GREEK PALAEOGRAPHY

By the same author:

L E SECOND LIVRE DE L 1 ECONOMIQUE D'ARISTOTE

PINDARE AU BANQUET, LES FRAGMENTS DES SCOLIES

 SHORT MANUAL
OF

GREEK PALAEOGRAPHY
BY

B.A.VAN GRONINGEN
Emeritus /'rr!fessor in the University of Leyden

WITH 12 PLATES AND I 5 FIGURES

FOURTH PRINTING

A. W. Sl.JTH OFF - LEYDEN

1967
No part of this hook may be rt"pro<lucnl in any form, hy print, photoprint, mirrof1lm
or any other means without written permission from tlw publisher.
Printed in the Netherlands
 TABLE OF CONTENTS
Pages
PREFACE 7
CHAPTER I: Introduction 9
§ 1. Definition and Aim of Greek Palaeography 9
§ 2. History and Development • • 9
§ 3. Short Bibliography. IO
CHAPTER II: The Oldest History of Greek Writing • 13
§ 4. Aegean Writing • • • • • • 13
§ 5 . Northern Semitic and Greek Alphabet • 14
§ 6. Groups of Greek Alphabets 18
CHAPTER III: Writing-material and Book-form . 20
§ 7. Papyrus, Parchment and Paper 20
§ 8. Pen and Ink; Ruling 22
§ 9. The Roll • 22
§ IO, The Codex • • 23
§ II. CoJlections 25
CHAPTER IV: The Writing. 28
§ 12. Typology • • • 28
§ 13. Literary Inscriptions 29
§ 14. Literary Papyri • 30
§ 15, Uncial Writing • • 32
§ 16. Minuscule Writing . 34
§ 17. Codices Vetustissimi ' 35
§ 18. Codices Vetusti • • 37
§ 1 9. Codices Recentiores 41
§ 20. Codices Novelli • 43
§ 21. Abbreviations and Symbols. 43
§ 22. Expansion of Greek Writing 47
CHAPTER V: Various Items. • 49
§ 23. Ornamentation • • 49
§ 24. Numbering; Stichometry and Colometry 50
§ 25. Subscriptions and Personal Remarks. 50
§ 26. Accentuation and Punctuation • 51
§ 27. Corrections 53
§ 28. Glosses and Scholia 54
§ 29. Errors • • 54
INDEX AND LIST OF FIGURES 60
ADDENDA 65
PLATES at the end

5
 PREFACE

The character of this Manual was described in the first edition (1940) as
follows: "This booklet has no pretence to scientific originality. It reproduces
the courses which I have given for years to the graduates (candidates) of
Leyden University. It is meant in the first place for the student; if any one
else can use it, he is welcome to it. I emphasize the fact that Greek Palaeo­
graphy is not treated here as an autonomous branch of knowledge-a thing
which it certainly is-but as an ancilla philologiae, taking philology in the
strict sense of the word. So I have not written for the epigraphist, nor for
the papyrologist either, but for those who study Greek literature and who
wish to know the principal styles in which it was written and handed down.
That, nevertheless, incriptions as well as documents are mentioned here and
there stands to reason. One may also expect every young hellenist to have
some insight into the oldest history of the Greek alphabet. The theory
exposed here only turns to full advantage by reading manuscripts or, if need
be, reproductions".
This second edition differs from the first in the following respects. Errors,
omissions and misprints have been corrected. A new paragraph (29) lists
the most common mistakes of copyists. Two plates (I, II) have been added
reproducing two literary papyri of different periods. Finally, the description
of the plates is followed by a transcription of a few lines, wherever the deci­
phering seems difficult for a beginner.
In the first edition I had to thank Dr. T. P. Sevensma, sometime director
of the University Library at Leyden, for allowing me to reproduce parts of
Leyden manuscripts. For this second edition I could profit by useful obser­
vations presented by reviewers of the first, by Dr. E. Hulshoff Pol, assistant
keeper of manuscripts in Leyden, and my colleagues Dr. J. C. Kamerbeek
(Amsterdam) and Dr. W. J. Verdenius (Utrecht).

PREFACE TO THE THIRD EDITION

This new edition is an offset-reproduction of the second. A small number

of Addenda proved necessary. I thank Dr. K. A. de Meyier of the Depart­
ment of Manuscripts of the Leyden University Library for some useful re­
marks.
Leyden, August 1963
 PREFACE TO THE FOURTH PRINTING

Except for a few corrections this fourth reprint is identical to the third.
Leyden, Spring 1967

8
 CHAPTER I

INTRODUCTION
§ t. Definition and Aim of Greek Palaeography

We generally call palaeography the branch of knowledge which studies

methods of writing used in former times, It is born from the necessity to
read texts written in earlier centuries and to ascertain as exactly as possible
the contents of those texts, It is closely related to the so-called codicology,
which studies form, material and history of the manuscripts themselves.
If Greek texts are concerned, we speak of Greek palaeography.
It is not interested in modern writing, even though it may incidentally
make use of it. The time-limits depend on the preconceived aims: the
student of ancient Greek literature need not go any further than the XVIth
century A.D.
Greek palaeography aims at three purposes: first of all practical ability in
reading the manuscripts; secondly insight into Greek writing as such, its
characters, types and styles, the influences to which it has been submitted and
which it has exercised, its history; thirdly, in combination with codicology,
the accurate description of the ancient book-formL

§ 2. History and Development

There have always been scholars who, by natural disposition or by practice,

were more or less expert in reading old manuscripts. The Greek man of
letters was often obliged to search for his sources and data in old books. The
oldest literary texts themselves were, for a large part, not written in the
Milesian alphabet which became normal later on. For a transposition into
other Greek characters (µe-raxaea,n71ernµo() a certain skill in reading old
handwritings was necessary. Later on, copying was often based on old
books, written in characters different from those which the copyist used
himself. The humanists of the Renaissance were, as a rule, able to read with
great ease manuscripts d:iting from all periods. The same may be said of
later hellenists, but men like Rciske, Bekker and Cobet were palaeographers
only in the second place, one would he inclined to say, by mere necessity;
their chief aim was philological. The first person who studied Greek palaco-

9
 graphy for its own sake was Bernard de Montfaucon, who lived in the
XVIIIth century (see § 3).
Subject for research is in particular the duct us of the separate characters or
groups of characters, this means the successive movements by which the
hand moves the pen of stilus; then the style of the writing as a whole; the
influence of the material, of the school and the method; the routine; the
aesthetic taste of the period and of the individual. All these investigations
have not yet made much progress. It still is a hazardous task to determine
with accuracy the time of an undated manuscript: differences of one, even
more centuries, do occur. Neither has a systematic study of the numberless
abbreviations ever been published. The present handbooks limit themselves
as a rule to a description of the different hands. General or particular lines
of development, typical characteristics for certain periods and certain places
can hardly be given. The main reason is that the art of writing has been,
except for one important brea k in the IXth century, astonishingly conserva­
tive in Byzantine times. The uniformity of culture in the whole realm, the
strong sense of tradition, the often recurring tendency towards convention­
ality and mannerism make systematizing very difficult.

§ 3. Short Bibliography
On writing in general:

The Alphabet, 2 vols. London 1899.

I. Taylor,
H. Jensen, Geschichte der Schrift. Hannover 1925; new edition in 1935 with
the title: Die Schnft in Verg angenheit und Gegenwart.
D. Diringer, The Alphabet, London 1947.

On Greek palaeography in general:

B. de Montfaucon, Palaeographi a Graeca, six vols. with a seventh as appen­
dix. Paris 1708. (Still useful on accoun t of the amount of material
brought together).
W. Wattenbach, Anleitung zur Griechischen Palaeog raphie 3 • Leipzig 1895.
V. Gardthausen, Griechische Palaeographie2• Leipzig 1911-13. (Rich in
data, but not very clear and insufficient as regards the papyri).
E. M. Thompson, An Introduction to Greek and Latin Palaeography. Oxford
1912. (Recommendable).
A. Mentz,Geschichte der Griechzsch-Romischen Schrift. Leipzig 1920.
P. Maas, Griechische Palaeographie, in Gercke-Norden, Einleitung m die
Klass. Altertumswissenschaft, I 9, pp. 69--81. Leipzig 1927.
A. Sigalas, 'lar:oela Tijc; 'EAA:,1yi"iJc; I'e acpij<;. Thessaloniki 1934. (The
be t and clearest work which now exists. The illustrations, however,
are bad).

IO
 See also H andbuch der Archiiologie I (in I. von Muller -W. Otto's H andbuch
der klass. Altertumswissenschaft). Munich 1939: Die Schrift und die Schrift-
zeugnisse p. 147-354 (v. Bissing, Rehm, Pernice, Arntz), especially p. 189-206
(Rehm).
On Greek books:
Th. Birt, Das antike Buchwesen (I. v. Muller's Handbuch d. kl. Alt. -wiss. I 13)
Munich 1913.
W. Schubart, Das Buch bei den Griechen und Romern 2 • Berlin 1921.
H. L. Pinner, The World of Books in Classical Antiquity. Leiden 1948.
On inscriptions:
W. Larfeld, Griechische Epigraphik 3 (in I. von Muller's Handbuch d. kl.
Alt.-wiss. I 5. Munich 1914. Especially the chapters IV: Ausfiih-
rung, and VIII: Schriftzeichen).
On papyri:
F. G . Kenyon, The Palaeography of Greek Papyri. Oxford 1898.
W. Schubart, Griechische Palaeographie (in I. von Muller's Handbuch,
I 4, I. Munich 1925. 155 of the 184 pages deal with the papyri
and are excellent, the others are of little importance).
M. Norsa, La scrittura letteraria greca del secolo IV a.C. all' VII d.C.
Florence 1940. (With plates. Important).
Facsimiles and Reproductions:
The Palaeographical Society (since 1892 New Pal. Soc.) edits complete
facsimiles of important texts and documents. London 1873 etc.
H. Omont, Fae-similes de manuscrits grecs des XVe et XV/e siecles. Paris
1887.
H. Omont, Fae-similes des manuscrits grecs dates de la Bibliotheque nationale
du /Xe au X/Ve siecle. Paris 1891.
H. Omont, Fae-similes des plus anciens manuscrits grecs .••• de la Biblio-
theque nationale du /Ve au Xlle siecle. Paris 1892.
S. G. de Vries, Codices Graeci et Latini photographice depicti. Leiden 1897
etc. (The first volume was edited by N. G. Du Rieu).
S. G. de Vries, Album palaeographicum. Leiden 1909. (Gives separate
pages taken from the former publication).
E. Steffens, Proben aus Griechischen Handschriften und Urkunden. Trier
1912.
P. Franchi de' Cavalieri et J. Lietzmann, Specimina Codicum Graecorum
Vaticanorum 2 • Berlin 1929. (Very useful).
W. Wattenbach, Scripturae Graecae specimina 4 • Berlin 1936.
II
 L. Th. Lefort- }. Cochez, Palaeographisch Album van gedagteekende
Grieksche minuskelhandschrzftenuit de /Xe en Xe eeuw, met enkele
specimina van handschriften uit de Xl - XV/e eeuw. Louvain 1932.
K. and S. Lake, Dated Greek minusculemanuscriptsto the year 1200 A.D.
Boston (Mass.) II vol., 1934-1945.
0. Kern, InscriptionesGraecae. Bonn 1913.
W. Schubart, Papyri Graecae Berolinenses.Bonn 1911.
M. Norsa, see above.
Remark: In order to illustrate the arguments, I refer not only to the plates
at th e en d of the book, but also to the reproductions in the works of Thomp-
son ( = Th.), and Cavalieri-Lietzmann ( = CL.), occasionally also to Norsa
(N.), Omont (0.d. = Omont, Mss. date s; 0.nd. = Omont, Mss. non dates),
de Vries ( - V.), Lefort and Cochez ( - LC), Wattenbach ( = W.) and
Steffens ( ·= St.), whereas only few references are made to Kern's Inscrip-
tions ( K.) and Schubart's Papyri ( = Sch.).

Additi onal Note.

I draw attention to Prof. A. Bataille's very interesting book Pour une
terminologie en paleograplzie grecque, Paris 1954. The palaeographers by
profession will have to decide whether or not they accept his proposals.

J''
 CHAPTER II

THE OLDEST HISTORY OF GREEK WRITING

§ 4. Aegean Writing
During the second millennium B. C. the important centres of culture in
and round the Aegean Sea, Crete, Tiryns, Asine, Pylus, all practised the
art of writing. The Cretan scripts (the so-called Scripta Minoa) are best
known owing to the excavations made by A. J. Evans in Cnossus. It is
likely that the first stage was ideographic ; the second syllabic development
went through three stages, the first being ideographic, the second syllabic
and the third even alphabetic.
The following types are distinguished:
I. Hieroglyphic A: oldest form, only on gems and the like; about go signs.
B: often used since Middle-Minoan II (1900-1750);
about 95 signs of which some 50 in common with A.
II. Linear A: generally used since Middle-Minoan III (1750-1580);
about go characters.
B: special chancery-hand used in Late-Minoan II ( 1450-1300);
about 88 characters, of which nearly half in common with A.
This Linear Bis probably deciphered now; for 65 signs a phonetic value
has been proposed. Most unexpected result of all: the language appears to
be Greek (cf. M. Vestris and J. Chadwick, Evidence for Greek Dialect in the
Mycenaean Archives, J. Hell. Stud. LXXIII, 1953 p. 84-105).
The syllabic script which was still used in Cyprus in the Vth cent. B.C. goes
back to a similar writing; it possesses 50 characters for consonants followed
by a vowel, 5 for vowels only, besides a vertical stroke for the separation of
words or word-groups. It distinguishes neither quantities, nor tenues and
mediae; it has no separate signs for the aspiration or for aspirate consonants;
it never doubles a consonant, and often neglects the nasals; if in the same
word two consonants follow each other, the vowel of the second is also
written with the first. E.g. the first words on the bronze table of Edalium,
dating from 445 B.C. (K2), are written: o te ta po to li ne I e ta li o
ne I ka te vo ro ko ne ma to i I , which means: lfn: -idv n-i6liv 'Ec5a,hov
XO'iefOltrov Mdc5o,.
 All these manners of writing have nothing to do with the later Greek
alphabet. With the sudden fall of the Aegean civilization the art of writing
was lost. At a given moment, the Greek had to learn afresh a new method
to fix his thoughts.

§ ;. Northern Semitic and Greek Alphabet

The Greek alphabet is nothing but a Phoenician or better Northern

Semitic alphabet adapted to the needs of the Greek language. The Greeks
themselves have never forgotten that they learned the characters from the
Phoenicians (see e.g. the myth of Cadmus 1)), They presupposed, be it
wrongly, the existence of an original alphabet of 16 or 18 characters:
a p yd,; m , "A. µ v one a-iv (91). Even without this tradition it would
be clear that it was directly taken over. This appears from the names,
the order of the alphabet, the direction of the writing and partly even
from the phonetic value of the signs.
The names are almost identical, The Greek has only adapted the foreign
words to his own pronunciation: practically unmodified are waw = fafl,
pe = ;,i,;i (only later called ni), taw = -iav. A final muta, which Greek
does not know, is followed by an a in l£A.cpa = aleph, Pfi-ia = beth, -q-ia
(formerly 17-ra) = kheth, i>fj-ia = teth, l6ha = yod, xanna = kaph, Aaµ(P)da
= liimed, ,conna = qoph. Neglect of aspiration explains he = ,;l (only
much later called e V',A.6v). .dtha is still closely related to daleth. Instead
of zayin we find tiJ-ia under the influence of �-ra and l>ij-ia. Nun becomes
vii and under its influence mem becomes µii (still called µiJJ by Democritus
fr. 19 D-Kr.). Ros (the Canaanite name of the Aramaic and Hebrew res)
becomes ero.Siimekh perhaps became alyµa (aiyµa) via the form *alµ xa 2),
The transformation of gimel into yaµµa (Democritus 1.1. calls it ytµ µa),
and of sin into aav remain unexplained, whereas the names 'ayin and ssiide
are not known in Greek.
The sequence of the Semitic alphabet appears in acrosticha as the Psalms
I II, 119 and 145 and is as follows: aleph, beth, gimel, daleth, he, waw, zayin,
kheth, teth, yod, kaph, lamed, mem, nun, siimekh, 'ayin, pe, ssiide, qoph,
res, sin, taw. The resemblance to the Greek is clear.
The Semite wrote and still writes from right to left. This was also the
original direction of Greek writing. Soon, however, it changes into the way
which is called Povar:eocp'1Jd6v, the manner in which a plough goes through
the field, first e.g. from right to left, then from left to right etc. (K. 5a). A
change in direction at the same time changes the form of the letters: written
from left to right they reproduce the original form as seen in a mirror.
Already towards the end of the Vlth cent, it fell into disuse; from that time

1) Herodotus, V 58. Cf. Rhys Carpenter, Letters of Cadmus, Am. J. of Philol.

56 (1935) P· 5-13.
2) Unless it simply means 'hissing' cf. ulCw.

