly but hope-
fully concluded, “que nous restons pour le moment
dans l’incertitude sur la date des premiers
95
Price 1984, 211–4, 221; 1989, 9, 10.
96
Vickers 1985, 8, 9; Carradice1987, 80, 81, 91, 92; Descat
2000.
97
Le Rider 2001, 101–21 (ch. 3).
611
créséides. . . . Mais attendons de nouvelles
découvertes.” 98
It goes without saying that the gold and silver
croeseids from pre-Persian contexts at Sardis are
just such a discovery. In an instant they remove the
post-Croesus hypothesis from the realm of possibil-
ity and strengthen as never before the traditional
association of the Lydian lion-and-bull gold staters
with the Kroiseioi and Croesus. In retrospect, of
course, if one may judge from other famous gold
stater coinages of the Greek world that took their
name from the monarchs who instituted them,
Dareikoi and Philippeioi,99 this is only as it should be.
For, in his day, Croesus was as well known to his
Greek contemporaries as Darius and Philip II of
Macedon were known to theirs.
Admittedly, at first sight the condition of the pre-
Persian, heavy gold 12th from Sardis might cause
some unease about attributing to Croesus the lion-
and-bull gold as a whole. The coin looks as if it had
experienced a considerable amount of wear, and if
it could be established that the dulled relief of the
98
“[T]hat for the present we remain uncertain about the
date of the first croeseids. . . . But we await new discoveries”
(Le Rider 1998, 672).
99
Cf. Poll. Onom. 9.84; Melville Jones 1993, no. 450.
612 NICHOLAS CAHILL AND JOHN H. KROLL [AJA 109
Table 2. Structure of the Lydian Lion–Bull Coinage
Phase Subphase Gold Stater Silver Stater

I (relatively naturalistic A heavy: 10,7 g with fractions of 10.7 g with fractions of

rendering of the obverse 1/3, 1/6, 1/12 1/3, 1/6, 1/12, 1/24
animals) a
B light: 8.06 g with fractions of 10.7 g with fractions of
1/3, 1/6, 1/12 1/2, 1/3, 1/6, 1/12, 1/24

II (stylized rendering of – light: 8.06 g half stater: 5.35 g

the obverse animals)b
a
Nimchuck Styles A–C = Naster’s “massive” and “nervous” obverse styles
b
Nimchuck Styles D–F = Naster’s transitional and fully stylized obverses

obverse is proof of 15 or more years of circulation,
the coin would have to have been minted before
Croesus came to the throne, very much in keeping
with Giesecke’s idea that only the later, lighter gold
staters should be identified as Kroiseioi. So far as
we know, however, there is no way of determining
how long a coin like this must have circulated prior
to deposition100 or even of being sure that its dulled
relief resulted from wear since minting from a worn
obverse die (as was not uncommon in many archaic
coinages) or from hasty, weak striking or a combina-
tion of both factors could have produced the same
effect. The coin’s weight at least suggests that it has
not appreciably lost any metal from extensive
abrasion.101
Apart from the problematic question of wear,
Giesecke’s pre-Croesus dating of the heavy gold has
always involved two difficulties of its own. One is the
great implausibility that the name “Kroiseioi” could
have been meaningfully applied to one group of
well-known gold staters that weigh 8.06 g but not to
the coins in an earlier group that were identical in
all physical respects save that they happened to be
25% heavier. Anyone who has handled these coins
can attest that it is not easy to tell a heavy gold stater
from an early light one without weighing them; and
in the case of the small fractions, weighing is virtu-
ally the sole means by which corresponding denomi-
nations in the two series can be distinguished. The
term “Kroiseioi” ought to apply to the gold coinage
in its entirety. Second, from their study of the ob-
verse dies and the reverse punches of the earlier
gold staters of both weights, Naster and, in far
greater detail, Nimchuck have shown that a num-
ber of dies and punches used in the manufacture
of the heavy staters were employed also for striking
the light staters; this implies a fairly tight continu-
ity between the coinage of both weights and has
suggested to most scholars who have reviewed the
evidence that the heavy phase of the coinage, which
is survived by many fewer specimens, was probably
of short duration.102
This deduction is understandable in light of the
exchange equivalencies that seem to lie behind
the two standards of the gold coinage. The 10.7 g
mass of the heavy lion–bull staters corresponds very
closely to the intrinsic value of a Lydian electrum
stater of 14.1 g, if the latter were made of natural,
unadulterated electrum;103 whereas the 8.06 g mass
of the lighter gold staters corresponds to the actual

100
Being 99% pure (Cowell and Hyne 2000, 171, table 7.4,
sample nos. 46863–46866), the gold of croeseids is soft, as coin
metals go. Moreover, a small fraction like this probably remained
in more continuous use than larger coins and would have been
subject to a more intensive amount of abrasion when carried
about in purses with other coins.
