Elhanan Winchester.

Herondaea online

. (page 1 of 4)
Online LibraryElhanan WinchesterHerondaea → online text (page 1 of 4)
Font size
QR-code for this ebook


Shelf No






Boston Public Library

Do not write in this book or mark it "with pen or
pencil. Penalties for so doing are imposed by t
Revised Laws of the Commonwealth of. Massachusetts.
This book zcas issued to the borrower tin the date
last stamped below.


FORM NO. 609; 10.2.34; 224M



John Henry Wright


Vol. IV., 1893.








By John Henry Wright.



I. The Spaces. — It is an important peculiarity of the papyrus
manuscript of the Mimiambi of Herondas recently discovered (Papy-
rus No. CXXXV., British Museum), that, while as a rule the letters
of the several verses are written continuously, without break or pause
between the different words, now and then — in about twelve per
cent of the verses — slight breaks or blank spaces do occur, never
amounting however to more than the space ordinarily taken up by
from one to two of the letters of average breadth. The significance
of these breaks for the punctuation of the text was first emphasized
by Blass, and has been recognized by several critics of the poet;
but thus far only sporadically. In this article I propose to present
all the examples, and to discuss the doubtful ones, not neglecting at
the same time the examination of a few related topics, important in
their bearing on the text-criticism of our author. 1

It should be observed, in the first place, that these breaks are
never intended to mark words as words, nor to suggest the proper
combinations of letters into words in ambiguous instances : this work
is performed, but without system, incompletely, and only very rarely
directly, by the marks of the rough breathing, 2 the accents, 3 the coro-

1 For convenience ordinary type will be regularly used in the notes for the
readings of the manuscript. It is to be regretted that in the text the font of
inscriptional type so imperfectly represents the cursive majuscules of the papyrus.

2 Only the rough breathing is written, and always in an angular form ( H ),
except in the late ovd'iv, VI. 3. The cases are : II. 70 (upayrjcr) ; V. 20 (orev-
kk); VI. 25 Qf) BitcLtoct); VI. 68 (d/xt[\X]7;); VII. 46 («).

3 The accents, acute, circumflex, and grave, exhibit interesting peculiarities.
The circumflex and acute are used with many proper names, but not with all ; also
to distinguish between words spelled alike but differently accented, and to indicate
the correct grouping of letters into words, etc. : e.g., I. 29 (dial, not deal); I. 85


170 John Henry Wright.

nis, 1 and perhaps once or twice by the use of the dot or point. 2 The
last, however, has a more extended use — in perhaps a dozen exam-
ples, collected below — as a sign of punctuation (ony/x^), having here
a value not wholly unlike that of the spaces, but not so strong as that
of the Trapaypa.<j>os.

Punctuation within the verses is indicated mainly by these spaces :
indeed, these spaces have no value except as signs of strong inter-
punctuation, and they always have this value, when not accidentally

(/na, not /*d); ibid, (oaaov = os gov, not 'oggov); IV. 42 {avr-q = olvtt], not
avrri); II. I (ko-ri = iari, not eVre). At V. 41 (o5?j) perhaps the accent may
indicate odrj. At II. 9 (^/xeatr) the accent seems to suggest the synalcepha of the
final syllable (Crusius) : and in V. 49 the acute on the ultima (aKrjKovKas) may
be intended to indicate a rising tone of voice, necessary in a question (Diels).

The grave accent regularly appears to be used to provide against misapprehen-
sions. The examples are I. 60 (raraXi . . . : to show that this is not rara; it
also shows that we have a longer word here than rara) ; I. 76 (Jlvdew 5£) ; I. 70
(\avayrjs: 6 + dvay^s, not wv /crX.); II. I (J=gt£: tare, not eVre) ; II. 24 (fyi' :
ep.t, not £/J.a); III. 74 (irtpvas : i.e. ivepvas, not irepvas); IV. 91 (jrkXavov : i.e.
ivekavbv) ; VII. 46 (hi: perhaps taken for article — wrongly; hardly "aporiae
indicium," Crusius).

