Digitized by the Internet Archive in 2019 with funding from Kahle/Austin Foundation

https://archive.org/details/oxyrhynchuspapyrO060unse

HBS £982.

* — ~

THE OXYRHYNCHUS PAPYRI

VOLUME LX

THE OXYRHYNCHUS PAPYRI

VOLUME LX EDITED WITH TRANSLATIONS AND NOTES BY

RAS COLES M. W. HASLAM P. J. PARSONS

With CONTRIBUDLLONs BY

G. BASTIANINI HoeM: COCKLE J PUENAER TS D. LUHRMANN Ie ois ih. GelLURNER AND

MEMBERS OF THE.ISTITUTO G. VITELLI

Graeco-Roman Memoirs, No. 80

PUBLISHED FOR THE BRITISH ACADEMY BY THE BGY PL EX PhORATION SO GIETY 3 DOUGHTY MEWS, LONDON WCIN 2PG

1094

Thomas J. Bata Library TRENT UNIVERSITY PETERBOROUGH, ONTARIO

PRINTED IN GREAT BRITAIN By THE CHARLESWORTH GROUP, HUDDERSFIELD AND PUBLISHED FOR THE BRITISH ACADEMY BY THE EGYPT EXPLORATION SOCIETY (REGISTERED CHARITY NO. 212384) 3 DOUGHTY MEWS, LONDON WCIN 2PG

ISSN 0306-9222 ISBN 0 856098 117 6

PREFACE

Much the largest contribution to this volume has been made by Dr Coles. He himself edited the great bulk of the Roman and Byzantine documents, and collabor- ated on the rest (4063-7, introd.): in addition, he has supervised and contributed to the publication of the large group of Aeschines papyri.

The theological texts (4009-11) comprise a fragment probably of the apocryphal Gospel of Peter, published in collaboration with Professor Dieter Luhrmann of Marburg University; and two liturgical texts from the everyday life of the Christian community, edited by Dr Kurt Treu, whose early death was a sad blow to us and to many other colleagues. Section II includes newly identified pieces of Euripides’ Phoenissae and Orestes, edited by Professor Haslam; and another handbook of Euripidean hypotheses, covering the Bacchae and other plays, edited by Dr H. M. Cockle. Section III consists of Menander: mostly fragments which provide small supplements or variant readings to known portions of his text, but one novelty with (probably) the opening scene of Leucadia (4024). Section IV contains all the papyri of Aeschines so far identified in the collection; the editing is mainly the work of members of the Istituto Vitelli of the University of Florence. Among the documents we single out those from the Arabian nome (a rare provenance); and, of unique interest, the long and difficult accounts relating to the two Oxyrhynchite mansiones on the Roman route running north and south along the edge of the Western desert (4087-8).

For the indexes we are indebted to the skill and perseverance of Juliane Kerkhecker. The Charlesworth Group have again set the text with wonderful precision.

February, 1994 P. J. PARSONS J. R. REA General Editors

ate

Pe —_ ~~ ree <€ a = A Gm « nls ees — oe 7a % atten ee ee | vi - tile edt.) 15 See ss . a ‘ ae © a > & ohne re ees S Gass sae o> > pales ar oy 86 uae. wy onl » ee -\0p2enl ate ® One > ~ itm oem ob Gud ~-en OeReiom -

+. > Gas Gat - . > => (0 Gaze) - Se oe) at <—ptay Y, Sa Vy o = =a © ats 0 ge —a@ be — is Oamies ex ion oe) 2 =i- ees Conga 6). Co Bn ee ¢ (wy a0? heat :

| = =~g bw = > & > 2 6 am «= =) ap’ = _ mn =n nee wr eaten

CONTENTS

PREFACE y,

TABLE OF PAPYRI ibs

List oF PLATES xi

NUMBERS AND PLATES Xi

Note ON THE METHOD OF PUBLICATION AND ABBREVIATIONS X11

TEXTS

I. THEOLOGICAL TEXTS (4009-4011) I

II. EURIPIDES (4012-4017) 9

Il. MENANDER (4018-4026) 24

IV. AESCHINES (4027-4055) 49 V. DOCUMENTS OF THE ROMAN AND BYZANTINE

PERIODS (4056-4092) 11g

INDEXES

I. THEoOLocIcAL TEXxTs 233

Il. MENANDER 234

Ill. Hyporueses 236

IV. RuLeRs AND REGNAL YEARS 227

V. ConsuLs 238

VI. INpicTIoNs 238

VII. Montus 238

VIII. Dates 239

IX. PERSONAL NAMES 239

X. GEOGRAPHICAL 245

(a) Countries, Nomes, ToPARCHIES, CITIES, ETC. 245

(b) VILLAGES, ETC. 246

(c) MiscELLANEOUS 246

XI. RELIGION 246

XII. Orricta, AnD MiLitary TERMS AND TITLES 247

XIII. Proresstons, TRADES, AND OccUPATIONS 248

XIV. MEAsuRES 248

(a) WrEIGHTs AND MEASURES 248

(6) Money 249

XV. TAXEs 249

XVI. GENERAL INDEX oF Worps 250

XVII. Corrections TO PUBLISHED TEXTS 258

4009 4010 4011

4012 4013

4014 4015 4016 4017

4018

4019 4020 4021 4022

4023 4024 4025 4026

4027 4028 4029 4030 4031

TABLE OF PAPYRI

LP LHEOCOGICAL TEXTS

Gospel of Peter?

Pater with Introductory Prayer

Hymn (Psalm 75, intercalated)

Euripides, Phoenissae 430-7, 461-7

Euripides, Orestes 314-20

Euripides, Orestes g86—1002

Euripides, Orestes gq0—-93

Euripides, Orestes 1233-52 Hypotheses of Euripides’ Bacchae

and other plays

Menander, Dyscolos 529-31,

557-61

Menander, Dyscolos 740-50 Menander, Epitrepontes: Hypothesis Menander, Epitrepontes 150-164 etc. Menander, Epitrepontes 290-301,

it:

DLu./PJP Second century

KT KT ll; BURIPIDES

MWH MWH

MWH MWH MWH HMC

MENANDER PJP

PJP PyP PJP PJP

338-345, 376-400, 421-447

Menander, Epitrepontes 655-65 etc.

Menander, Leucadia? Menander, Misoumenos? Menander(?), Progamon

EGT/PJP PJP PyP PyP

IV. AESCHINES

Aeschines, In Tim. 3 LSab. Aeschines, In Tim. 14-15, 17-18 LSab. Aeschines, /n Tim. 38-43 GM Aeschines, In Tim. 43-52 IA Aeschines, /n Tim. 79 LSab.

Fourth century Sixth century

Fifth century

First century BC/first century AD

Second century

First century

Second century Second century

Fourth-fifth century

Third century Second century Third century Second century

Third/fourth century First Century

First century

Third century

Second or third century Second century

Third century?

Late second century Second or third century

Se (Sr _

4032 4033 4034 4035 4036 4037 4038 4039

4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052

4053 4054 4055

Aeschines, Jn Tim. 131-2, 134 Aeschines, /n Tim. 190-192 Aeschines, /n Tim. 194-6 Aeschines, De Fals.Leg. 43-5 Aeschines, De Fals.Leg. 64-5 Aeschines, De Fals.Leg. 134-5 Aeschines, De Fals.Leg. 171-2 Aeschines, /n Ctes. 6—7

Aeschines, Jn Ctes. 8 Aeschines, Jn Cies. 15-17, 29—2 Aeschines, Jn Ctes. 33-4, 35-6 Aeschines, Jn Cles. 39 Aeschines, /n Ctes. 56-8 Aeschines, Jn Cites. 57-9, 60-1 Aeschines, /n Ctes. 80-81, 91 Aeschines, Jn Ctes. 98 Aeschines, Jn Cles. 101 Aeschines, /n Ctes. 110-115 Aeschines, Jn Ctes. 157-8 Aeschines, /n Ctes. 160-161 Aeschines, Jn Ctes. 195-6

Aeschines, In Cites. 213-14, 215-16 Aeschines, Jn Cites. 216-17 Aeschines, Jn Ctes. 87-92, 220, 229-—F, 290, 233-4, 240, 242, 248-9, 252

RB RB RB AM AM AC AC EB

EB

IC CF CF RAC ai RAC salve RAC DLi SR LSalv. FM

VB VB RAC

Second century

Second or third century Second or third century Second century

Late first century Second or third century Second or third century Late first or early second century

Second or third century Second or third century Second century

Second or third century Second century? Second or third century Fifth or sixth century Late second century First century

Third century

Second or third century Late second century Late first or early second century

Second or third century Second or third century Third century?

75 af

79 80

82 82 83 84

86 87 gi 92 93 95 98 99 100 101 104 104 106

107 108 109

V. DOCUMENTS OF THE ROMAN AND BYZANTINE PERIODS

4056

4057 4058 4059 4060 4061 4062

4063

Receipts for Refund of the Price of mupoc cuvayopactiKoc

Report to the Strategus

Official Correspondence

Offical Correspondence

Official Correspondence

Official Correspondence Document Addressed to a Strategus

Sworn Declaration of a Liturgist

RAC

RAC RAC RAC RAC RAC RAC

GMS

154/5°

154/5?