101
The mass of the 0.88 g Sardis piece corresponds to the
weights of the three heavy gold 12ths recorded by Walburg
(1991, 16, nos. V.1–3): 0.90, 0.88, and 0.87 g (the last being
very worn or worn looking). The ideal weight of a heavy gold
12th is 0.89 g. The Sardis silver 12th at 0.72 g and 24th at 0.35
are both lighter than normal, to judge from the 12 silver 12ths
that were known to Naster (1976, 133; median weight 0.77 g),
the 10 listed by Walburg (1991, 16–17: average weight 0.80
g), and the three silver 24ths listed by Walburg (1991, 17), to
which can be added the specimen in Athens (Regling 1915,
112); average weight of the last four is 0.40 g. As indicated by
the condition of their cleaned obverses, however, the two Sardis
specimens appear to have been damaged by corrosion, which
probably resulted also in some loss of metal. The above figures
and Naster’s tabulation of silver thirds (1976, 133) imply, inci-
dentally, that many fractions of the silver croeseids were struck
intentionally below ideal weight.
102
Naster 1965, 29–32; Kraay 1976, 31; Carradice 1987, 74–
5 (“These coins [heavy and early light gold] are obviously closely
connected”); Walburg 1991, 10–1. Nimchuck (2000, 11–8, 20–
2), however, thought that the heavy and early light gold pieces
were struck concurrently rather than in sequence.
103
According to Herodotus’ (1.50) comparison between the
gold and electrum bars that Croesus dedicated at Delphi, elec-
trum was reckoned as having a gold content of ca. 73% (Head
1887, 302), a figure that approximates the amount of gold in
2005] NEW ARCHAIC COIN FINDS AT SARDIS
gold-and-silver content of existing Lydian electrum
staters, which, as is known from laboratory analyses,
were regularly debased with added silver and other
metals. Accordingly, the two standards of the lion–
bull gold were likely derived from the two probable
means of valuation of the electrum coins they re-
placed, the heavy standard from the putative ideal
or “official” value of an electrum stater, and the light
standard from its true, intrinsic value.104 If, as is usu-
ally understood,105 the electrum coins were highly
overvalued, one suspects that when they had to be
called in for the minting of pure gold and silver
croeseids, they could not be readily redeemed at
their lower, intrinsic value; in that case people
would refuse to bring them in for redemption.
Hence, it is tempting to conjecture that they were
exchanged at a highly favorable 1:1 ratio with the
10.7 g heavy gold until enough electrum had been
withdrawn from private ownership and the Lydian
treasury could afford to cut its losses by reducing
the weight of the stater to a sensible 8.06 g; this way,
a 1:1 exchange could be maintained without fur-
ther disadvantage to the state. In other words, the
25% weight reduction of the gold stater would have
transferred the burden of unequal exchange from
the state to anyone who might still be holding onto
electrum coins, which in effect had now become
devalued. One result of this reconstruction is that
it strengthens the general perception that the heavy
lion-and-bull gold was tied specifically to the intro-
duction of the coinage and that the reduction in
weight ought to have followed soon thereafter.
