New Fragments of Ancient Greek Poetry NAOUMIDES, MARK Greek, Roman and Byzantine Studies; Fall 1968; 9, 3; ProQuest pg. 267
New Fragments of Ancient Greek
Poetry
Mark Naoumides
THE MANUSCRIPTS of the lexicon attributed to St Cyril
(Lexicon Cyrilli) have thus far yielded a small number of other
wise unknown fragments of ancient writers.1 These fragments
come chiefly from two closely related Mss, the Codex Hauniensis 1968
(XII cent.) and the Codex Messanensis S. Salvo 167 (XIII cent.). A third
Ms, related to these but until now completely overlooked,2 has pre
served additional new fragments, which deserve the attention of
philologists.
This Ms (which I call Z), Codex Matritensis Bib!. Univ. 2-22 no. 116
(olim E.l no. 61), is a parchment codex assigned to the X/XI century
and written by at least three distinct hands.3 The second of these
exhibits the characteristic style which R. Devreesse calls "en as de
pique"4 and suggests a South Italian origin for our Ms. The text of
the Lexicon Cyrilli bears numerous corrections, additions and other
notes by various hands (some even writing in Latin) which cannot be
clearly distinguished from one another on microfilm.5
1 Cf. R. Reitzenstein, "Inedita poetarum Graecorum fragmenta," Index lectionum in Aca
demia Rostochiensi semestri hiberHO a. MDCCCXC/XCI., 3-18; A. B. Drachmann, Die Ober
lieferung des Cyril/glossars (K. Danske Vidensk. Selskab, Hist.-fil. Meddelelser XXI.5, Copen
hagen 1936) 33. D. L. Page, "The Sources of Stesichorus Fr. 74 (Bergk) and Sappho Fr.Z.5
(L-P.)," CR 73 N.S. 9 (1959) 193-4; W. Biihler, "Ein neues Wort fUr Aeschylus' Glaukos
Potnieus," Philologus 110 (1966) 306.
Z Despite Ch. Graux's cross-reference to it in his deSCription of the Hauniensis in "Rapport
sur les Mss. grecs de Copenhague. Notices sommaires de la grande Bibliotheque Royale
de Copenhague," Archives des missions scientifiques, me SER. 6 (1880) 198-9.
• For a summary description see Ch. Graux / A. Martin, "Rapport sur une mission en
Espagne et en Portugal. Notices sommaires des manuscrits grecs d'Espagne et de Portugal,"
Nouvelles archives des missions scientifiques et litteraires 2 (1892) 130-4. It is beyond the scope
of this paper to give a detailed description of this Ms.
'us manuscrits grecs de ntalie meridionale (Studi e Testi 183, Vatican 1955) 34-6.
6 Special acknowledgement is due to the Biblioteca de la Universidad de Madrid for mak
ing the Ms available to me in microfilm, as well as to the University of lllinois Research
Board for generous financial assistance in connection with my study of the Mss of the
Lexicon Cyrilli, the first fruits of which are presented in this paper.
267
268 NEW FRAGMENTS OF ANCIENT GREEK POETRY
The two related Ms, the Hauniensis and Messanensis, which I call
hand m respectively after Drachmann,6 are not copies of Z. Of these,
h displays a text close but inferior to that of Z, especially in the second
half of the Lexicon Cyrilli where its text is, to some extent, epitomized;
indeed the scribe of h considerably shortened the longer entries of Z,
while omitting some others entirely. This explains why the new
quotations (all of which occur towards the end of the lexicon) are
absent from h. However, it seems safe to assume that h was not
copied directly from Z, since in the entry omuov, h is clearly copying
from a Ms in which the explanation was omitted (see below, gloss 8).
Furthermore, the readings KaA€LT(U (vs. AaA€LTa, in Z), s.v. A07Tas-, and
TOVTOVS- (vs. TOUS- in Z), s.v. 7TPOYOVO' , further indicate that h is not
derived from Z. The Messina Ms agrees closely with h but only in
the latter half of the lexicon (i.e. from the entry I-'TJAov61-'oS- on),
whereas in the first half it follows a different recension, the one indi
cated by K. Latte with the siglum n.7 Despite the close agreement
between hand m (which is plainly demonstrated in the glosses pub
lished here) the two Mss are not directly related but seem to depend
on a common source.s It is interesting to note that both these Mss
also seem to come from southern Italy.
The new fragments are listed below in the order in which they
appear in Z. Since the folios of the Ms are unnumbered and the
signatures of quadernia have been cut out, more precise reference
cannot be made. Scribal errors and corrections are given in paren
thesis, accompanied by the sigla Z (for the first hand) and ZC (for all
subsequent corrections). The readings of h, m, and parallels found in
other lexica, are given subsequently, the latter only insofar as they
are judged to bear directly on the glosses edited here.9 To the best
6 Op.cit. (supra n.l) 14 and 24-5.
7 Hesychii Alexandrini Lexicon recensuit et emendavit Kurt Lane, I (Copenhagen 1953)
xlvii-xl viii.
8 Cf K. Lane, op.cit. (supra n.7) p. il (xlix) n.1. The independence of m from h is well
illustrated in at least one case in the glosses published here (cf infra, gloss 4).
8 Unless otherwise indicated, the following editions were used both for quotations from
and references to the pertinent texts: ApoHon.Soph., Lexicon Graecum Iliadis et Odysseae, ed.
H. Toll (Leiden 1788); Etym.Gud., ed. Fr. G. Sturz (Leipzig 1818); Etym.Mag., ed. Th.
Gaisford (Oxford 1848); Hesych. Lex., ed. K. Lane (Copenhagen 1953-66) for the sections A to
0, ed. M. Schmidt (Jena 1858) for the sections II to Q; Photo Lex., ed. S. A. Naber (Leiden
1864-5); Pollux, ed. E. Bethe (Leipzig 1900-37); Suda Lex., ed. A. Adler (Leipzig 1928-35)
Evva'YWY~' ed. L. Bachmann (infra n.lO); Zonaras (or rather Ps.-Zonaras), ed. Tittmann
(Leipzig 1808).
MARK NAOUMIDES 269
of my knowledge these glosses do not occur in any other Ms of the
Lexicon Cyrilli and can safely be considered as interpolations from a
more ancient source.
1 4> a t 8 p 0 s . YEY7J()cfJS (YEY7J()WS Z), "Iwv (tWV Z, ;'cfJv ZC) 'ApYE{OtS
( zc, cipylo,~ Z) cPav€p6~.
cf 11, m: 4> a t D P W s (4)atDpcfJs h)· YEY7J()WS.
Hesych. r;p 40: 4> a t D p 0 S • Ka()apos, YEY7J()cfJS, cpavEpos.
It is clear that Hesychius has condensed the entry by omitting the
reference to Ion's work and has at the same time combined it with the
gloss cpatDpos' Ka()apos, which occurs as an independent gloss in the
Evvaywy~ M~EWV XP7JUlfJ-WV,lO Photius and the Suda. There is no other
close parallel to our entry in the extant lexicographic works, with the
possible exception of the Etymologicum Gudianum, which connects
etymologically the word cpatDp6s with cpatvw and 4>avEp6s (p.547.47).
The reference to Ion is made, I believe, for cpatDpOS in the meaning
4>avEpos, i.e. the reference here precedes the explanation. This infer
ence is supported not only by the punctuation of the Ms, but also
by the fact that of the two meanings attributed to the lemma, the
first is the common one and as such needs no confirmatory reference,
whereas the second is rare-indeed with the exception of the present
fragment, it is completely unattested.H The reference to the ehian
poet comes as no surprise to those familiar with Ion's diction and
style.12
Very little is known about the 'ApYELOt. It is generally assumed
that it dealt with the expedition of the Seven against Thebes.13 It
would be interesting to know, but vain to speculate, to what subject
Ion applied the adjective cpaLDpos with this meaning.