14
 1
I 2 )
I' I f
' 1

Clbpli ~ ~ ~ k.O. A A
8a1, .J 1 :i 3 s B
ji.mtl 1 /' /\
cW.lh 0
1
/)> ,
J 'J
6 0,. D.
Ri 3~ ~ n '\ t E
ll.l'll..r 'i '( r 1 y "( -
~;:,,. I z l I 1 I
Hitl. es ~ n BH 8 H H
t..~h EB ® u IB EB Q

~M % j
., z ~ I
~a.pl. '¥ r :i t ~ ~
fanwi Ll t ';, l t,. l,..
11\im
~
;
~
~
,
II) Y.)
l
"/
'1
/V'
/v
M
N Fig. I,
Col. 1 gives the Semitic
~ =p i 0 ~ \. - names; col. 2 the cha-
racters written on the
'4:1•n
. 0 0 y 0 0 0 sarcophagus of Ahiram,
king of Byblos (XIII-
pt 1) 1 8 7
,.,,.
r r XI cent. B.C.) ; col. 3
those of the inscription
ss.a.li /4' ~ vi - of Mesa, king of Moab
(:!: 850 B.C.) ; col. 4
p cp ~ cp 9 -
qoph
,u; ?
f

" , "
~ f p
the usual Hebrew cha-
racters; col. 5 old Greek
letter s, mainly from
Thera, written from
"·
6Ln w w iV \N' N\ JV1 i right to left; col. 6 the
same written from left
t......., 1" x n T T T to right; col. 7 the nor-
mal Attic characters.

on the Greek writes definitely and without exception from left to right.
Even at a cursory comparison of the two alphabets (see fig. 1) we see that
the forms and values are for the greater part identical. But an absolute
conformity was impossible. The Semitic alphabet has only consonants;
as a rule, vowels are not written and, if so, only by means of auxiliary dots
and strokes. Besides, the Semitic words, all based on roots of three conso-
nants, are not unsuitable for such an alphabet, whereas that structure has, at
the same time, sharpened the faculty for distinction of consonantic phonems.
The existence of the aleph is a clear proof, for it indicates the glottal occlusive,

15
 of which the Gre eks 1) were never clearly aware. To a language rich in
vowels and forms as the Greek language is, this system was very inappropriate.
On the other hand the Greeks did not need everything which the Semitic
alphabet offered: they did not want the aleph, nor the diversity of sibilants,
nor the different aspirations. So they had to adapt it.
In order to understand this adaptation, as far as the phonetic value is con-
cerned, one should know that the Gre ek first learned the names of the letters
in a row and only afterwards the characters and their value 2 ). That is how
he acted with the Semitic alphabet also. In order to write a word, he recited,
as it were, at every sound of which he became conscious, the series of names
till he found the corresponding sign or the sign which seemed to suit best.
In doing so he proceeded according to the principle of acrophony, which had
also guided the Sem ites 3 ). In ord er to write e.g. the word li.xea, in which he
heard three different phonems, he could use the first letter of the alphabet
twice (aleph-li.Acpa), for, according to him, the word did not b egin with a
consonant, but with a vowel; for the two other phonems he had to repeat the
alphabet respectively till kaph - xanna and rM- ~w. In this way the follow-
ing characters can be easily explained: a p y iSF {; t (the consonant y being
already lost in Greek) ,d µ,v n e u.
In addition, two other principl es are applied, regarding the valu e as well as
the form of the letters. The former is the principle of differentiation. The
Semitic teth and taw differ in so far that the former is a more emphatic dental
than the latter. The Greek differentiates according to his own needs and
uses teth for the aspirate dental and taw for the tenuis. •Ayin becomes omi-
kron and related sounds, because aleph became alpha. Because epsilon and
related sounds were written with he, th e Greeks who still used the aspiration
had to use kh eth for it 4 ); in its Semitic name the h-sound was much more
audible than in he. The Ionic dialect and oth er dialects which had lost this
phonem, differentiated two e-sounds: for the short one they took he, for the
long one kh eth. The latter additional prin ciple tends to normali zatio n and
comes especially into force when the Semitic alphabet gave too little. Because
th e Greek alphabet had a separate sign for th e aspirate dental, it seemed right
to form also separate characters for the two other aspirate mutae 5 ). The
wish to create new signs for the combination of a guttural with a sibilant m
') Th e smooth breathing notes it, it would seem, at the beginning of a _word.
Even this does not imply a conscious apperception, it is only an outward sign of
the absence of the rea l aspir ation.
•) Cf. D ion. Halie. Demosth. 52 : .neww .. µ.i .. Ta o..6,..aTa TW.. cn:o,xcfo, .. ,:fir; 'P"'"ii
·
o..noAaµ/Jchn:m - - bu:n:a rov, Tv.nov, rt: UV'CW'Y><a• .5v.. &,..,,,..
Quin ti!. I, I, 24
does not agree.
') Cf. M. P. Nilsso n, Uebernuhm ~ und En tu,icklung des Alphabets durch die Grie-
chen, Hi st. - fil. M edel., K gl. Dans ke Viden sk. Selskab I, 6, 1917-8.
4 ) Not only as a separa te phonem, but also in combination with the mut ae, n H
and K H. As seen above, th e case of the aspirate dental was different.
' ) A further going normalization then gave birth to e H for the aspirated dental
next to K H and n H.
and of a labial with the sibilant (1P) may be 1 2
attributed to the existence of the t, combination
of the sibilant with the dental media or the reverse. A <t:AA.A
At any rate the distinction of the long and short B
o-sound (o and ro) is a direct consequence, as we fl "" c " t.l1
r /\11'( C
have seen, of the existence, in some dialects, of 71
A O t>
next to e. The new signs are probably variations
of existing ones, but nothing is sure in this res- E f=E~~X
pect ).
1 They were naturally added at the end F P~tt=N
of the series, but the phonetic value is not always z I
the same. H BOB'){
There are unsolved difficulties: waw was natur- e ®(90~
ally destined to become the consonant F and the
vowel 11, but it is strange that we find the variated
5~ i r
K I<
sign, for the F, in the original place, and the old
sign, for the 11, at the end. The story of the three A I' Id' I-
Semitic sibilants siimech, ssiide and sin, is still full M /""/N '\I\
f'N7
*
of riddles 2). N
The forms of the letters were very uncertain,
nor do they always represent the same sound:
-:E .K itt
0 ocoo
so-called homographs occur rather often (see
fig. 2).
n rr c
As to the time of the adoption, no direct indic - 9 'P Q y ~
ation is at our disposal, nor is it possible to arrive p p~ D
at a reliable conclusion on the strength of indirect i ~~sf
considerations. The oldest written documents T
quoted by the Greeks go back to the Vlllth y VYY
cent., e.g. the list of victors at Olympia, whir.h q> a, m 4'
starts in 776; the Athenian lists of archons are
more recent and begin in or about 685. In keep-
x +
ing with this terminus ante quern is the probable
chronology of the Iliad and the Odyssey: we can
safely fix their origin in the Vlllth cent. Writing Fig. 2 .
is perhaps not essential to the spreading of such Col. I shows normal
characters of the classic
voluminous works, but it is indispensable for their period ; col. 2 various
creation. Homer does not mention the art of other forms found in
inscriptions.
writing 3 ); he apparent Iy thought that it was a
recent invention. The Dorian invasion which ruined Mycen ean culture,

1)An attempt is made by Sigalas, p. 84.

') Besides olyµa we find the name <10.-(cf. Herod . I, 139, Pind. fr. 6r,2 Bowra) and
the old Ionic had a special sign (T) for the hard sibilant, which disappeared when
consonants were doubled. Certain alphabets have assumed the siimckh, for which no
direct use was possible, for the combination ks.
•) Not even in Z 168 ff., where he speaks of o,)µarn .lvyea instead.
2 17
 is commonly attributed to the XIIth or Xlth cent. Therefore the
Semitic alphabet was adopted by the Greeks somewhere between rooo
and 8oo B.C.

§ 6. Groups of Greek Alphabets

The Semitic alphabet was not adapted everywhere in the Greek world
in the same way. We can distinguish Jifferent groups of alphabets related
to each other; the added letters form the decisive factor. But whatever
classification one may give, it will always be approximative. Here follows
Sigalas' classification 1):
I.Cretan group: an alphabet of 21 letters 2), viz. a/Jyr5tfCh{}
i"). µ v on 9 ea i- v; used in Crete, ·Thera, Melos, the Cyclades in
the Northwest, Attica and Thasos. v is the only non-Phoenician sign.
2. Corinthian group: an alphabet of 25 letters, viz. a Py r5 e F Ch{} i"
x
). µ,..on� 9 ea -r v <p V'; used in Corinth (with its colonies, especially
Corcyra), Argos, Sicyon, Megara and Phlius.
3. C halcidian group: an alphabet of 24 letters, viz. a fJ yr5 e F Ch{}
x
i"). µ. v on 9 ea i- v � <p 3); used in Euhoea, the Peloponnesus as far as
it was not mentioned among the Corinthian group, Central and Northern
Greece (Boeotia, Phocis, Locris, Thessaly) and the numerous Euboean
colonies in the North and the West.
4. Milesian group. Originally it included the Western coast of Asia­
Minor with the islands belonging to it, but the great influence which the
Ionian land exercised by its trade and industry as well as by its prose and
poetry, and the excellent adaptation of this alphabet to the needs of the
Greek language has had the remarkable result that it became, in the end, the
common Greek alphabet: AthenY adopted it officially under the archonship
of Euclides (403-2); hut it was already used in daily life bcfore 4 ). So the
development of this type is more elaborately described here. We must not
forget, however, that 11 is for the greater part a theoretical reconstruction.
There are seven stages:

') Pp. 72 ff. and go ff. The classification proposed by A. Kirchhoff, Studien
z1ir Geschichte des Gr. Alph.', Gutersloh 1887, kept its ground for many decades.
He distinguished : I. The Cretan group (Crete, Melos, Thera); 2. the Eastern group,
Eastern h alf (Asia-Minor, Argos, Phlius, Megara, Sicyon, Corinth with their colon­
ies); 3. the Western half (Cyclades, Attica, Aegina); 4. the Western group (the
continent and Euboea with its colonies). The 2d group of Kirchhoff and the 2d
of Sigalas, as well as the 4th of K. and the 3d of S. are almost identical. But in the
other groups the case is different.
') I fake into consideration only the value and Ihe place in the series, not the
fornt of the characters. So some deviJfions from Sigalas arc unavoidable.
') Here � w.1s written with I he sign X and x with 'I'. Cf. the Litin x and its
pl.ice in the series.
') The measure had become inevitable, bec.111se the letters gave rise to confusion:
/\ w.1s i11 Attic y, in Ionic J..
 a) The oldest alphabet of 23 letters, still very similar to the Semitic
one, viz. a Py de (also for 1J and the pseudodiphthong tL) Fl; h If od. µ,.,
samekh o (also for w and the pseudodiphthong ov) n T (sharp a)
9 e Mn -iv.
b) The §in is kept for the single sibilant sigma and siimekh becomes ~.
whereas 'P x 11' are added.
c) The aspiration is lost and h becomes 11·
d) In consequence w is added.
e) Sharp consonants are doubled (xx next to ", nJt next to n, etc.).
The sharp sibilant being written aa, T becomes useless. In this stage the
well-known cipher-system came into existence 1), in which F counts for 6
(today printed !., the later form of the minuscule combination of a and
1:, see fig. 7, p. 44), 9 for go and T, now removed to the end, for goo 2 ).

/) The f disappears as the sound is not pronounced any more (this

must already have happened in the VIIIth cent.).
g) The 9 disappears because the " seemed to be sufficient for both
gutturals (this happened not later than in the VIth cent.). The final
result was the alphabet of 24 letters which is employed up to now, viz.
a Py (5 el; 1JIf i ,C .t /J, Y ~ 0 n {! a T: V 'PX 11'w.

') a'-c'
= 100--800;
1-5;
?\ =
"' =
900;
6; (;'-,(},'
,a = 1000;
= 7-9; "=
,, = 10.000;
10; ><'- n'
,e =
=
= 90; e'-o,'
E.g. 1955 =
20-80;
100.000, etc.
9'
,a?\" c', A different system was the acrophonic one used in Attica in which I was
1, II two etc., then fl ni..-.1:, f;. di><a, H hc><a~6.. , X x<A,o,; besides there
were combinations as (iii 50, etc.
•) Much later its form was changed into the so-called sampi (?\).
 CHAPTER III

WRITING-MATERIAL AND BOOK-FORM

§ 7. Papyrus, Parchment and Paper

A few literary texts have been preserved, it is true, on stone 1), on
wood 2), or on a potsherd (ostracon 3)), but in the main only three materials
must be considered here.
1. Papyrus (minve oc; or PvPAoc;, later pipAoc;) is a marsh-plant
( Cyperus papyrus L.), described a.o. by Theophrastus (hist. plant. IV,
8, 3): <pVE:T:aL oe d .nanveoc; OV)(. tv pMh, 'fOV iJOaToc; dAA' O<JOV tv ovo
:rl1JXE:<1LV, tvia x ov oe "al tv tAdnovt. IIa x oc; µev ovv -cijc; elt71c; 71A[)(.01'
)(.aenoc; XE:L(!O( dvoe oc; EV{! W<J'JOV, l'fi"oc; oi: vnt e 0£)(.a .n?JXELc; • cpvnai ot
{mee 1ijc; y ijc; av'fijc; nAaylm; eitac; de; T:OV n71Aov -xa{J.iE:i<Ja At:nTdc; )(.ai
nvxvcic;, li.vw ot 'fOVt; nanve ovc; �) )(.a.tovµevovc; 'f{!Ly wvovc;, µtyd}oc; we;
-rE:-rean1 x £ic;, "6 µ71 v £x ov-rac;, axeeiov, da{}£vij, "aenov oe oAwc; ovMva ·
'JOV-COV<; o' dvaolowa, )(.a'fa noHd µte11 · In Antiquity it was very com­
mon in Egypt and some parts of Syria; nowadays it is not found there
any more; it still grows in the Sudan and in a few marshes of Sicily.
To the old Egyptians it was very valuable: the bark with its rough
fibre was used for all sorts of cordage and wickerwork 5); the pith of the
stem was made into writing-material. Pliny the Elder gives an unclear
description of the manufacture (nat. hist. XIII, xr-2), which on the whole
took place as follows: the pith is cut into long, flat strips in which the
fibres run parallel with the length. These strips are pressed together in
two layers in such a manner that the fibres in the upper layer run horiz­
ontally, and those in the lower vertically. These separate sheets are
pasted together, as a rule twenty at a time; they are put on the market in
the shape of rolls, of which the inner side, the so-called recto with the
horizontal fibres, is destined to be written upon, the outer side or verso not.

1) See § 13.
•) E.g. the fragments of Callimachus' J-lecale (fr. 260 Pf., p. 156 Cahen), and the
fragments of Babrius on waxen tablets in Leyden, the ,,Tabulae Assendelftianae"
(cf. Hesseling, J. Hell. Stud. XIII 1893, p. 293).
3) E.g. the fragment of Sappho in Florence (Arch. f. Pap. Forsch. XIII, 1938, pp. go f.)
') Viz. the almost triangular plume of leaves.
') Cf. a.o. Herod. VII 34: one bridge over the Hellespont was built with this
material.
20
 The sizes and the quality may differ. Pliny writes 1.1.:magna in latitudine
ear um (scil. plagularum) difjerentia, tredecim digitorum optimis - - - ; em-
poriticae (scil. chartae, the smallest size and worst quality) sex digitos non
excedit. Praeterea spectantur in chartis tenuitas, densitas, color, levor.
This writing-material, called by the Greeks nanveoc;, pvpA.o<;,ptpA.or:;,
p,pA.to,, 1) or xae-i71<;(Latin charta), was used by the Egyptians from
times immemorial 1). It was doubtless the ordinary material on which
the Greek book was already written in pre -alexandrian times. The Ptol-
emies and later the Roman emperors propagated the use of it as much
a., possible, because they monopolized the manufacture of papyrus, which
yielded large profits.
2. Parchment, or vellum (d,q,IHe.a, dteµ.a; Latin membrana or charta
Pergamena) 3) is the prepared skin of an animal (sheep, goats, donkeys,
calves), smoothed on one side, later on both. As writing-material leather
is more natural than papyrus : the production does not depend on
local circumstances. It was in use everywhere: in Egypt perhaps already
during the Old Kingdom, in Persia for the royal annals 4 ), in Israel for
its Holy Scriptures. Greece was no exception: Herodotus tells us that
the word d,<pl>te.ahad once been used for papyrus, though the added
explanation is not exact (V 58: ,ml Ta( pvpA.ov<; d,q,IMear::,c:aUova, d.no
-ioii naA.a,oii ol •I rove<;,6n xo-re tv anav, PvPA.rov txetrov-r:o d,g,l>ee.ya,
alyeyal -r:exal 01.tya,). The manufacture of parchment was, on the whole,
not promoted by the Roman emperors on account of the reason mentioned
above. Sometimes the parchment was dyed, especially in purple.
A comparison of the two materials shows that papyrus is cheaper, but also
less durable. It was possible to wash away the writing on papyrus, but
on parchment the original text could be rubbed out and a new one put in
its place. Such a manuscript is called a palimpsest, e.g. LC. 79 (first text
written in 953, second text in the Xlllth cent.). Finally parchment can
be manufactured everywhere.
3. Paper. It is a well -known fact that long ago the Chin ese used
a kind of paper made out of herbacious fibres such as hamp, cotton and
flax. In the middle of the VIIIth cent. A.D. the Arabs learned the

') Bvpl.o!; is the oldest word ; the diminutive • pvpl.io.. has ch:inged in Attic
into p,/Jl.lo .. (influence of the second vowel on the first) ; as p,pl.lo .. was the most
common word, the vowel was changed in the related words too.
1 ) There are Egyptian papyri which go back to the Old Kingdom (sixth dynasty) ;
the roll of papyrus written as a determinative of abstract ideas is found in the fifth
dynasty, but it is undoubtedly older.
1 ) Probably so called because Pergamum has been famous for its vellum. Pliny
the Elder found in Varro an explanation which is wrong (nut. hist. XIII, 11 :
aemulatione circa bibliothecas regum Ptolcmaei (VI) et Eumenis (II), supprimente
chartas Ptolemaeo, idem Varro membrunas Pergami tradidit repertas).
•) Cf. Diod. Sic. II, 32, 4: £>< rwY paa,l.,xwY 11,tplJrew .., £Y al!; ol lll:eaa, ra!;
naAa,d!; nea{t:&!; xard Tl't'U 't'Of&O"£lxo .. Utl't'J:Hayµl:Ya!; .

21
 technique of paper-making at Samarkand 1), Rags were their usual
material. This Arabian rag-paper is distinguished from the European
paper by its stouter substance and a glossy surface. Moreover western
paper often shows rulings and watermarks 2). Paper is, though less strong
than vellum, much more supple and cheaper. The Greeks called it
Evloxaetlov or Evl(fr:ev,nov (Latin charta Damascena or bombycina, this
last word possibly derived from the Greek Baµ./Jvxl-.,11), The codex
Vaticanus Graecus 2200 (Th. 52, CL. 5), which is believed to have been
written in Damascus in or about 800, is already made of paper, but it is
not before the XIIth cent. that paper manuscripts become numerous.