158/9 Between 159 and 163 c. June/July 161 goth July 163 c. 163

October/November 183

119

121 122 124 127 142 143

14°

4064 Nomination to a Liturgy GB 4065 Nomination to a Liturgy SR 4066 Nomination to a Liturgy PP 4067 Substitute Nomination to a LP Liturgy 4068 Imperial Rescripts RAC 4069 Official Correspondence RAG 4070 Offer to Contract for Work on RAG Trajan’s Canal 4071 Petition to an Acting Strategus RAG 4072 Declaration by an émipedAnric xpibyc RAC 4073 Official Correspondence RAC 4074 Petition RAC 4075 Daybook of the Curator Civitatis RAC 4076 Report of Property Registrars RAC 4077 Official Document (Proceedings?) RAC 4078 Nomination to a Liturgy RAG 4079-4080 Nominations to a Liturgy RAC 4081 Summary of Prices declared by RAG Guilds 4082 Petition to an Ekdikos RAC 4083 Document Addressed to an RAC Official 4084 Document Addressed to an RAC Official 4085 Sworn Declaration to the Logistes RAC 4086 Declaration to the Strategus RAC 4087 Mansio Accounts: Tacona and RAC Oxyrhynchus 4088 Mansio Accounts: Tacona and RAG Oxyrhynchus 4089 Financial Report to the Strategus RAC 4090 Petition to the Riparu RAC 4091 Report to the Strategus RAC 4092 Lease of Land RAC

IA= I.Andorlini RB=R.Barbis

MWH= M.W.Haslam JL=J.Lenaerts

VB=V.Baroncelli EB=E. Bassi GB=G.Bastianini PC=P.Carrara AC= A.Casanova HMC=H.M.Cockle CF = C.Foches

DLi= D.Limongi DLu. = D.Luhrmann GM=G.Menci FM=F.Morelli

AM =A. Moscadi

LP=L.Papini

15 December 183 22 December 183 24 December 183 16 January 184

March—April 200 Early third century c. 208

pyar Cee Oia Late third century a7 4-13 June 318 320? 325-337! 9 February 327 3 May 328 Fourth century

g September 330 ood

6 May 339

7-25 May 339 345

BIG, 327,0951 On 394)

Cao Oo

October/November 351

12 April 352 B02 1 October 355

PJP=P.J.Parsons PP= P.Pruneti SR=S.Russo LSab. = L.Sabini

LSalv. = L.Salvador1 GMS=G.M.Savorelli

KT=K.Treu EGT=E.G.Turner

153 155 157 160

163 164 166

167 168 170 rz 173 WS) ye) 178 180 183

185 187

187

189 190 E95

214

221 226 228 230

LIST OF PLATES

I. 4009 recto, 4018 verso, 4022 VI. 4017, frr. 6-27, 4026, 4034, 4053 recto, 4023 flesh side VII. 4041 Il. 4009 verso, 4018 recto, 4022 VIII. 4068, 4076 verso, 4023 hair side IX. 4078 III. 4010, 4019, 4020, 4024, 4025 X. 4087 cols. i-ii IV. 4011, 4021 XI. 4088 col. ii V. 4017 frr. 1-5 XII. 4089 col. ii

NUMBERS AND PLATES

4009 recto I 4023 hair side II 4009 verso II 4024 III 4010 II] 4025 III 4011 IV 4026 VI 4017 frr. 1-5 V 4034 VI 4017 frr. 6-27 VI 4041 VII 4018 recto II 4053 VI 4018 verso I 4068 VIII 4019 Ill 4076 VT 4020 Il] 4078 IX 4021 IV 4087 cols. i-ii X 4022 recto I 4088 col. i XI 4022 verso I] 4089 col. ii XII

4023 flesh side I

NOTE ON THE METHOD OF PUBLICATION AND ABBREVIATIONS

The basis of the method is the Leiden system of punctuation, see CE 7 (1932) 262-9. It may be summarized as follows:

apy The letters are doubtful, either because of damage or because they are otherwise difficult to read Approximately three letters remain unread by the editor

[apy | The letters are lost, but restored from a parallel or by conjecture

P| Approximately three letters are lost

ba Round brackets indicate the resolution of an abbreviation or a symbol, e.g. (aptaby) represents the symbol —, crp(arnydc) represents the

abbreviation crp§ [apy ] The letters are deleted in the papyrus The letters are added above the line <aBy> The letters are added by the editor {apy} The letters are regarded as mistaken and rejected by the editor

Heavy arabic numerals refer to papyri printed in the volumes of The Oxyrhynchus Papyrt.

The abbreviations used are in the main identical with those in J. F. Oates e¢ al., Checklist of Editions of Greek Papyrt and Ostraca, 3rd edition (BASP Suppl. No. 4, 1985). It is hoped that any new ones will be self-explanatory.

a _ ; < Sry SSE ae -_

ee —_—_ 7°, Geaveeet =e aw Am

Led Be MRORLCO RG MOo Beit. 1b

4009. GosPEL oF PETER? 62 6B.82/C(1-3)a 2.9xXg cm Second century

A scrap from a papyrus codex. For convenience, we refer to the side on which the writing runs parallel with the fibres as ‘recto’, and the other as ‘verso’; there is no telling which precedes which.

A column of writing held at least 21 lines, each line (to judge from the plausible supplements in R 5—8) had 18—21 letters: written height at least 8 cm, written width (reconstructed) c. 4.7 cm. Surviving lower margins of 0.8 cm; surviving right-hand margin 0.5 cm (R), left-hand margin 0.8 cm (V). If there was only one column to the page, we have a miniature codex, with a page width of (say) 7 cm. For similar books, see E. G. Turner, Typology of the Early Codex 22 (papyrus) and 30 (parchment). In that case, the original page height may have been no more than 10cm, so that relatively little text would be lost at the top. But it remains possible that we have here one of the rare examples of a two-column papyrus codex (Turner 36), so that the extent of text lost would be much greater.

Paragraphing by blank line-end (R 10) and perhaps by ecthesis (V 10); punctu- ation by middle stop R 4?, 11, 15?, by blank space R 3, 9g, 14?, 17. Nomen sacrum xe V 13. The correction at V 8 seems to be by the original scribe.

The text is written in a small round informal hand with a tendency to lean to the left. There are many ligatures (note especially R 11 Aeyeipor); the cursive touch shows also in the letter forms—looped a; » €v in one movement. Note e with the upper loop nearly closed, wide «, low-bellied y, v with a right side that does not touch the line, 7 with strongly curved right side, ¢ flattened with extended cap, wide 7, w in two movements. Ornament takes the form of hooks at the head and foot of uprights. With an informal hand, and a small sample, dating presents particular problems. For datable parallels see: Schubart, Pal. Abb. 81 (PLond I p. 132 ff., horoscopes later than the death of Titus); Norsa, Scritt. Doc. XVc (Edict of Petronius Mamertinus AD 133/7); Schubart, PGB 22B (135?) and 24 (148). For similar scripts in literary texts see PGB 28; 31 (Theaetetus Commentary); Norsa, Scritt. Lett. gD (Menander, Theophoroumene); Roberts, GLH 13b (BM Hyperides)—all normally assigned to the second century. These parallels suggest, at least, that 4009 might be dated to the second century rather than to the third.

The nomen sacrum identifies this as a Christian text; and R 4 ff. preserve key-words of four logia of Jesus. In fact, R can be reconstructed in detail from synoptic and non- synoptic parallels. We have not found any similar basis for reconstructing V.

We have two clues to the precise provenance. (i) R 11 suggests a first person narrative. (ii) R 9 ff., the logion of the wolves and the lambs, shows an extended text

2 THEOLOGICAL TEXTS

that recalls the version quoted (from an apocryphal gospel) in 2Clem. 5.2—4. There we have a third person narrative, which quotes a dialogue between Jesus and Peter. If it is again Peter who speaks in our text, but as narrator himself, we could assign 4009 to the Gospel of Peter; the Akhmim fragment, PCair 10759 (Van Haelst 598), shows that this text took the form of a first person narrative (xiv 60). Another fragment from Oxyrhynchus, XLI 2949 (Van Haelst 592), has been plausibly attributed to the same gospel (Luhrmann, <NTW 72 (1981) 216-26, accepted by C. H. Roberts & T. C. Skeat, Birth of the Codex (1983) 44); it is not part of the same manuscript. For further discussion of the place of 4009 within the Gospel, see Luhrmann, ‘POx 400g: Ein neues Fragment des Petrusevangeliums’, forthcoming in Novum Testamentum 35 (1993).