To sum up, although the gold 12th from the Sardis
fortification recess may owe its worn appearance to
actual wear, a worn die, imperfect minting, or a com-
bination of these factors, there seems to be no plau-
sible case for placing its time of manufacture before
electrum flakes that may still be panned from the Pactolus River
(70–80% gold; Cowell and Hyne 2000, 172). Using this 73:27
gold-to-silver ratio for natural electrum, it follows that a 14.1 g
stater of natural electrum ought to have contained 10.29 g gold
and 3.8 g silver, which at the conventional gold-to-silver ex-
change rate of 1:13 1/3, yields an intrinsic total value of 10.58
g in pure gold, a figure that very closely approximates the 10.7
g of a heavy gold croeseid.
104
The Lydian electrum coins analyzed by Cowell and Hyne
(2000, 172) contain very close to 54% gold and 44% silver, in
both cases with a tolerance of only 1–2%. Accordingly, a 14.1
g stater of such adulterated electrum contained ca. 7.6 g of
gold and ca. 6.2 g of silver. As the value of the latter is equiva-
lent to 0.46 g of gold, the total intrinsic value in gold of the
adulterated electum stater is identical in weight to an 8.06 g
light gold croeseid. So far as we know, Walburg (1991, 12–3)
was the first to recognize this equivalency between the light
croeseid and coined electrum. Previously, scholars had routinely
assumed that the lighter croeseids were introduced merely to
613
Croesus. We accordingly may return with renewed
confidence to the conventional attribution of the
earlier (Phase I) croeseids—heavy gold, earlier light
gold, and earlier silver—to the 14 years (Hdt. 1.86)
of Croesus’ reign.
conclusion
Reliable chronological pegs for anchoring the
development of Archaic coinage in western Asia
Minor and Greece are far more limited than is of-
ten realized. Evidence from near the end of the
sixth century includes the several coins sealed in
the Apadana foundation deposits at Persepolis,106
impressions of coins on cuneiform clay tablets dated
to 500 or 499 also at Persepolis,107 and, in the West,
the production of coins at Sybaris before its destruc-
tion in 510. Earlier than these, the next reliable
point of chronological reference is the coinage
identified with Croesus, previously a matter of plau-
sible—though by no means decisive—inference
from textual references and surviving coins but one
that is now confirmed by the recovery of specimens
of this coinage from beneath destruction debris that
dates to the 540s.
Concerning the most challenging question in
Greek numismatic scholarship—how far back did
the preceding electrum coinage of Lydia exist be-
fore the time of Croesus?—widely diverging opin-
ions have appeared in print over the past several
decades, all of them focused on the electrum coins
and other valuables beneath the Artemisium at
Ephesus in deposits that some scholars have dated
as early as mid seventh century,108 while others have
urged a dating in the third quarter of the seventh
century109 or in the fourth quarter;110 others have pre-
ferred a date around 600111 or one even as late as ca.
560.112 The earliest coins in these deposits must go
allow a simpler exchange between gold and silver (1:10 instead
of 1:13 1/3) (Kraay 1976, 31; Naster 1976, 132–3; Melville
Jones 1998, 259–60).
105
Melville Jones 1998, 259, 263; Le Rider 2001, 85–6, 94–
100.
106
See Meadows (2003) and Zournatzi (2003), both with
good bibliographies.
107
Root 1988.
108
Kagan 1982, following Gjerstad 1937.
109
Weidauer 1975, 72–80; Williams 1991–1993; Weissl 2002,
315–21.
110
Holloway 1984, 5–9, 18.
111
E.g., Robinson 1951; Kraay 1976, 20–1; Stingle 2000–
2001, 41–4.
112
Bammer 1990, 149–50; approvingly noted by Price (1989,
10 n. 3) and excellently summarized in Le Rider (2001, 62–6).
But Weissl (2002, 315–9) emphatically disputes Bammer’s in-
terpretations of the “Central Basis” architecture and deposit.