§ 8. Pen and Jnk ; Ruling

First the Greeks wrote on parchment and papyrus with a reed cut
aslant, sometimes also with a tiny brush. But probably already in the
Hid cent. B.C. this reed (xalaµ.oi;, dovaE, Latin calamus, canna) was
cut like our pen so that it was easier to alternate the thickness of the
strokes. Pens of metal (bronze) also are still to be seen in our musea.
The quill is not named before the Vth cent. A.D. (nt-.,va, yeatpl<;, Latin
penna), but it must have been used earlier, especially on parchment, for
which it was very suitable.
The ink (µ.tla-.,, ly xava-r:o-.,, Latin atramentum) used on papyri is
generally beautifully black; it was made of soot. On Byzantine papyri
it is often of lesser quality and faded. On vellum manuscripts good
black ink is normal. For headings and initials red ink (,eo,e,ewo-., µ.tlav,
Latin m inium) is often used. Gold or silver ink was used in very precious
manuscripts, e.g. on p1,1rple vellum.
There is, as far as I know, no example of a ruled papyrus. On parch­
ment the scribes often marked the horizontal lines as well as the vertical
margins with a sharp stilus, since the Xlth cent. with a small leaden
wheel. It was sufficient to draw the lines on one side of the leaf, since
they were at the same time visible on the other side (see e.g. Pl. III, V,
Th. 63, CL. 24, V. 1-4). In the beginning the letters were written
on the lines, since the Xth cent. often under them (see Pl. VI, Th. 63,
CL. 18, 24 etc., V. 17-20).

§ 9. The Roll
The roll made of papyrus (,ev,t,,,deoc:, Latin volumen) was at once
suitable to be written upon. As mentioned before, it consists of sheets

1) The oldest d ated Arabian manuscript n paper (n ° 298 in the Legat um War­
o
nerianum at Leyden) is from 866.
) Knowledge of the watermarks is extremel y useful in codicology.
2
See the
extensi�e list in C. M. Briquet, Les Filigranes. Dictionnaire historique des marques
du pap1er des leur apparition vers 1z8z jusqu'en 1600. Paris 1907.
 pasted together ("oU,jµaTa 1), Latin plagulae). The text was written
from left to right along the fibres in broader or narrower columns
(<1ellde<::), which do not correspond with the 'XoU1µarn. Afterwards it
was rolled round a wooden stick (1<ov-i-oc;), the writing coming on the
inside 2) and the end of the text being rolled last. This roll, pvfJAor; or
fJlPAoc;, but as a rule called {Ji{J).lo,,, was kept in a parchment cover
(dupt'1te a) or in a box ('Xt{Jcm6r;). It was provided with a label of
parchment (alUv{Jo<:: 3) or alnv{Joc;, later also .nina'XtO'>', Latin index,
titulm) bearing the title and the author's name which appear also at
the end of the roll (N. 6, 14a = Th. 14). A roll 4) which forms part
of a larger text is called T6µor;; the whole of the text i-evxor; 5). To
roll the book is called el'lew (eLAei'>'), Hlaaew (Latin volvere, plicare);
to unroll it d,,e,..tei'>', dvdi<1aew (Latin evolvere, explicare).
The roll was manufactured of papyrus or parchment. It has remained,
till the IVth cent. A.D., the usual form of literary manuscripts. Then
it had to give way to the codex (§ 10), though liturgical manuscripts
continued for a long time to be written on rolls, but then parallel to the
height of the roll (CL. 35).
Length and height of the roll vary according to needs and taste. For
the height the extreme measures are about g cm. and 22 cm. (3! and 8!
inches). The papyrus-roll of Bacchylides in London was at least 5 m.
(17 feet) long, Aristoteles' Constitution of Athens more than 5 m. 6); the
non-literary papyrus with the Revenue Laws of Ptolemy Philadelphus
measures, however, 14 m. (47 feet). Zonaras, the historian from the
XIIth cent., mentions (XIV 2) a vellum roll of Homer measuring some
40 meters (130 feet), but this was just a curiosity.
This book-form had some inconveniencies: the writing only on one side
meant a considerable loss of material; in order to start the reading one
had to unroll the whole book; the consulting of a different passage was far
from easy; regular use caused considerable wear and tear, especially with
the brittle papyrus.

§ 10. The Codex

This Latin word (originally caudex) meant first the trunk of a tree,
then a special book-form. The Greeks called it again {Ji{JlloY, but also
awµanov. The codex is formed either of wooden boards bound

') The first sheet, on which in Byzantine times often an official mark was stamped,
was called newr6><o.Uo.. (hence our protocol), the last with title and author's name
11uraro1<6,U101'.
•) The verso with the vertical fibre is more elastic than the recto. Sometimes
both sides were written upon. Then it was an o:n:,uD6yeacpo...
) Cf. Cic. ad. Att. IV 4".
0
') And afterwards the codex also.
') Cf. d Jlcncinvxo,, the five books of Moses in the Old Testament.
0
) This was perhaps a niixo, of two r6µo,.

23
together and written upon on both sides, or of waxed tablets. But the
usual material was papyrus, vellum or paper. The codex in its sim-
plest form is made by folding one or more leaves put together and
by fastening them in the folding 1). As a rule, however, a certain number
of these gatherings are again bound together on the fold as is the case
with our modern book. Then the codex consists of a number of so-called
quires (n-ielicJe~ or -ine111h.11,Latin quaterniones; these names being
due to the fact that the normal number of sheets was four) 2). In each
quire the number of leaves (9111U11,z11ed11, Latin /olia) is twice the
number of folded sheets 3 ). Since the papyrus fibre runs differently
on recto and verso, and since the flesh-side in vellum is always a little softer
and whiter than the hair-side, it is habitual, in forming a codex, to arrange
the leaves in such a way that in opening the book the two pages are
similar. Finally a binding keeps the whole together.
On a page the text may be written in one or more columns. Pages with
one or two columns are normal. In two of the oldest codices of the Bible
we find three and even four columns on one page 4 ): the aspect of such a
codex, when opened, strongly resembles a partly unrolled volumen.
The advantages which the codex possesses in comparison with the roll
are obvious. Because the cover and the book form a whole, there is less
wear and tear; the material is constantly written upon on both sides; with
one movement of the hand the book is opened at any place which the
reader wants and comparison of passages is easy. The use of the codex
may have started in the schools (and perhaps on journeys).
As papyrus easily breaks on a fold, it was not very convenient for
the codex. Yet it has often been used, e.g. in the famous Chester -
Beatty papyri of the Old and New Testament, which date from the Illd
and IVth cent. 5 ). Parchment, on the contrary, formed excellent material
and it may be said that in the IVth cent. A.D. the vellum codex has
become the ordinary form of the literary book 8 ). This form was especially
favoured by the Christian church for religious texts. St. Jerome (Epist. 141)
writes about the library of Caesarea: quam ex parte corruptam ••• . in mem-
branis instaurare conati sunt, and St. Eusebius tells us (Vita Const. IV. 36-7)
that the emperor Constantine ordered fifty Bibles to be copied for fifty
new churches b, cJ,91IMe11~.
In tha;e days scholars also started to copy the literary texts which

1) Thi s happens still with our small pamphlets and brochures.

') The codex Vaticanus of the Bible has quires with 10 sheet s = 20 leaves.
•) For the numbering see § 24.
•) The Codex Vaticanu s ha s three column s (Th . 44, C.f.. 1), the Sinaiticus four
(Th. 45, W. 6), the Sarravianus (Pl. III) two. Th e Alexandrinu s has only one (Th. 46).
The Vaticanu s ~r. 1288 of C assius D10 (Vth cent.) has three columns (CL. 2, St. 5).
') Cf. H. I. Bell, Recent discoveries of biblical papyri. Oxford 1937.
•) See however next paragraph letter B.
 they valued, from the frail papyrus-roll on the stronger vellum-codex. This
explains at the same time that the archetypes of later manuscripts often
seem to belong to that period. This explains also, to a great extent,
why some authors have come down to us and others, e.g. the bulk of the
Hellenistic authors, not: the literary taste of those days determined what
was worth copying and what was not, and by doing so it decided at the same
time what would be spared for later centuries.

§ 11. Collections
A The most recent list of literary papyri is given by R. A Pack, The
Greek and Latin Literary Texts from Greco-Roman Egypt, Ann Arbor 1952.
A choice of the most important texts follows here 1):
1) England. London (British Museum):
Aristotle, Constitution of Athens (Ist cent. A.D.) (W. 5)
Bacchylides (1-lld cent. AD.) (N. rna, Th. 6)
,, Herondas (1st cent. AD.) (W. 1)
Hyperides (1-lld cent AD.) (Th. 4, 9; W. 2, 3)
Oxford (Egypt Exploration Society):
Alcaeus (I-IId cent. AD.)
, Daimachus (?)2), Hellenica (II-IIId cent. AD.)
Callimachus, Aetia and lambs (I-lld and IVth cent. AD.)
Euripides, Hypsipyle (II-llld cent AD.)
Ibycus (Ist cent. B.C.)
Pindar, Paeans (IId cent. A.D.)
Sappho (II-IIId cent. AD.)
Sophocles, lchneutae (lld cent. AD.)
2) Germany. Berlin (Museum) 3):
Aristotle, Constitution of Athens (IVth cent. AD.)
Callimachus, Aetia (Illd cent. AD.)
Corinna (lid cent. A.D.)
Timotheus, Persae ([Vth cent. B.C.) (Pl I)
Heidelberg (Library):
Phoenix, Choliambs (lid cent. B.C.)
3) France. Paris (Louvre):
Hyperides, Against Athenogenes (Ild cent. B.C.)
(National Library):
Aleman, Partheneum (lst cent. B.C.)

') Only new texts are mentioned. There are numerous fragments of extant clas­
sic.11 authors besides.
1) According lo F. Jacoby, Fr. Gr. Hist. no. 66. Others have thought :!bout
Theopompus or Cratippus.
1) At least before the war.
 4) Italy. Florence (Papyrological Institute):
Callimachus, Aetia, Hecale (1st, lid and llld cent. A.D.)
Erinna, Distaff (1st cent. B.C.)
Sappho (11-Illd cent. A.D.)
Naples (Museum):
Epicurean texts from Herculaneum, mainly by Philodemus and
Metrodorus (Ist cent. B.C. and A.D.) (N. 3b, 6, Th. 5)
Rome (Vatican):
Favorinus, de exilio (Illd cent. A.D.) (N. 13)
5) Switzerland. Geneva (University Library):
Menander, Agricola (Vth cent. A.D.)
6) Egypt. Cairo (Museum):
Menander, varia (Vlth cent. A.D.) (N. 16).
B. As regards literary vellum codices,a few fragments have been found
from time to time together with papyri, e.g. a fragment of Euripides' Cretans
(before 1945 in Berlin) and fragments of Menander's Disceptantes (Oxford)
and Circumtonsa (Heidelberg and Leipzig). But special attention sho11ldbe
given to the rich collections of manuscripts formed in the days of the Renais-
sance and afterwards 1). The principal ones follow here with the Latin
names of the codices preserved in them 2 ):
1. Italy. Rome (mainly Vatican): Vaticani, Barberiniani, Chisiani, Otto-
boniani, Palatini (see also Heidelberg), Reginenses, Romani, Urbinates.
Florence: Fiorentini, Laurentiani, Medicei.
Naples: Neapolitani, Borbonici, Farnesiani.
Milan: Mediolanenses, Ambrosiani.
Ravenna: Ravennates.
Venice: Veneti, Marciani.
Modena: Mutinenses.
Verona: Veronenses.
2. France. Paris: Parisini, Coisliniani, Colbertini, Sangermanenses (now
added to the Fonds grec), Sarraviani.
Montpellier: Montepessulani.
Strasbourg: Argentoratenses.
Tours: Turonenses.
3. Germany. Berlin: Berolinenses.
Heidelberg: Heidelbergenses, Palatini (see also Rome).

') See V. Gardthausen, Sammlungen und Kataloge griechischer llandschriften,

Leipzig 1903, and 0. Schissel, Kataloge griechischer Hczndschriften, Graz 1924.
•) A complete list of Latin names is given a.o. by F. W. Hall, A Companion to
Classical Texts, Oxford 1913, pp. 291-357.
 Leipzig: Lipsienses.
Munich: Monacenses, Augustani, Bavarici.
Wolffenbiittcl: Guelferbytani.
Konigsberg: Regimontani.
4. England. London: Londini enses, Britanni, Harleiani, Burneiani, Cotto -
111a111.
Cambridge: Cantabrigienses.
Oxford: Oxonienses, Bodleiani, Clarkiani, Dorvilliani.
5. Netherlands. Leyden: Leidenses, Lugdunenses, Vossiani, Scaligerani,
Perizoniani, Vulcaniani.
6. Belgium. Brussels: Bruxellenses.
7. Denmark. Copenhagen: Haunienses.
8. Spain. Madrid: Matritenses, Scorialenses (or Escorialenses).
9. Austria. Vienna: Vindobonenses.

27
 CHAPTER IV
THE WRITING
§ 12. Typology
It is necessary to distinguish different types of Greek writmg, bu t it
is as necessary to realize beforehand that the lines of demarcation are not
always clear. The types depend on the kind of material, the nature of
the text, the taste of the period and even of the individual writer. The
inscription chiseled in stone is naturally composed of loose letters 1); words
are seldom separated; changes of the letter-forms are only accidental (see
K. passim). As soon, however, as the text is no longer chiseled in the
material, but written on it, it is possible to simplify or round off the characters,
and to connect or to combine consecutive letters: especially in the latter cases
the forms are exposed to great modifications. The text which is destined
to come into the market and to be preserved as a book will be written with
gr eater care than a note for personal use or a document which may soon
lose its interest; the nature of the contents will therefore also influence the
letterform. Finally the popular or individual fancy will sometimes prefer
the simple, then again the flourished style; now it will like harmony, now
variety; it will stress now the horizontal, now the vertical or oblique line.
A t school one learned first of all the type in which all the letters are
written apart and keep their origin al form. For convenience's sake we
shall call it the majuscule or rather uncial type 2). One may disting uish
a severe, straight, bold form and a more fluent and loose style. Those
who had not gone far in the art of writing used this uncial as best as
they could. But on the other side it was used in a regular, elevated and
elegant style by the professional scribe to whom the copying of an
important text was entrusted.
For daily use this way of writing took too much time: almost spontan­
eously a cursive writing developed from the uncial and alongside of it; it
presented more convenient and often simplified forms and the characters
were repeatedly combined. We may assume that already in classical
Greece this simplified and accelerated writing was in constant use: the

') If the number of lellers is steady in each line, the inscription is wrillen
urn,x'1•'°" (K. 14a, h).
') We must not think here of the capitals which we put at the beginning of a
new sentence etc.
 size already of the work of a Plato or a Democritus proves it. The variety
of the cursive hands was infinite: the papyri testify to this in a most eloquent
way. Beside the slow and stiff cursive of the man of little practice and the
hasty note of one who has no time to lose, we find the fine calligraphy 1) of
the professional scribes. The cursive used at the offices of kings, emperors
and prefects shows a special character of elegant loftiness. It was not
allowed to use this so-called chancery-style for other purposes, but it often
exercised a great influence (an example in Sch. 35).
By minuscule is meant a special form of the cursive which was normalized
and conventionalized in a later period (§ 16) and which became for many
centuries the usual book-writing.
Finally stenography (shorthand) should be mentioned (see e.g. W. 17).
It is not of direct, but of indirect importance, because stenographical signs
hold their place as abbreviations both in cursive and minuscule hands (§ 21).

§ t J. Literary Inscriptions

A few remarks will suffice. Literary inscriptions may not be neglected

by the hellenist, but they show no special interest from a palaeographical
point of view: in contrast with the papyri, literature and documents are
written on stone with the same letters and in the same style 2). Moreover
the literature preserved on stone is, as a rule, not very important.
First of all, the countless epigrams should he mentioned, mainly written
in elegiac distichs 3). Next to them are religious texts 4).
We have already seen that the letters stand apart, but that separation
between words is rare. As regards the letter-forms, in brief outline the

') The word calligraphy should be avoided when speaking of book-writing in

contrast with the writing of documents etc. ; it gives an aesthetic appreciation but
is of no value for typology.
•) Metrical inscriptions of course being written in the convenient form.
3) The old collection of G. Kaibel, Epigrammata Graeca ex lapidibus collecta,
(Berlin 1878) is still useful as long as the new edition prepared by W. Peek is not yet
ready.
•) The most important (all to be found in I. U. Powell, Collectanae Alexandrina,
Oxford 1925, partly in E. Diehl, Antho/ngia Lyrica Graeca II, Leipzig 1925) are
in chronological sequence :
r. Anonymus, Paean in honour of Asclepius (about 36 0 B.C.), found at Erythrae,
Ptolemais, Athens and Dium. Cf. P. Biilow, Ein vielgesungener Ask/epiospaean, Xenia
Bonncnsia 1929, pp. 35-49.
2. Philodamus of Scarpheia, Paean in honour of Dionysus (about 335 B.C.) from
Delphi. Cf. W. Vollgraff, Bull. de corr. hell. XLVIII 1924-LI 1927.
3. Anonymus, Hymn to the Curetes (IVth cent. B.C. but the copy is from about
200 B.C.) from Palaecastro in Crete.
4. Anonymus, Hymn to the Idaean Dactyli (end of IVth cent. B.C.) from Eretria.
5. Isyllus of Epidaurus, Paean in honour of Asclepius (about 300 B.C.) from
Epidaurus. Cf. U. von Wilamowitz-Moellendorff, lsyllos von Epidauros, Berlin 1886.
6. Aristonous of Corinth, Paean to Apollo (222 B.C.) from Delphi.
7. The same, Paean to Hestia (222 B.C.) from Delphi.
8. Limenius of Athens, Paean and Prosodium to Apollo (138-129) from Delphi.
9. Anonymus, Hymn to Apollo (about 130 D.C.) from Delphi.