‘Recto’ ‘Verso’

Tells wiles aT lees che! aca ere | J.ve.[

] Gepucpoc . [ cvde__ _[

5 | ,atocwcar[ 5 mapecx | Jauppoviw, [ Bovrun[ JececBew, [ Kaco.a, | ] ovAvxer[ ont $l Jol ],ov eavov [ Aavapya[

15 Juew [..] gol 15 J. auf ], van[ .7p9. | Je xa. | 77. [

Rey a ne eee [ 20 lpreal s00m 0 tan oun [ Jwee[ foot foot

] ] BY Jenne ec al ] ]

4009. GOSPEL OF PETER? 3

‘Recto’

1 ].[, descender, more ink above, on displaced fibres, and to right on edge 2 |,, curving foot (a, «, A, pw; 6, &, x?) [, left-hand arc as of 0, ¢ (¢?) 3 ],, right-hand tip of horizontal level with letter-tops [, long oblique descender (v? but more ink to top right) 4 ],, ink on crumpled strip of projecting fibres {, point on projecting fibres 5 ]., ink high in the line, then heavy dot a little below mid-height 6 _[, ink on edge at mid-height 7 .[, oblique back as of 0, c, w? 8 ]_, point on edge, just below letter-tops (tip of horizontal, e.g. of overhang of c?) g ]., trace just below tops of letters 13 av, ., first, y 77; then daguler loop, o or second of w 14 , ,[, oblique top as of aA; then top of upright? 15 ,[, trace at mid-height ]., ink level with letter tops 16 ]., perhaps parts of w 7, hook at beginning of horizontal anomalous? —_ but enough remains of the right-hand vertical to exclude 7? 17 .[, upper part of sloping upright (e.g. ¢) ii@@) |] aay Babies high oblique descending from left to right, joining top of EPH second, probably extremities of w 20 ]_., oblique foot below, more ink to top right (x, y? or even c?); then y? [, displaced fibres

‘Verso’

2 __[, left-hand are and cross-bar of €, 0; foot of upright 3 |., possible trace (end of horizontal) just below left-hand curl of v {, high horizontal joining top stroke of ¢ (unless the whole thing is the top stroke extended) 4 ...[, high horizontal with curving upright below, y, 7? or part of 7?; then right-hand arc of small circle, ro would suit spacing; then high dot, or left-hand end of horizontal, on the edge 7 .[, high dot, or left-hand end of horizontal, on the edge 8 _¢, perhaps part of the loop, and the beginning of the oblique tail, ofa — [| Ja[, a overwritten on (less probably by) a rounded letter? 12 _[, vertical traces on edge 14 ].[].,, first, high point of ink; second, flattened tail as of a, A, p etc. {, high dot on edge 15 |]... ]w,ae or |upar? 16 ue high trace joining cross-bar of 7? or simply extension of that cross- thar {hs ink on edge 7 _[, upright with junction at mid-height (», «, p?) 18 ]_, left-hand arc, middle damaged, « or e _ {, ink at mid- level (stop?), more traces to right TG al eee [, loop of p or ¢? then v? then p (but unexplained ink to left)? then a sloping back, followed by an upright, e.g. av?

‘Recto’

Line-ends are visible in g-15, and can be reconstructed in 6 (where the rules of syllable-division exclude ¢poviu|[oc); the line-length can be estimated from the plausible and consistent supplements in 5—8.

4 ff. offer key-words of four logza:

(i) 4 Bepicudc cf. Matt. 9.37—8/Luke 10.2.

(ii) 5-7 cf. Matt. 10.16b (quoted, in the singular, by Ignatius, Epzst. ad Polycarp. 2.2), where the serpents come before the doves; IV 655 ii b 19-23 (Van Haelst 595) (=Gospel of Thomas 39.).

(iii) 7 f. Matt 10.16a/Luke 10.3; cf. 2Clem. 5.2 (quoted below, 9-15 note).

(iv) 11 ff. 2Clem. 5.4; cf. Matt. 10.28/Luke 12.4 f.

Thus Luke juxtaposes (i) and (ili); Matthew (11) and (iii) in reverse order.

To anticipate the following discussion, we suggest a reconstruction of 4 ff. on these lines:

] 6 Depicpde-[

5 yeivov b€ axé |patoc we ai [me- pictepal Kai dpdvipo[c we ot ddeuc: | écecbe we [ pvia ava pé |cov AvKwr[- 10 crrapaxOe® |pev;

[

[a

[efrov mpoc ad|rov' éav od(v) [

[6 5€ amoxpiBeic] Aéyer pou ot [AvKou crrapa |€avtec TO [apviov ob |xérs avTa@ ov—

[

dev d¥vavtat| toujcar. du-

4 THEOLOGICAL TEXTS 15 [0 éyw Aéyw b |pety’ [u]7) po— [Beicbe amo tT] @v ar[oxtev— [vovtwy tua |c, Kal [pera TO [amoxreivar] pnKe|te trou [Acar duvapeé |vwv [wndev.

(These supplements give an approximately even left-hand margin, on the assumption that, as often, that margin sloped leftwards as it descended the column. It is possible that new sentences, or sections, were indicated by ecthesis, cf. V 10.)

3-4 Ifthe supplements suggested for 5 ff. are correct, Gepicudc must end its clause, for there is no space to continue (we therefore take the final trace in 4, a dot on projecting fibres, as a middle stop). That clause may begin at 3 xa, where a clear blank (punctuation) precedes. If this is heavy punctuation, the clause was very short. At the end of 3, we see a long descender suggesting « or ¢ or possibly p (not normally so long). Against v, the further ink visible to the top right (unless we could imagine, say, ec in ligature, as usual, with the cap of e projecting to the right); against ¢, the position at the line-end. In any case, the text cannot be identical with Matt. 9.37—8/Luke 10.2.

5-7 —avoc in 5 shows that this injunction was in the singular (addressed to the narrator), and we have restored it accordingly (Se rather than ovv to suit the spacing). In axe]pacoc, the trace suits p well enough; but we cannot explain the heavy dot, most suggesting a middle stop, which precedes ac.

7-8 The number reverts to the plural: Peter speaks for the disciples.

g-15 The reconstruction is based on the version preserved in 2 Clem. 5.2—4 Aéyer yap 6 Kupioc: EcecBe dic dpvia év pécw AVKwv. dmoKpiHeic Se 6 Ilérpoc abt@ d€éyer av obv diacrapagwew of AdKor ta apvia; etrev 6 Tncotc t@ Tlétpw un PoPetcbwcav ta apvia tobe AVKovc jeTa TO aTroHavely adbra: Kal bpeic yn PoPeiche Tovc dmroxrévvovtac tac Kal undev vyiv duvapeévouc moreiv KTA. This dialogue between Jesus and Peter is a part of the extra-canonical tradition, which the author of the letter has taken over. It is true that the /ogia in 5.2 and 5.4 have parallels in the rest of the tradition; but 4009 now represents the prime parallel for the text as a whole.

gf. As in 2Clem. 5.3, éay od(v) introduces an objection to Jesus’ words. The clause ends with Jwev (10), ie a form of the first person plural: we could supply crapay@]a@pev, after 2Clem. 5.3 (dva— looks too long). The first part of 9 should mark the change of speaker: e.g. efov mpoc av |rov.

11-14 ] Aéyex wou introduces a new speech; supply e.g. 6 6€ doxpGeic]. Thus the first-person narrator is addressed, as already in 5~—7; it follows that it is the same narrator who raises the objection in 9; therefore the narrator is Peter, as in 2Clem. 5.3. The other speaker must be Jesus, although his name does not survive in what remains of the text. This version is not identical with that in 2Clem. 5.4, but the general run can be restored with reasonable certainty: here as there Jesus dwells on the relations of lambs and wolves, but in a direct statement instead of in an imperative clause.

12 cmapa |€avrec continues the reconstruction proposed for 10. There is a palaeographic doubt: one might have expected to see the tail of a] showing on the preserved papyrus to bottom left of €.

13-14 avtw suits the trace and the space; but avo too has something in its favour (if w is right, we might expect to see a trace of its first loop on the narrow strip of fibres to the left). wocncar suits the traces well; after it, a short gap before 61; there is no room for another word, we therefore take it as a punctu- ation-blank.

14-19 Few letters survive, but a plausible reconstruction is possible on the basis of 2Clem. 5.4b+c¢ and its variants in early Christian literature (cf. Matt. 10.28/Luke 12.4—5; Ps-Clem., hom. 17.4; Justin, apol. I 19; Hermas, mand. XII 6.3; Irenaeus, haer. III 18.5; Clem. Alex., excerpta ex Theod. 14.3 and 51.3).

15 0 wet, the next trace can be taken as a middle stop; [Jy fills the gap.

19 The traces are much damaged, but ]vwyr[ suits well.

20 f. We expect a continuation corresponding to 2Clem. 5.4c or its parallels. But we have found no plausible reconstruction.

‘Verso’ We have found no parallel from which to reconstruct this side. That its content has something in common with the Recto is suggested by the similarity of its ‘synoptic’ vocabulary, and by recognisable

4009. GOSPEL OF PETER? 5

hints of dialogue structure. The precise sequence remains unclear. We may assume that in 5—6 Jesus is speaking. In 13 Jesus is addressed: we may guess from the Recto that the speaker is the first person narrator, Peter. But it is difficult to be sure where the speaker changes (9-102), and whether it changes again after 13.

2 |Pey[ possible.

4 E.g. cd d¢: new clause. But there is a patch of damage after the apparent sigma, so that o (e.g. ovdé) may not be excluded. :

5 mapecx—. Perhaps a first person singular (of Jesus), then r@ mpoceA]|@dvre pfou. A reconstruction exempli gratia, based on Matt. 8.21~2/Luke 9.59—60: 038é 76 O[ diac rarépa] | rapécx|nKa Ta mpoced]|Aovre u[or.