2 4> a A a v () 0 v . 7ToA,6v, E04>OKA-fts 'Axatwv uvAAOyCfJ. El DE (malim
01. DE) cpaAaKpov, 4>aAav()ov NEU-rOpOS Kapa (Kapav Z).
cf Hesych. r;p 91: cp eX A a v () 0 v . 7ToAtov. Kat TJ NEuropos KeXpa. 01.
DE cpaAaKpov.
10 Cf L. Bachmann, Anecdota Graeca I (Leipzig 1828) 1-422; also known as Lexicon Bach
mannianum. For brevity's sake it is here referred to simply as L'vvaywy-r7.
11 Cf LS] s.v. The genuineness of the meaning .pavfpo, is supported by the etymology of
q,aL'8p6s; cf E. Boisacq, Dictionnaire etymologique de la langue grecque4 (Heidelberg 1950), who
connects it with the words .pal'8Lp.o, and .paLos and translates it 'brillant, serein, clair.'
12 Cf W. Schmid / o. Stahlin, Geschichte der griechischen Literatllr LIT (Munich 1934) 518.
13 Cf A. von Blumenthal, Ion von Chios (Stuttgart/Berlin 1939) 32.
270 NEW FRAGMENTS OF ANCIENT GREEK POETRY
The word &.Aa.v(J0~ is quite rare. Except for our new fragment it
is not attested before the third century B.C., but it is listed as a dialectal
gloss from Corcyra in the brieflist referred to as rAwuu(u KO:7'U 1T6AE",~.14
So far as can be judged from the extant passages where the word
occurs and from its meaning in the Corcyraean dialect, it was com
monly used in the sense o:'A6~) and o:AO:Kp6~.
Hesychius' entry, although clearly related to that of our Ms,
differs from it in a few significant points. First, Hesychius has omitted
the reference to Sophocles; he has also condensed the quotation and
converted it to the nominative case, thus making Nestor's head the
equivalent of aAa.v(Jov;16 the quotation thus condensed appears after
the word 1ToA,6v and before the second or alternative meaning
o:'AO:Kp6v; finally, instead of the reading E"L 8' of our Ms, Hesychius
reads 0;' U. Notwithstanding obvious blunders, Hesychius' authority
should not be rejected in all these points without closer scrutiny.
To the contrary, I believe that both in the reading o[ 8' and in the
position of the quotation, Hesychius offers a much more attractive
reading than our Ms. The phrase E"L 8' ... Kapo:, which in the Matri
tensis takes the place of a quotation from the Assembly of the Achaeans,
has a characteristic Sophoclean flavor. Indeed the use of a periphrasis
with the word Kapo: and the genitive of a proper name (with or without
an additional adjective such as lA7'O:7'ov, Kpa7'tU7'Ov, KAE"w6v, o:v7'a8E"Aov)
for a person physically or mentally present or addressed is a well
attested mannerism in Sophocles intended to convey affection or
respect on the part of the speakerP The phrase as it stands in our
1& Critical ed. by K. Latte, "Glossographica," Philologus 80 (1924) 137-8. On the reliability
of the collection and the antiquity of the glosses of that list, see C. M. Bowra, ''rAwaaaJ.
ICaTIX 'II'OAEtS'," Glotta 38 (1960) 4~0.
16 For the meaning av8oS' = 'hair: cf EvvaywY17 (s.v. cf>&Aav8ot) 402.29; Photo 11.255; the
Suda 11141 (if. also A 2518); Etym.Mag. p.786.57; and Ps.-Zon. c.1794. This meaning is not
listed in LS].
16 For the declension of IC&pa as a feminine noun of the first declension in late antiquity
and Byzantium, if. Stephanus, TLGI IV (Paris 1865) col. 956 and LS] S.V.
17 Cf Ant. 1, OT 40,950, 1207 (addressing a person phYSically present); Ant. 899,915,
EI. 1164 (addressing a dead brother); OC 321 (introducing Ismene); OT 1235, OC 1657 (a
messenger's reference to a character off stage). In all these cases the disposition of the
speaker towards the person addressed or referred to is favorable. Not so in Euripides
MARK NAOUMIDES 271
Ms, however, seems to be nonsensical both on account of the juxta
position of the two etymologically cognate adjectives and the absence
of a verb. Even if we bracket the word cpaAalcpbV as a gloss that crept
into the text and emend further El 01 to EloE, we still will have to
explain Hesychius' reading. Besides, we are faced with metrical
difficulties; for, while such phrases are frequent in the dialogue parts,
the resulting phrase could not be accommodated in an iambic line
despite its obvious iambic clausula.
Another possibility would be to insert a comma after the word
cpcx.>..cx.KpbV and translate: "But if (the word means) bald (then cf the
phrase) cpa>..cx.v(Jov Nla.,.opo~ Kapcx.." The result is clearly far from satis
factory. (1) The quotation would be left without a reference as to its
source. Such practice is limited (so far as our Ms is concerned) only to
glosses originating from the Homeric lexicon of Apollonius the So
phist and quoting Homer. This phrase is non-Homeric. (2) To the
best of my knowledge there is nowhere a reference to Nestor's bald
head. On the other hand the meaning 'whitish', which is attributed
to Sophocles in the first half of the entry, would be much more
suitable for old Nestor.
Against these two alternatives one could follow Hesychius' lead
and transpose the words cpa>..cx.v(Jov Nla.,.opo~ Kapcx. immediately after
the reference to Sophocles' play. Disturbances inside an entry are
indeed not unparalleled in the interpolated glosses of our Ms (cf
s.w. >"07T(X~ and cpo'iVtg). With the quotation restored to its proper
place the change of El 01 to ol 01 is inevitable. The latter phrase
was commonly used in the lexica and the commentaries to introduce
an alternative but less plausible (in the eyes of the author) view
or interpretation. Since cpa>"cx.v(Jo~ with the meaning 'gray' is hapax
legomenon, the virtual rejection of the meaning 'bald' by the lexicog
rapher can be understood only with regard to the passage from which
the quotation was taken. The possibility that this gloss originated
from a commentary on Sophocles should be seriously considered.
If this theory is correct, vi\:. that the quotation was taken from
Sophocles' Assembly of the Achaeans, we may further speculate (on the
analogy of the use of the periphrasis with Kapcx. in Sophocles) that
Nestor was one of the characters of the lost play and that he was at a
certain point addressed respectfully by another character with the
(if. Hipp. 651, Tr. 1024) who also has the tendency to omit the proper name. The peri
phraSiS occurs once in Aeschylus (Ag. 905; I{>lAoV Kapa).
272 NEW FRAGMENTS OF ANCIENT GREEK POETRY
words quoted in our entry, which can tentatively be completed so as
to give a complete iambic line as follows: cPaAavOov N/.aTopos
Kapa. Perhaps Nestor was called by the speaker to mediate,
as in Homer, in one of the frequent quarrels which erupted in the
assemblies of the Achaeans. This would accord both with the little
that is known about the plot of the play and the characteristically
Homeric treatment of the myths by Sophocles, which has prompted
the characterization ·Op,"f}pLKciJTaTos. Violent quarreling on stage with
some attempt at mediation is again not unknown in the extant plays
of Sophocles: cf Odysseus' mediating between Teucer and the
Atreidae in the Ajax; Jocasta's mediation in the shouting match be
tween Oedipus and Creon in Oedipus Tyrannus; Theseus' mediation
in Oedipus Coloneus, and (to some extent) Neoptolemus' mediation in
Philoctetes.
3 1> 0 LV t g (cPOlVLg Z)- TO O€VOPOV, "cPOlVLKOS V€OV epvos" [Od.6.163].