29
 development of the style has been as follows: the classical period strives
after simplicity and harmony; the inscriptions are samples of good taste
and artistic feeling; a tendency to archaism is often to be noticed (K. 18,
Athens, Vth cent. B.C.). Towards the end of the IIId century there
begins a dear preference for adorning the characters with small strokes,
curves or hooks (K. 35a, Tenos). In the period of Augustus the angular
forms of E, 2, later also of Mand SJ begin to give way to the round forms
(K. 41, Anisa). After a short strongly archaizing reaction under Hadrian
(K. 44b, Athens), the style of the inscriptions begins more and more to
show the characteristics of the usual uncial-writing (K. 48a, Athens IIId
cent. A.D.).

§ 14. Literary Papyri

Only in very exceptional cases it is possible to date literary papyri with
certainty on the ground of external data alone. Our knowledge of the
development of the book-hands used on papyrus, as far as it goes, is due
to the experience of a few papyrologists. The calligraphical element
which in these papyri so often comes to the fore, is an additional difficulty.
Before giving a short sketch of the development we must point out that
more than one style of book-hand may be in use at the same time. Different
Jines of development often run parallel to one another. Secondly it must be
understood that the periods may only be indicated approximately: the transi­
tion from one style to another always takes at least one generation.
The oldest literary papyrus now extant is the text of Timotheus' Persae,
it dates from the IVth century B.C. (Pl. I); each letter is separately written
exactly after the manner of an inscription; deviations are the result of the
difference in material. The forms of the letters are simple, angular, dear;
there is no trace of cursive influence. During the IIId century the style
remains practically the same, but the letters are often small and fine,
though of different size; the horizontal line predominates, and the letters
are often broad (Sch. 40: Euripides, Phaethon; Th. 2: Plato, Phaedo).
In the Ild century a rounder, more cursive form begins to appear, a pro­
cess which continues in the Ist; moreover we observe a tendency to equalize
the size of letters (Sch. 6o: Euripides, Hippolytus; Th. 4: Hyperides, against
Athenogenes; N. 6: PhiJodemus, on Death).
During the reign of Augustus a new type begins to develop; it strives
after elegance and adds small oblique Jines at the end of vertical and some
horizontal strokes (Sch. nb: Anonymus, Anapests; 19a: Hesiod, Catalogue;
19c: Iliad; Th. 8 and St. 3: Od ssey). Still at the end of the Ild cent.
y
A.D. this style appears in the Hawara-lliad ( Th. 18).
Next to it we find in the beginning of our era a simpler type of round,
easily written letters (Pl. II; Sch. 18; Anonymus, Romance of Nin us), which

30
gradually develops into the severe style characteristic of the IId century
A.D. Here the letters stand upright, with predominantly horizontal lines
and a great variation in thickness of the strokes: 2
1
now it is a lighter (Th. 6, N. 10a: Bacchylidcs, Odes)
then a heavier type (Th. 17, W. 4: the famous
Ilias Bankesiana). Towards the end of this cen -
tury under the influence of the cursive, a secun-
r
dary form comes into existence with sharp angles
A
and sloping lines (Th. 15: Iliad).
In the IIId cent. the influence of the cursive r
is still greater; different forms of the same letter r. z. z
occur next to each other on the same page. But H H 11 n
the reaction arises in the IV th century: the 0 e-c- s
influence of the cursive is intentionally avoided; f 7
form and ductus of the letters are regular, har - /< )< LL
monious and clear: the fundamental forms are A A"- A
the circle and the square. From this century,
M
however, date the oldest complete vellum-codices,
which offer a better foundation for study; see
next paragraph.
He who has carefully studied the examples 0 0
mentioned above, will have noticed that the dif- {"' n 7T n
ferences do not come out so much in the real
t' Pe
forms of the letters, which very often remain ( c.-
remarkably alike, but in the character of the hand-
writing as a whole. The table given in fig. 3 has
,. T y Y'
..., y y '( y
therefore not the intention to mark a historical
development of letter-types, but to illustrate th.: 'f 'f '? ,j
great uniformity of the essential forms, and to x .x )C
show, on the other hand, the influence of the r trf
cursive on the more rounded and simplified forms.
In addition, he will have observed the scriptio
continua which he already knew from the inscrip- Fig. 3.
tions: the letters are normally detached from each Letters found in literary
papyri; col. 1 the charac-
other, but the words are not separated; the last ters used in Timotheus'
letters of a line, however, are more than once Persians (Pl. I) ; col. 2
characters in other papyri.
written smaller in order to finish a word on the
same line; in poetry the verses stand apart, but prose, as a rule, goes on
without interruption; so-called capitals do not occur; prosodic signs
(see § 26) only incidentally. The columns are rather narrow (see e.g.
Sch. 18 cited above; Th. 9, W. 2: Hyperid es, for Euxenippus, Ist cent.
A.D.); they are seldom larger than a dactylic hexameter.

31
 § 1;. Uncial Writing
As we noticed in § ro, the parchment codex became, in about the IVth
cent. A.D., the usual book-form, though the roll of papyrus disappeared only
gradually. As examples to be copied on this expensive and durable material
the scribes of course took the best and most carefully written papyrus-texts
available. As a matter of fact, the letter-form and the style are essentially
the same, but on parchment, which is heavier but smoother, the strokes are
somewhat heavier and more regular also. Once again scriptio continua is
the rule.
The history of uncial writing may be divided into two periods, the
line of demarcation being the period in which the minuscule originated,
viz. the beginning of the IXth century. Before this date, the uncial is
the normal type used for a book, later it is exceptional.
In the first period we find a type of uncials which, on the whole, shows
very few divergences: the fundamental forms of most letters are the square
and the circle; they are upright and, as far as possible, of the same size
(fig. 4); the general aspect of the column or page is harmonious; prosodic
signs are rare; separation of chapters or
paragraphs is either not indicated at all,

NI On CT Yr 4' f
X
or very discretely. This style is mainly
UJ known from manuscripts of the Bible:
Fig. 4. the Vaticanus, the Sinaiticus, the Alexan­
Uncial characters in the Codex drinus; and the Sarravianus (or Colber­
Siniaticus of the Bible (Th. 45,
w. 6). tinus) of the Old Testament. The first
mentioned (CL. 1, Th. 44) is the oldest
and dates from the middle of the IVth cent.: it shows the enumerated charac­
teristics in full purety 1). The Sinaiticus, from the second half of the same
century, only differs in so far that a new pericope is indicated by a new line of
which the first letter projects to the left (Th. 45, W. 6). This is also
the case with the Alexandrinus (Th. 46) written in the beginning of the
Vth cent.: its style, however, is a trifle more elegant. As to the Sarra­
vianus, from the Vth cent., it shows again pure characteristics (PL Ill,
0.nd. 21 V. 1-4) 2).
Uncial codices of the same time, bearing other texts, are also handed
over, though in a fragmentary state. First of all we draw attention to the
fragments of Homer in the Ambrosian library at Milan (Ambrosianus
F 205 inf., new number: 1019) 3): they are 58 illustrations on 52 sheets cut
out from a magnificent manuscript, of which the verso preserves parts of the

1) In later times all the letters have been rewritten with darker ink and accent
s
etc. have been added.
•) The Vaticanus is in Rome ; the Sinaiticus and Alexandrinus are in the British
Museum; the Sarravianus partly in Leyden, partly in Paris and Leningrad.
3) Complete facsimile edition by A. Ceriani and A. Calderini, Milan 1953.
text. We are not quite sure about the date: scholars hesitate between the
Hid, the IVth and even the Vth or Vlth century. According to specialists
the style of the illustrations points to a date between 350 and 450. The style
of this uncial is perfectly pure; a few accents, though, have been written
(Th. 43). Next to it comes the mutilated Cassius Dio in the Vatican {CL. 2,
St. 5) written in the Vth cent. (Vat, gr. 1288): the only deviation from
strict regularity is to be found in the v and 'I' which now and then are rather
large. This series may well conclude with the famous Dioscurides in Vienna
(Vindob. med. gr. 1), with its many illustrations, a.o. a portrait ofthe proprie -
tress Juliana, the daughter of Flavius Anicius Olybrius, Roman emperor in
472; Juliana died in 527-8, which gives a clue to the date of the codex. The
writing differs in so far from the severe style that there is an obvious endeavour
after elegance: thick and thin lines alternate; there is a tendency towards the
lengthening of horizontal lines which, as a rule, end in a thick point or a dash.
For the rest, the characteristics are identical ( Th. 47, St. 6, V. 5-8).
Another style which we do not know from old examples, shows the
influence not of the severe book -writing, but of the chancery -cursive:
the writer is not intent on evenness of form, but on contrast: some letters
show the fundamental forms of square and circle, others of the upright
rectangle and oval. An example is the Marchalianus of the Prophets
in the Vatican library (CL. 4) from the VIIlth cent.: note the initial of a new
section projecting into the left margin as well as the consistent addition of
accents and punctuation (cf. also N. 19).
All these codices owe their preservation, in the main, to their costliness
and there are only few of them left. Therefore they may not be taken as an
absolute standard. We may assume the existence of another style of writing
which was influenced by the cursive, especially in cheaper manuscripts,
with sloping letters, abbreviations and supple forms. See e.g. Th. 48, a
treatise of mathematics from the VIIth cent., the fragment of Euripides•
Cretans mentioned above (p. 26) and a leaf from the Gospel after St. Mark
in the Freer -Collection at Washington (the so -called Washington -gospel) from
the IVth or Vth cent, They all show these special characteristics 1).
In the second period of the uncial which begins at about 8oo, we always
find an element of mannerism, because henceforth the normal type for
books will be the minuscule. We refer to CL. 8 where on one page an up-
right and a sloping variant arc to be seen of a style 2 ) which is akin to the
writing in the Marchalianus (CL. 4), and to CL. 13 ( St. g); mannerism
and unnaturalness have led to a very unsatisfactory result. But in both
cases it is a Christian text. Texts which belong to Greek literature are, as

') A distorted type is shown in CL. 6, a text from St. Gregory the Great from
8oo A.D., where for the first time we meet with an adorned initial.
1 ) This is sometimes called the "slavonic" style, because the "Cyrillian" alphabet
(see § 22) falls back upon it.

3 33
 1 1 a rule, written in minuscule. Parisinus gr. 2179
of Dioscurides, and Parisinus gr. 2389 of Ptol­
a, a4\�ap. emy's Almagest, both date from the IXth cent.
u !Bl.JC{!> and hardly form an exception to this rule. The

y iv rr
different styles of the uncial, however, are still
used there, either for initials, titles and subscrip­
� �AA
Cr c- E € c 1v� tions or for scholia, which, in this way, are easily
distinguished from the text itself 1).
zt
(,
.3 } Z, § 16. Minuscule Writing
k h.u.L.nH
-e- e9 As we have seen(§ 12), the cursive existed already
I l i for many centuries next to the uncial. In the eighth
ll k I< KI(_ x century this cursive writing assumed definite
forms 2) which have been set upright, normalized
X 7' �
)- and equalized 3) so as to form a new type: since
P' }),- J.)..
H-V v
the ninth century it came into general use for the
p N books. The oldest dated minuscule codex in Greek
J
0
t�s is the so-called Tetraevangelium Uspenskij written
in Jerusalem in 835, now at Leningrad (L.C. 4).
z;:r lT � Some scholars have been inclined to trace the origin

r
'{J]'

ee of the minuscule to Syria or Constantinople');

for the time being, it seems safer to attribute it to
CT c ( r; a general tendency and modification of the taste.
T T'" 't There may be a connection with the great cultural
v v-uv revival after the extinction of the Iconoclasts dynasty.

l cp The usual forms of the minuscule are given in fig. 5,

x x+-
4> column 1. It resembles the uncial writing in so

t r t/-
far that consecutive words are not separated, but
the great difference is that the letters are often
a> QC) LAI W combined according to certain rules 0). Most letters
Fig. 5. may be connected on both sides; only with the
Characters in minuscule preceding letter, however, the t ,,11, t
o, e,
'P
codices. and c»; only with the following thee, '1/, Hand e1 8).
Col. I gives the pure
forms, used in the vetus- These differences only refer to the pure forms and
tissimi; in so far as there are no chance connections; the
col. 2 altered minuscules,
cursive and uncial forms. scribe of Plato's Vaticanus (CL. 9), for instance,
1) CL. 10 shows a title, a subscription and notes. •) See CL. 5.
•) In papyri a similar chan�e is apparent. One may compare the change of sloping
writing into round writing which took place in the last two generations.
•) According to T. W. Allen, The origin of the Greek minuscule hand, Journal of
Hell. Stud. XL 1920, pp. 1 ff., the famous monastery of Stoudion had a decisive
influence. •) See § 21 and fig. 7.
6) This last letter has a second form only used in combinations, which is con­
nected with the preceding letter. See fig. 7. If a preceding letter ends with a
horizontal stroke, this is normally drawn till the left part of r, '1 and "·
 connects " with a following t. It goes without saying that a letter which
may be connected of its own, is never linked to a character which admits of
no connecting. That is the reason why the v stands loose in the word
ov,c, though by nature it may be linked on both sides. The spaces in this
writing are the result of the nature of the letters, not a means to separate words
in order to make reading easier. Combinations of certain letters are a normal
feature of the minuscule (see § 21). Accentuation already occurs in the
vetustissimi; in later manuscripts it is usual (§ 26). The accents are small.
The history of the minuscule is determined by the growing influence
which the uncial and, on the other hand, the not normalized cursive exercised.
But a regular process is here out of the question. Severe minuscules are
contemporary with a style strongly affected by other types. Much depends
here on the intention of the author and the destination of the codex. The
dating of undated manuscripts is always a hazardous undertaking. Fortun­
ately we dispose of a certain number of dated codices, which give a firm
basis to our study: from the IXth century only there are almost thirty
dated manuscripts.
For practical reasons we distinguish four periods:
1. From the origin of the minuscule till the middle of the Xth cent.:
codices vetustissimi.
2. From the middle of the Xth cent. till the middle of the XIIIth cent.:
codices vetusti.
3. From the middle of the XIIIth cent. till the middle of the XVth cent.
when printing begins: codices recentiores.
4. From the spreading of the art of printing onward: codices novelli.

§ 17- Codices Vetustissimi

Well-nigh all the manuscripts of this group excel in the extraordinary

regularity and care with which they have been written. This regularity
appears not only in the shape of the characters, which always display in
perfect harmony the same fonn, size and position, but also in the whole
aspect of the pages. The characters stand upright, sometimes slanting
a little to the left, and are practically of the same height. Uncials, the
use of which is a sign of weariness or carelessness, are rare.
The oldest dated specimen of a classical text, written in 888 for Arethas,
1
later bishop of Caesarea in Cappadocia, is the Oxoniensis D'Orville X 1 [A] ),
a vellum manuscript of Euclides. It gives a splendid example of a pure,
elegant minuscule written with perfect care and sloping a little to the left
(Th. 53, LC. 6).
Other vetustissimi important to the student of classics are the followin g:

') We put between square brackets the letter by which the manuscript is designated
in the critical editions of the author.

35
 Clarkianus (Bodleianus) 39, a codex containing Plato's first six tetralogies
with scholia [BJ, also originating from Arethas' library and written in
895. The writing is a little rounder and closer than the Euclides codex
(Plate IV, LC. 9 with the scribe's subscription, Th. 54, V. 9 - 12). It is
written in one column; a new section is marked by a letter projecting into
the left margin, a new dialogue by titles and initial. Impure forms are rare
(see e.g. Pl. II 1.4 e, I. 6 .t; in the margin two abbreviations of <1f'/µ,da>aa,).
Parisinus gr(aecus) 18o7 containing the eighth and ninth tetralogies of
Plato [A], probably still from the ninth century. The pages are written
in two rather narrow columns (O.nd. 28).
Parisinus gr. 2951 (917) together with Laurentianus LX 3 {already go6-7 ?)
and again part of Arethas' marvellous library, form the oldest manuscript of
Aelius Aristides (LC . 15).
Vaticanus gr. go of Lucian [I'] from the IXth or Xth century; with
scholia. The minuscule is a trifle less carefully shaped and contains some
impure forms (CL. IO: e.g. an uncial I' in 1. 10, I in I. 27, Kin 1. 36; a
combination of uncials, viz. -OZ in 1. 22 etc.; a cursive e in I. 33; in I. 35
two different ways of writing ~ei>, the second being a case of superposition,
see § 21).
Urbinas gr. III of Isocrates [I'] written in the IXth or Xth century;
the minuscule is of a large size; initials and subscriptions are in uncials.
Urbinas gr. 6I with Theophrastus' botanical works [U]; same time.
The minuscule is less elegant than the preceding ones and distinctly slopes
to the left. It contains a good deal of uncial forms. At the beginning
of a new book there is an initial. The space between two successive lines
is small (CL. 51).
Vaticanus gr. 1 of Plato is of little use for the constitution of the text,
but interesting from a palaeographical point of view (CL. 9: from time
to time, e.g. in 11.5 and 8, the " is linked with the following letter); it is
written by the same hand as the
Harleianus 5694 of Lucian [E] (Th. 56: same phenomenon in I. 12;
other irregularities in 11.4, 6, IO, 14, LC. 17). Both codices date from the
beginning of the Xth cent. (the Harl. before 913) and have been written
for Arethas.
Parisinus gr. 2934 of Demosthenes [S or I], from the beginning of the
Xth cent., written in two columns with brown ink, in a delicately formed
minuscule; the letters are distinctly quadrangular; an enlarged letter in
the margin denotes the beginning of a paragraph; at the beginning of a
new speech there is an ornam ented initial (O.nd . 29).
Parisinus gr. 1853 contains a.o. the Physics and the Metaphysics of
Aristotle [EJ with scholia (O.nd . 30).
A comparison of the writing of all these codices shows that it differs
only in details; the general character and the letters themselves remain
 practically the same. Exceptions confirm the rule; see e.g. Paris. gr. 14 0
7
with Lives of Saints etc, written in 8go, in which the writing is distinctly
narrow, somewhat akin to S of Demosthenes.
To the same group we may add seven manuscripts written in the same
style, but which are not decidedly from the first half of the Xth century:
°
Ambrosianus L 93 (parte) sup(eriore) (n 4go i� the new catalogue),
containing a.o. the Analytica and Categoriae of Aristotle with scholia of
the same time and notes from the XVIth cent. (W. 14).
Laurentianus (pluteus) LXX (codex) 3, oldest manuscript of Herodotus
[A], written in two hands, the second being of a more cursive type
(W. 18-19).
Laurentianus LXIX 2, oldest manuscript of Thucydides [C) ( Th. 57:
in I. 10 a new section is marked by the projecting into the left margin of
the first word which happens to begin a line in this section).
Laurentianus conv(enti) soppr(essi) 206: contains fourteen Biographies of
Plutarch [LJ (Th. 58: same phenomenon in 1. 7; rather numerous uncials;
abbreviations in II. 3 and 5).
Parisinus suppl(ement) gr(ec) 388, the oldest and most important codex
of Theognis [A]; it gives the Greek text in one column and an old Latin
version between the lines (O.nd. 38c). This is also a codex of Colluthus'
de Raptu Helenae [M].
Venetus 454, the famous codex A of the Iliad, published in 1788 by
D'Ansse de Villoison, "best example of a learned copy of an author that
has come down to us" (T. W. Allen). It gives the text in one column
with the critical signs of Aristarchus in the margin 1), further copious
scholia on all sides and shorter ones between the lines; the lemmata are
written in uncials. The minuscule slightly slopes to the right and has
rather numerous uncial forms. Initials at the beginning of a book are
painted in colours and in gold (Plate V, V. 13-16, W. 21).
Venetus (Marcianus) 447 of Athenaeus [A], incomplete archetype of all
the extant codices of this author; written in two narrow columns.