1 The coronis ('), usually written at the top of the line (at I. 15 and II. 83, at
the bottom), always appears to indicate elision at the end of words. It seems to
have been put in by the first hand, except at VI. 3 (ovd'ev). The other cases are :
I. 15, [ivifiaov = /JLVia 8gov; II. 24, t/u? ov = ep!ov = ip.Z ov; III. 49, KaXyQiv' =
K&Xrjdlva; IV. 5, n'uvTrep = KiSvirep — kcll wvirep; IV. 1 6, aXeKrop' i-qrpa = dXeK-
rbpa irjTpa; IV. 41, KvdiW'iovaa = KvdiWa lovaa. The only doubtful case is II.
83, KavTo<TTa<r,avTov : this must be ical avrbs ra era avrov, hardly rd creavrov, cer-
tainly not, in the light of all our examples, rds avrov (Doric short a) . The cases
show that the mark is not necessarily used, as in the Codex Alexandrinus, to sepa-
rate words as words, nor like the Siao-roX^ (yirooiao-ToX'q) : orav diaa-recXai Kal
diaxupl<rai 6(pet\u)p.iv riva \i^LV {e.g., iariv, d^ios, not ecrrt Nd£ios : Bekker,
Anecd. Graeca, II. p. 675) ; some of the examples, however, might be explained
as instances of the diastole: e.g., II. 24 {hfxov, not p.ov), III. 49 (KaKTjdiv', not
KaXijO'iva), etc. Cf. Gardthausen, Griech. Palaographie, pp. 273 f.

2 In Proem, n (Cr.) TaKvXX-aidiv, the point appears to be intended only to
mark off the words ra kvXXo. and aeldeiv : it can have no force for punctuation
here; cf. IV. 50 (e<T(reT-7)p.epa), but see p. 182, note I. (In I. 3 [rts - ] the dot is
merely a part of the sigma : likewise at II. 6 [/cVavcrcu], the mark above the X is
part of an unfinished a, begun too near the X; cf. I. 51 ; at VII. 48 [3/cws], the
mark on is part of a <p in the preceding line.) For some remarks on the use of
the signs (" , ") see below, pp. 177, 178 notes.

Herondaea. 171

An examination of the photographic facsimile of the manuscript
discovers about one hundred and ten cases of such intentional spac-
ings. Of these not more than from two to six are in any way ambig-
uous, and a fair consideration, it seems to me, would deny ambiguity
to all. 1 The remainder, over one hundred and four, are nothing but
indications of punctuation, for which we have in our modern editions
our various signs. A classification of these examples according to
the punctuation adopted in the latest text-edition of Herondas (Bib-
liotheca Teubneriana : ed. Otto Crusius, 1892), which errs by no
means on the side of excessive punctuation, yields the following
groupings : —

a. — The breaks accompany a change of speaker- in the dialogue at
I. 7* (io-Tiv, TvXXts), 20, 82 ; II. 48; III. 58, 78*, 81, 82, 87, 93 ;
V- 31, 55 ', VI. 15, 17, 19*, 22*, 23*, 25 ; and are thus represented
by our period, colon, dash, or question-mark (the last indicated by
the *) .

b. — They stand at the end of a question in the examples starred
above, and in the following additional cases where no change of
speaker occurs after them : I. 9, 48 [?] ; III. 43, 60; IV. 57 (per-
haps an exclamation, ola epya;) ; V. 10, 18, 41, 75 ; VI. 10, 44, 45,
75; 76 ; VIII. 4, 5. In these cases they are represented by a

c. — Many have the value indicated by Crusius by a period: most
of the unstarred cases under a, and the following additional examples :
I. 8, 79 ; II. 68; III. 59 ; IV. 33 ; V. 20, 56, 66, 67, 74 ; VII. 4, "7-

d. — They have the value of a colon at I. 15, 66, 82 ; III. n, 26 ;
IV. 21, 55, 92, 93; V. 6; VI. 5, 31, 61; VII. 65, 128; VIII. 11;
and of something like it at IV. 58.

1 The apparently exceptional cases are discussed below; see pp. 173 f. Occa-
sionally, but extremely rarely, when the large bulk of the writing is considered, we
find other slight breaks. In most of these instances the letters of a verse have
been written more sprawlingly than usual, and thus give the appearance of spac-
ing where no pause is intended. I have observed only these examples : ya.pfip.ewv
(I. 46); irodjwv (I. 60); ixereKOetv TJv^Ovprjv (II. 50); Ke7vov / pk (IV. 30);
oi jos (V. 43); tovjov (V. 58); rjfJ.£w A v (VI. 82). (In the apparent ^tXaiv^ov,
I. 5, the letter iota has disappeared, leaving only a slight trace.) 'Such is not the
explanation of the pause in I. 55, discussed on pp. 186 ff.