" t , \ f ) f \ _ , ft ~ , \ "\ ,
Kat 0 Kap7TOS Kat 0 7TVPpOS Tep xpwp,aTL, os TO p'EV al\l\o
T6aov (Toaov Z) 1>OLVL~ (c/>OtVL~ Z) -ryv, EV OE p,ETciJ7Tep AEVKOV
c:rijp,' ET€-rVKTO" [Il. 23.454-5]. Ka~ TO cP0LVLKOVV avOos, EV
cP Td: c/>OLVLKa ( cP0LVLKd: Z) {Ja7TTETaL, •• <1s 0' OTE ( OT/. ZO) Tls
(TLS Z) T' EAEc/>avTa (ZO, EAEcpaVTa Z) yvvT] CPOlVLKL (cPOLVLK~
Z) p,L~Vr( [ll. 4.141]. Kat. EOVOVS ovop,a, "O~ T6TE CPOLVL~
(cpOlVLg Z) ~AOEV av~p" COd. 14.288]. Kat KVPLOV ovop,a,
"CPOLVLg (cpOtVL~ Z) aTTa (ZO, aTTa Z) yEpaL€ (ZO, yEpaLE
Z)" [Il. 9.607= 17.561]. Kat 0PVEOV lEP6V ~Atov (ZO,
"f}ALOV Z). Ka~ opyavov Eloos p,ovaLKov, 07TEP EVLOL p,aya
oa, <1<; £ocpoKMj<; Bap,vpq. (Oap,vpaL Z, Oap,Vpd:L ZO), Ka~
• AAKaLos (ZO, aAKaLos Z) CPOLVLKWVOS ( cP0LvlKWVOS Z) ovop,tX
(fortasse Kat' AAKaLo<;. cP0LVLKWVOS ovop,a).
cf h, m: cP 0 { V L g (c/>olvvg m) . T6 O€VOPOV CPOlVLKO<;.
Apollon.Soph. p.695-6: cP 0 LV L g. TWV 7ToMd: o"f}Aovawv ~
\ 't. \ \ \ ~ I ~ "n;. I .,,, ,,~\ ~ \ I\Es LS. TO p'EV yap OEVOpOV, '¥OLVLKO<; E"f}V EPVO<;. 0"f}1\0 t Kat
\ '" \ ~ 'A \ \ I .I.. I .. n;. ~ t. " TO KVPLOV oJ.'op,a, TOV TOU XLI\I\EWS Tp0't'Ea' '¥OtVLs aTTa,
", , '" 'f\ .... I tfl'J'1'" YEpatE. G'YJp,aLVEL Kat TOV 7TVPpOV Tep Xpwp,aTL' 1. 0 p'EV
"\ \ '" I .J.. ~ t..,. • ~ \ I , \ ~,. I .. al\l\o OEp'tXS 't'0 LVLS "f}V, EV OE p'ETW7Tep I\EVKOV a"f}p, ETETVKTO.
\ \.J.. \" 0 .. • n """ , ,.\ I.J.. \ Kat TO 't'0LVLKOV av OS' ~t:S 0 OTE TL<; T EI\E't'aVTa yvV"f}
I " \ \ '0 I .. A \ I n;. ~ t. .3:\8 " p'L"f}V[J. Kat TO E VLKOV' LJ"f} TOTE '¥OLVLS '11\ EV aV"f}p,
• I, .~ I " a7TaT"f}Ma EWW<;.
MARK NAOUMIDES 273
Etym.Gud. p.556.19-24: (j:J 0 ;; v, g . 'TO 8€vSpov.
(/> 0 { v £ K 0 S . VEOV €pvos. KCXt <> KCXpmJS, KCXt <> 1TVPPOS 'TOU
, ,,\ \ "\ \ , ,I.. - C 'I' , ~ \ , XpWfLCX'TOS· os 'TO fLEV CXlV\O 'Toaov '1'0 LVt." YJV, EV OE fLe'TW1Tlp
\ \ -'" \ \,1.. -, , ~\ \ I\EVKOV UYJfL E'TE'TVK'TO· KCXt 'TO 'l'0tVtKOVV· E'TE'TCXK'TO OE 'TO
CPOLVtKOVV (sic) av()os J iv cfj 'TO: cpolvtcx {1a1T'TE'TCXL.
Hesych. (j:J 710: cp 0 LV t g . 'TO OEVOpOV· "CPO{VLKOS VEOV €pvos."
, t \ \ ( )f\ ..... ,
KCX£ 0 KCXp1TOS. Kat 0 1TVPpOS 'Tep XpWIWTL.
This entry is made up of two easily distinguishable parts. The first
and longer one (from SEVSpOV to YEpcxd) lists the various meanings
of the word CPOLVLt in the Homeric poems with supporting quotations,
thus reproducing with some variations the text of the Homeric lexicon
of Apollonius the Sophist. The latter part deals with some unusual
meanings of the lemma and comes from a different source, possibly
a lexicon of the kind referred to as EEVWS EipYJfLEva. Although both
Hesychius and the Etymologicum Gudianum leave off much before the
end of the first part of the entry, their close agreement with our Ms
in this part as against the text of Apollonius, together with their close
ness to the interpolated glosses of the Matritensis in general, suggest
that their ultimate source contained the full entry as we find it in our
Ms.
The two references which appear in the second half of the entry
are both problematic. Sophocles, in his Thamyras (fr.238 Pearson),
used the word /wyaDtD€S, and prima facie this seems to be what the
lexicographer is referring to.IS Since it was a common practice,
however, to quote from or refer to works which attested the various
meanings of the lemma, it would be more natural to connect the
reference to the Thamyras with the words opyavov e180s fLOVULKOV and
take the phrase CJ1TEP €VLOt Iwya Sa as parenthetical. This is not
impossible, since our lexicographer is shown to be a mere compiler
who excerpts and adapts from a more extensive source and is not al
ways accurate or precise (see below, s.v. A01Tas). The view that the
reference is to the use of the word CPOLVtt is further supported by the
consideration that Thamyras was a king of Thrace and CPOLVLg, accord
ing to Athenaeus 14.637B, was an instrument used at the banquets
of Thracian kings. The fact that Sophocles also used the word fLcxyaoLO€S
in the same play does not invalidate this hypothesis, since Sophocles
18 The correction of fLCly&8a to fLCly&8t8a seems inevitable. The error is a typical one due to
haplography.
274 NEW FRAGMENTS OF ANCIENT GREEK POETRY
mentions a number of different musicalinstruments in the same play,
which after all dealt with a famous musician and his musical contest
with the Muses: Avpat (fr.238 P.), Tptywvos (fr.239), p.ovavAo, and
1I7}KT'iOES (fr.241), besides p.ayaOtOES (fr.238).
There seems to be some confusion in the very last sentence of the
entry. The phrase Kal. 'AAKaios CPOWtKwlIOS ollop.a, in my understand
ing, introduces an entirely new meaning of the lemma that equates
the name of the date tree with a grove of date trees. Alcaeus is cited
as a source for this rare meaning, which seems to be completely
unknown elsewhere. The normal word order in this case would be:
Kat CPOtlltKWIIOS ollop.a, WS 'AAKa'ios. Another possibility should be
reckoned with, however, that a connective particle (e.g. Ka{) following
immediately after the reference to Alcaeus and introducing the last
meaning was omitted by the scribe. In this case both references would
support the meaning 'type of musical instrument'.
It is not clear whether Alcaeus in the second reference is the lyric
poet from Lesbos or the Athenian comic writer, for the latter is often
referred to simply as Alcaeus without the attributives 0 KWP.tKOS or
o KWP.cpOtoypacpos. The balance of probability is perhaps in favor of
the former, since our lexicographer, when referring to poets of the
Attic stage, always quotes the play in which the word occurred.