614 NICHOLAS CAHILL AND JOHN H. KROLL
back, of course, before the deposits were assembled
and closed, and now that the Austrian excavators of
the Artemisium have recovered several coins with
pottery from the last quarter of the seventh century,113
a dating of the earliest electrum coins sometime in
the second half of that century now appears to be quite
firm. The Sardis destruction horizon sheds no light
on this issue. However, by providing the sole archaeo-
logical anchor between these early Ephesus deposits
on the one hand and the increasingly secure numis-
matic chronology of the later sixth century on the
other, its importance can hardly be overstated.
department of art history
university of wisconsin–madison
800 university avenue
madison, wisconsin 53706
ndcahill@facstaff.wisc.edu
department of classics
the university of texas at austin
1 university station c3400
austin, texas 78712-0308
jhkroll@mail.utexas.edu
Appendix: The Silver Coin of Miletus
RICHARD STONE, KENT SEVERSON, DYLAN
SMITH, NICHOLAS CAHILL, AND CRAWFORD H.
GREENEWALT, JR.
At the time of its discovery, the appearance of the
silver coin of Miletus was very different from that of
the two silver croeseid coins. The croeseids were
covered by a thick layer of dark-gray silver chloride
corrosion products and were hardly recognizable
as coins. The coin of Miletus had only a small
amount of compact earth on the surface and was
immediately legible. The earth was removed, re-
vealing a silver metallic surface with areas of black
corrosion product in the interstices. Based on its
color and morphology, the thin black corrosion on
the coin is presumed to be silver sulfide, but the
material has not been analyzed.
It is unlikely that the seemingly slight variations
in the alloy from which the croeseid and Milesian
coins were minted could account for their differ-
113
Kerschner (1997, 100), with references to pottery in the accompanying catalogue; the lion’s paw coin is Karwiese (1995, 137,
no. 74). Weissl informs us that several unpublished electrum coins found in recent excavations in the Artemisium also come from
deposits that date to the late seventh century (see Bammer and Muss 1996, 90).
[AJA 109
ent preservation. Diverse burial environments are
also not likely to produce such dissimilar corrosion;
indeed, the two silver croeseids come from rather
different burial environments, one in earth, the other
in brick fall, but show very similar corrosion states.
Coins and other silver objects from Lydian levels
at Sardis are typically covered with silver chloride
corrosion products. The two croeseid coins show
deep intergranular corrosion characteristic of sil-
ver chloride attack. The coin of Miletus appears to
have undergone similar corrosion, revealing its
surface grain structure as black lines surrounding
bright metallic silver. The manifest granular sur-
face structure of the Milesian coin reveals that it
too was attacked by silver chloride corrosion, and
that the corrosion products were subsequently re-
moved. This cannot be the natural, uncorroded
state of the coin. We have been unable to discover a
natural process that could remove the silver chlo-
ride layer and leave the coin in the state in which it
was turned in. The conclusion is that it was cleaned
artificially and, therefore, must be intrusive. There
is no trace of mechanical or abrasive cleaning on
the surface of the coin. There are, however, many
commonly available substances that could remove
the chloride layer without destroying the silver be-
neath, including acids, bases, and thiosulfate solu-
tions. (RS, KS, DS, NC)
The workman who “recovered” the Miletus coin
in 2002, after being questioned in March 2005 by a
family relation (brother-in-law) who was also an ex-
pedition guard and after being confronted with
evidence that the coin had been cleaned before its
alleged recovery during sieving, said that he had
placed the Miletus coin in the sieve because he
feared punishment by authorities if his possession
of the coin became known, and that his wife had
found it in a field. Subsequently, in May 2005, he
denied having said that he had placed the coin in
the sieve and that it had been recovered elsewhere.
Many villagers at Sardis and other parts of Turkey
fear that possession of antiquities will bring trouble
and punishment from government authorities.