2 Change of speakers is usually indicated, but with many omissions, by the
irapdypcMpos; see pp. 178 ff.

172 John Henry Wright.

e. — Crusius represents them by a comma at I. 13 {pis), 67, 89;

11. 2 2, 49, 77; III. 49, 81 (7raw-ai A u<avat) ; IV. 43, 46, 90 ; V. 9,
25, 34, 42 {rovSe^Kal en;), 53, 69 (tclti A dXXa) , 70 ; VI. 3 (curnyi^cri;) ,

12, 18, 49, 77, [96, after ™] ; VII. 57 {bis), 58 (before Kawafilo-Ka) ,
60 (after anpoo-^vpia.) , 61 (after e<pr)f3oi), 98, no.

y. — In the following verses, where the spacings are indicated by
the sign of caret, Crusius inserts no mark of punctuation ; but no one
can deny that at least a strong phrasing, if not punctuation, was dis-
tinctly intended : II. 2 {ovk iare | ^//.etov Kpiral SrJKOvOev^ovSe 1-775
So^tis) ; III. IO {tov [xmtOov aLTCL^Krjv to. NavvaKou kXolvo-u)) ; IV. 83
{ei{A€V7]S £1775 J KaAais eV ipois TatO"8e A K£t Ttves ruivSe | lacr' oirvuqTai) ;
IV. 42 {ov o~ol AeycD A auT?7 rfj . . . )(acrKova~r] ;), and 55 {avrrj o~v K
fxeivov) ; III. 25 {rpiOr/fxepa Mapwva ypap-pLaTi^ovros | toS irarpos avTio A
tov Mapwva iTrocrjo-ev \ ovtos St'/^wva) j IV. 24 (ot^ 0/0775 Kava | eV rfj
/3ao-et A Ta ypd/xfiara ; i.e. ' Don't you see those [things] on the pedes-
tal, the letters?') ; perhaps also IV. 59 (tov 7ratSa St7 a (tov) yvp.vov).

At III. 80 there is a pause which taken in connexion with the cor-
rections at this point is extremely significant. As first copied, uncor-
rected, vv. 79, 80 read (in part): (79) ITICOIZUUHN (8o)*GP OC-
ACANHKAKHCQGNHIBYPCAI. The corrector, evidently the
first hand, having previously designated 80 as corrupt (by an oblique
line in the margin opposite ^GP ; see p. 181) sets himself to correct
it. He puts a mark of erasure over the N of ZUJH N (superior dot) ;
writes in, in the upper part of the space after 4>6P, the letters GIN ;
and draws his reed through the two I's (at C0GNHI mistakenly; at
BY PC A I apparently correctly ; probably these I's were earlier can-
celled by the copyist, as he wrote). These facts show that the original
manuscript from which the papyrus was transcribed read something as
follows, of course metrically an improbable reading :

Metrotome {to the master), « rl 0-01 ^urjv,
<pip , — ocras av 77 kcikt) aOevr) fivpera :

i.e. 'If I am anything to you (cf. V. 70), come, — [give him] all the
blows his vile hide may bear.' The corrector, however, so radically
modifies the text that it seems obvious, either (1) that the manuscript,
at the time it was copied (for the hands in GIN and <!>GP are the
same), was corrected by comparison with another manuscript contain-
ing different readings, or (2) that our copy was made from dictation,

Herondaea. 173

the scribe not distinctly hearing the words. The second alternative
cannot be adopted; see p. 183, note 2. The first alternative is sup-
ported by other inserted readings. Now the reading <pepav could not
have been in the original : otherwise the space after <p £ p would not
have been made. We infer accordingly that in the manuscript used
for correction the text read :

Metrotime {to her son) . el tl crot £<ar},
<pepav ocras av rj KaKrj crdevr) fivp&a.

' While your life holds out, you'll have to get all the blows your vile
hide can stand (or that the cowhide is good for).'