8 ort. If age: is to be read, one thinks of a part of ddim, cf. 13. The next letter was overwritten by way of correction; given the weight of ink, it seems likely that a was the final version, written over a rounded letter that could be c, or perhaps (since it is rather small) o. If we accept a, ad(e)ia[cw is a possibility; if we accept 0 or ¢, af(e)io[uev, ab (e)lo[ucw etc., ddete or ddeic (or an itacistic spelling of adjc)? This assumes that the suprascript ¢ is an addition to, not a replacement of, « (i.e. dec).

If this verb is recognised, one could think of dua[préac, dwa[prrjarta in the next line; before that, if A is rightly read, wa ]|Aau, cf. 2Peter 1.9? But e.g. év Padc]|Aala pal[Oyr77c is equally possible.

10 avr@ projects into the left-hand margin. Presumably the ecthesis marks a new section, or a new speech. Since adr@ itself cannot begin a clause, the break must come in the line before.

II-I2 ev (t@) 6]|vopari?

13 Apparently ddeic or ddeic, cf 8; then «(vpcje, an address to Jesus (probably by Peter, unless it belongs to quoted direct speech). 11~13 might be combined in some such sense as ‘Lord, do you remit sins in the name of God?’; but clearly other meanings of ddvévar (‘allow’, ‘let go’) are available.

14 «[o]Aov@ [ possible, a part of dxoAovdety or the like?

D. LUHRMANN—P. J. PARSONS

4010. Pater with INTRODUCTORY PRAYER 20 3B.36/H(1—-3)a 11.5 X15 cm Fourth century

This prayer has been copied in a handsome hand, with ample margins (to the left and below) of c. 3 cm. The writing runs with the fibres; the black is blank. Since the text ends, or could end, with the last line, we are probably dealing with a single column on an individual sheet, rather than with the beginning or continuation ofa roll.

The script is a version of the Severe Style, written upright with a thickish pen and some attempt at differential shading, to be assigned to the fourth century (compare Turner, GMAW 49 and 70; Cavallo & Maehler, Greek Bookhands of the Early Byzantine Period 12a—b). No lectional signs, except for the suprascript stroke in nomina sacra. There is some tendency to separate words; and short blanks were used to separate clauses (12 etc.). The scribe, though a competent penman, was careless enough to omit a clause in 13, and to duplicate one in 19.

The Pater Noster (11 ff.) is introduced by a preliminary prayer, see e.g. Liturgy of St Mark pp. 135-6 Brightman. For other examples of the Pater circulating separately on papyrus, parchment and other materials see van Haelst nos. 345-9 and PKoln IV 171. Many of these have been thought to be amulets; the physical size of 4010 seems to make that unlikely (there are no clear signs of folding).

6 THEOLOGICAL TEXTS

Lilie ba eter ] «Ae, ], vpo,[

dectroTa TT QAVT [ TTY) P TWV OLKTLP[LWV

Kat Oc macyc malpakAncewc = C. 8-11

10 Kat eAencov Kat KuB... .[ Cc. Q-12

KaTaguwcov nuac . | c. 6-9 TEP NILwv

O €V TOLC OVVOLC aytacOntw [To OVvOLa COU

eMetw 7 Bactrera cov we e[v OVV@ Kat Emre

yc TOV APTOV HYUWV TO [v €7T LOUCLOV doc

15 nw cnuepov Kar adec y[puw Ta ogerdn

pata nuwy werep Kale] nulerc adnKaperv Tots ofiAetaic nuwy Kall un ELceveyKNc

pas ec Tetpacuov adda [pucar qpac a 70 TOU TT OV1) POU pucat nelac

foot

2-5 Fibres stripped where the initial letters should come. In 5, isolated horizontal trace: paragraphos or part of letter?

6 ff. The lines which are certainly restorable have 29 to 32 letters.

6 pu. ,[, lower arc of circle (e 80 ¢ w?); tip of upright descending from left to right (a dA vx). |, foot of upright (if word end, 7 or v?). Perhaps pov [Ka]l eAéy[cov, cf. g—10.

7 |., perhaps a lower right-hand arc. eAe[{yco]y would not suit the trace, and in any case looks a little too long. Perhaps ¢Ae[oc c]ov, which would just fit: the phrase is common in LXX. E.g. éAén[cov nude xara TO péeya] |eAe[dc cov, after Ps. 50.3?

po [, traces (of an upright?) on the edge.

8-9 2Cor. 1.3 6 matyp Tay olkTipudy Kal Dedc mdcnc mapakAncewc. At the beginning of 8, perhaps mavTo, perhaps 7avtw; the traces following, on a narrow strip of fibres, are vestigial; wavtoxpa[twp would be possible, but looks long for the likely space. g end, e.g. émdxoucov (Ev. Barth. 4.49.8 émaxoucdv pov Kai €héncov Tove apwaptwAoic).

10 «vB, not puc, seems to suit the remains; the final traces, on straggling fibres, indeterminate. Probably xuBepyy[cov, then space for c. 4—7 letters. Joh. Chrys. 63.926.38 kuBépyycov ro bmdAourdv pou THC Cwijc.

1r __[, indeterminate ink on straggling fibres. We expect a verb of speaking to introduce the Pater, as at Liturgy of St Mark p. 135.31 Brightman katagiwcov Hac ... roAuav emuxadreicBal ce ... Kal A€yew KTA. But Aéyew itself looks too short (unless followed by a punctuation-space).

11 ff. Ev. Matt. 6.9—13.

13 cov: the following clause yevnO77w 7o A€Anpd cov omitted by homoeoteleuton.

16 weep: we Matt.

4010. PATER WITH INTRODUCTORY PRAYER q

17 «xa[v, nothing remains but indeterminate traces. 1g pucar nul[ac repeated. After that, more than enough room for dy7v, but not for a doxological formula (see van Haelst 345-6; PKéln IV 171), unless the text continued into another column.

(Ke TREO

4011. Hymn (Psalm 75, intercalated)

22 3B.16/F(3—-4)a 15.5 X 13.5 cm Sixth century

On one side of this piece, written across the fibres, stands the upper part of a memorandum, beginning bropvynctix(dv) Ta evAaB(ectatw)|Xpi[cto|dwpw dvax(dvw) PoiPaypwv|BonP(oc) [],, (perhaps y[a]ip(ew)?) (vac.)|Karak&ia[c]n 7 cy edAaBra eveyxeiv ... (the items to be transported include ¢axaAra Svo).

On the other side, written parallel with the fibres, stands a Christian hymn-text. To the left is a heavy sheet join, overlapped by the line-beginnings, with fibres running at right angles to the rest: this is the joint between the protokollon and the roll (Turner, The Terms Recto & Verso 20); the papyrus seems to have been cut or broken off down the left edge of the overlap. The text, punctuated only by section marks in the form |_, represents a cento. The two complete sections, lines 1-9, derive from Psalm 75, with interpolations both from LXX and from NT.

The memorandum was copied in a sizable and handsome cursive script, the hymn in a smaller and more rapid cursive with many phonetic misspellings. Both texts can be assigned to the sixth century.

XMT + yvwctoc ev TH Llovdea NV w Oc ev Tw icpanA peya peya TO ovoua avTou KaTwKiCcOy EV cLwY ELpnYnV m[o]AAnv evavyeAcCopevoc ade erapaxOycav 7 ACHVETHTE TH KapOla KaL ECTAUPwWCaY AUTwWY 5 ave[c]tn de ex vexpov Kat Lou Taca y yn Wcnxale TEC peta hoBov Kat TpanTyToc Evxac avTw amTwoLdw Kat Owpa mpocpepovtec | (vac.) L_ exe cuvetpupev ta Kpatyn Twv To€ov QuTov wraw Kat poudea Kat TOAE WY KaL ECTAUpwCaV 10 ave|[c|T €k veKpov 0 Oc dvackopmuicac TavTac

TOUC €xOpouc QUTOV ATO TTPOCW7TOV AUTOU

8 THEOLOGICAL TEXTS

ve nl... ].onl.......], Pavarw nar erap[a}x| peel ler leno ter |

XMT: see most recently LVI 3862 1 note; R. W. Daniel &F. Maltomini, Supplementum Magicum 11 62.2 note.

1-2 Ps. 75 beginning, yrwerdc év 7H Iovdala 6 Bedc, ev rd Icpand péya 76 dvopa adbrod. 3-4 ib. 6 érapdyOncav mavrec of dcbverou TH Kapdia. 8 ib. 4 exet cuvérpupev Ta Kpdty TaY TOEwv, OmAov Kal poudalav Kal mOAELov.

2 katwxichy, the last two letters very cursively written, but we have not found a better reading (katwxitar, for Katorxeirat, fails on tr, which would take the cursive form not found elsewhere in this piece; Katwxicato too fails on 7, and in any case looks too long, and the middle unexpected).

3~4 paraphrase Ps. 75.3 kal éyevjOn ev elprvy 6 témoc adrod Kai 76 KaToLKnTHpLov avTod ev Civ. Is. 40.9 6 edayyeAlouevoc Ciwv, 52.7 evayyeALopevov axony eipyvyc, Acts 10.36 evayyeArlopevoc elpnvny.