Recovery of an Archaic coin from modern ground
surface at Sardis is unlikely, because ancient sur-
face remains are typically Late Antique, but not
impossible, because here and there they do coin-
cide with Archaic strata. Very puzzling is the appar-
ent circumstance that a coin “planted” in a mid
2005] NEW ARCHAIC COIN FINDS AT SARDIS
sixth-century B.C. context of excavation at Sardis
should be as close in date to that context as the
Miletus coin. Workmen employed by the expedi-
tion appear to be totally unfamiliar with the devices,
styles, and chronology of coins and other artifacts,
apart from some pottery, and the worker who recov-
ered the coin, during the three seasons that he was
employed by the Sardis Expedition, showed no such
familiarity. He had not been employed by the expe-
dition before the season in which the coin was re-
covered, and he was not hired in 2005. (CHG)
Works Cited
Balcer, J.M. 1972. “The Date of Herodotus IV.1 Darius’
Scythian Expedition.” HSCP 76:99–132.
Bammer, A. 1990. “A Peripteros of the Geometric Period
in the Artemision of Ephesus.” AS 40:136–60.
Bammer, A., and U. Muss. 1996. Das Artemision von Ephesos.
Mainz am Rhein: von Zabern.
Becker, F. 1988. “Ein Fund von 75 Milesischen Obolen.”
SNR 67:5–33.
Borrell, H.P. 1840. “An Inquiry into the Early Lydian
Money, and an Attempt to Fix the Classification of
Certain Coins to Croesus.” NC, 1st ser., 2:216–23.
Briant, P. 2002. From Cyrus to Alexander: A History of the
Persian Empire. Winona Lake, Ind.: Eisenbrauns.
Brijder, H.A.G. 1983. Siana Cups I and Komast Cups.
Amsterdam: Allard Pierson Series.
Burkert, W. 1985. “Das Ende des Kroisos. Vorstufen einer
Herodotischen Geschichtserzählung.” In Catalepton.
Festschrift für Bernhard Wyss zum 80. Geburtstag, edited
by C. Schäublin, 4–15. Basel: Seminar für Klassische
Philologie der Universität Basel.
Burstein, S.M. 1978. The Babyloniaca of Berossus. Malibu:
Undena Publication.
———. 1984. “A New Tabula Iliaca: The Vasek Polak
Chronicle.” GettyMusJ 12:153–62.
Buttrey, T.V., A. Johnston, K.M. MacKenzie, and M.L.
Bates. 1981. Greek, Roman and Islamic Coins from Sardis.
Archaeological Exploration of Sardis, Monograph 7.
Cambridge, Mass.: Harvard University Press.
Cahill, N.D. 2000. “Lydian Houses, Domestic Assemblages
and Household Size.” AASOR 57:173–85.
Cargill, J. 1977. “The Nabonidus Chronicle and the Fall
of Lydia.” AJAH 2:97–116.
Carradice, I. 1987. “The ‘Regal’ Coinage of the Persian
Empire.” In Coinage and Administration in the Athenian
and Persian Empires, edited by I. Carradice (BAR-IS
343), 73–95. Oxford: British Archaeological Reports.
Cook, R.M., and P. Dupont. 1998. East Greek Pottery. Lon-
don: Routledge.
Cowell, M.R., and K. Hyne. 2000. “Scientific Examina-
tion of the Lydian Precious Metal Coinages.” In King
Croesus’s Gold: Excavations at Sardis and the History of
Gold Refining, edited by A. Ramage and P. Craddock,
169–74. Cambridge, Mass.: Archaeological Explora-
tion of Sardis.
Descat, R. 2000. “Remarques sur les origins du
monnayage achéménide.” In Mécanismes et innovations
monetaires dans l’Anatolie achéménide: Numismatique et
615
histoire, Actes de la Table Ronde Internationale d’Istanbul,
22–23 mai 1997, edited by O. Casabonne, 1–8. Paris:
De Boccard.
Drews, R. 1969. “The Fall of Astyages and Herodotus’
Chronology of the Eastern Kingdoms.” Historia 18:1–
11.
Dusinberre, E.R.M. 1999. “Satrapal Sardis: Achaemenid
Bowls in an Achaemenid Capital.” AJA 103(1):73–102.
———. 2003. Aspects of Empire in Achaemenid Sardis. Cam-
bridge: Cambridge University Press.