4 X E L P 0 f3 0 UK 0 II • Tall OLd: TWII XELPWII ~WIITa. 1:0c/>OKATJS LlaLOeXAcp
(ZC, OaLOaAW Z). yauTp{XELpa (yaTpl. XELpa Z) OE 'AIIT{
p.axos Eo/r}.
cf h, m: X E L P 0 f3 0 UK 0 II • Tall (TWII h) OLd: TWII XELPWII ~wlITa.
Hesych. X 286: X E L P 0 f3 0 U K 0 II' Tall OLd: TWII XELPWII
~WvTa.
Etym.Gud. p.564.25-26: X E L P 0 f3 0 U K 0 S • 0 Otd: XEtpoS ~WII,
WS cp'YJu't EOCPOKATJS [fr.lI13 P.].
To the Sophoclean use of the word XELp0/30UK6s, known from the
Etymologicum Gudianum, our Ms adds the name of the play in which the
word occurred. It also attests the case in which the word was used,
for it was a well-established method of the old lexicographers to
use in the lemma the exact form of a word as it appeared in the
passage from which they excerpted it.19 The close parallel with
Hesychius leaves no doubt that the accusative has more authority
18 Cf my article "Fragments of Greek Lexicography in the Papyri," in Classical Studies
Presented to Ben Edwin Perry (Urbana 1968) 194-5.
MARK NAOUMIDES 275
than the nominative of the Gudianum. Save for this unique reference
the word is known only through the works of lexicographers and
grammarians20 and can safely be classified as hapax legomenon.
As for the context in which it occurred, very little can be guessed,
since the remains of the Daedalus are scanty and the actual plot quite
uncertain. It seems reasonable to consider it as a derogatory epithet
for Daedalus; or, if the title of the play indicated Hephaestus,21 it
could have been used in reference to the smith-god. The derogatory
sense of the word is attested by Pollux (cf. n.20). The known com
pounds with -{30UKOS are not earlier than Sophocles, who used the
words avOo{3ouKos(fr.31 P'),YTJP0{30UKoS (Ajax570) and EPPTJV0f30UKOS (or
apTJvof3ouKo~; cf. fr.655 P.) besides X€Lp0f30UKO~. In most cases the
second component has an active meaning, but AW'T0f30UKOS (Trag.
Adesp. 236 N.) shows that the passive meaning of XELp0f30UKOS was
not unique.
The reference to Antimachus is entirely new, as is the word
yaU'TptXELP attributed to him in our entry. It seems likely that Anti
machus coined it by reversing the word XHpoyaU'Twp, first used by
Hecataeus (fr.367 J.). Our lexica list two more words comparable
to the above, yau'T€pOX€LP€S and JYX€LpoyaU'TOpES,22 and paraphrase
them .. Ot a7Tb XELPWV ~WV'TES." The second of these words is attested
by Athenaeus 1.4D,23 while the first occurs only in Strabo, in a pas
sage which I quote, because it seems to have some bearing on the mat
ter discussed here: Tn fJ-EV ovv TLpVVOL opfJ-TJ'TTJplctJ xp~uauOaL OOK€L
n ~ , , ~'rT \ I ,,< \ 't \ ~ 0 ~, POt'TOS Kat 'TELXLUaL uLa nVKI\W7TWV, OVS E7T'Ta fJ-€V E vaL Kal\€tU at U€
I A... I , ~, ,,~, I •
yaU'TEpOXELpa~ 'TPE'f'0fJ-€VOV~ €K 'TTJS 'T€XV7J~, TJKHV OE fJ..E'Ta7TEfJ..7T'TOVS EK
AVKtas ••• (8.6.11 Meineke).
Although we do not know from which of Antimachus' works the
word yau'TptXELpa was taken, we may assume that, like all the new
fragments, it was excerpted from a poetic composition. Since, how
ever, so far as the evidence goes, Antimachus employed exclusively
20 Cf Pollux 7.7, TO yap xnpo{10GK,)r; • •• -?TTOV.xV nr; ,"POGOiTO; and Etym.Gud. p.572.49,
XEIP0{10GK,)<;' & 31a TWV XEIPWV {10GK,)fJ-€VOr;, which is a mere variant of the previous entry of
this lexicon.
21 Cf A. C. Pearson, The Fragments of Sophocles I (Cambridge 1917) p.llO, and C. Robert,
"Daidalos," in RE 4 (1901) 1995ff.
IS Cf Hesych. r 191; Lex. Bekkerianum no.5, p.230.13; Suda Lex. E 141,181; Etym.Mag.
p.221.52. 313.29; and Zon. c.598.
13 CJ. also the title of a comedy by Nicophon which is given variously as X€lpoyaGTop€<;
or iYX€lpoyaGTop€<;.
3-G.R.B.S.
276 NEW FRAGMENTS OF ANCIENT GREEK POETRY
dactylic hexameter and elegiac couplet, we are immediately faced
with a serious difficulty, because this word (with a short syllable
between two long ones) is unfit for dactylic meter. The parallel of
'HcpataTOTEVKTOV~ (fr.62 Wyss) from the Lyde, which is assumed to
have been in elegiacs, does not help, because that word has long
been suspected by scholars as a false reading. Unless we are ready
to suppose that Antimachus occasionally employed other meters as
well (such as the iambic), we must resort to emendation in order to
obtain a word that will fit a dactylic verse. The simplest way would be
to change the unparalleled yaaTp{XEtpa to the attested yaaTEp6XEtpa.
Strabo's passage can then give us a clue as to the context in which the
word may have been used. Indeed, I cannot resist thinking that
Strabo, who quotes Antimachus a number of times, had him in mind
when he referred to the Cyclopes who built Tiryns. The difference
in form (accusative singular vs. nominative plural) is trivial and may
be explained by the different contexts in which the word is quoted.
Neither Strabo nor our lexicographer pretends to reproduce the exact
form. The latter obviously changed the word so as to agree in form
with his lemma. Strabo's plural, however, has more claims to genuine
ness because of the number of Cyclopes and because our lexica (see
n.22 above) use the same form as the lemma.
We have no way of deciding as to which poem was the source of
the word, but perhaps the Thebais would be the most natural choice.
5 x,,\ t & y p a (x,;"{aypa Z)· ~wvCPWll, W~ LS"
(fr.69 B4.), KPLY~ (fr.50 0 3 .) and VKTJ (fr.136 B4.).
MARK NAOUMIDES 277
comparison with the attested compounds of aypa indicates that the
formation of XLAuxypa is unique. Indeed in all instances of such com
pounds the first component functions as a complement of aypa (e.g.
Kp€aypa, d8ovTCxypa, 7Tvpaypa),25 while the first component of XLAuxypa
is the numeral xtfl.Wt. Although one could press the point that
XLfl.taypa is the animal that preys on a thousand different animals, a
more likely interpretation may be of the animal with a thousand
'catches', or claws, or feet. 26 As such it would refer to a member of the
class of myriapoda, probably the one called UKOA07T€v8pa (our centi
pede) in ancient technical writings and which in the spoken Greek of
today is characteristically called uapavTa7To8apovua, 'one with forty
feet'.27 This myriapod (whose number of legs varies from species to
species) is known for its poisonous bite (actually inflicted by the fore
most pair of feet) and its name could have been used by Hipponax
either literally or metaphorically in a reference to a greedy person
one with a thousand claws or venomous "bite."
The second part of the explanation (Ka~ vOILLUILa)28 poses a difficult
question, since there is no important coinage featuring a myriapod
or any other animal with many appendages.29 The Eretrian coins
with cuttlefish on the reverse can be safely excluded, both because
the cuttlefish was commonly known by the name uTJ7Tla (which has
also survived in spoken Greek of today in the form of uov7TLa) and
because Hipponax mentions this very animal with its traditional
name (cf fr.68b B.4). In view of these difficulties Miss Margaret
25 Cf P. Chantraine. Etudes sur Ie vocabulaire grec (Etudes et Commentaires 24, Paris 1956)
44. The author does not discuss the word XLALaypa nor the words ovaypa, op€o{3a'aypa
and 'Tavaypa, which are apparently of obscure origin.