There are now left six cases, which upon first examination appear
to be exceptions to the law that spacing always indicates punctua-
tion. 1 These are I. 1 and 64 ; V. 68 ; VII. no and 118 ; and VIII.
3. (1.) In the difficult and corrupt passage I. 64 we are not shut up
to one reading ; the space after 7rp^eis favors a reading like that first
proposed by Crusius : a irprj&is, lySe' eo-ri ktX., or Blass's Soia Trprj£cis,
■fjSovrjv ktX., rather than Bucheler's or Crusius's in the text-edition,
although it is not wholly impossible with the latter. 2 (2.) At VII. 118,
the papyrus reads : YUJPH (118) APHPGNOTTAH BOYCOAAKTI-

CACYMAC. Bticheler's if/upr) | aprjpev birXrj, fiovs 6 AaKTtcras vp-as —
' scabra congruit ungula, bos pressit vos cake ' — gives excellent
sense, besides preserving the punctuation, and therefore may be pre-
ferred to Crusius's ^u>prj \ aprjpev owXr] /Sous 6 AaKTicras vp.a<; — 'Der
Ochs der euch versohlt hat, fuhrt eine raudige Klaue.' The cases
VII. no, V. 68, and VIII. 3 belong together. (3.) In VII. no (ex«s
yap^ovxi yXaa-crav, TjSovfjs K yj6p.6v), there is a strong rhetorical pause
before ovxi whereby oix<- yASo-crav becomes parenthetical. (4.) At V.
68 (^Kar-qpTiqa-Ooi ovroi Kara /avos uHnrtp $ Aaou Tiprf), the interesting
pause seems to be a rhetorical one, due to the verb that must be
supplied, of which Tip.rj is subject. (5.) In VIII. 3 (^ 7rpo<rp,ev€is cm,
p.€XP L vtVftfXios OdXif/ei I [tov kJvo-ov ecrSvs), the space is not large and
may be accidental, but a rhetorical pause is quite probable here also,
especially if we read p-e^pis ev. It is not unlikely that the passage is

1 Among the exceptions I should not include I. 55 (&6uctos is KvO^pi-qv *<r<ppTi-
yts) ; see below, pp. 187 ff. The text at VIII. 28 ([a]va\{<76ai A K7) : Frag. 2. 7), and
at Proem. 9 (SevrepTi^yv) is too fragmentary to be taken into consideration.

2 Diels proposes irp^eis, ijdiws 8r) rep<pdei<rri.

174 John Henry Wright.

slightly corrupt, and that the pause may be not original. 1 (6.) There
remains I. i (apaa-ua rrjv dvprjVfji'i' ovk oif/a, | et kt\.). 2 If we are to
adopt this punctuation as one originally intended in thought, we may
suppose that the scribe, by whom the original was made of which our
manuscript is a transcript, was misled by the form of ns : he took
the pronoun with the ov as an interrogative beginning the sentence,
and therefore spaced it off from the preceding word (cf. IV. 21 ;
VI. 18). Our scribe merely copies what he has before him.

Of the six doubtful cases, then, one is probably due to a copyist's
mental confusion ; two cease to be exceptions upon the adoption of
otherwise approved readings ; and three, if not purely accidental,
likewise cease to be exceptional if we admit the possibility of the

1 The synizesis of ev + 77, across a pause in the sense, is not an objection : cf.
III. 81 (jravvai, Uavai), and IV. 50 (p.apTvpop.ai, <pi\fxi • evaer' i/pi.e'pr) Kelvrf). But
the place is otherwise open to criticism, and from several points of view appears
to be corrupt. As it stands it would probably be better to take it as ^XP LS €V
^Xtos 6&\ipei (for ev compare VII. 123, — where read ttjv . . . ^alrrjv \ 8a\irov<ra.v
ev dec'vdov . . . ical pdirreiv — and for the position of ev compare also Dem. Cor. 144,
ev irpdyfia cvvrede'v, or Plat. Rep. I. 329 C, ev ovv /mol Kal rbre edol-ev iKelvos elwelv') ;
or, possibly, /x^XP' * ev 8d\\peu But the synizesis ev + V is perhaps too harsh to
be allowed even to Herondas. The papyrus has only one other instance, to /xeu
al/xa. (V. 7), but a similar synizesis in kereuw (III. 71) was avoided by the cor-
rector by erasing the v. And at II. 43, where /^XP' S °<5 — a not dissimilar diph-
thong, though elsewhere freely suffering synizesis — is used, hiatus is permitted
(/*expis ov etirrj). Perhaps even to /xev alp.a, in V. 7, is an analogical form, and
should be written to p,eo aTp.a : compare t£o in VIII. I (reO, II. 98) and aiu
Trp^ifts (= <rto i] Trprj^is, Cr.), VII. 96, if reference may be made to so problemat-
ical a passage. If, now, we reject the present reading because of its extraordinary
synizesis, the words will be seen to be an easy palaeographical corruption of
M G X P I CO Y H A I OC (cf- tfxp" ov eftrtf, II. 43. and & X P" V^os 8vy, II. 88), or,
since that combination is objectionable because it made hiatus at II. 43, and must
not here, MGXPICOH A IOC (cf. t[ov i)\]lov divros, II. 13). In the latter case
the CO might have been taken for £0 (Cobet, Nov. Led. pp. 1 78 f.) , and this
easily written into the more familiar GY- This process was, of course, helped by
the MG X P ICG Y> a few lines below (p.i%P L °~ ev > but perhaps m^XP is e ^)- I n this
line (VIII. 3) we can hardly make the letters = p^XP' s e ^> as an hyperionism for
M^XP' J °^> under the influence of the foregoing H^XP 1 Te ° ( v - 0*