Eph’al, I. 1984. “The Assyrian Siege Ramp at Lachish:
Military and Lexical Aspects.” Tel Aviv 11:60–70.
Evans, J.A.S. 1978. “What Happened to Croesus?” CJ
74:34–40.
Francis, E.D., and M. Vickers. 1983. “Signa priscae artis:
Eretria and Siphnos.” JHS 103:49–67.
Giesecke, W. 1938. Antikes Geldwesen. Leipzig: K.W.
Hiersemann.
Gjerstad, E. 1937. “Studies in Archaic Greek Chronol-
ogy II: Ephesus.” AnnLiv 24:15–34.
Grayson, A.K. 1975. Assyrian and Babylonian Chronicles.
Locust Valley, N.Y.: J.J. Augustin.
Greenewalt, C.H., Jr. 1971. “Fikellura and ‘Early
Fikellura’ Pottery from Sardis.” CSCA 4:153–80.
———. 1973. “Ephesian Ware.” CSCA 6:91–122.
———. 1979. “The Sardis Campaign of 1977.” BASOR
233:1–32.
———. 1990. “The Sardis Campaign of 1987.” BASOR
Suppl. 27:1–28.
———. 1992. “When a Mighty Empire Was Destroyed.
The Common Man at the Fall of Sardis, ca. 546 B.C.”
Proceedings of the American Philosophical Society 136:247–
71.
———. 1994–1995. “Sea Serpents at Sardis.” Harvard
University Art Museums Review 4:1, 6.
———. 1997. “Arms and Weapons at Sardis in the Mid
Sixth Century B.C.” Arkeoloji ve Sanat 79:2–20.
———. 2003a. “Sardis: Archaeological Research and
Conservation Projects in 2002.” Kazi Sonuçlari Toplantisi
25. Ankara: Kültür Bakanligi Milli Kütüphane Basimevi.
———. 2003b. “Archaeological Exploration of Sardis.”
Harvard University Art Museums. Annual Report 2002–
3. Cambridge, Mass.: Harvard University Press.
———. 2005. “Sardis: Archaeological Research and Con-
servation Projects in 2003.” Kazi SonuçlariToplantisi 26.
Ankara: Kültür Bakanligi Milli Kütüphane Basimevi.
———. n.d. “The Sardis Campaigns of 1999–2003.” Un-
published.
Greenewalt, C.H., Jr., N.D. Cahill, H. Dedeoglu, and P.
Herrmann. 1990. “The Sardis Campaign of 1986.”
BASOR Suppl. 26:137–77.
Greenewalt, C.H., Jr., N.D. Cahill, and M.L. Rautman.
1987a. “The Sardis Campaign of 1984.” BASOR Suppl.
25:13–54.
Greenewalt, C.H., Jr., and A.M. Heywood. 1992. “A Hel-
met of the Sixth Century B.C. from Sardis.” BASOR
285:1–31.
Greenewalt, C.H., Jr., A. Ramage, D.G. Sullivan, K. Nay
r,
and A. Tulga. 1983. “The Sardis Campaign of 1979
and 1980.” BASOR 249:1–44.
Greenewalt, C.H., Jr., C. Ratté, and M.L. Rautman. 1993.
“The Sardis Campaigns of 1988 and 1989.” AASOR
51:1–43.
———. 1994. “The Sardis Campaigns of 1990 and 1991.”
AASOR 52:1–36.
616 NICHOLAS CAHILL AND JOHN H. KROLL
Greenewalt, C.H., Jr., and M.L. Rautman. 1998. “The
Sardis Campaigns of 1994 and 1995.” AJA 102(3):469–
505.
———. 2000. “The Sardis Campaigns of 1996, 1997, and
1998.” AJA 104(4):643–81.
Greenewalt, C.H., Jr., M.L. Rautman, and N.D. Cahill.
1987b. “The Sardis Campaign of 1985.” BASOR Suppl.
25:55–92.
Greenewalt, C.H., Jr., M.L. Rautman, and R. Meriç. 1986.