26 For the use of the word XlALOL as a first component to indicate a large number, cf
XLALOcf>VAAOS, xtALoSvvafLos. Cf also the compounds with fLVpLOL and £Ka'TOv.
27 The word UKOA01T€VSpa is also used in Modern Greek, restricted for the most part to
the formal idiom (the katharevousa). In the vernacular of Cephallenia, however, according
to an oral communication by Mrs Rene Kahane, the word has been preserved in the form
uKOVA01T€Tpa-the change clearly effected by the observation that the little animal lurks
under stones (1TETpaL) in search of necessary humidity.
28 Both the depiction of animals on coins and the custom of referring to particular coins
by the figures consistently depicted on them are well attested. Hesychius, to whose Lexicon
the Matritensis is closely related, has preserved a number of examples: aYKvpa (A 577).
{3ov, (B 968, LI 551, E 3183), YAavg (r 610, 615), K6pULOV (K 3663), fLEALuua (M 717), 1TEA€KV,
(H 515, n 1313), UlAcf>LOV (B 350), X€AWV'T} (K 495), and possibly also r1T1TO, (1848).
29 For information pertaining to the Greek coins I am deeply indebted to Dr Margaret
Thompson, Curator of Greek coins of the American Numismatic Society, who was kind
enough to answer my queries on the subject.
278 NEW FRAGMENTS OF ANCIENT GREEK POETRY
Thompson has suggested a possible corruption of the word XIMAIPA
to XIAIArPA. To quote from her letter, "It is certainly strikingly
close. In that case the coinage is undoubtedly that of Sicyon. where
the Chimaera is an almost invariable obverse type. This was a very
extensive coinage, comparable to the turtles of Aegina, the owls of
Athens and the colts of Corinth, and might well have been given a
popular designation." The emendation is indeed compelling and I
adopt it here without hesitation.so Since, however, the word XLAHiypa
cannot be entirely dismissed. I am inclined to think that we have here
an example of the conflation of two neighboring glosses into one,
caused by the similarity of their lemmata. It is a mere coincidence, but
a very characteristic one, that an entry xll-'aLpa (this one a genuine
Cyrillean gloss) appears in the Matritensis immediately after the entry
XLAuxypa. A comparable fusion of two successive entries is shown in
the entry cpaKTov (see below). Unfortunately Hesychius has preserved
no trace whatsoever of the second gloss, and consequently we lack
any decisive proof for this theory.
6 ifJ a A a a a w v (ifJaAa.,)aov Z)· KLVO.,)/LEVOS, EOCPOKAfjS 'AA£gavSp a,pavuTos.
The simple verb ifJMaaaw had not been attested before the Hellen
istic period, although 7Tpo,paA[ag?7sJ occurs in Soph. Ichn. 241 and
v7ToifJaAauu£T£ is found in Ar. Lys. 84. Note also that the verbal adjective
aifJaAaKTos is attested in Soph. fr.550 P., Ar. Lys. 275 and Crates Com.
fr.46 K. Our entry offers two more instances of the simple verb from
two fifth-century tragedians, each of whom used it in a different
sense.31 The explanation KLVO.,)/L£VOS, which is not listed in LS1 but is
confirmed by Hesychius' E,paAagaTo (E 7693, translated ;,pavu£v and
10 Other possibilities cannot, of course, be entirely discarded, since the evidence available
to us is indeed very slim. Hesychius offers a striking example, which shows how absurd the
reduction of an already abbreviated passage can be. The word Kpu.'TTu:raJ..o. (K 3971) is
explained as follows: 1Tapa 1TO,uOt. & p,wpo •• ~ vop,IC1p,a. It is only with Pollux' assistance
(9.83) that the nature of that strange "coin" is fully revealed.
81 For the vocabulary shared by Sophocles and Ion, cf T. B. L.Webster, "Sophocles and
Ion of Chios," Hermes 71 (1936) 268ff.
MARK NAOUMIDES 279
EKtv-rJ87J) , suggests that Sophocles used it as an intransitive verb.32
Hesychius' KtVEL (s.v. if;aAauUEt) is not decisive, since his is clearly a
composite entry compiled from two or more related glosses. It is at
least conceivable that in Hesychius the voice of the verb was changed
to active so as to agree with that of all the other verbs in the explana
tion. The second meaning, here ascribed to Ion, is that of a transitive
verb and is more common (cf also (nToif;aAauuw and &if;aAaK'TOS). The
use of the aorist infinitive (if;avuat) in the explanation seems to suggest
that Ion used it in that form, but it is also possible that the form
€if;aAaga'To (€if;aAauua'To cod.) in Hesychius may have been the one used
by Ion.
Our entry contributes nothing towards a better understanding
of the plot of either of the two plays.
7 if; E cp a t a s v v K 'T 0 S • UKO'TEtVfjS (UKO'TtVfjS Z), EOCPOKAfjs TpwtAq>
(ZC, 'Tpw'iAw Z). ypacpE'Tat DE Kat if;Ecpavyovs (if;Ecpavyovs
Z) ws AVKOCPWS (sic, fortasse AVKOcppWV).
cf h, m: if; E cp a { a S v v K 'T 0 S • UKo'Ttv6s.
Hesych. 1]1 134: if; E cp a t a s v v K 'T 6 s . UKO'TEtVfjS'.
1]1 135: if; E cp a v y 0 fJ S' • UKO'TEtVfjS'.
Hesychius has clearly split the entry into two while omitting the
reference (or references). As with XtAtaypa, this is the only occurrence
of the word outside Hesychius.
As is clear from the construction of the explanation, if;ECPavyovS was
a varia lectio for if;Ecpatas, and therefore this gloss must have originated
ultimately, like the entry cpaAav(Jov, from a commentary. The end of
the entry, which attempts a justification of the other form, is un
fortunately corrupt. So far as I can see, there are two ways to explain
the unintelligible AVKOCPWS; either an authority was quoted for the
rare form if;EcpavyovS', possibly Lycophron; or the form was explained
on the parallel of the adjective AVKavy~s 'of the grey-twilight', which
the scribe subsequently confused with the more familiar word AVK6cpws
'twilight'. In view of the practice of abbreviating the names of autho
rities in the grammatico-lexicographic works, I am inclined towards
the first possibility. It should be noticed, however, that whereas
Lycophron is known for his work on comedy, there is no evidence
for any critical work of his on tragedy.
as The intransitive sense of the verb may have been the original one. Cf E. Boisacq.
op.cit. (supra n.ll) 1074 s.v . .p&>J.w.
280 NEW FRAGMENTS OF ANCIENT GREEK POETRY
The lemma as it stands forms part of an iambic line (most likely
~ - t/;Ecpalas VVKTOS - ~ -" ~). It is not necessary, however, to
assume that the words stood in the text in the same order and
proximity as in our entry. The reference to the dark night is con
sistent with the story of the ambush of Troilus by Achilles.33 As to
the context from which it was taken, one could suggest a number of
possibilities, e.g. in the planning of the ambush or in a messenger's
speech announcing the circumstances of Troilus' death.
In addition to the fragments of ancient authors our Ms contains
a number of dialectal glosses, of which the following are, to the best
of my knowledge, new in the sense that their dialectal origin has not
been attested by any other source.
8 " \ ~ \,~ ~ M ~ I
o 7T tao v . TO T"f}S yvvatKOS awo LOV, aKEOOVES.
cf h: () 7T tao V' '7}T( Et}.