2 Can we take dpdo-aei impersonally, and read the verse : Qpi'C<r<ra, dpavo-ei t^v
dvpTjv. tLs; ovk 6\j/ei kt\.? (Cf. Kiihner, Ausf. Gramm. II. p. 30.) It is per-
haps better, however, to explain it as above, if after all the pause be not an
accidental one, like those in o5 a tos (V. 43), and tov^tov (V. 58) mentioned at
p. 171, note I.

Herondaea. 175

use of the space to suggest merely a slight rhetorical pause, such as
was regularly indicated by the anyfir) iiiarj in the writing of the
Roman period. 1

It may, therefore, be reaffirmed with emphasis, that in the Heron-
das papyrus the blank spaces between certain words in the verses
always have the value of strong interpunctuation, and must be care-
fully heeded by all who would seek to construct the text or to inter-
pret the poet.

But the punctuation by spacing goes only a little way. If it had
been applied consistently and completely, we should have had more
nearly a thousand than a hundred cases to register.

II. ^Tiy/xaL — Punctuation is also indicated in the papyrus by the
use of the dot or point in the line. This method of punctuation has
not the significance of the former for purposes of text-criticism, since
it may be in large part the arbitrary work of later correctors or
readers, whereas the spacing must have been made by the original
scribe, and can have been only a reproduction of what he had before
him. Punctuation according to spacings may go back to Herondas ;
but that by points hardly.

Some of the points or dots in the text may be mere blots, — for we
find others like them in the middle of words, or hanging on the tips
of thickly-written letters, — or even parts of letters detached from
the body of their letters on fibres of papyrus slightly shredded off.
Again, very frequently the intentional dot does duty in Herondas to
indicate omission or erasure, and is then regularly placed above the
letter or letters to be rejected : occasionally it is also placed, both
above and below, and once in a while at the right side or on both
sides 2 of the objectionable word or letters. In the latter position it
may lead to a confusion with the use of the point for punctuation.
Actual cancellation is effected by drawing a line obliquely, or some-
times horizontally, across the undesired letters, syllables, or signs : it
is sometimes combined with omission as indicated by a superior dot

1 On the use of the n£<rt) see Blass, Griechische Palaeographie, in I. Miiller's
Handbuch, I 2 , pp. 311, 312, 323.

2 For example at I. 50 (TlaraiKiov-TvWov-), where the marginal Tpv\(Kos)
is to replace the word in the text. Possibly the point in II. 98 (after $oi/3t?) has
a like value; but its mate is not visible at the beginning of the word, nor has any
substitute or gloss been written on the margin.

176 John Henry Wright.

(see IV. 67). It is not unlikely, though not certain, that this cancel-
lation was done by the first hand, in the progress of his writing : see
IV. 83, where in GMTT the M is cancelled, and the correct TT placed
just after, though it is possible that the scribe here wrote out G MTTP
at first. The dots indicating omission were added on the revision by
the corrector, who was apparently the first hand (see on III. 80,
above, pp. 172 ff., also p. 184).

The points as distinctly used for punctuation 1 may be grouped as
follows : cases where they are by Crusius represented by periods, by
interrogation-points, by colons, and by commas. Where the any^r}
falls at the close of the verse it is designated in my list by an asterisk.

a.— Periods: I. 3 (GCWAG. ; point at middle) ; I. 4 (ACCON'*) ;
I. 8 (AOYAH-; middle); I. 82 (TTGI0I-; middle); II. 98
(<t>OIBH.; perhaps middle); VI. 5 (MGTPGUU* with H- written
above G ; unless the point here merely indicates erasure of H on
second thoughts [so Crusius, and cf. III. 62], it means that we are
to read MGTPH. with full pause, and not MGTPHUU) ; VII. 76
(TTPH5P: if this be a o-ny//,^; if a line, there are no similar uses in
the papyrus) ; VII. 113 (OUUMGN').

b. — Question-marks: I. 3(0YPHN.; interrupted question) ; I. 3

c — Colons: IV. 21 (ArAAM ATWN") ; VII. 114 (TTA5- ; mid-

d.— Commas: I. 8 (TI-; middle) ; IV. 37 (BATAAHN.).