§ 18. Codices Vetusti

In this group we find two different types. The first is in close relation
to the vetustissimi and differs only by its lesser harmony and eurhythmy:
not all the characters are so regular in size, nor are they so upright;
the letters are more spaced and the whole displays some r estlessness;
uncials appear in larger numbers, first the ", A., µ., v, n, e, 1/, later also the
p, �. v, rp and ru. When the writer does his utmost, the result is, never-

') The 6pdot; (-) denotes a spurious verse ; the 8m:J.ij ()) draws t he attention
to something; the t5udij n1:emn,yµi"'1 Of) marks errors made by Zenodorus; the
�o
0.(HF.(!l<T><oc; ( is written next to passages which reappear elsewhere. There are
still other signs.

37
 theless, not harmonious, but rather stiff and awkward. This style is found
especially in liturgical texts and church-books, which often indulge in a
conservative style. Good examples are the Vaticanus gr. 2155 from 981
and the Urbinas gr. 20 from 992, both of St. John Chrysostome (CL. 15
and 18, LC. 54 and 65). Closely related is also the Urbinas gr. 102 with
important extracts from Polybius' Bk. I-XVIII [F], written in the Xth or
Xlth cent.: it is a fairly regular, upright minuscule, though the characters
are not always written with great care ( CL. 55: the accentuation is far
from perfect).
Beside it, however, we find a second type which is of greater importance
to the hellenist, because Greek literature is, as a rule, written in that
way. It is nearly always a sloping minuscule jotted down in a flowing,
loose fashion; the lines are almost everywhere of the same thickn ess,
because there is no endeavour after calligraphy. These codices are
articles for every day use. The influence of the cursive in the shape of
the letters and in the abbreviations is clear: characters of larger size mar
the harmony; the circumflex and the abbreviations also become larger;
combinations are bolder. Column 2 of fig. 5 shows cursive forms which
occur between the severe minuscules . The two types are used alongside
each other, though the cursive type remains still comparatively pure, in
CL. 33 (Vaticanus gr. 1431, a Christian anthology from the XIlth cent.)
and in Th. 51 or LC. 47 (Bodleianus Laud. greek 75, a codex of St. John
Chrysostome written in 976). We shall now successively mention some
reproductions showing the cursive type. They are not arranged chronologi -
cally, but so as to show progressive deviation from the pure style.
Vaticanus gr. 2369 of H erodotus [DJ probably from the Xlth cent. (CL.
52): we observe a certain carelessness in the drawing of the characters
and the whole lines. Apparently the scribe wrote rather hastily; there
is a slight inclination to the right. But the general form of the characters
is still the square or the circle and similar to the genuine minuscule;
the accents are still small.
Laurentianus LIX 9 of Demosth enes [P], from the Xth or the beginning
of the Xlth cent . (Th. 64), written in the same hand as the famous
Ravennas 137, 4a, the best manuscript of Aristophanes [R]. It is the
hand of a trained scribe, a little sloping and with large spaces between
the letters; impure forms are not numerous. The scholia are in uncials
(Plate VI, V. 17-20).
Laurentianus (Mediceus) XXXII 9 of Sophocles (usually called Lau -
rentianus [L]) , Aeschylu s and Apolloniu s Rhodius (here commonly styled
Mediceus [M]). It was form ed at a later period by the joining of L
and M. The writing of L is probably the sam e as that of th e Ravennas
of Aristophanes. M, which is of the same age, is written in a finer and
weaker minuscule, less sloping, but less pure also (W. 26).
 Burneianus 86, important codex of the Iliad, commonly called Town-
leianus [T], written in 1059; it gives the text in one column with scholia
in the margins as well as between the lines. The writing (Th. 66) is
rather regular and strikes by its roundness; abbreviations occur in the
text itself; the letters have varying forms (in 11. 2 and 3 four different
forms of £).
Vaticanus gr. 130 from the Xlth cent. contains book I-IV of Diodorus
Siculus [C]. The writing forms by its pointedness an interesting contrast
with the Townleianus. It shows that these less conventional minuscules
often bear a personal note. The deviations from the pure forms are more
numerous (CL. 27).
Vaticanus gr. 1335 of Xenophon [B], which is perhaps a few years older,
shows a more cursive writing in which especially the little height of the
characters is remarkable (CL. 54).
Urbinas gr. 105 of Dionysius Halicarnassensis [B], from the Xth or
Xlth cent., shows all the characteristics of this group to a still higher
degree. Here also the height of the letters generally is small, but the
number of abnormally sized ones is large (CL. 53: in I. 8 a combination
of uncials; numerous abbreviations).
Vaticanus gr. 126 of Thucydides [B] is a clear specimen of a private
copy. We draw attention to the supple, agile, sloping writing which, by
its numberless abbreviations and simplified letter-forms sacrifices much
of its clearness. The size of the letters which, it is true, still belong in
the main to the minuscule type, varies very much; some abbreviation-
marks are abnormally high; there is no real difference between the writing
of the text and of the scholia; a certain reserve, however, is still evident
in the absence of every embellishment: from time to time a line projects
into the left margin in order to mark a new section in the text; the
breathings are written in three different forms; the circumflex is sometimes
directly connected with an abbreviation (CL. 26).
Vossianus gr. F(olio) 77 III from the beginning of the XIIld cent. By
far the best manuscript of Julian's Speechesand Letters [VJ(Pl. VII), written
in a broad, very characteristic minuscule, the jauntiness of which strikes
the eye; abbreviations are numerous; the size of the letters is so different
that the whole makes an unquiet and unharmonious impression.
Barberinianus gr. 70 from the Xlth cent. contains the Etymologicum
Gudianum [d]. Text and scholia are full of abbreviations, hardly one
word has been completely written; the letters stand upright. The whole
is distinctly a copy made by a scholar for his personal use (CL. 56).
To this group belong a great number of manuscripts of Greek authors:
a choice of the most important follows here:
Alciphron: Vindobonensis phi!. gr. 342 [B], XII-XIIIth cent. Neatly
written in 12°. Best manuscript of this author.
39
 Anonyrnus (Pseudo-Longinus) de sub/imitate: Parisinus gr. 2036 [P], end
of the Xth cent. Archetype of all the manuscripts (0.nd. 31a).
Anthologia Palatina: Palatinu s 23 and Parisinus suppl. gr. 384 [P] (two
volumes, the former is in Heidelberg and contains books I-XIII, the
latter is in Paris and contains books XIV-XVI), finished about 98o
(O.nd. 33, W. 22).
Appian: Vatic;anusgr. 141 [VJ, two parts, the former written in the Xllth, the
latter in the Xlth cent. Contains only portions of the author's work.
Aratus: Marcianus 476 [M], Xlth cent.
Aristotle , Organon:Marcianus 201 [BJ, 954 (LC. 34).
Poetics: Parisinus gr. 1741 [Ac], Xlth cent. No other manucript
gives the complete text. It contains also the Rhetorics. (O.nd. 37a).
Arrian: Vindobonensis hist(oricorum) gr(aecorum) 4 [A], XII -Xlllth cent.
Archetype of all extant codices.
Epictetus, Diatribes: Bodleianus (or Saibantinus) 251 [SJ end of Xlth or
beginning of Xllth cent. Archetype of all the codices.
Euripides: Hierosolymitanus 36 [HJ, late Xth cent. A palimpsest of great
value.
Marcianus 471 [M], Xllth cent. Apparently the best manuscript.
Parisinus gr. 2712 [A], Xlllth cent.
Parisinus gr. 2713 [B], XII-XIIIth cent. With valuable scholia.
Hesiod, Theogony and Shield: Parisinus suppl. gr. 633 [C], XIIth cent.
Works and Days; Parisinus gr. 2771 [C], Xlth cent. Numerous
scholia (0.nd. 31b).
Homer, Iliad: Venetus 453 [B], Xlth cent. With scholia.
Odyssey: Florentinus conv. soppr. 52 [B], end of Xth cent.
Laurentianus XXXII 24 [G], Xth cent. These are the two best
codices of the Odyssey, but neither is in any sense comparable to
the Venetus A or even B of the Iliad.
Lycophron: see Aratus [here called A].
Parisinus gr. 2723 [C], from 1282. Very beautiful manuscript
(0.nd. 63).
Lysias: Palatinus 88 (in Heidelberg) [X], XIIth cent. Archetype of all
extant manuscripts.
Maximus of Tyre: Parisinus gr. 1962 [R], XIth cent. Also an archetype.
Nicander: Parisinus suppl. gr. 247 [II], Xlth cent. Interesting on account
of its illustrations (O.nd. 38d).
Orators (viz. Andocides, Isaeus, Dinarchus, Antiphon, Lycurgus, Gorgias,
Alcidamas, Lesbonax, Herodes): Burncianus 95, generally called
Crippsianus [A], first half of Xlllth cent. For some of these authors
no other manuscript is available.
Pindc:::ar: Parisinus gr. 2774 [CJ, XIIth cent.
Vaticanus gr. 1312 (BJ, end of same cent.
Plut.:L.rch, Moralia: Laurentianus LXIX 13 [L], about 950. Palimpsest
rewritten with Diogenes Laertius etc.
Parisinus gr. 1957 [F], end of Xlth cent. (0.nd. 43) 1).
Vitae: Laurentianus LXIX 6 [L], 997 (LC. 71).
Strabo: Parisinus gr. 1397 [A], Xlth cent. (O.nd. 40).
Theophrastus, Characters 1-15: two Parisini gr., 2977 [A] from X-Xlth
cent. and 1983 [B] from Xth cent.
Thucydides: Parisinus suppl. gr. 255 (also called ltalus or Cisalpinus) [A],
XI-Xllth cent. With scholia (O.nd. 41).
Britannus 11.727 [M], Xlth cent.
Xenophon, Cyropaedia: Parisinus gr. 1635 [A], Xllth cent.

§ 19. Codices Recentiores

This group contains the manuscripts dating from the two centuries
between the middle of the XIIIth cent. and the moment when prin­
ting comes into being. This is generally fixed at about 1450, though the
first book completely printed in Greek, Constantinus Lascaris' Grammar,
was not published before 1476 (in Milan by Dionysius Paravisinus). The
wide interest shown in ancient authors, and next to it the desire to form
large libraries, caused the number of manuscripts to increase considerably.
The scholars, as a rule, had a private collection, partly written in their
own handwriting. Again with the exception of biblical and liturgical
texts, the minuscule is mostly written in a very cursive form, the
letters, nevertheless, standing upright. The abbreviations are numerous;
the shape of the letters varies considerably; the handwriting has often a
strong personal character. That which was an exception formerly, now
becomes normal: accents and abbreviations are immediately linked with
the letters themselves: the acute accent is merely a stroke starting from
the vowel and pointed upward; breathing and accent form a whole. In
short, one writes with a purely practical view, without minding aesthetics.
The book is an object destined for daily use and the writing shows the
obvious signs of it; calligraphy is rare.
The number of Greek literary codices from this period is so large that it
is impossible to give even a selection: this group forms by far the greatest
number of the codices on which our knowledge of antique Greek literature
is founded.
Nevertheless the peculiarities of this group ma y be shown by some

1) There are many other vetusti with the text of the Moralia.

41
 examples. We can distinguish two types, the calligraphic type and the
ordinary one.
The calligraphic type, more frequently found at the end of the
period, strikes by its regularity, also by the omission of abbreviations;
but uncial and cursive forms, combinations which formerly were unusual
in calligraphy, and the connected accents and breathings are here quite
normal. See e.g. the Londiniensis 11728 of Polybius [B] from 1417
(Th. 78): the letters borrowed from the uncial writing (so ", o, e) have
been completely assimilated by the general trend of the script, which is
supple and regular; the most frequent form of the e is the one adopted
from the cursive. Libanius' Vaticanus gr. 84 from 1425 (CL. 45) shows
similar characteristics, but the writing is a little stiffer; moreover the
letters are formed more after the example of the conventional minuscule.
Parisinus gr. 1671 with Biographies and Moralia of Plutarch [A] written
in 1296 (O.d. 67-8), Vaticanus gr. 1334 with Opera minora of Xenophon,
Aristotle and others, written in the first half of the XVth cent. (CL. 6o)
and Urbinas gr. roo with Moralia of Plutarch, written in 1402 (St. 21)
are also models of calligraphy. These handwritings are especia11y im-
portant because the first typographical alphabets of Greek have been
cut on their pattern. He who compares a Greek incunabulum or even
a post-incunabulum with e.g. Libanius' above-mentioned Vaticanus, will
find more points of resemblance than of difference. The characters which
were used for Lascaris' Grammar were probably drawn by the Cretan
ca11igrapher Demetrius Damilas. Up to the XVIIIth century Greek is
usua11y printed with a certain number of ligatures and abbreviations; the
combination of a and -i- (the so-ca11ed stigma, cf. p. 44) did not disappear
before the end of the XIXth century.
We can illustrate the second and most important group by Plates VIII-XI.
In the Vossianus of Aeschylus and Pindar (Pl. VIII, second half of the
Xlllth cent.) the characteristics of a private copy are fu11y evident and the
result is not very satisfactory; the clumsy efforts to ornamentation are
piteous; at a glance one observes a striking variety in size and shape
of the letters and signs. Notice that the two columns must not be read
one after the other but next to each other (cf. Laurentianus XXXII r6,
a manuscript of Hesiod [D] written on paper in 1280 = Th. 72). Next
comes the Vossianus of Aristophanes (Pl. IX, second half of XIVth cent.),
which doubtlessly belonged to a scholar thoroughly interested in his
author; but again it may easily be seen that it belongs to the same
group: it is the writing of a different scribe with all its characteristics,
but the general aspect is nevertheless identical. Plate X reproduces a
page from the famous Mosquensis of th e Homeri c Hymns [M] written in
the same period but different in type; there is no pretence to ca11igraphy,
as is clear from the numerous abbreviations especially at the end of verses,

42
and from the great economy with which the paper is used; but the writing
is so regular and the abbreviations as well as the impure forms are so well
adapted to the rest of the minuscule that the aspect of the page as a
whole is very satisfactory (cf. e.g. Herodotus' Laurentianus LXX 6 [d]
from 1318 =-= Th. 75). The Vossianus of the Iliad (Pl. XD written in
the XVth century is again a copy meant for the scholar: a certain care
is unmistakable, red ink being used for titles and initials, and the whole
aspect of the page is not bad. But the handwriting itself is very cursive;
accents and breathings are constantly combined. In the scholia abbreviations
are countless. Finally we mention as a manuscript very characteristic for
this group the Barberinianus gr. 164 from 1294, a codex of Aristotle with
commentary (CL. 41) 1).

§ 20. Codices Novelli

The last group, dating from about 1450 onward, draws its importance
mainly from two facts. In the first place, even after the birth of the art of
printing, many a fine copy has been written by expert scribes for rich
Maecenates, and private copies for their own commodity. The style
of these two types hardly deviates from that of the former group: see e.g.
Leidensis of Pausanias (Pl. XII, about 1470), a very beautiful manuscript
on smooth white vellum, written in a neat and clear upright handwriting
which in faultless regularity forms page after page, and Harleianus 5658
of the Odyssey ( Th. 81), written in 1479, which both represent the calli­
graphical style. Parisinus gr. 1816 of Plotinus [F], 146o (LC. 99), written
in a loose, upright hand, and Palatinus gr. 223 from 1495 which gives a
commented text of Euripides ( CL. 47) are examples of copies written for
private use.
Secondly a manuscript of this class may still be ·.,aluable for the constitution
of the text, because it represents either the best tradition or a tradition which
may not be overlooked.

§ 21. Abbreviations and Symbols

The scribes have invented different methods to save space and time
when writing. It is quite natural that when a text is written in calli­
graphy, the number of abbreviations will be less. There are six different
ways of abbreviation: superposition, combination, suspension, contraction,
symbols, various signs.
1. Superposition,, as the word itself indicates, means the placing of
letters above each other instead of next to each other. This happens
by preference at the end of a word and at the end of a line. The last

1) Similar, though differing from each other, are also St. 18-:;?3, in the main Ambro­
siani, and W. 30-34, two Ambrosiani and three Laurentiani.