3 ade, the last character apparently the cross-bar and lower curve of epsilon in the literary shape, the cross-bar cut by an upright descending from above the line. We might interpret this as (i) aAAle], ie. add’ with a dittography of the following ¢ deleted, or perhaps even overwritten with a small a; or (11) aAAe, with something, possibly eta, suprascript above epsilon. Perhaps (ii) might represent aAAn(Aoua), but we can produce no parallel for the abbreviated form. (i) would be an addition to the psalm text, but intelligible enough in itself.

4 7 acnvernre: the next trace, suggesting sigma, is merged with the tau of 77, suggesting a correction. The Psalm text has of ac’verot, of which 7 acnvern could be an itacistic spelling. But re then remains to be accounted for. On the other hand, it seems unlikely that the copyist intended at dcuvérnrec, or of dcuvérnToL, words otherwise not attested.

«at, below alpha a long trailing stroke, as if the copyist had written a in ligature and then added the iota separately.

5-6 Za. 1.11 mdca 7H yh KatouKxeirar Kal Hcvyaler. [Peter 3.16 wera mpaityntoc Kai PdBov. Ps. 49.14 amddoc T@ vipictw Tac edxac cov, and often.

6 amwéidwrec: apparently —dw, not —doy or —d@ (for —dwr).

7 After mpochepovtec, a long thin oblique descending from right to left into the next line (unless it is an unusually extended branch of kappa in 8 xparn), then a short thick oblique descending from left to right. A section mark? or a, 1.e. a(AAndAoua), cf. 3?

8 Here the right-angle paragraphos opens the line, and touches the initial € at half height; in 10 it comes between the lines in the usual way.

10-11 Ps. 67.2 dvactijtw 6 ede Kal diackopmicOjtweav of €xOpol avtov, Kal duyétweay of pLcobvTec avTOV amo TPOCWTTOU QUTOV.

11 e€xpouc or perhaps exOpouc.

ee , fourth perhaps «. |, top of upright. ]., possibly right-hand curve as of omicron? etap[a]x[@ncav, cf. 3?

13 av[t]wy enrn€ay (the final nu represented only by upright traces on the edge) seems possible: IMace. 12.28 édoByOyncav Kal ertnéav tH Kapdia adtav.

TK. TREU

Ue PU REED Eas

In volume LIII were published such manuscripts of Euripides’ extant later plays—those constituting vol. iii of the OCT—as had at that time been identified (3712-19). The texts presented here are a supplement to that group. Again thanks are due to Dr James Diggle for additions and corrections.

4012. Euripipes, Phoenissae 430-7, 461-7 95/69(a) 4.2X 4.2 cm Fifth century

A scrap of a parchment codex, written in a medium-sized sloping hand of the type illustrated by G. Cavallo and H. Maehler, Greek Bookhands of the early Byzantine Period, nos. 15a and 15b and assignable to the fifth century. The contrast between thick and thin strokes is pronounced but not extreme. a is sharp, the midstroke of « descends and is kept short, o tends not to attain its full oval height and shape, 7 has dots at either end of the thin top-bar. Much if not all of the lectional apparatus, which includes extensive accentuation, appears to be by a second hand. On the rather unsafe assumption that between 437 and 462 there was no discrepancy of line-count between the papyrus and the medievally transmitted text, the depth of the written area may be calculated as c. 18 cm, occupied by c. 30 lines.

Two notable points of textual interest: a nearly new reading in 434, and substan- ual discrepancy at 436 f.

Recto (flesh side)

430 toAAo be davalw| mapeict Aupav] xapw [a]va[ didovrec: em ylap THv euny [ moAw: Oeouc 6’| emapmoca we | tou diAtatoulc Texovcw n[

435 add’ exc ce Tewer| Twvde dia[

437? Jexape[

10 EURIPIDES

Verso (hair side)

Pe Sey Steel aly opt ed owcw | Ker TadTa xp Kakwv de] Twv mplv pl 465 Aoyoc pev oluv coc trpocb[ cv yap ctpa|revpa dava[

aduka tremov |Owe we[

433 é€mapoca scriptio plena (so too R).

434 texodcw. The paradosis is roxedow: editors read éxodaw, attested only as a yp. variant in schol.® (roxedow schol.M, rexotow schol.Y°, but that these are corrupt for éxodow in the scholium is guaranteed by the exegesis that follows, r@ ddeAp@ prov xtA.). The halfway house texodow, apart from its occurrence in the corrupt schol.¥°, is found only in Vr (Palat. gr. 343)—not a significant conjunction, I take it.

436 £ as transmitted run prep, diadAdéacav opoyeveic pidouc | madcar Tovwy pe Kal ce Kal Tacav TOAW. The papyrus had something different: I imagine 437 in a slightly different form, watca: (or Wecklein’s Tavoov) movwy cle Kaue [Kal macav moAw. 436 is apparently absent: corroboration of Nauck’s suspicion of it. (On the presumptively interpolated status of verses absent from papyri see CQ 26 (1976) 4-10.) ce Kape, which was in fact conjectured by Elmsley, has what seems to me the distinct advantage of allowing both pronouns to be emphatic.!

462 cove |Adw[v] not —ov7’.

M. W. HASLAM

4013. Evuripiwes, Orestes 314-20

95/68(b) 3.0 X 5.0 cm First century BC/ first century AD

A scrap written in a fair-sized round and upright serifed hand which I would date early in the first century or towards the end of the previous one; comparable hands are those of P. Vindob. G 19996 a and b (MPER n.s. 1), Pap. du Fayoum 1 (Bull. Soc. Alex. n.s. 3, pl. ix), P. Berol. 9775 (Schubart, Pap. gr. Berol. 11b). It comes from the same manuscript as P. Koln VI 252 (Or. 134—42: II* Diggle, 409.1 Mertens- Pack), to judge from the plate in P. Koln III (Taf. [Vc). The back is blank.

Orestes papyri are listed by James Diggle, The Textual Tradition of Euripides’ Orestes (Oxford 1991), 115 f, and in the 3rd edition of Roger A. Pack’s The Greek and Latin Literary Texts from Greco-Roman Egypt, revised by Paul Mertens (whom I thank for a

' Here as at e.g. Med. 327, Hipp. 1409, IA 677, fr. 898.4, I find no good reason not to prefer the orthotone form (contra Kuhner-Gerth I 557, Schwyzer II 187; cf. Diggle, CR 32 (1982) 134 n. 4). I am delighted to say that Dr Diggle now accepts this.

4013. EURIPIDES, ORESTES 314-20 II

preview of the relevant section); another, at Duke University, is reported by W. Luppe,

APF 37 (1991) 81. The Florence Orestes papyrus mentioned in the introduction to LIII 3716 is published at SCO 35 (1985) 13-23.

Kav 4) vocl 315 Kapalroc Bpo| iP. [ dp ]opadecl |rorviadec|

JaBaxxevro[

aa ]8ax [

316 ff. The lyrics were evidently indented by c. 1.5 cm. Likewise in P. Kéln 259, at least for vv. 140 f,; it is possible that v. 142 was not indented (so Diggle, Textual Tradition 132, following O’Callaghan and Gronewald), but I doubt it: we do not know the extent of the papyrus’ textual divergency.

316 aac was not written (Dr Diggle reports af af or ai at for all the manuscripts he has collated). ee, it would seem. Cf. e.g. Hipp. 595.

320 Consistent with daxpuc[c] xa[, as transmitted. Not Saxpvorce (OCr); failure to share trivial error in Cr does not compromise the possibility of Cr’s access to ancient tradition raised by its duws at 138, in apparent agreement with P. Koln 252 (Diggle, Textual Tradition 118; but Gronewald’s exclusion of w[ in the papyrus should not simply be swept aside).

M. W. HASLAM

4014. EuripipEs, Orestes g86—1002 104/23(a) 4.3 X 16.0 cm Second century

A fragment written in a medium-sized round and upright hand of the same type as XIX 2224+ XLIV 3152 and IV 664+ L 3544, assignable probably to the latter half of the second century. Back blank. The lower margin, if 17 is the last line of the column, as seems likely, was 4.5 cm. A scribal error in line 8 has been crudely corrected.

The papyrus is without at least some of the surface error of the later witnesses (it presents Porson’s dAdov in v. 1001), but does little to resolve the textual uncertainties of the passage, though it may have done more if there were more of it. As it is, textual reconstruction is unusually problematic.

There is slight textual overlap with 4015.

12 EURIPIDES

Jopacep[ Ie eee Joyparra| |vectoAwi[

on

JAaye cl [ 990 ITAovdol

Jovrou [

]pocy|a ]paucreatc[ Jov [ 10 ]. .ac [

Jepmotc [ 995

Ic |pacadocroxo|

|pvocotrote [

15 JoAoov [

| i [ 1000

|wrovadiov [

| [ ] [

No reconstruction is offered, for it has proved unexpectedly difficult to find supplements of consistently compatible length with one another. One can only assume there were not differing degrees of indentation, but to assume otherwise helps little. It may be that some of the lacunae conceal more textual discrepancy than is displayed among the medieval manuscripts. I number the lines 1-17 for convenience of reference.