“The Sardis Campaign of 1983.” BASOR Suppl. 24:1–
30.
Greenewalt, C.H., Jr., M.L. Rautman, and C. Ratté. 1995.
“The Sardis Campaigns of 1992 and 1993.” AASOR
53:1–36.
Greenewalt, C.H., Jr., E.L. Sterud, and D.F. Belknap. 1982.
“The Sardis Campaign of 1978.” BASOR 245:1–34.
Greenewalt, C.H., Jr., D.G. Sullivan, C. Ratté, and T.N.
Howe. 1985. “The Sardis Campaigns of 1981 and
1982.” BASOR Suppl. 23:53–92.
Hanfmann, G.M.A., and J.C. Waldbaum. 1975. A Survey
of Sardis and the Major Monuments Outside the City Walls.
Cambridge, Mass.: Harvard University Press.
Haslam, M.W. 1986. “The Fall of Sardis in the Roman
Chronicle.” ZPE 62:198.
Head, B.V. 1887. “Electrum Coins and Their Specific
Gravity.” NC, 3rd ser., 7:227–308.
Holloway, R.R. 1984. “The Date of the First Greek Coins:
Some Arguments from Style and Hoards.” RBN 130:5–
18.
How, W.W., and J. Wells. 1928. A Commentary on Herodotus.
Oxford: Clarendon Press.
Kagan, D. 1982. “The Dates of the Earliest Coins.” AJA
86(3):343–60.
Karwiese, S. 1995. Die Münzprägung von Ephesos. Vienna,
Cologne, and Weimar: Böhlau.
Kerschner, M. 1997. “Ein stratifizierter Opferkomplex
des 7. Jh.s v. Chr. aus dem Artemision von Ephesos.”
ÖJh 66:87–226.
Kraay, C.M. 1976. Archaic and Classical Greek Coins. Ber-
keley and Los Angeles: University of California Press.
Le Rider, G. 1998. “Le début du monnayage achéménide:
Continuation ou innovation?” In Light on Top of the
Black Hill: Studies Presented to Halet Çambel, edited by G.
Arsebük, M. Mellink, and W. Schirmer, 663–73.
Istanbul: Ege Yayinlari.
———. 2001. La naissance de la monnaie, Pratiques
monétaires de l’orient ancien. Paris: Presses Universitaires
de France.
Mallowan, M.E.L. 1972. “Cyrus the Great (558–529
B.C.).” Iran 10:1–17.
———. 1985. “Cyrus the Great (558–529 B.C.).” In The
Cambridge History of Iran, edited by I. Gershevitch, 392–
419. Cambridge: Cambridge University Press.
Mansfield, J. 1983. Review of The Chronicle of Eusebius and
the Greek Chronographic Tradition, by A.A. Mosshammer.
Mnemosyne 36:205–6.
Meadows, A.R. 2003. “The Apadana Foundation Deposit
(IGCH 1789): Some Clarification.” NC 163:242–3.
Melville Jones, J.R. 1993. Testimonia Numaria, Greek and
Latin Texts Concerning Ancient Greek Coinage. Vol. 1.
London: Spink and Son.
———. 1998. “The Value of Electrum in Greece and
Asia Minor.” In Studies in Greek Numismatics in Memory
of Martin Jessop Price, edited by R. Ashton and S.
Hurter, 259–68. London: Spink.
[AJA 109
Miller, M. 1963. “The Herodotean Croesus.” Klio 41:58–
94.
Mionnet, T.-E. 1833. Description de médailles antiques. Suppl.
6. Paris: Bibliothèque du Roi.
Mosshammer, A.A. 1979. The Chronicle of Eusebius and the
Greek Chronographic Tradition. Lewisburg, Pa.: Bucknell
University Press.
Naster, P. 1965. “Remarques charactéroscopiques et
technologiques au subject des crésèids.” In Congresso
Internazionale di Numismatica, Roma 1961. Vol. 2, Atti,
25–36. Rome: Istituto Italiano di Numismatica. (=
Naster 1983, 30–9.)