43
 @ e- <S>A> ly' ~ tN

M-J N-i Nl n-t fP Tu M-i yy

Fig. 6.
Combinations of uncial letters
Y"'
viz. o,, o,, o,, ae, av, ,,,,., ,,.., yo
f''I', "'I, '1'11' 1 11''/, :ne, T'/'I', f''/• c 't
Cf. CL. IO, margin 10, 21; 11,
10 ; 13 ii 13; 26, 30 j 52, 4 j 30( 1}$l-
Th. 51, 3; 64, IO; 81, I j
v. 6, 36. ov ~
Ot; ~
ov(~) 0 2"Vv ~ ~1"
1
ay rrD' 1>,IA-

yo tr l rP"-n}
lL rro
HT e""
tl1'
i 't rrw a.
tu ro< Tt:J. ·rJ
ro ol'T°b
TOU i'
re }l
Kt rw a; J:
rrir wv ~

Fig. 9.
Combinations of letters by
superposition.
<J't

UV

Fig. 7.
Fig. 8. ~ = { r~s 0

Combinations of let-
Usual combinations of ters used in later Fig. IO.
minuscule letters. minuscule. Two monograms.

letter is then generally written at a higher level than the penultimate one,
but the sequence is now and again also the reverse. Already in the papyri
and in the uncial codices we repeatedly see the N, as a rul e deformed into
a mere horizontal stroke, at the end of the line and above the preceding
letter. See a.o. Th. 14 (papyrus of Julius Africanus, middle of IIId cent.
44
A.D .) line 6 and 10, CL. 1 (Vaticanus of the ' rf
Yf"''(( t ...
Bib l e) i 30 etc . (cf. Th. 44 i 1), Th. 45 (Sinai - "tT
ticu. s of the Bible) i 6 etc., Th.46 (Alexandrinus ,n-i E
of th e Bible) 14 etc., Pl. III (Sarravianus of I
ltf.YT(_Ov' ~
the Old Testament) i 9 etc., CL . 2 (Cassius
Dio from the Vth cent .) ii 23 etc. Most com- >.~ros ->:
0
mon is the superposition of OY, see o.a. CL. 13 - - o~
ii 1. Later on, all the characters may be written .. E-
11'"
in this way . See e.g. many cases in Pl. VII - '"f'
XI,. then in Th. 81 (Odyssey from 1479), CL. 36 '
fl"'1J"t1W'11H 0
~
(Menaeum from 126o), with a repeated super- .-x_o~IOV II;\
pos ition in 1. 1, CL. 37 (Hymn-book from 1282),
f
CL. 43 (Aristotle from 1294), CL. 56 (Etymo-
logicum from the Xlth cent.), W. 31 (Oppian
from 1287).
X(,.-.-o')
we"' *f
Fig. II,
2. Combination of letters, or ligature, occurs Abbreviations through sus -
in two different forms: a. the letters are written pension together wit~ super -
position or combination. Cf .
in or across each other; b. they have one or Pl. II. title, schol . 2, I. ~ ;
more strokes in common. The first type is CL. 9, margin 16 ; 10, margm
15; II, 1, :! ; 52, 4; T~. -48,
spe cially found with the uncial. The old Greek 3 ; V. 18 ; W. 20, margm 28.

1/:v{J.erono<;I
cipher -system mentioned above (p. 19 n. 1) makes
ANOI use of it. The second type which may already
be seen on inscriptions from the Vlth cent .
.llavdd t:,,.At:,. B.C., occurs with the uncial, but afterwards
6e6<; eI regularly with the minuscule. For the uncial
see fig. 6. The minuscule has a number of
these combinations which belong to the normal
' Ieeov<1aA:qµ. It\ HM type of writing and which have their own
'l71<1ofi<; history (fig. 7). In § 18-19 we saw that they
became more and more numerous. The defor-
mation of the minuscule by the introduction
of uncial and cursive forms at the same time
o y No f implies the adoption of combinations of these
forms (see fig. 8). The superposition also gave
rise to new combinations (see fig. 9). In certain
:rcvdJµ.a nNA CJSes the combination may develop into real
IHP monograms (fig. 10). ·
3. Suspension means the omission of the end
ofa word. One may write either the initial only,
e.g. the 9 on coins and horses from Corinth (the
Xf!ia-r6<; so-called xon.mula,) and the 2 on the shields
Fig . 12.
Nomina sacra. Cf. Pl. Ill, of Sicyon 1) and the A on the shields of Lace -
ii 3, 12; VIII vs. 7 etc.
2 , 27,
LC. 7a 19, b4, 24; 20, 1,9 etc. 1) See the story related by Xenophon, Hell. IV, 4, 10.

45
daemon, or a few letters from the beginning of the word, as Af>E was
written on the Athenian coins and IY on the coins of Sybaris. In
manuscripts there is as a rule one additional feature: the last letter is
written above the preceding one, and a short oblique stroke is added
next to or through the last letter 1 ). Such a superposition in a word

l(<(Tal

• -J.U.
VO\
Q

·- - ~
s OIL ll ' I
£1/i'I

J.<l T,;. ~
tL 0(1' c.T
......
,
)(_Of;S
' ,, ( ' x
w
'A ''l
"""e£"
(o701
~
, ,
«VTI 'ls
Fig. 13. ...
Kol&
Abbreviations through
contraction together with
superposition, suspension
or combination. Cf. Pl. V,
schol. 2, 7; CL. 10, margin
.. ,' o,c;
14 ;47,3 ;48, 7; Th. 66,
margin 3 and 8 ; 70, 4 ;
j j \\ ,
72 ; 5, ii 6 and 14; LC. ,.,' . ov .o v
6 margin 1.

LIV t

1<.oa-,-.0,
'c..
,,, E.V

1<.i'.i
K Xo,_
7 7 l,JV
• ~ I
Of 0(,

<rL.A'l''l
/ f.' frr,' ..,...5 ws
Fig. 15.
Fig. 14. Symbols. Various abbreviations.
Cf. CL. II inf. 5; LC. N.B. u•and .,., are written as c,.,,
3, 6 etc.; n, 31, ,~ and e,~ as .,~.
abbreviated through suspension may be found on Pl. VI in the title, on
PI. V in many scholia, then in CL. g margin 23 ; IO margin 24 2); a
repeated superposition in CL. 11, 1. Strokes marking the suspension
are numberless: see e.g. t.he same plates and CL. g margin 22; Th. 81,
') Sometimes a horizontal stroke is drawn above the last letter.
' ) The last letter may be written double, if it is a plural. See e.g. V, 18, 25 at
0 0
the end, where ZIION means an:o.,dw.,,
 IO; 48, 2 and 9. Suspension combined with the two types of letter­
combination is frequent (fig. 11 and Pl. VI, 1. 2 of the scholia the name
�0µ71eof:). It should be noted that the different parts of compound words
may each be abbreviated by suspension, whereas for the combination
of suspension and superposition two successive words may be treated as
one single word (e.g. el{;r = el{; i-6v).
4. Contraction consists in omitting the central part of the word.
First of all we must mention the so-called "nomina sacra" reproduced
in fig. 12. The name has been given by L. Traube 1), considering the fact
that it is almost exclusively used for sacred terms; the origin, according
to him, is to be found in the Greek translation of the Old Testament
(the so-called Septuagint) made by the Jews in Alexandria. It is a fact
that these abbreviations are regularly applied only in religious texts, but
it seems to occur already in an Attic inscription from the IVth cent. B.C. 2),
it was probably known in Plato's time 3) and recent, but not yet published
investigations by Dr. A. H. R. E. Paap apparently ruin the theory. Deriv­
ations and compounds are treated exactly as the single words: e.g. 911.tdv­
(beom)oi;, ai(av)eovv, n(ai-)e"'fow (LC. 7r, 2r). The contraction is also
combined with other ways of abbreviating: a suspension like n:o l /, which
may be used for many words, is often determined by a contraction, e.g.
nolr/ = no.tlt71f;. For some other usual abbreviations consisting of con­
traction together with superposition, combination or suspension, see fig. I 3.
5. Symbols are conventional signs comparable to the ideographic
hieroglyphs, especially used for technical terms relating to mathematics
and astronomy. A small selection is given in fig. r4. Letters used for
ciphers are symbols too, though of a different kind.
6. The last type of abbreviations consists of various signs, mainly
deformed letter-combinations or adopted from shorthand writing. They
are practically countless and occur nearly in as many forms as there are
handwritings. They are mainly, though not exclusively, used at the end
of words. A thorough knowledge is only to be gained by long practice;
a small number of the most frequent signs are reproduced in their usual
form in fig. 15.

§ 22. Expansion of Greek Writing

The Phrygian, Lycian and Carian alphabets, at the same time also the
Etruscan alphabet (with its derivates, the Oscan, Umbrian, Faliscan and

') L. Traube, Nomina sacra. Munich 1907.

') Sigalas, p. 297 m entions IG. IP 1400 EK KE(P.AME)J2N, but the fact that
no other instance can be given, makes the case doubtful : it might be just a mistake.
') See the very old corruption in Hippius I, p. 289 a 4 (ANQN became AAAS2).
The case is probably similar in Pseudo-Aristotle, Oecon. II p. 1345 b 31 (ANQN
became AAA.ON).

47
 perhaps Latin 1) one) probably originated independently from the Semitic
alphabet. But in later times the influence of Greek civilization was so
great that their alphabet could not but be adapted to the needs of other
languages. The Coptic alphabet used by the Egyptians since the Ud
cent. A.D., when Christianity roused them again to spiritual life, is not
different from the Greek; only eight new characters have been added.
In the IVth cent. Wulfilas used for his Gothic translation of the Bible
an alphabet which is partly Greek. In the Vth cent. Armenians and
Georgians adopted the Greek letter -forms; St. Mesrop is mentioned as
the introducer. In the IXth cent. the Bible was translated in Slavonic
by Cyrillus and Methodius; the alphabet of forty letters which they started
using contains twenty-one pure Greek forms taken from the minuscule
and nineteen other signs mainly formed by alteration of pure forms. Soon
afterwards, however, a new alphabet came into use for Slavonic languages,
adapted from the uncial, containing forty- three characters of which twenty -
four are purely Greek. From this alphabet, wrongly called the Cyrillian,
derive the present Bulgarian, Servian and Russian writings.

') This has perhaps been influenced by Greek alphabets from Magna Graecia.
 CHAPTER V

VARIO US ITEMS
§ 2J. Ornamentation

A book is not only a thing of daily use; it is often also a thing of beauty.
From time to time an endeavour to beautify a page or a column is distinctly
perceptible. There are different methods, namely:
1. The use of initials at the beginning of a book, a chapter or a
paragraph; these are either larger or of a different colour, or more elaborately
drawn. Initials of larger size occur already in the Codex Alexandrinus
of the Bible (Th. 46) and in the Dioscurides of Vienna (Th. 47); later
they are more frequent (e.g. CL. 6); initials painted in colours may be
seen in CL. 16 and 36. From the IXth cent. they become more intricate
(CL. 38, Th. 67); sometimes they are interesting drawings: an A is
drawn as a bird next to a stem (St. 19), an H as two young men holding
each other's hands, an I as a flower-stalk, an Y as a trunk on which two
birds are perched (see also CL. 17 ii), an O as a human face (LC. 5ob).
We must mention here too the spreading custom to write titles in a
different letter-type; e.g. the title in uncials whereas the text is in minus­
cules (Pl. VIII, CL. 10, II, 13, etc., Th. 54, 64 etc., W. 4, 6). St. 21
shows an open space left for a title which has never been written.
2. The drawing of a more or less ornamented border at the top
and beside the text, sometimes also under it. This device is rare in
classical Greek manuscripts.
3· The placing of so-called ,,vignettes" at the beginning or at the end of
a section, book or chapter, or between two successive parts. They are
sometimes just a few simple strokes; sometimes they form complete minia­
tures. See Pl. IV, VfII, XII; CL. 20, 25, 35, 38, 46; Th. 46, 64.
4. The use of "monocondylia", i.e. the playing with letters or
letter-combinations, especially in titles and subscriptions (§ 25). Good
examples may be found in CL. 50 i, O.nd. 63 and W. 33.
5· The drawing of illustrations and pictures which refer to the
contents of the book. In mathematical and astronomical works we can
hardly consider the added figures as an illustration (Sch. 31, LC. 2, 6. See
also CL. r r), because they arc as necessary as the text. Illustrated literary
4 49
 papyri a re exceedingly rare , codices not. We have al ready mentioned the
Ambrosianus pictus of Homer (p. 32 f.) probably from the IVth cent. A.D.
CL. 21 gives an ill ustration with the name of the artist. The Venetus A of
the Iliad gives some drawings in the margin next to the Excerpt a Procli
(V. 13). See further the two manuscripts of Dioscurides (V. 5-8, O.nd. 8)
and the Parisinus suppl. gr. 247 of Ni cander from the XIth cent. (O.nd. 38d).
LC. 93 [1091] gives a portrait of S. Gregory of Nazianzus; see also CL. 25.

§ 24. Numbering ; Stichometry and Colometry

In the pa pyrus-roll the columns are very rarely numbe red. It had, as
a matter of fact, little meaning, because one could not turn up a special
part with one motion of the hand. In the codex this became different:
first of all the quires were numbered, then the leaves, which was more
practical; pages have not been numbered except in printed books.
In poetry the verses we re counted; often the hundreds were noted in
the m a rgin (see the Ilias Bankesian a, Th. 17, W. 4). This stichometry
also e xtended to prose-works. It was not only us e d to facilitate the
turning up of the verses and the finding of a special passage, but also
in order to know the size of a book and, consequentl y, its price. The
salary of the scribe depended on this.
As a standard for prose a line of sixte en syllables was the average, i.e.
about the l ength of a dactylic he xameter. Certain manuscripts note the
number of lines at the end of books or chapters, e.g the Laurentiani of
Sopho cles and Herodotus, the Pa risini of Demosthenes (O.nd. 29) and
Plutarch (O.nd. 67-8). The o ld Greek cipher-system is often used for
it and see ms to prove that sti chometry is very old. Diogene s Laertius
indi cates the total size of Aristotle's works (V 27): µve,adec; -ie.,;-uie t:<; ,cal
.,;enaea.Hov-ia neo<; .,;ol'<; :n;e,,.,;a,c,axillo,<; ,cal d1axoalo1<; tpdoµiJ,cona =
445.270 lines.
By colometry is meant the numbering of ,c@A.a, so-called "sense-lines"
in prose which help to facilitate reading. They count from 8 to 17
syllables.

§ 2;. Subscriptions and Personal Remarks

Not only the text w ith title and notes occur in the manusc ripts. The
scribes repeatedly added annotations of a di fferent nature and a different
purport in prose and poetry (often in adz o, nol,n,col, see CL. 50 ii).
It may be a personal remark in c onnection with the te xt, a well -known
example of this being the viole nt outburst against Juli an the Apo�tate which
penetrated even into the te xt of the only e xt ant manus c ript, the Vossianus
(see p. 39); see J. Bidez and F. Cumont, I uliani imperatoris epistulae et
leges, Paris 1922, p. 134.
50
 More numerous are the subscriptions added at the end of the book.
In religious texts it is mostly an instruction to later copyists to write with
care and to correct properly; sometimes an incantation is addi!d. The
scribe often mentions his own name (CL. 48 in Latin) 1) or that of his
employer (V. 12, I. 20). More important to the history of the text are
the chronological 2) and critical subscriptions (O.nd. 8, 0.d. 63): in the
former we find a statement of the time at which the copying began or
ended, in the latter data concerning the origin of the text or the author
of the recension (CL. ro), and other particulars. At the end of the Clarkianus
of Plato (LC. 9) the subscriptions mention the scribe's and employer's
names, '.the price of the work and the date 3).

§ 26. Accentuation and Punctuation

The accentuation is an invention of the Alexandrian scholars (Aristo­
phanes of Byzantium is mentioned by name). It was invented in order
to facilitate the reading of texts written in scriptio continua and to suggest a
solution in case of doubt.
On papyri dated before our era, accents are marked from time to time,
but their use is not regular. It is not always the accentuated syllable
which bears a mark: e.g. in the Bacchylides papyrus "(?<ho; is written,
whereas actually "eat:6; is meant, and in the Hawara-Iliad (Th. 18)
{}eoeid11 ; means {}eoeidfJ;. In the first centuries of our era the use
became more frequent. The oldest vellum codices are again quite without
accents, as the severity of style requires (the accents in the Vaticanus of
the Bible are of later date, see p. 31). Gradua11y, however, the scribes
began to write all accents according to the system which is still in use
now. We must, nevertheless, pay attention to the fact that, as a rule,
they put the accent on the first vowel of a diphthong and not on the
second, as we do, that, exceptionally, the accent is noted on the prepo­
sition in a compound word (CL. 7, 1), and that in cursively written
manuscripts the place of the accent may be quite fortuitous. In the vetus­
tissimi the accents are written precise and small, in the vetusti they begin to
grow larger and in the cursive recentiores they are assimilated by the writing
itself. Compare e.g. Pl. IV with Pl. IX.
As to the breathings, their origin and history are practically the same.
They were again introduced by the Alexandrian philologists who used
the two halves of the old H, viz. I- for the rough, and � for the smooth
breathing. It is comprehensible that, as long as the signs were not

') See M. Vogel-V. Gardthausen, Die griechischen Schreiber des Mittelalters und
der Renaissance (Zentralblatt fur Bibliothekswesen, Beiheft 33), 1909.
") According to the Byzantine time-reckoning: the aera begins with the creation
of the world, in 5508/9 B.C.
•) Cf. LC. 22 (name and date), 56 (date), 87 (name, date, request for intercession),
96 (date), 99 (names, date).