I ds €rexev €rexe yevérlopas seems too long a supplement. Perhaps without erexev; that would be compatible for length with 4, the next line whose text is anything like assured, though perhaps a little on the short side. For the medieval manuscripts Dr Diggle reports: érexev €rexe(v) plerique: érexe LRRwZmZuTp: érexev AdZbZd.

1-2 After yevéropas the paradosis is éuébev déuwr, of Kkateidov aras, the main manuscripts dividing after déuwv, MnSZ after éuébev (Diggle, Textual Tradition, 149). In the papyrus 2|douwv o Kare].d[o]v (v. sim., €.g. dofovs av) would seem to suit the space; |ovas kat. too short.

3 Going by 4, I would say zoravov (Porson) pev du]wyya has the edge for length on the transmitted To mTavov 4. 6., but there can be nothing like certainty.

4 I am presuming reOpirmoBapolv, as transmitted (—wv. MS).

5 medoy ote me]Aayeor seems a bit short: so perhaps omore (reported by Diggle for MnS and

4014. EURIPIDES, ORESTES 986-1002 13

V*Aa™An'*L?) or o7’em. Both 4014 and 4015 have —eo., lending no support to Diggle’s sympathy for —eoa., given by MBO of the vetustiores.

5-6 It is not clear whether or not ze]Aayeor ends its line (as in 4015: most codd. continue it to &e|: Diggle, Textual Tradition 138). Close inspection reveals that the surface is not quite blank, and abrasion hereabouts is severe. On the other hand, |d:eduppevoe xp |tAov would seem to be better suited to the space in 6 than would |é.¢—. The remains of 6 are rubbed and not readily identified, but ¢ is recognizable.

7 dixwyr es oda tlovrov.

8 The correction of the slip is entered in a large thick hand. Before ]poc, Aevkoxupocw (or —acw) may be a bit long: perhaps —c (which Dr Diggle reports for An).

9 movtiwy caA|wv seems on the short side, but perhaps not intolerably so. Diggle reports that division after yepaoriais is almost universal, but that the only manuscripts which isolate wovriwy odAwy are MBVACKRITT>p, most of the others combining it with what goes before or after.

10 Low (or Adow or sim.) appatevoas is the paradosis (4puardoas At [~At?]), but the traces on the papyrus are anomalous, hardly Jucac. v would be acceptable (only letter-top traces remain), but what follows cannot be read as c, and does not seem to be any regularly formed letter.

11-13 ofev doporce Toc] eworc |[NAP apa modAvcrovo]|c is unproblematic, but then Aoyevpa troysviorce] jracadoc seems rather long.

14 To xpucouaddov (or —padov) alpvoc.

14-15 The mid-stroke of the final € is extended. The papyrus evidently divided ozore|yevero, as most codd., metrically untenable. For 15 yevero (or eyeveto) tepac] oAoov is too short, and it may be that the papyrus had oAoov doubled, weakly attested among the medieval mss (MnPrRRwS, a homogeneous group: Diggle, Textual Tradition 36), though this looks rather on the long side.

16 There is no knowing what precise forms the papyrus had, but the longer forms (the transmitted Arpéws rather than Arpéos, Dindorf’s im7oBwra rather than —Bd7a or —Badra) seem best for length.

17 ofev epic to Te mTEp|wrov aAdiov. In the supplemented part I cannot exclude any of the variants. aXdtov Porson: deAvov (vel adeA— vel HeA—) codd. and Philoponus. The papyrus confirms the antiquity of the transmitted colometry (cf. Diggle, Textual Tradition 139).

M. W. HASLAM

4015. EurrprpeEs, Orestes g90—93 104/9(b) 2.4 X 2.4 cm First century

A scrap written in an ungainly plain informal hand assignable to the first century. Back blank.

JeAayece [ 990 ~—sr€Aos ore trl eAayect JupriAovd| dedigpevce |uptiAov p[ovov |watrov7[ duxwv ec 01d|ua movt[ov

ee | AevKoKkupoce mpoc] yepaluctvacc

990 a corrected from ¢ calamo currente. The ‘restored’ transcript at the right is what seems most suggested by the spacing. 7]eAayece ends its line (cf. the preceding number). The medievally transmitted colometry is the same except insofar that at 990 the best attested division is the metrically difficult dre | Sidpevae (see Diggle, Textual Tradition 138 for details of the manuscripts’ colometries); the papyrus’ division between the words presumably derives from this. édére or (better for space) dr’ emt instead of dre is not excluded, but the space requirements seem satisfied by the transmitted text.

993 Text not assured.

M. W. HASLAM

14 EURIPIDES 4016. EvurrprpEs, Orestes 1233-52 104/162(a) 6.3 X 10.2 cm Second century

A badly damaged and abraded fragment of a single column of a de luxe manu- script, written in an early ‘Biblical Uncial’ type of hand comparable with II 224 and P. Lit. Lond. 78 (plates 6 and 14 in G. Cavallo, Ricerche sulla Maiuscola Biblica) and assignable to the latter part of the second century, though I would not care to rule out the first half of the third. Some accents are in evidence, and a breathing (diacritical: 1244 efc), all I think by the first hand. Elision is apparently signalled where effected, but scriptio plena is preferred in cases of potential formal ambiguity (1236 azéAuca, 1241 dxovrilouce). In cases of antilabe the lines are divided. Paragraphi and speaker identifications do not survive. On the back are much damaged remains of apparently documentary line-ends.

The suspected v. 1245 is present. A recent conjecture in v. 1250 is confirmed. In 1246 ff. there is some textual overlap with XI 1370 fr. 9.

In the transcript I have effected identification of letters only where what remains seems sufficiently indicative to warrant it, though in some cases identification could not be arrived at without the aid of the known text; in contexts where the traces are wholly indeterminate dots alone are given.

w cuyyevera rat |po[c

ayapwepvov evc|aKouv|cov

12358 es 12356 nyaunv eyw €u|pouc. ].[. J... .aeduca olny 1237a | 1237b lee

ovKouv overdon T]ade KAvw|v 1239a = Oak puoic KaTalctre|v]dw ce 1239b ie 1240 mavcacbe Ka] mpoc epyov e€opy.|wucba euep yap evcw] ync axovtilouc: al pac kAvew cv 6 w Clev mpoyove Kar dix [Nc dot evtuxn|cat Twd’ Ewor TE tpiccoic prdot|c yap efc ayw[v 1245 n Snv amacw] n Oavew o[decrerar aed el TQ TPWTA KATA 7 |eAacy[ov

twa Apoe|ic avday trorva [

4016. EURIPIDES, ORESTES 1233-52 15

1250 ett cou TOO ev Savardav | oAec

cTNTE aL pev yuwly Tovd’ aulaénpn at 6 evbad addo|v oor ec [

1235a Perhaps exrewa unrep|a, followed by a low stop as in the next line,

1236 The initial traces are so scant as to be unusable. x is hardly suggested before a, but if kav was written (for the scriptio plena cf. aeAuca o[xv—) I cannot verify it. What preceded is still less recoverable. am itself looks somewhat dubious, and I was tempted to read 07, but close inspection gives assurance of the pi if not of the alpha.

1237a Ifthe surface had not suffered such damage the end of cou tatep apnywy should be in view, but now there are no clearly visible traces.

1237b The assignment of such vestiges of ink as remain is quite uncertain, but there is nothing to suggest that the text was anything other than the transmitted ovd’ eyw mpovSwxa ce.

1239b eyw 8 orxrowcr] ye presumable but unverifiable.

1241 axovriCover alpar cf. 1236 above for the scriptio plena. At Bac. 1131 in XXII 2223 I read Taca op[ov.

1245 The letters are sparsely represented but the reading is in little real doubt.

1246 The surface is mostly blank, and the few scattered specks give no basis for decision between pirat (codd.) and ¢iAvac (Hermann). Despite 1370’s aberrant layout we would expect the lyrics to be uniformly indented with respect to the trimeters regardless of speaker, and the evidence is consistent with this.

1247/8 I see no certain traces of ink before ¢. ta mpwra Kara fits the space well enough before reAacyov (I am using the next line for a fix on the line-beginning, see on 1250) but slightly shorter supplements (e.g. Hartung’s zpwr’ ava) are not out of the question.

1250 On the assumption that the previous line began twa Opoerc and that this line was ranged with it, the space will nicely accommodate er: cou 7o]8 (or ta]8), as transcribed. Division at this point is well attested among the medieval manuscripts, see Diggle, Textual Tradition 139. I expect rod’ ev was written with apostrophe, now gone; rod]e ev is perhaps not quite excluded, but ]é rather than Je better suits the base-line trace.

davaiday accords with Diggle’s conjecture: Javaida@v codd. The upper part of the letter remains, defin- itely not w.

1251, 1252 No reliance can be put on the supplements with regard to elision and scriptio plena.