Hesych. 0 1020: 0 7T tao v . AeXXavov aypLOv. ~ TpWKTOV. ~
t a7TtaTov.
The reading of h suggests that its scribe found the explanation miss
ing in his exemplar, which apparently omitted it for reasons of mod
esty. It becomes therefore clear that h was not copied directly from
Z. On the other hand m left the entry out altogether, depriving us of
any more precise evidence of the degree of its relationship with h.
The word 07TtaOV is not listed in LSj. Its genuineness, however, is
confirmed by Hesychius, who in turn does not record the meaning
preserved in our Ms. There is no doubt, however, that a word indicat
ing a type of vegetable or plant could also be used metaphorically
for the membrum muliebre.34 The sexual connotation may in fact be
responsible for the apparent disturbance in the latter part of Hesy
chius' explanation.
9 ,/.. I I "A I 1\ 'A \ ( \ 1 .,., a K T 0 V . ,.U.TpOV 7Tapa pKaatV, KOTVl\at T7"£Kat KOTVI\EaTTEt
Kat Z) TpE LS. cp a V A W S • KaKws. EVtOt OE oAoaXEpws,
cf h, m: cp eX K T 0 V . ,d.TpOV.
Hesych. rp 76: cp p a K TEL v . CppeX7"7"EtV. cp p a K T 0 S
\ .
yap 0
,/.. , '" ,/... , .,.,paYfLos. Kat TO fLETpOV .,.,aKTOV.
cp 74: cp eX K Tat' A7Jvol, at7Tvat, 7TVEAot.
sa Cf Dio Chrys. 11.77-78 and A. C. Pearson, op.cit. II (supra n.21) p.253ff.
a4 Cf. the similar case of u£>"tJlOJI recorded in Hesychius and Photius. For more examples
see E. A'I]Jlalov (Ch. Charitonides), 'A1T6pp'I]TCI. (Thessalonica 1935) 20ft".
MARK NAOUMIDES 281
The form cpaK'rov points toward the adjective cpaKwTos, which was
given to lentil-shaped bottles. Apparently the measuring unit referred
to in our entry had (at least originally) such a shape. On the other
hand, Hesychius' cpaKTaL (([> 74 Schmidt) supplies additional evidence
in behalf of the genuineness of the form cpaKTov, which is otherwise
unattested.
The latter part of the Ms entry is clearly an independent gloss that
was confused with or absorbed by the entry cpaKTov. This gloss does
not appear in other Mss of the Lexicon Cyrilli and must, therefore,
have found its way to our Ms from the same source as the other
entries peculiar to Z. Hesychius supports such a view by offering a
close parallel to the latter gloss (cpavAws· OAOax€pWS Ka~ T~ ofl.ota, ([>
249). The fusion of the two entries into one indicates that the source
from which the interpolated glosses were taken listed the entries
continuously and not in a line-by-line arrangement as was common
in the oldest Mss of the Cyrillean lexicon.3s
Two more entries, although already known from the Mss related
to the Matritensis, deserve to be discussed here more fully, because
of the special problems which they present and because the readings
of our Ms throw new light on these problems.
10 A 0 1T a S (ZC, A01Tas Z)-l:vpaKoaWL TO T7Jyavov (ZC, Tlyavov Z). 1Tap~
SE eE01TOf1-1TCfJ (ZC, (}W1T0f1-1TW Z) EV 'AOf1-7JTCfJ Yj aopos
(awpos Z)- Ka~ 1Tap~ TOtS KWf1-£KOtS. KaAEtTaL (AaAE'iTaL Z)
SJ OVTWS Ka~ 0 EV Tn t EAAaDL (ZC, €AAaD1] Z) YLVOf1-€VOS
(y€vof1-€VOS Z) AL(}OS.
f h \ / ~ / ,/ \ ~'a / • C • : 1\ 0 1T as· L.JVpaKOVaWL TO nyavov. 1Tapa OE It!I€01TOfl-1TCfJ €V
'AOf1-7JTCfJ Yj aopos. Kai. 1T€Pi. TO'iS KWfl-LKO'iS. KaA€'iTaL SE
" ,t. ~'\\/~ / \/() OVTWS KaL 0 €V TTl a/\l\ao1] YLVOfl-€VOS I\L OS.
Suda A 674: A 0 1T as· 1Tap~ l:vpaKovaLOLS TO T7Jyavov· 1Tap~
~, a / • / , \ ~ ~ \ A ~,
OE It!I€01TOf1-1TCfJ 1] aopos, Kat 1Tapa TOLS KWfl-LKOLS. Kal\€LTaL O€
" \ <, ~ J..cfo,. Our Ms dispels any doubts as to which
reading is the «genuine" one, i.e. the one that goes back to the com
mon source of all three lexica. The reading is nonetheless erroneous
and ought to be emended. A. Meineke's correction of €V fE>J..&.o, to
€V fH't..,a{qf36 was undoubtedly suggested by the Aristophanean pas
sage (Vesp. 508-511) quoted in the Suda immediately after t..{f)O!).37
It is, however, completely unwarranted. Far superior is Koraes'
emendation of the Suda passage to 0 €V Tfi €'t..a{~ YWOf.LEVO!) ~AO!),38
which is amply supported by Theophr. HP 4.14.3. A slight improve
ment can be effected by substituting for €Aa{~ the form €Acfa, which
not only is the Attic form of the word and the form used by Theo
phrastus,39 but also explains better (palaeographically) the change
of EAAAI to EMAI1I (possibly through an intermediate form
EAAAAI).40
The identification of the writer cited in the gloss with Theopompus
Comicus cannot be seriously challenged. F. jacoby's41 preference for
the historian Theopompus dates from a period when the testimony
of h was not known, while Edmonds42 has apparently overlooked the
inconspicuous reference to the playwright in Adler's edition of the
Suda.
36 Cf Philologus 13 (1858) 508-9. The correction was adopted by M. Schmidt in his ed. of
Hesychius.
37 There can be no doubt that the quotation from the Wasps is not a part of the original
gloss, but was added by the compiler of the Suda, who freely inserted quotations from the
extant comedies of Aristophanes. Cf V. Coulon, Quaestiones criticae in Aristophanis fabulas
(Diss. Philo!. Argent. XIll.I, Strassburg 1907) esp. 9-10 and 265; A. Adler, "Suidas," in RE
4A (1931) 698. Meineke apparently missed the broad joke effected by the playwright's
fanciful metaphor and interpreted the Aristophanean passage too literally and prag
matically. Besides, he left the word 'YLV0I'EVOS in the Suda unexplained.
3S Cf A. Koraes, BEVOKpa:rovs Ka~ ra.A7Jvov 7TEP~ rijs a7To 'T'WV lvtJ8pwv 'T'po¢>~s K'T'>'. (Paris
1814) 156.
89 The form l>.cfa is also used by Hesychius; cf s.v. 'YEP'YlpLf.'Os (r 412).
'0 The tendency of the scribes to duplicate letters is well attested. Duplication of lambda
occurs frequently in the Mss, e.g. in the word 'A7Tpl>.LOS (spelled 'A7Tpl>.>.LOS).
H FGrHist. 115 F 408.
n The Fragments of Attic Comedy I (Leiden 1957) p.876 (fr.92).
MARK NAOUMIDES 283
There can be little doubt that Photius' second gloss is related to
the one discussed here. The reading T~V (}EOV is simply a scribal error
for T~V (J'opov effected by the proximity of the word eE01TOP.1TOS.
Photius in fact, who clearly draws here from a different lexicographic
source, seems to have preserved the exact form in which the word
occurred in Theopompus, although he has omitted the name of the
play.