A glance at this list shows at once the futility of attempting to
identify these points with any ancient system of o-ny/wu {nXda.,
vTToo-Tiyfxr] [and /xe'o-^]). They are inserted with little discrimination.
Thus all three are used to indicate a strong pause; the "TeAeta" at
I. 4 has very strong force ; less at IV. 21. The " v-noo-Tiyixrj " is weak
at IV. 37, less weak at I. 3, and rather strong at VII. 113. Prob-
ably the papyrus is not carefully enough written to justify us in very
nice distinctions between the " fxia-q " and either of the others, but a
difference is certainly to be observed between the top and bottom of
the line as places to receive the points. And we must also bear in
mind that some of these cases may well be those of accidental

1 The following cases appear to be accidental: in V. 21 the point under /jl of
/jlvcmt ; and the points on both sides of X in VIII. 42 (o-\ - t;i), where the sense
demands ov\r).

Herondaea. 1 77

blots. Probably some early owner of the papyrus began with the
good intention of putting the points in (Siaori&u tov "HpwvBav), but
soon gave up the task. It will be observed that eight out of the
thirteen or fourteen oriy/W are found in the first mime, and most of
these near the beginning.


THE ■7rapdypa(jiO<s AND o/SeAos. 1

I. Ilapay/aa^os. — A short horizontal line, drawn distinctly, firmly,
and usually with full reed, is frequently met with in the papyrus, and
has various values. 2 Within the verses and between the lines, where
it occurs rarely, it is placed close above certain letters, regularly
vowels : in this position so miscellaneous seem to be its functions
that we cannot speak more definitely of it than to say that it calls
attention to something noteworthy in the letters or words marked. 3

1 On this name see p. 180, note 4.

2 The sign (") is used five times in Herondas, and, as — with perhaps one
exception — it is always over short syllables, it may be identified with the sign
invented by the Alexandrine metricians to indicate a short syllable (/3paxeia,
irpwTos x/do'tos). The sign cannot be taken as a rhythmical sign, since while
ordinarily in the Apans, at VII. 108 it stands on one of the resolved feet in the
6£<jls. The cases are I. 50 (6 ~M.aTaK[i]v7js) ; I. 56 (Mt'cnjs) ; IV. 30 (jbv yipovrd ;
7T/30S Moipiuv); VII. 108 ([8v]vo.lt6 /jl iXaaai); and the puzzling IV. 62, which
has given rise to a spirited controversy (see Crusius, Philol. 50 (1891), p. 446;
Ludwich, Berl. Phil. Woch. S., 1892, pp. 642, 1349, and L. Miiller, ibid. p. 995).
Here the original draft had TTYPACTON, or TTYPA r TON : over T a P is
written and upon T and A stand the marks",". Meister's Trvpacrrpov is now
adopted by both Kenyon and Crusius. The first syllable of the word, contrary to
usage, is here metrically long : hence it is marked; it also has the acute accent.
The second sign perhaps refers to the original or natural quantity of the syllable
in irtipaypov, which — on this theory — the scribe must have thought he had before
him, in his original: otherwise the sign is unintelligible to me. For irvpaypov,
cf. wpdyprj, Anth. Pal. VI. 117.

3 The examples of this sign (") in the papyrus are the following: over iota,

III. 74 (i<7 = els) ; III. 79 (1 = el, followed by enclitic) ; V. 5 (n pocpaa'tcr =
■7rp0(pd<xeis') ; V. 18 (^4>epl(T = <j)ipei.s, Cr. : probably (pip' eh) ; Proem. 1 1 (^eTrtovcri— ?) ;

IV. 43 (p.arlv, a short vowel: perhaps a mark of cancellation?); VI. 25 (BItcltoo;
a short vowel); and perhaps in the obscurely written I. 82 (5eI£oj/=?), unless
here it be meant for the superior dot indicating erasure, the scribe mistakenly
thinking of 5e"|o. The only other cases are : over alpha, III. 79 (rara), and

1 3 4

Online LibraryElhanan WinchesterHerondaea → online text (page 1 of 4)