M. W. HASLAM

4017. Hyprorueses ro Euripipes’ Bacchae AND OTHER PLAys 88 X5V Wi 2527 ~ jem Second century

Twenty-seven papyrus fragments (of which only three contain more than six lines of text) containing parts of hypotheses to four or more plays of Euripides. They are written on the back of quotations from the /liad and Odyssey written out as prose. The identification of these helped in making nine joins. The hand is a rather ungainly, sometimes backward sloping, semi-cursive of a type typical for hypotheses of this kind. It is similar to XLII 3013 (hypothesis to a Yereus) and LII 3653 (hypotheses to Sophocles’ Nauplios Katapleon and Niobe) and may be dated to the second half of the second century, cf. C. H. Roberts, Greek Literary Hands, pl. 14a and in particular 17a=

16 EURIPIDES

VI 853. There are no punctuation or lectional signs but the scribe has sometimes indicated word breaks by a wider space between letters. There are c. 27-30 letters per line, judging from fr. 2 11 4, which can be firmly supplemented.

The present text follows the usual pattern and layout for papyrus hypotheses (see the introduction to 3653). It has previously been referred to in Pap. Congr. 1X (Oslo, 1958), 7.

The identifiable parts are fr. 1, part of a hypothesis to Eur. Andromache, fr. 211 3 ff., which contains the first part of a hypothesis to the Bacchae, and fr. 4, almost certainly from a Bellerophon hypothesis. Above and before fr. 2 ii 3 is part of another hypothesis, which, given that such texts are ordered alphabetically only to the first letter of the title, will have been of'a play in A or B. Fr. 5 has the beginning of a fifth but unknown hypothesis. Fr. 2 contains the right-hand ends of nine lines from the lower part of a column, followed by lines of a second column surviving for the greater part of its width. Since fr. 2 11 1 1s on the top edge of the top margin of the text on the front, it is probably the first line of its column.

The text of the Bacchae hypothesis is close to that of the hypothesis preserved in the Codex Palatinus (P), but like other papyrus hypotheses, where the medieval text survives and can be compared, has many minor discrepancies.

I am greatly indebted to Dr W. Luppe (Halle) for several helpful suggestions and ideas, and also to Dr J. Diggle, especially for some re-readings of P.

Frat Fr. 2 : : see pp. 18-19 Shea Ipaver[ |v weca[ rome 5 ] acl

ea Coe

4017. HYPOTHESES TO EURIPIDES’ BACCHAE AND OTHER PLAYS 17

Bri

Ri

5 PAO)

Top ]udopa., [ ].¢.[¢.5 Ne [ peas re |e yl eR cHttema |

Lara ol... Jel Jectevepoc 2[ 8121

]roqoucrov.[..]....

Jace. [.. vf |v. ].¢[ Frey aul JAogp| Ieee. peor ] Aer, [ Jeul

Fr. 22

18 Fr. 2 col. 1 ll, dos |nTowa | .Aeyov 5 Jo Jevwe ]8ov

].a Ju

20

EURIPIDES

col. 11 Top (?) mpetrov[| Cc. 24 outouc ve. [| ei8

Bal Giaat. © poier: Ty x9ova noutrobect| _L. Jvv.ov 0, pecnrovrec evn Bl. . “| eGEer €lVal 0 aah odavToicTipw| by: _.).mectycernump, .o |, Javer[ PR oo), p emoncev tac{ | [, Jaro |. | .eKac wy acadul. |v arep|.. |. nyovevarrouchi| Jcwol.. | .y.[ oie Jocrov

eoceT,, |e. Peres | Beovopye [....].

Ft ewes Ue AL

BactArave [

ager ee ]xwvevvAa| edncev ena, [ 6-8 |] [.. ]r[

amrectetAev|

4017. HYPOTHESES TO EURIPIDES’ BACCHAE AND OTHER PLAYS

20

col. u

mpetrov| ovtouc ve. [

Ba|kyat dv apyy:

Rai teiata ible enaac

xOova 7 O° UTobec{c: Ai[6 |vucov of mpecyKovtec ev OnB[ atc ovK ed|ycev etvar Oedv: 6 6’ adrotc Tyww|pi— av| eméctynce THY mpétov(clav. év[ua— veic yap erroincev tac [O|n[Blatwy y[v— vjaixac, dv at Kadpy[ov] buyarep|[ec a |dnyovpevat Tove B[a]cuolvc] nya[yov T[ploc tov Kibarpava. Kadpoc pelv y|npadrgoc [,, nd... Cewedn[ 4-5 aro een eed oboe Beov opyt [....].[ ©. 11 nik a ollie ance Bacirtav €d[uchdper Toic ywopevorc Kai twac pelv Trav Bax|y@v cvvAa[Bav edncev, em’ avl[rov de tov] O[€6 |v [aAAouc

drécrevder|

9

20 EURIPIDES

Fr 23 Pe, 24 Fr. 25

cal 8 Foot? Jacl 0¢ Foot? Fr, 26 tree i al [

Le a 4 ..[, 7 or xa possible

de, Op

1 ]., trace only 2 ]..0., possibly foot of descender, then low right trace followed by base of rounded letter and lower part of vertical 4 ],, thick stroke of ink, compatible with top right arc of rounded letter 5 0 preceded by deliberate space 7 Ifv, the two verticals are very widely spaced 8 ]., cross-bar linking to top of a

Fr. 2 i

2 After c, a deliberate blank space of a letter’s width followed by base of a vertical; then lower part of descender, with possible trace of horizontal linking to preceding vertical; perhaps te Then another letter space. [, foot of descender 4 Jt[, low trace only |z, top right of cross-bar After 7, two low traces, probably separated by a letter 6 _.[, high trace followed by top of vertical After v apparently top of vertical, and then top right arc of rounded letter o,,, 0 or @, then foot of vertical; third letter has feet of two verticals with right end of high cross-bar as of 7 7 ]., mid trace; ¢ and v fairly certain 8 ],, horizontal linking to top left of 7 — After p high trace; then oblique sloping down to right like the first 7 in the line After o top left of letter compatible with v g After v, part of rounded letter followed by low slanting stroke, compatible with cursive ec ligature; before p high cross-bar and low oblique __ tac, cross-bar of t written over twice Then right of 7 rather than cross- bar of @ linking to left of 7 | B completely lost [, speck only 10 First five letters very cursive and only identifiable because the text is known —]_, mght of @ with cross-bar linking tov Of y only right end of horizontal 11 Before 7 traces followed by right of letter, compatible with ¢ The combination yo is certain, cf. the exactly similar sequence of strokes in 12, kadpoc _y .[, two verticals, the first with a leftward serif at the foot as of y or v, then vertical with cross-bar at top; third, pointed tip of a letter 12 _[, high ink, compatible with top right of 7 — After roy ligature of ax just discernible 13 ],, vertical plus trace of cross-bar? After 7 uncertain whether a deliberate space or abrasion _ Before e a likely suggestion could be confirmed; after, mid ink on either side of a hole 14 ,.[, perhaps e followed by y, 7 orz ]_,, .[, jumbled traces, perhaps including e at the end 15, [, ink on broken vertical edge —]_[, trace only 16 7 [, ty, 74, or possibly z[ After a high ink to left and right and low ink between; then mid horizontal _—_]_[, possibly a 17 _[, top of letter 1g _[ and ].{, high ink — After Jy[, probably a space

4017. HYPOTHESES TO EURIPIDES’ BACCHAE AND OTHER PLAYS 21

jhe, 8}

3 After a: top of rounded letter and high ink to right, followed perhaps by cursive 7 — After o and e traces only [, vertical ink 4 ]., low trace After yw left of rounded letter, e.g. e 5 After v two letters ligatured together, second perhaps a or ; then vertical as of « or right of @ |p oo or 7, then o or c, followed by 7 |.., two specks of high ink, as also after c and 7 Fr. 4

1 |v ora [, low ink joining preceding a 2 |, low right ink, perhaps a [, right-facing curve 3 .[, perhaps left of # |, _, » or m followed by o orc — After a possibly c, then mid and high ink 4 ,.[, first, a rounded letter, € or o? [, top half of vertical with top hooked to left, as in k Or 7 5 ]., foot of vertical and horizontal joining y, perhaps y or tr After » rounded letter, possibly ¢? o[, perhaps yu, then o or € 6 Badly rubbed = After ¢ perhaps @ — €_0, two verticals At end perhaps 7 7 tov [, rest of fragment has mostly jumbled traces only Pr

1 Descender only: ¢ or p 3 o.[, small circle of o might be top of p; then right-facing curve with

extra ink at top, e.g. w

F7

1 _[, vertical and specks to right —c or v? 3 |,, high rounded ink [, right-facing curve Fr. 10

Stripped area on left of fragment 1 Left-facing curve 2 _[, right-facing curve Bre 12

Remains of 2~3 letters, first perhaps v

Preis 1 First and fourth have long descenders, i.e. p or ¢ Pr rs Before x horizontal at mid-level Prize 1 Feet of two verticals close together 3 ]., small o or ¢? 4 |., ink joining A at mid height, as of € or A 5 Tops of letters, mostly rounded Fr. 24 1 ],, nght foot of perhaps A or pw [, long vertical of p or ¢ visible above next line Fr. 25

1 __[, left- and right-facing feet; then perhapso orc Below, either bottom margin or part of blank area in heading of a new hypothesis

Pese7 3 worw

Pret

Identified by Luppe as part of a hypothesis to the Andromache on the basis of overlaps between Il. 2-3 and 5 and the text preserved in L and other mss. A papyrus text of this hypothesis survives in LII 3650. Although the line-length of c. 27—30 letters cannot allow a text in precisely the same order or necessarily as complete as that in L or 3650, the coincidences are convincing.