The fact that in three of the four glosses43 in which the word A07TCXS
is equated with the word (J'opos reference is made to Theopompus,
together with Pollux' silence about such a meaning of the word,44
seem to suggest that Theopompus was the sole source for such a mean
ing, or at least that this meaning was extremely rare. Hence the phrase
Kat 1Tapa TOLS KWP.LKOLS in the Suda and in our Ms (both of which
clearly draw here from the same lexicographic source) seems to me
extremely suspicious. Since on the other hand the meaning 'frying
pan' is attested from at least two comic writers,45 we may speculate
that a transposition of the phrase Kat 1Tapa TOL') Kwp.LKoL') took place at
some stage of the transmission of the gloss. In that case we would be
justified in transferring the phrase before the reference to Theopom-
t d t"' I 1/ 1\,. ,. I pus so as 0 rea : "c.,vpaKO(J'WL TO TYJyavov· KaL 1Tapa TOLS KWP.LKOL').1Tapa
DE eE01TOP.7TCtJ KTA.46 The dislocation may indeed go back to the original
compiler, who in excerpting and perhaps compressing the work
presumably of the lexicographer Pamphilus (see below) committed
an occasional blunder. Our entry offers another instance of an error
which can be detected with the assistance of Athenaeus. The reference
to the Syracusans for the use of the word Aomxs in the sense 'frying
pan' (T~yavov) is indeed contradicted by the corresponding passage
in Ath. 6.2298, who seems to draw here from Pamphilus and who
unmistakably states the opposite, i.e. that the Syracusans used the
word T~yavov not for a frying pan (the sense which the word com
monly had in Athens) but for the dish or platter for which the Attic
employed the word AomXs. A hasty reader, however, could have mis
understood Athenaeus.
43 The fourth reference found also in Photius seems to be of the same origin as those
discussed here.
"Pollux lists a number of words used in reference to a coffin, such as ClOPO" 7TV€AO"
Kt{3WTO" A"7VO, (cf 7.160,8.146 and 10.150).
U Cf Eubulus fr.l09 K. and Archedicus fr.2.4 K.
46 In view of the general structure of the entry, the reading 7Tapa kvpaKOVCllot, in the
Suda may be considered as that of the original gloss.
284 NEW FRAGMENTS OF ANCIENT GREEK POETRY
11 X a p- a L 7T E T W S . waTE p-~ Els To';;Sa4>os ( TooSa4>os Z) p'it/JaL
(ptt/JaL Z). AlaxuAos rAaOKCfJ lIoTVLE'i.
ef m, h: X a p- a L 'IT E T W s ·AlaxoAos yAavKo7ToTVE'i.47
Hesych. X 135: X a p- a L 7T E T W ~. waTE p-~ Els To';;Sa4>os
pit/JaL.
M. Schmidt, in his edition of Hesychius, bracketed the puzzling
negative P-7] of the explanation, while suggesting as an alternative
that the original entry was p-~ xap-a£7TETws. Our Ms confirms the
authenticity of the reading of the Marcianus and calls for a new con
sideration of its meaning.
While the second component of the adjective xap-aL7TET7]S is derived
from the verb 7Tt7TTW . Since, however, the semantic difference between 'falling
down' and 'being thrown down' is slight (the one indicating the result, the other stressing
the process), one should not press the point too far.
48 H. Frisk, Griechisches etymologisches Wlirterbuch (Heidelberg 1954-), derives the adjective
-TTt7'TJ> from TTtTOftat and -TT~n/> from TTtTOftat as well as TTtTTTW (cf. 522, 543).
50 Cf W. Schmid, Der Atticismus in seinen Hauptvertretern IV (Stuttgart 1896) 660ft'.
MARK NAOUMLDES 285
containing the new fragments. All, except aJLwoyt-TI'ws which is termed
IrI]TOpLK~ MgLS by the Etymologicum Genuinum and the Etymologicum
Magnum, are found in the latter half of the lexicon. Since their pre
sence in the Ms bears on the question of the origin of the new frag
ments, I edit them here, indicating briefly their closest parallels in the
extant lexicographic works :51
1 < , (' I Z) ~ ,,, 0'" a JL W 0 Y E 7T W S aJLwayE7Tws . TOVTO EV LOlp, Ka ovnva
(onva Z) ovv Tp07TOV. €VLOL (€VLOt Z) OE TO JLETptWS. 7Tapa
llA&'TWVt [Charm. 175c etc.] TO JK 7Tav7' 6 V' ava{oOTJToV. JLETacf>OptKWS TO pot'ov (pOL'OV Z) JL~
7TOLOVV' uKWcf>0V (KWcf>OV Z) yap fNAOS avopo~ av&'AKt8os
, ~ A" [II 11 390] , \ ~ \ - I ...~, " OVTtoaVOLO . . . E7T' OE TOV KVJLaTO~' ws 0 OTE
(0' TE Z) 7Topc/nJPTJ (7TOpcf>VPEt Z) 7TEAayo~ JLEya KVJLan
Kwcf>0 (KWtcf>WL Z)" [no 14.16J, 1'0 JLTJ8E7TW (JLTJ8E7TWt Z)
Kax'\&sovn dpxoJL,fvcp (dpXOJLEVO' Z) DE fL€yaXovw(}aL
aIji6cf>ws.
Also in h. Cf Etym.Gud. 359.6, Hesych. K 4902, Apollon.
Soph.427-8.
3 K W X E V 0 VOL' oXOVOt, JLETEWpl'ovow' EOcf>OKAfj~ KafL'KO'iS,
"moTot JLE KWXEVOVaLV Jv cf>opq. (Evcf>opaL Z) 8,fJLa~"
[fr.327 P.].
Also in h. Cf Hesych. K 4905, Etym.Gud. 360.5.
4 K E 7T cf> 0 ~. 0PVEOV KOvcf>OV, W~ • ApLOTOTEATJ~ cf>TJo{v [HA 593b
14, 620a13], 7TEpL T~V B&'Aaooav Tpl{3ov. alOE A&.pav (Aapav
Z).
Also in h. Cf Hesych. K 2242, schol. Ar. Pluto 912 (ed.
Diibner, p.372), Etym.Mag. 504.1, Zonaras 1183.
5 v v v 8 E 0 E a L JL &. K apE ~ • TOVTO cpaotv eg6otov E tvat Tpaycpowv
(lege paljicp8wv)- "vvv OE BEoL JL&.KapE~ TWV eoBAwv ac/>wvol
(lege acpOovoL) eOTE."
Also in h, m. Cf Hesych. N 730, Ael.Dion A 76 Erbse.
51 For breYity's sake no distinction is made between Z and zo, except when some parti
cular problem is involved.
286 NEW FRAGMENTS OF ANCIENT GREEK POETRY
6 g v I-' rp 0 p eX • uvvTvx{a ( uvvTvxda Z). Kal. €1TI. aya80u TeXaa€Tat, ws
1Tap' AlaxvAcp EV Kafl€{pots [fr.49 Mette], Kal. E1TI. KaKOU
1Tapa EOrpOKAd.
Also in h, m. Cf Hesych. 8 114.
7 0 A € 8 pO v . €VtOt (EVtOt Z) 1Tapa • AptaTorpaV€t [cf fr.320.3 K.]
8
'\ , '(' Z) , , ~ "" \ €1Tt Koal-'0V yvVatKHOV yvvaLKtOV , ov KaI\WS. a/\/\OL TOV
fla8vv (flaOv Z). €aTt OE OA€OpOV. a1TOOVCT1T€TOUVTa yap
€l1T€'iV TOV 1Tapa T0 (TO Z) 'AptCTTorpaV€L' OA€OpOS yap <>
8avaTos.
Also in h, m. Cf Hesych. 0516.
" '\" , (' Z)'~ o I-' a v 0 v . 0l-'0KOLTOV, aVYKoLTOV aVVKOLTOV , Of.J.OV
I-'EVOV' EOrpOKAijS LVEL (rp{VEL Z) [fr.717 P.].