———. 1976. “The Weight-System of the Coinage of
Croesus.” In Proceedings of the 8th International Congress of
Numismatics, New York–Washington 1973, 125–33. Paris,
Basel: Association internationale des Numismates
professionels. (= Naster 1983, 68–75.)
———. 1983. Scripta Nummaria: Contributions à la
méthodologie numismatique. Louvain-la-Neuve: Société
royale de numismatique de Belgique.
Nimchuk, C.L. 2000. “The Lion and Bull Coinage of
Croesus.” Journal of the Classical and Medieval Numis-
matic Society, 2nd ser., 1:5–44.
Pedley, J. 1972. Ancient Literary Sources on Sardis. Cam-
bridge, Mass.: Harvard University Press.
Price, M.J. 1984. “Croesus or Pseudo-Croesus? Hoard
or Hoax? Problems Concerning the Sigloi and
Double-Sigloi of the Croeseid Type.” In Studies in
Honor of Leo Mildenberg, edited by A. Houghton, S.
Hurter, P. Erhart Mottahedeh, and J. Ayer Scott, 211–
21. Wetteren: Cultura.
———. 1989. “Darius I and the Daric.” RÉA 91:9–13.
Ramage, N.H. 1986. “Two New Attic Cups and the Seige
of Sardis.” AJA 90(4):419–29.
Regling, K. 1915. “Dareikos und Kroiseios.” Klio 14:91–
112.
Robinson, E.S.G. 1951. “The Coins from the Ephesian
Artemisium Reconsidered.” JHS 71:156–67.
Root, M.C. 1988. “Evidence from Persepolis.” NC 148:1–
12.
Rotroff, S.I., and A. Oliver. 2003. The Hellenistic Pottery
from Sardis: The Finds through 1994. Cambridge, Mass.:
Harvard University Press.
Schaeffer, J.S., N.H. Ramage, and C.H. Greenewalt, Jr.
1997. The Corinthian, Attic and Lakonian Pottery from
Sardis. Cambridge, Mass.: Harvard University Press.
Schaudig, H. 2001. Die Inschriften Nabonids von Babylon
und Kyros’ des Grossen samt den in ihrem Umfeld
entstandenen Tendenzschriften: Textausgabe und
Grammatik. Münster: Ugarit-Verlag.
Sealey, R. 1976. A History of the Greek City States, 700–338
B.C. Berkeley and Los Angeles: University of Califor-
nia Press.
Smith, S. 1924. Babylonian Historical Texts. London:
Methuen.
Stingle, T. 2000–2001. “Barren oder Münzen? Über-
legungen zum Beginn der Elektronprägung in
Westkleinasien.” Boreas 23/24:35–52.
Storck, H.A. 1989. “The Lydian Campaign of Cyrus the
Great in Classical and Cuneiform Sources.” AncW
19:69–76.
Ussishkin, D. 1990. “The Assyrian Attack on Lachish:
The Archaeological Evidence from the Southwest
Corner of the Site.” Tel Aviv 17:53–86.
Vickers, M. 1985. “Early Greek Coinage. A Reassessment.”
2005] NEW ARCHAIC COIN FINDS AT SARDIS 617
NC 145:1–44. im Artemision von Ephesos.” ÖJh 71:313–46.
Wade-Gery, H.T. 1951. “Miltiades.” JHS 71:212–21. West, S. 2003. “Croesus’ Second Reprieve.” CQ 53:416–
Walburg, R. 1991. “Lydisch oder Persisch? Ein Goldobjekt 37.
aus der Frühzeit der Münzprägung.” SNR 70:5–17. Williams, D. 1991–1993. “The Pot Hoard from the Ar-
Weidauer, L. 1975. Probleme der frühen Elektronprägung. chaic Artemision of Ephesus.” BICS 38:98–103.
Freiburg: Office du Livre. Zournatzi, A. 2003. “The Apadana Coin Hoards, Darius
Weissl, M. 2002. “Grundzuge der Bau- und Schichtenfolge I, and the West.” AJN 15:1–28.