51
 systematically written, the rough breathing occurs more often. Th e original
forms (see Hawara-Iliad, Th. 18, 8; further Plate IV, V; CL. 4, 6 etc.,
Th. 53 and 54 = V. 9-12) are reduced to L and J (Th. 15, CL. 12 etc.
and already from time to time in the Marchalianus of the Prophets, CL. 4).
The cursive forms• and' have prevailed, though not exclusively (Pl. VI shows
the three different forms); they are normal from the XII th cent. onward, but
are doubtless as old as the breathing-marks themselves (see e.g. Th. 49 a
psalter from 862, CL. 30 etc.). In codices recentiores they are often combined
with accents and letters (see Pl. VII, VIII, XI; CL. 41 from 1294, Th. 72
from 1280).
The Cyprian writing used an upright stroke in order to separate words
or groups of words (p. 12). Old Greek inscriptions used a single, double
or treble point (. : : ) to the same purpose. One of the oldest Greek
papyri, the so-called Curse of Artemisia (N. 1b) dating from the IVth
cent. B.C., uses the colon as an interpunction; this is again the case
with various literary papyri written in the IIId cent. B.C. Later on,
however, it is practically reserved for the change of persons in the dialogue,
either in the drama or in prose. Aristophanes of Byzantium may be
credited with the invention of a method in which the high point (') was
equivalent to a full stop, the middle point (-) to our comma and the point
on the line (.) to our semi -colon (PL V; CL. 6, 9 and 10 are accentuated
in this way). This precise method is generally used in choice manuscripts;
elsewhere one or two of the available signs are considered sufficient (Th. 6,
N. ma, the London Bacchylides -papyrus, constantly uses the high point which
is the most normal sign; cf. e.g. PL VI, V. 17-20, the Ravennas of Aristo -
phanes). The comma (,) makes its first appearance in the minuscule
codices (e.g. already CL. 12 from 916, Th. 63 from 1040); the note of
interrogation (; or ;) 1) probably also (Pl. IV, Th. 54, 14 and V. 9, IO, the
Clarkianus of Plato; PL VI, V. 20, vs. 288, the Ravennas of Aristophanes).
The naea.yeacpor; (sci!. yeaµµ,j) may also be considered to belong to
the marks of punctuation. It is a dash drawn in the left margin under
the line of the text which finishes a section or when, in a dialogue, there is a
change of speaker (Pl. III, V). A paragraphus adorned with one or more
flourishes is called "oewvlr;, when it marks the end of a separate part of
the text (Pl. I has a coronisin the shape of a bird; cf. also N. 3b).
There are many other signs which are destined to facilitate reading.
We mention the principal ones:
r. a horizontal stroke above proper names (Pl. VII, 6, 10, 16, XI, vs. 5;
Th. 57, u) and, in geometrical texts, above letters designating points,
lines, angles and the like (LC. 2, CL. 11);

') It is perhaps a transformed t, suspension of t11rw.

52
2. an apostrophe between consecutive consonants or vowels (Pl. VI; Th.
47, 2 and 6; V. 5, 6 and 17) or used to mark the elision of a vowel
(Pl. V, VI; Th. 6 passim; V. 5, 28);
3. the diaeretic point or double point above , and v (Pl. III; V. 1-4; Th.
15, 7; 18, 7; 46, 1; CL. 6 i 17 etc., 51, IO etc.), often too where no
diaeresis is meant (Th. 6, 1; CL. 41, 1 etc.);
4. quantity-marks (_ and ..,) in poetry (Th. 6, 1 etc.);
5. µtv and de sometimes get a double accent in order to draw the reader's
attention to the antithesis (CL. 26, 1);
6. the arrow-head (d,.nM;) is written next to quotations (CL. 20 ii 18-9;
23; LC. 23; Th. 6i ii 9-10) and elsewhere to fill up a line (Pl. III, W. 3).
We may also consider the iota adscript or subscript as a prosodic
mark, because it was no longer pronounced in the last two millennia and
only helps to distinguish certain word-forms. In the oldest papyri iota
adscript is constantly added, in later papyri (e.g. Th. 14 and 15 from the
Illd cent, A.D.) it is, as a rule, missing. The same happens to the uncial
codices of vellum: a iota adscript is an exception; a iota subscript is
unknown. With the minuscule the iota adscript is regularly used again:
the Clarkianus of Plato e.g. (Pl. IV, V. 9- 12, St. 8, Th. 54) often gives it.
In the course of the Xllth cent. the iota subscript begins to appear: it
is nothing else than a iota adscript written lower and lower.

§ 27. Corrections

There are three kinds of corrections: addition, deletion, alteration.

Addition occurs either between the lines or in the margins; in the latter
case a reference-mark is often used: addition between the lines e.g. CL.
2 ii 27; Th. 3 ii 2; in the margin e.g. Th. 45 ii 14; CL. 17 ii 17; with
reference-mark e.g. (viz. the word llvco) Pap. Oxyr. 1231 fr. 1 - Sappho
fr. B 2 Lobel, 55a Diehl. A single letter is, of course, often inserted between
the other ones (Pl. IV, 1. 3 of the text).
Deletion is possible on papyri by wiping out, on vellum by erasing ( CL.
18 i 5 etc., 31); on both by striking out (CL. 31; W. 6 iii 14), or by the
use of parentheses viz. () or [ ], or by putting dots 1) over, under or
over and under the wrong letters, see e.g. Pl. VI 1. 15 of the text. In CL. 4,
14 the three methods have been used on the same page.
Finally the right letter is sometimes written at once across the wrong
one (St. 21, 9).

') Cf. Latin expungerc.

53
 § 28. Glosses and Scholia
There is a difference between glossae and scholia: the former are syn­
onyms of words used in the text, usually written between the lines; the latter
more extensive annotations concerning the text, as a rule put into the margins.
Two cases are to be distinguished: either the notes have been added
against the original intention of the first scribe, or they belong essentially
to the manuscript from the very beginning. The first kind of notes is
written wherever space is available. In the second case space is at once
left free for the annotations, either in the margin (CL. 13, 26) or in the
margin and below (Pl. VIII; CL.9, 47) or on three or four sides (Pl. V, VI, XI;
CL. II, 40; LC. 13, 44; V. 14-16; O.nd. 31b). Explanations between the
lines may still be added (CL. 47; V. 14-16).
References to the text and to the scholia are made by ciphers (CL. 28),
or by conventional marks (CL. IO, 12, 26; LC. 44).
Annotations are often written in a type of writing which differs from
that used for the text (Pl. VI; CL. 8, IO; LC. 13; W. 26, 27).

§ 29. Errors
It may be useful to list the most common types of palaeographical errors 1)
which appear in manuscripts and to add a few examples 2).
1. Confusion of similar letters. Aesch. Suppl. 95 lan-c,:i Ji: dmJwv
(�EAITIC)UN), read with Hermann&' tAnlJrov (C)E/\m�.nN). Lysias 30,
17 {}val� -cd; b, -cwv ,cve /J1:rov ,cal -cwv evnlrov (EY n /\ .n N), read with
Taylor ai11 Awv (CTH/\UN). Also ciphers and letters may be confused.
Lysias 25, 14 ofJ,: (o IC) E), read with Markland ol ie uh,ov-ca (o I/\').
2. Neglect or wrong interpretation of abbreviations. Aeneas Tact. 31, 28
,cal ov,c 1:IJnoe ov, read with Valckenaer 1:lJnoeov lfv (tvno{i ov). Xenoph.
Oec. 5, 12 ,) y fj {J,f1ovaa, read with Stobaeus {hoc; ovaa (oc, nomen
sacrum, read as 0 €; then {J,fovaa, which yielded no sense, corrected into
.i-tAovaa).
3. Loss of letters, words, sentences or verses; so-called lipography. Aesch.
Suppl. 4 dno n e oo-coµlrov 11:n-coµa{}wv Nellov, read with Pauw Aen-co­
tpaµa{},o.,v. Ibid. 317 Ai/Jv 11 , µeyia-cov yiji; ,caenovµev-17, insert with Burges
ntJov after yfji;. Eurip. Hel. 561 'EH11 vli; el ni; I} 'mxw el1J y vv71 was
lost in the direct tradition because next line began with the same word
('E1111vli; • dHd ,cal -co aov {}t).ro µa.i-1:iv). Restored by Markland from
Aristophanes' parody in Thesm. 907. When instead of two identical words

') Th�re. are also many errors of a non-palaeographical nature, due to similarity _of
pronunc1atwn, to the copyists iBnorance or false knowledge of the language or mis­
understanding of the contents. They are not taken into consideration here.
") More will be found in F. W. llall's Companion to Classical Texts, Oxford 1913,
p. 153 ff.

54
or parts of words only one is written, there is a haplography. Herod. 9, 78
.AdJ'-1'0>1'6 Hvl>eo,, Aly&1f1Jd0>11Td nei.i>Ta, read with Cobet Aly&1f1JTI0>1'
M,-, Td neruTa. Aeneas Tact. 15, 2 a haplography is combined with the
confusion of two similar letters, TOV<;de aTean1yovc; nae611Ta<;, read with
Hercher aTean1yovc; Tove; nrie6nac;.
Initials to be added later on by the rubricator may be forgotten and a
further corruption of the text is often the result. Stobaeus cites Eurip.
Phoen. 392 as av d' ov Tad' 1:lnac; 1''1/.Uyn11 II. nc; rpeo11El. The direct
tradition preserved the right text dovlov etc, In the anthology where
Stobaeus or his source found the verse, the initial had been forgotten and
consequently OY A.OY read as CY t:.OY.
4. Addition of superfluous letters, words, sentences or verses. Lucian
Dial. deor. 4, 5 xaTal>elYa& el!; "K",)11Yil", read with Co bet xa6e,.,,a,.
Aeneas Tact. 10, 12 T«p de YavxA,jecp dY0Ax71.,,xal xal>oA,x,j,,, read with
Casaubonus xa6oA,c'l" (a well-known word of daily use, such 3*a6oA,x6c;,
often comes in stead of a rarer one). When a syllable, a word or a
number of words are written twice, there is a dittography. Athenaeus XII
52, p. 537 b TOVH1)1' oliY 6YTO>YT(»1' ll!7Jl'4TQ)Y •••• xve,oc; fEY6µEYoc;,
delete with the Epitome 6n0111, being a dittography of o/i.,, (misread as
01') and T(»1'.
An explanatory gloss or even a scholion may wrongly enter into the text.
Demosth. 01. 2, 20 al yde wnea~la, dt:wal avyxev1Pa& Hal avax,ciaa,
Td To,aiiT" 61'dd71 is the reading of most manuscripts, but the Parisinus
S (see above p. 36) rightly omits xal avax,ciam, the explanation of
av7xevv,a,. Xenoph. Symp. 4, 26 t'«Ja>c;de xal ol ~(!<iJTE<; TO~lha, duz
TOVTOxaAoii1'Ta& 6n xal ne6arul>E1' ol xaAol nTeroaxova, rightly expunged
by Dindorf as a marginal note. Many manuscripts gave the choice
between different readings, a variant being added either between the lines
or in the margin. From time to time such a duplex lectio is considered
part of the text. Herod. 3, 4 Tro1' wYovzo:,y TOY nwT6TaTo1' dnoaTElAac;
- - xaT ' atiTcS,, is the right reading preserved in mss. A, B, and C. R and
S, on the contrary, have elc; TatiTOY which is manifestly wrong since
Herodotus writes le;, not elc;, and since TatiT6Y yields no sense. V reads
xa.eic; TmhcSv, copied from a ms. which had xaT' atiToY with elc; written
as a variant above xaT'; whereas the other mss. chose between the two
variants, one group rightly, the other group wrongly, V mechanically wrote
down the two.
5. The gloss may even chase the original word from the text and take its
place. The reading of Rand S in the preceding case is an example, Another
is Aesch. Agam. 282 dn' dyyeAov nveoc;, reading of all the mss. But
indirect tradition (Suida-; and the Etymologicum Magnum) preserved the
original reading d77aeov.

55
 6. A usual word tends to replace a less common one. See already above
,ca'l?-oAuojv for ,ca'l?-oA,c,jv. Xenoph. Symp. 4, 59 d dvvaµ.uo c; ll.eu11:ovc;
.no,elv, read with Brodaeus deea-iovc; (one ms ., A, already conjectured
deea-iovc;). Harpocr. s.v. Zev~ic; has deia-iodA71c; -iwv ,ca-i-' l,celvov -iov
xeovov ~ooyeaq,c.w, read with Goesius H.eia-ioc;.
7. Transposition of letters, words or verses . Aesch. Eumen. 727 .na..ta,dc;
dalµovac;, read with Cobet d,avoµa c;. Plato Rep. 437 D tv 6..tlyq,, read
with Cornarius tvl A6ycp. Aesch. Suppl. 120 .:no..tAci,c,d' tµ.nl-ivro ~vv
Aa,cld, Aivoa,vel, read with von Wilamowitz Aa,cld, avv. Ibid. 310 and 311
are rightly transposed by Murray.
8. Wrong separation of words written in script io continua . Eurip. El.
801 .nveav ~.n'tOV is against th e metre; read with Canter nve dvij.n-iov.
g. Wrong accentuation. Eupoli s fr. 276, 1 K . cSwcU,ca-i-o c; d -iv'PAoc;,
-ielwc; d -i11v ,caA17v lxwv, read with Cobet ,c6.A71v('tumour' or 'hernia').
 I ND Ex 1)

abbreviations, 38ff., 4If., Barberiniani (codices), 39, 43 colometry, so

43ff. Bavarici (codices), 26 columns, 22!. , 32
accents, 32, 35, 38f., 5If. Bible (mss. of the), 24, 32, 41 combination of letters, 34£.,
acrophony, 16 bibliography, uf. 43ff.
adaptation of Semitic alph. pt.{JAo, (PvfJAo,, p,fJAiov ), contraction, 43, 46f.
15f. 21, 28 Coptic alph., 47
addition, 53 Bodleiani (codices), 27, Corinthian alph., 18
Aegean writing, 13f. 36f., 40 correction, 53
Aeschylus, 38, 42 bombycina (charta), 22 Cottoniani (codices), 27
aims of palaeogr., IO book-form, 2of. Cretan alph., 18
Alciphron, 39 Borbonici (codices), 26 Cretan writing, 13f.
aleph, 15 order, 49 Crippsianus (codex), 40
Alexandrinus (codex of the Po1un:eoq,.,lil,v, 14 critical signs, 37
Bible) 32, 43, 49 breathings, 32, 39, 4 1, 51f. cursive, 2 8f., 31, 35, 38f., 41
alteration, 53 Britanni (codices), 27, 41 Cyprian writing, 13f., I7, 51£.
Ambrosiani (codices), 26, 32, Bulgarian alph., 48 Cyrillian alph,, 48
37, 50 Burneiani (codices), 27, Cyrillus, 47
0>'£thi,-, 23 38, 40
av£,liaa£iv, 2 3
Damascena (charta), 22
Anthologia Palatina, 39 Cadmus, 14 Damilas (Demetrius), 42
Apollonius Rhodius, 38 calamus, 22 definitions, IO
apostrophe, 53 calligraphy, 29f., 38, 41 ff. deletion, 5 3
Appian, 39 cunna, 22 Democritus, 14
Aratus, 40 Cantabrigienses (codices), 27 Demosthenes, 36, 38, 50
archaism, 30 Carian alph., 47 �l:eµa, 2 1
Arethas, 35f. Cassius Dio, 3 3, 43 diaeretic point(s), 52
Argentoratenses (codices), 26 caudex, 23 differentiation, 16
Aristides (Aelius), 36 Chalcidian alph., 18 Diodorus Sic ulus, 39
Aristophanes of Athens, 38, chancery-style, 29, 33 Diogenes Laertius, 4 0
42, 52
chapters, 32 Diogenes Laert. V 28, 50
Aristophanes of Byzantium,
xa e 1:tJ,, charta, 21 Dionysius Halie., 39
51f.
xaerl.ov, 24 Dioscurides, 3 3, 49{.
Aristotle, 23, 36f., 40, 42f.,
Chester-Beatty papyri, 24 J,q,,9,1:e a, 2 l, 2 3
44, 50
Chisiani (codices), 26 Jin:Afj, 52
Armenian alph., 47
ciphers, 19, 45, 47, 50 direction of writing, 14f.
.Arrian, 40
Clarkiani (codices), 27, 36f., Jova§, 2 2
Artemisia (Curse of), 52
52f. Dorvilliani (codices), 27
Athenaeus, 37
codex, codex, 2 3f., 32
il.lramentum, 22
,4ugustani (codices), 27 coins, 45 Ed.ilium (table of), 14
Coisliniani (codices), 26 £Alaa£tv , 22
Bacchylides, 23, 30, 51f. Colhertini (codices), 27, 32 ci'AE,v (clAciv), 2 3
Baµpvxl.vfJ, 22 collections, 37 cy><avato>', 23
1) References arc lo pages. Latin wonts are In Ualir.s.