Also in h, m. Cf Hesych. 0 681.
9 ''''..I.. t \ 'A '''' \ , \ "" < > 1T 0 P K as' El\a. 462, Etym.Gud. 553.37.
15 cp W S' • dgv'TovwS' fLEV av( BPW7T )oS' a7TO TOV CPWTttELV T0 '\oy4-' 7TavTa,
7TEPL07TWfL€VWS (7TEPLa7T0J.L€VWS Z) 8E TO 7TVP Kat fLE'Ta
cpOpLKWS ~ xapa. oiov' "Tpwwv pijgE (ptgE Z) cpa,\ayya,
cpows (cpows Z) 8' £'TapOLatV EBTjKEV" [II. 6.6J.
Also in h, m, abbreviated. Cf Etym. Gud. 560.38, Apollon.
Soph. 701-2, Hesych. l/> 1119.
16 X t 'T W V (XLTWV Z)· E7T!. fLEV 'TOV ov~eovs, "TOV 8E (Tov8E Z) XL'TWV'
EvoTjoa 7TEpL Xpo t (Xpw t Z) oLya'\oEvTa (ZC, ya'\oEvTa Z)"
[ad. 19.232J. E7T/' Oi 'TOV (JwpaKos ((JopaKoS' Z) "'EKTOPEOV
8€ Xt'Twva (XEt'TWVa Z) 7TEP'l, a'T~eEa(n (aT~(JEUtV Z, O'T~
BEaatV ZC Y' [II. 2.416J.
Not in It, m. Cf Etym.Gud. 567.1, Apollon.Soph. 711-2,
Hesych. X 486; cf. also Etym.Mag. 812.9.
17 X ,\ ij S 0 S' • apoEvLKOV (apOEVLKWV Z) Kat. 7TEpL07TWJ.tEVOV (7TEpumo
J.tEVOV Z). KVptWS' SE 0 OwpOS TWV '\{Bwv. KpaTrjs Sf
288 NEW FRAGMENTS OF ANCIENT GREEK POETRY
Tp07T'£ICW~' .. apyvplov X>.fjSov (x>'7]Sova Z. x>'7JS6va ZC)
>'af3wv (Zc, >.af36v Z)" [fr.28 K.].
Also in h, m, abbreviated. Cf Etym.Gud. 567.43, Hesych.
X 510, .EvvaywYl7 419.6.
18 X >. t S W V (XAv8wv Z)- xv87Jv, uwp7JS6v (uopLS6v Z). E7T't (ZC, [ ... ]
Z) yap (fort. /L~V) 7T'A7]8ov~ E/Lc!>a.UEWS, cfJs Aluxv>.os
"07T'AWV KpluEL [fr.290 Mette]. Kat X>'LSWVTa (XE>'LS6vTa
Z) aVTt TOU 7T>'7JOVovTa.
Also in h, m, abbreviated. Cf Etym.Gud. 567.34, Hesych.
X 509.
As with the entries contammg the new fragments, the closest
parallels to these glosses are almost exclusively found in the lexica of
Hesychius, Apollonius the Sophist and the Etymologicum Gudianum.
The agreements with Hesychius are especially noticeable both in
number of parallels and degree of closeness and are indeed of special
importance for determining the source. Since it is known, on Hesy
chius' own testimony, that he has absorbed the dictionary of Apol
lonius the Sophist through the medium of Diogenianus' lexicon, and
since our Ms agrees with Hesychius whenever its text deviates from
that of Apollonius, it becomes clear that the latter was not the direct
source of the pertinent glosses. The relationship with the Etymologi
cum Gudianum is also very striking. Since, however, the archetype of
that dictionary is contemporary with, if not later than the Matritensis,
it cannot have influenced the latter. At any rate, the independence
of our Ms is convincingly shown from such entries as, e.g., c/>oiv,g and
XE'p0f30UIC6~. It becomes clear, therefore, that both the Madrid Ms
and the Gudianum drew independently from a common source. This
source, which was very closely related to Hesychius, must have
reached southern Italy some time before the supposed date of the
Matritensis (i.e., roughly speaking, ca. A.D. 1000).
So far the tendency among experts in the history of ancient lexicog
raphy has been to attribute such glosses to the lost lexicon of Dio
genianus52 (the main source of Hesychius' lexicon), which is occa
sionally quoted by the Byzantines down to the XII century.53 The
argument for Diogenianus as a source for such glosses rests on the
U Cf. R. Reitzenstein, op.cit. (supra n.1) 3; and K. Laue, Hesychius (supra n.7) pp.x-xi.
51 Cf. Lane, Hesychius pp. xlii-xliv.
MARK NAOUMIDES 289
observation that the lexica which quote him display a number of
glosses (anonymous to be sure) which are similar to, yet more exten
sive or complete than the corresponding entries in Hesychius. This
argument, however, may be countered by the following considera
tions; (1) the express testimony of Hesychius himself, who, in his
dedicatory letter to Eulogius, states that he has absorbed in toto
Diogenianus' dictionary and that the latter lacked eTTLypacpdS, i.e.
references to the sources of the glosses, those found in Hesychius'
lexicon having been supplied by Hesychius himself; (2) the complete
agreement between Hesychius and Diogenianus in the entries in
which the latter is expressly referred to and the absence of any refer
ence to him in precisely these glosses which are used to support the
theory of a fuller Diogenianus; (3) the testimony of PSI 892 (Pack2
2125); this papyrus, dated earlier than Hesychius by at least one cen
tury, has preserved a fragment of a lexicon much like that of Hesych
ius, yet entirely bare of quotations or references and for this reason
rightly attributed to Diogenianus. To counter these objections, the
exponents of the theory of a fuller Diogenianus have postulated the
existence of several versions of that lexicon, one of which was much
more extensive than the other. Some even have gone so far as to
theorize that Hesychius supplied his own dictionary with quotations
and references he found in the fuller version.54 Despite this argument
I believe that the balance of probability is against the existence of a
fuller Diogenianus.
On the other hand it is well established that Hesychius' text, as
we know it from our unique XV-century Ms, has been interpolated
and (at least occasionally) also abbreviated.55 It seems, therefore, con
ceivable that our interpolated glosses may have been taken not from
a supposed fuller version of Diogenianus but from a somewhat fuller
version of Hesychius' lexicon, one that contained longer entries as
well as more references and quotations than the Marcianus. This
possibility, which I advance here as a mere hypothesis, receives some
strength from the well-established fact that a copy of Hesychius (an
ancestor of our XV-century Ms) was in the vicinity of southern Italy
(where the Matritensis originated) from the time of the Arabian occu
pation of the neighboring island of Sicily.56 On the other hand, there
54 Cf esp. Latte, Hesychitts pp. xi-xii.
56 Examples are given by Latte, Hesychius pp. xvii-xviii, xxi, xxiii-xxiv.
58 Cf Latte, Hesychitts pp. xxii ff.
290 NEW FRAGMENTS OF ANCIENT GREEK POETRY
is not the slightest evidence for the parallel existence of a copy of Dio
genianus' lexicon in that area. One may even go so far as to conjecture
that Hesychius was occasionally referred to as Diogenianus, on
account of the express and unreserved acknowledgement of his debt
to his predecessor. This would explain not only the existence of
the supposedly "Diogenian" glosses occasionally found in Byzantine
lexica, but the almost complete absence of references to Hesychius
before the end of the Byzantine period.57
UNIVERSITY OF ILLINOIS
May, 1968
57 Latte, Hesychius p. xxi, mentions only one reference to Hesychius besides the two found
in the scholia of Arethas to Pausanias (cf Fr. Spiro, "Pausanias-Scholien," Hermes 29 [1894]
143-9).