57
 Epictetus, 40
epigrams, 29
errors, 54
Etruscan alph., 47
Etymologicum Gudianum,
39, 44
Euclides, 35(.
Euripides, 30, 33, 40, 44
Eusebius (St.) vita Const. IV
36--7, 24
evolvere, 23
explicare, 23
facsimiles, 12
Faliscan alph., 47
Famesiani (codices), 26
Fiorentini (codices), 40
/olium, 24
Georgian alph., 47
glosses, 53
Gothic alph., 47
rea.,,,,;;, 22
groups of alph., 18f.
Guelferbytani (codices), 27
Harleiani (codices), 27, 36, 43
Haunienses (codices), 27
Herodotus, 37f., 42, 50
Herodotus V 58, 21
Hesiod, 30, 40, 42
Homer, 17, 23, 3of., 37f., 40,
42f., 44, 5of.
Homeric hymns, 42
hymns, 30
Hyperides, 3of.
illustrations, 32f., 49f.
index, 23
initials, 36f., 49
ink, 22
inscriptions, 29f.
iota adscriptum and
scriptum, 52f.
lsocrates, 36
Jerome (St.) Epist. 141, 24
John Chrysostomu s (St.) 37f.
Julian the Apostate,40,42, 50
Julius African us, 43
Ka.J.aµo,;;, 22
.. ,pu,r:6,;;, 23
x.Sx,ul'01' µiJ.a,,, 22
...su.,,..a,22f. novelli (codices), 30, 43
xonnar:la,;;, 45 numbering, 50
xoew•l,;;, 52
...a,.IJeo,;;, 22 Olympian victors (list oO, 17
&µ.,,aJ.6,;;, 23
label, 22 23
&n,alJ.6yea'l'o•,
Lascaris (Const.), 41f. Oppian, 44
Latin alph., 47 Orators, 40
Laurentiani (codices), 26, ornamentation, 49
36(., 4<>,42, 50 Oscan alph., 47
leaves, 24 Ottoboniani (codices), 26
Leidenses (codices), 27, 43 Oxonienses (codices), 27, 35f.
Leningrad (codices at), 32, 34
Palatini (cod ices), 26, 39{,,43
Libanius, 42
palimpsest, 21
ligature, 44f.
Lipsienses (codices), 27 paper, 2If .
papyri, 25(., 30{.
liturgical mss., 23
nanveo,;;, papyrus, 2of.
Londinienses (codices), 41
naearea.,,o,;;, 52
Longinus (Pseudo-), 39
parchment (see vellum)
Lucian, 36
Parisini (codices) 34, 36(.,
Lycian alph., 47
L ycophron, 40 39ff., 42, 50
Pausanias, 43
Lysias, 40
pen, 22
ni••a, penna, 22
Marchalianus (codex of the
Bible), 33, 51 IIcnanv;ro,;;, 23
Marciani (codices), 26, 37, 40 Pergamena (charta), 21
material (writing), 2of. Perizoniani (codices), 27
Maximus of Tyre, 40 Phaestus (disc of), 13
Medicei (codices), 38 Phoenician alph., 14{.
µila•, 22 Phrygian alph ., 47
membrana, 21 .,,,,uo,,,24
Mesa, king of Moab, 15 Pindar, 40, 42
Mesrop (St.), 48 n,r:r:ax,o", 23
µn:azaeaxr:'7(!&aµ6,;;, IO plagula, 22
Methodius, 48 Plato, 35f., 52f .
Milesian alph., 10, 18f. plicare, 23
miniature, 49 Pliny, nut. hist . XIII 11-2
minium, 22 20{.
minuscule, 29, 34f. Plutarch, 37, 40{., 50
Monacenses (codices), 27 Polybius, 38, 4If.
sub- monocondylia, 49 printing, 35, 41 ff.
monograms, 45 proper names, 52
Montfaucon (B. de), 11 Ptolemaeus (Claudius), 34
Mo squensis (codex), 42 punctu ation, 51 f.
Mutinenses (codices), 26
qu antity -marks, 53
quaternio, 24
names of letters, 14
quill, 22
names of mss., 26f.
Ni cand er, 40, 50 quire, 24
nominu sacra, 46f. quotations, 52
normalization, 16f. Ravennates (codices), 38, 52
 recentiores (codices), 35, 41f,
recto, 20, 23f,
reference-marks, 53
Regimontani (codices) 27
RegiNmses (codices), 26
Revenue-laws (papyrus), 23
roll, 22f.
ruling, 22
Russian alph., 48
Sangermanenses (codices), 26
Sappho, 20, 53
Sarraviani (codices), 26,
32, 43
scholia, 36f., 42f., 54
Scorialenses (codices), 26
scripta Minoa, 13
scriptio continua, 31 f,
oe}.l~, 23
Semitic alph., 14f., 47
Septuagint, 47
sequence of alph., 14
Servian alph., 48
shorthand, 29, 47
signs, 43, 47
oiUvpo~, ol-r:n,po~, 23
Sinaiticus (codex of the
Bible), 32, 43
size of roll, 23
Slavonic alph., 47f.
awuanot>, 23
Sophocles, 38, 50 Umbrian alph. 47
stenography (see shorthand) uncial, 28, 32f., 35
or:iio• no1cr:ucol, 50 Urbinates (codices), 26, 36ff.,
stichometry, 50 42
stilus, 22
or:o,z'l66., 28 Vaticani (codices), 22, 33,
Strabo, 41 36ff., 39f., 42 43f. 49
subscriptions, 5of. Vaticanus (codex of the
superposition, 43ff. Bible), 32, 43, 51
suspension, 43, 45ff. vellum, 2 I, 32
syllabic writing, 13f. Veneti (codices), 37, 40, 50
symbols, 43, 47 verso, 20, 23f.
vetusti (codices), 35, 37(.
Tetraevangelium Uspenskij, vetustissimi (codices), 35f.
34 vignettes, 49
rer:ealfcot>, nr:ea~, 24 Vindobonenses (codices), 27,
nii.ro~, 23 33, 39f., 49
titles, 42 volumen, 22
titulus, 23 volvere, 23
Theognis, 37 Vossiani (codices), 40, 28,
Theophrastus, 36, 41 42f., 50
Theophrastus, hist. plant. Vulcaniani (codices), 27
IV 8 3, 20
Thucydides, 37, 39, 41 Washington gospel, 33
time of adoption of Semitic waxed tablets, 20, 23
alph. 17(. Wulfilas, 47
Timotheus, 3of.
Xenophon, 39, 41f.
r:6µo~, 23
fvlon:v,cr:011, 22
Townleianus (codex), 39 fv10,caer:io11, 22
Turonenses (codices), 26
types of writing, 28f. Zonaras XIV 2, 23
59
 LIST OF FIGURES
Page

r. Semitic and Greek Characters. 15

2. Greek Characters used in Inscriptions . 17

3. Greek Characters used in Literary Papyri 31

4. Uncial Characters in the Codex Sinaiticus of the Bible . 32

5. Characters in Minuscule Codices. 34

6. Combinations of Uncial Letters • 44

7. Usual Combinations of Minuscule Letters 44
8. Combinations of Letters used in Later Minuscule. 44
9. Combinations of Letters by Superposition . 44

ro. Two Monograms • . . 44

Ir. Abbreviations through Suspension together with Superposition or
Combination . 45
12. Nomina Sacra • . . . 45
13. Abbreviations through Contraction together with Superposition,
Suspension or Combination. 46
14. Symbols . . . . . 46
15. Various Abbreviations

6o
 SHORT DESCRIPTION OF THE PLATES
I. Berlin papyrus 9875, containing Timotheus' Persae. Height of the
roll 18,5 cm, length II1 cm. Written in the IVth cent. B.C. The plate
reproduces col. V, vs. 187-247. The writing has many features in common
with the types used in inscriptions. In the left margin a coronis and a
paragraph us.
II. Leyden papyrus n° 22, upper part of two columns, containing Plato,
Phaedo 65 A-66 B. Actual size u,8 X 16,8 cm. Originally the height
was about 27 cm. If the roll contained the whole Phae do, its length was
about IO m, so it is safer to assume that the dialogue covered tw o smaller
rolls. It is written in a nice, regular hookhand which may be assigned
to the 1st cent. A.D. A colon marks a change of the speaker. The text
appears to have been a fairly good one.

III. Vossianus Gr. Quarto 8, Parisinus Gr. I7, and Leningradensis Gr. 3,
extant parts of a manuscript of the Old Testament (Genesis to Ruth) on
more than 400 leaves; 130 are in Leyden (Sarravianus), 22 in Paris (Colber­
tinus), 1 in Leningrad (size : 24,7 x 23 cm). Written in the Vth cent.
The plate reproduces the verso of the 19th leaf of the Leyden part and
contains Leviticus X 4-10. It is a good example of the severe uncial.
Some lines, e.g. r, 8, 13, are filled up with the Jinlij, in others, e.g. 3, 7 , 9,
the last letters are written in a smaller size.

IV. Bodleianus ms. E. D. Clarke 39 ( Clarkianus) of Plato. Written

in 895. Contains 424 original vellum leaves, preceded by 6 of later
date (size: 32,5 x 22,5 cm). The plate reproduces the recto of leaf
263, containing the end of Alcibiades I (from p. 135 c 8 onward) and the
beginning of Alcibiades II (till p. 138 c r). Specimen of the bea utiful vetus­
tissimi which Arethas ordered for his collection: the minuscule is practically
free from uncial or cursive elements; the titles, one at the end, one at the be­
ginning of each dialogue, are in uncials; smaller sections are indicated by a
letter projecting in to the left margin. On other pages, especially in much
read dialogues, scholia are added. They are written in uncials.

V. Venetus (Marcianus) graecus 454 of the Iliad. Vell um codex

of 327 leaves (size: 39,3 x 27,8 cm.), written in the Xth cent. Beside a
carefully written text which has been corrected by a contemporary hand in
61
 the extreme margin, it contains prolegomena, arguments, precious scholia
and a continuous series of critical signs. The plate reproduces leaf 253
verso with T 76- 100. The minuscule is fluently written but very regular;
strange elements do occur; the scholia themselves are also in minuscules,
but the lemmata, as a rule, in uncials. In the scholia the number of abbre-
viations is not so large as one might expect. The first lines in the upper
margin read as follows: t Toi'a, de xal ,u:Thui(n-) c1,,~ d,.deru(1') Aya-
µt(µ,..ow) · avTo.i-(n-) t~ Ede(71c;), 6 d' [read ovd'] t,.. µlaaola'& d1'a0''1'~ ·
oil-r(roc;) .11aeaA(!&aTocpa(1'e,). t,.. de Tij, Maaaali(A)TI.Xij, xal 12 XLa,.
Tola, d' d1'UJ'Taµe(yoc;)µedcp7J H(!el"'" Ayaµl(p1'0J1') • µij'l'(n,) d1'aO'TE:1'azro1'
xal vcp' Elxeo(c;) cllyea naazro'l' • OVTOJC: 6 L1ldvµo(c;).
VI. Ravennas 137, 4,A of Aristophanes. Vellum codex of 191 leaves
(size: 33,2 X 21 cm.) dating from the end of the Xth or the beginning of the
Xlth cent. It gives the text with scholia on four sides. The plate reprod-
uces the beginning of the Frogs on the recto of leaf 34. The minuscule is
regular, but slopes to the right. Impure forms are not numerous. The
scholia are written in a cursive form of the uncial with many abbreviations.
Changes in the dialogue are indicated by means of abbreviated names.
Reading of the first lines of the scholia: - 6 Savi}[~ tnl lf7ov naeayeT(a,)
xa.i-e,oµn•ok) lzro1' tnl T611' ropOJ7 d1'acpoeo1' 8.nov ;1' Ta aTemµa-,;a
J2 - T:W1'£lro{J,6u,,v: - dPd T(oii) T(t»1') l:{}[µc,w. ,) µ£TOJ:(7/)d1'T(t) TQ))'
6voµa-r;ruv· xal "0µ('1/)e(oc;). TO'l' d' ali T7JUµazo(c;) .ne.n1'11J
3µl1'o(c;) dn(l)

-r;(
oii) .nwv-i;oc;.
VII. Vossianus Graecus Folio 77 III. Manuscript of 139 paper
leaves (size: 26 X 16,5 cm.) dating from the first half of the Xlllth cent.
Main codex of some works of Julian the Apostate. The plate shows the
recto of leaf 129. The upper part has suffered from damp. The page con-
tains the end of letter 188, then letter 108 and the beginning of letter 112
(according to the edition of Bidez and Cumont, Paris 1922). It is a model
of a fluently written minuscule with many cursive elements: superpositions,
abbreviations, combinations of accents and breathings. The size of the
letters differs considerably; compare e.g. the o and e which are often very
small, with p, e, x and ro. As a rule, a horizontal stroke is put over proper
names. Letter 108 begins: 71 µ(b) naeo,µla cp7J(a)(l7) • tµol av dujye,
[read d,71yei']-,;ovµ(01') lf7ae. tyw de lo,xa aol TOa( 01') 12 vnae dcp7Jyew{}a,.
VIII. Vossianus Graecus Quarto 4. Paper manuscript of 70 leaves
(size: 25,3 X 17,5 cm.), bound in two parts, the former (A) containing
Aeschylus' Prometheus, Persae and Septem, the latter (B) Pindar's Olym-
pian Odes, some of them with scholia. It dates from the second half of the
XIIIth cent. The plate reproduces leaf 38 verso on which one sees: the main
part of the metrical analysis of the strophe and epode of the second Olym -
pian, written in two columns, the "cola" being numbered (see A. Drachmann,
62
 Scholia vetera in Pindari carmina I p. 57, 9f.); then a rough ornament in red
ink, the title of the ode in red, showing some combinations of letters, the
beginning of the ode written in two columns which must be read together
(cf. plate IX), the initial again being in red ink. In the right margin and
below are scholia with lemmata in red (see Drachmann p. 59, 6 f.). The
introduction reads: 'fO P' (= devu:eov ,ewAov). lrovi,eoy dlµneov 11:a-ia-
d.no a-ieox() [read -ieox(ai:,eiji;)]. It ends in the second column
..t17,en11:(011)
with: Tt..toi; -ioii fJ71palov 'ltero1101;. The scholia in the right margin:
11.va~i<poeµiyy(e<;) iJµvoi' 0£1][ read ol] -iij<;cpoeµiyyo<;dvci/aaovn:<;• .nethov
(yde) -id lf.aµar:a avvr:l??-e(v-i)ai,elr:a oih(ro<;) .neo(i;) avr:d (,ea,) 7) Avea
deµoter:ai, etc.
IX. Laurentianus XXXI 15, forming together with Vossianus Graecus
Folio 52 one single codex of in total 177 (162 and 15) paper leaves
(size: 30,5 X 22,5 cm.), written in the XIVth cent. The former part
contains four tragedies of Euripides (Hippolytus, Medea, Alcestis, Andro-
mache) and six mutilated comedies of Ari~tophanes (Acharnians, Women in
Parliament, Knights, Birds, Wasps, Peace), the Leyden part has the begin-
ning of the Lysistrate (vs. 1-1034) and the end of the Birds (vs. 1492 etc.).
The two columns are arranged as in plate VIII. The scholia are numerous.
The plate reproduces leaf 14 recto of the Vossianus and shows Birds vs.
1639-1675. Beginning of the scholia in the upper margin: t L1eve' wt;
fµ' d.nOXW{!1JaOV !, 8n lfµoiOV -iq>-io.ncp fKElY'{) • ()£V(>' t,t{}j; ai,y aol -il£AAa
PovAoµa, q,eaaai :, diapciA..te-,;alae d freiot;. on 'fO[read -,;q,]d,apa.Al2Ae-i-ai
xeoovrai t.ni WV t~a.nar:iiv, Wt; (,eal) :4.exi.n.not; tv IIAov-i'{), etc.
X. Leidensis Bibliothecae Publicae Graecus (B.P.G.) 33 H. Paper
manuscript from the first half of the XVth cent., now containing 50
leaves (size : 19,5 X 21 cm.) with part of the Iliad (fJ 435-N 134) and
Homeric Hymns of which it is the most complete and most reliable manuscript.
It is commonly styled Mosquensis, since it has been brought to Western
Europe from Moskow (the first part of the Iliad may still be there).
The plate reproduces leaf 40 recto with the Hymn to Apollo vs. 447-498
written in two columns. In each column one letter projecting to the left
indicates a new episode in the narrative. There are many abbreviations in
the text.
XI. Vossianus Graecus Folio 64. Paper manuscript from the be -
ginning of the XV th cent. It contains on 493 leaves (size: 28,2 X 20 cm.)
the Iliad with countless glossae written above the lines in red ink, and
many scholia in the margin written in black but with red initials. For the
constitution of the text the manuscript is worthless, but it gives a fair
example of a working copy which may have belonged to a school-master.
The plate reproduces the recto of leaf 314 and shows: a double im6{}eaii; in
prose, a second in one single dactylic hexameter written in red, the verses
1 - 10 of O with a red initial, and scholia. Beginning of the hypothesis:
'Y.n:diJ-eai,; -iijc; o 'Oµ:fJeov ~av,rpolac;: t- Zevc; tyf:e{}dc; 1t(al) {}1:a11aµ(1:v)oc;
rnvc;: Tewac; vi\ 2 ,eroµtvovc;, "Heav t.n:i.n:.A:fJae1ei,
1tal 'Ieiv µe-iantµnt:t neo(c;)
Iloae,/;wva, ,cel1:vwv dcplcaaalJ.a, Toii .n:olt1iov.

XII. Leiden.sis B.P.G. 16 L. Fine vellum codex of 213 leaves (size:

33,7 X 23 cm.). Written about 1470 in a careful calligraphic minuscule
with red titles and initials: the x is the only letter which mars the regularity
of the writing. It is a manuscript of Pausanias and the plate reproduces
the verso of leaf 188 with the end of book IX and the beginning of book X.
Between the two books there is a simple ornament in red. The margin
contains a short note in red which draws the att ention to an int eresting item
in the text, viz . .ne(el) T(ijc;) ~w,ct(ruv) ,cal IJ.e [these two letters expunged
by dots under them] 0eaaa1(~v) (the e is corrected] µax71c; Tijc; neo -iwv
M71(foc(wv).
 ADDENDA

§ J. Short Bibliography

Page IO: On writing in general:

M. Cohen, La grande invention de l'ecriture et son evolution. Paris 1958;
3 vol.
Page 11 : On Greek palaeography in general:
Collective Work, Geschichte der Textiiberlieferung; Band 1: Antikes und
mittelalterliches Buch- und Schnftwesen. Zurich 1961.
Page 11 : On Greek books:
F. G. Kenyon, Books and Readers in Ancient Greece and Rome 2 • Oxford
1951

Page 11 : On papyri:
C.H. Roberts, Greek Literary Hands 350 B. C. - A. D. 400. Oxford 1956
(24 plates).

Page 12: Facsimiles and Reproductions:

J. Bick, Die Schreiber der Wiener griechischen Handschriften. Wien 1920
( 52 plates).

§ 6. Groups of Alphabets

Page 18-19: L. H. Jeffery, The Local Scripts of Archaic Greece. Oxford 1965.

§ 10. The Codex

Page 23: note to the paragraph title:

cf. C. H. Roberts, The Codex. Proc. of the British Academy, 1954, 169-204.

§ 14. Literary Papyri

Page 30: after Phaedo (12th line from the bottom):

Excavations at Derveni, north of Thessalonica, brought to light in 1962
a carbonized, but readable papyrus with a commentary to the Orphic
Theogony, dating from the middle· of the IV th cent. B.C., and, consequent­
ly, older than the Timotheus-papyrus.
 Page 31: after same page (12th line from top):
A good example of a literary text of this period written by a moderately
experienced scribe is the famous Geneva codex of Menander's Dyscolus.

§ 15. Unical Writing

Page 33: after 450 (3rd line from top):
R. Bianchi Bandinelli, Hellenistic-Byzantine Miniatures of the Iliad, Olten
1955 (cf. Studi Miscellanei del Seminario di Archeologia ... della Univer-
sita di Roma I, 1961, 1-10) argues that the codex dates from the reign of
Anastasios (491-518).

§ 18. Codices Vetusti

Page 40: after scholia (21st line from top):

Reproduced by ]. A. Spranger, Paris-Florence 1938, 2 vol.

Short description of the plates

Page 63: No. X, after XVth (17th line from bottom):
The reference to this ms. on p. 42 has to be corrected in accordance with
this date.
Page 63: No. XI, after from the (8th line from the bottom):
For beginning read second half.
Page 64: No. XII, after 23cm. (6th line from top):
For written about 1.170 read written by George Tribizius at Venice between
1469 and 1473

66
GREEKPALAEOGRAPH�

PLATES
V
XII