2-4 malpayev[opevoc ... TavTnv ev amjnyayely meicalc Neortodeuw 8 emeBovdAcevce|y dv cali ..., with ‘Opéctyc perhaps preceding 7apayevopevoc.

5 ITnAe[i.

22 EURIPIDES

Fr, 2 ii

1-2 E.g. row]/ovrouc.

3-5 The heading for the next play, the Bacchae, is indented in the usual way for such hypotheses. Spacing suggests that #j«w could easily have fitted into |. 3, but that the line was intentionally indented from the right as well as the left.

3 Bal«xat: the title as preserved by P and most sources; [TevBevc L.

4 Since the low trace of « can only be from Aiéc, the word must have been widely separated from the preceding and following words.

TH[v]d[e OnBaiwy: L P; OnBatav grammatici quidam: OnBalav «ara fere codd. Prisciani.

6 1. mpocyKovtec.

6-7 Ardvucov of mpochovtec od« Epacav elvar Hedv P. The first word of the papyrus hypothesis can hardly be other than Avédvucov, as in P, but the antepenultimate letter is not obviously c. év O7Baic om. P.

A main verb with a plural subject is required at the beginning of |. 7, but since the e is clear, the pap. did not have éfacav. The cev ending appears to be singular, which must be wrong; perhaps od |/« &f|ycev, |. éfacav; the spacing at the line-beginning would probably be too tight for od« to start the line.

g-10 rac Trav OnBaiwy yuvaixac, dv ai rod Kadpou Ovyarépec P: trav and rod om. pap.; see J. Diggle, KPE 77 (1989) 9-10.

11 O[a]cuwo[dc], wrongly: Apacouc P, Aidcouvc Elmsley.

jyalyov: etcnyov P, e€jyov Kirchhoff.

12 q[ploc: emi P. ely odv Diggle, cf. <PE 77, 2 ff.

12-16 Kadpoc ev ... not paralleled in P, which after Kifaipdva carries on ITevfevc de ....

13 [av] 7dn ... CewéAn[c matHp Luppe.

14 én 7[7 cuyye]veca Luppe.

15 opye.[: part of dpy.a or dpyradfw. Cadmus is presumably still the subject here, so that this refers to his reverence for Dionysus, cf. Bacchae 178-89. Shortly after, however, the subject must change to Pentheus, as in P.

16 It is tempting to read [TevOedc 8€, 6 /tH[¢ Ay]avy[c malic, tapakaBwy tHv, but the remaining ink and the spacing in the first half of the line are not convincing. tapaBwv P, tapaAaBwv Elmsley. Dr Diggle suggests as an alternative t]avt[yc, with Agave mentioned in the preceding sentence; having omitted the preceding sentence, P would have had to insert the name here.

17 1. BactAeiav. The line, if supplemented from P, is too long by g—10 letters; perhaps tovroic instead of roic ywopevorc.

19 Dr Diggle points out that P has aAAouc, not dAAwe as previous editors have claimed.

Ene The text on the front indicates that this fragment may come from the upper part of fr. 21. 2 aluroc de, Diggle.

Fr. 4

Identified by Luppe from the names in ll. 6 and 8 as part of a Bellerophon hypothesis. This would fit the A—B sequence of plays represented in these fragments; a fragment of a Stheneboia, another Euripidean play dealing with the same theme, is less likely among these fragments. Parts of a Bellerophon also survive in LIT 3651 (see introd. there), but there is apparently no overlap in the texts.

1 E.g. ric culudopac [.

4 peravon| ? Cf. weravorjcac in hyp. Andr. 4, wetevoncev ibid. 12, Diggle.

6 CbeveBora[

7 totouc TovT[ou]c?

8 Bed}Acpol do |vry.

Lives The blank spaces at the right of 2 and 4 show that this is the beginning of a new hypothesis. If the layout were the same as in the Bacchae hypothesis, 1. 1 would be the last line of the preceding hypothesis,

4017. HYPOTHESES TO EURIPIDES’ BACCHAE AND OTHER PLAYS 23

c in |. 2 would be the last letter of a title (perhaps a masculine singular name), followed by od, Ac or av apx7y some distance to the right. Ll. 3 and 4 contain the first line of the play—unfortunately not identifi- able—with 7 6’ trodecic presumably again some way to the right.

Frr. 6, 8, 9, 11-12, 14, 16, 17, 19 and 21 are all blank.

Pr 13 2 |BAérw ?, Diggle.

H. M. COCKLE

ITD MENANDE Ix

This section includes items which can be assigned, certainly or plausibly, to known titles of Menander.

In referring to the plays, I use Sandbach’s line-numbering. ‘B’ indicates the Bodmer Codex, ‘C’ the Cairo Codex.

I am indebted for help, advice and ideas to Dr C. F. L. Austin; and for various corrections and suggestions to Mr P. G. McC. Brown and Professor R. Kassel.

4018. Menanver, Dyskolos 529-31, 557-61

86/104(a) 6.2 x 6 cm Fourth-fifth century

This scrap represents the top outer corner of a page from a parchment codex, with an upper margin of at least 2.5 cm, and a side margin of at least 2 cm. The parchment is fine and papery. The main hand used an ink which has now turned dark brown; in places it has penetrated right through the parchment, or even eaten it into holes. A second hand (blacker ink) added the numeration on the recto, a third (thin pen, ink that is now pale brown) added the lectional signs, which include grave accent, rough and smooth breathing, elision mark and diastole (558). The script is small (1.5 mm high), upright and foursquare, with heavy shading and occasional ornament (blob-finials on the upper strokes of « and y; elegant sloping finials on the foot of p and the apex of a). It has some likeness to the Biblical Uncial in its later phases (note « with the branches separated from the trunk); but some letter forms (looped a, uw with rounded bow), and infringements of bilinearity (p breaks the lower line, as often; but o floats above it), are alien to the classic form. A closely comparable hand is that of III 411 (GMAW 71), which Cavallo assigns to the beginning of the fifth century (Ricerche sulla maiuscola biblica 1 73).

The recto contained Dyskolos lines 529 to 556. That makes a written height of c.14cm, and a page height of c. 20cm; the written width can be estimated at 11-12 cm, and the page width at 15-17 cm. This sort of format has parallels enough: see E. G. Turner, Typology of the Early Codex 28.

A new (but inferior?) reading in 529.

4018. MENANDER, DYSKOLOS 529-31, 557-61 25

The recto carries the numeral ‘27’, set off by curlicues, at the top right. This might be a page number (nothing similar survives on the verso, but if the numbers were consistently to the right of the page, it would be lost there), or a quire signature. The earlier part of this play, 528 lines, would have taken up some 19 pages; if then ‘27° refers to the page, and assuming that the pages were numbered consecutively from the beginning of the codex, Dyskolos must have begun on page 8. But the remaining seven pages would be too little for another play, and too much for prefatory matter. That makes it more likely that ‘27° identifies the quire; see on this Turner, Typology 77.

529 petectpedouny te mhvixa B. 4018 had érnvixa, as the rounded trace and rough breathing make clear; before that, v preceded by an upright on the edge, allowing ]yv (the apparent vertical bisecting nu is the stem of rho on the other side; the dots above are probably accidental, not deleting). The indirect interrogative suits the grammar; but it seems a pity to lose the liveliness of tv.

558 The trace allows a[xazpoc.

559 The trace allows peip[axcov.

P. J. PARSONS

4019. MeNnanDER, Dyskolos 740-50 118/54(f) 6 xX 12.5 cm Third century

On one side, parallel with the fibres, remains of cursive. On the other, across the fibres, line-beginnings from the foot of a column, intercolumnium of at least 3 cm to the left; the text is easily identified as from Dyskolos. The hand is a rapid informal degeneration of Severe Style; no lectional signs are visible, not even paragraphoi. The nota personae in 748 is, or may be, by the main hand.

It might be guessed that a badly written copy on the verso, with wide margin preceding, might be just an extract from the play copied for practice. But the presence

26 MENANDER

of a character name goes against that. For other complete plays in verso copies, see LIX 3968 introduction.

The only other source for these lines is B, in which a tear has removed the first few letters of 740-45. In almost all cases 4019 confirms the supplements commonly printed; but not, seemingly, in 740.

inh ae eee [

740 |taxAwovpe B, c. 6 letters lost at the beginning. xa]rdxAwov seems inevitable, with a trochee (or its resolved equivalent) missing at the beginning. Editors generally supply aAAa (Fraenkel: viv dé ed. pr.). But in 4019 the first trace looks like y or better 7 (the cross-bar seems to project slightly to the left of the upright; any further leftward extension would be lost in a patch of damage); then perhaps the lower part of an oblique ascending from left to right; then scattered traces on increasingly damaged fibres, the first perhaps part of an oblique descending from left to right, the rest so uninformative that they are left out of account in what follows. rax[d cd Austin (and rayd <c¥> in B). rad[ra Rea (which would suit the spacing in B), a single-word clause meaning ‘That’s it’, ‘That’s all I have to say’: he compares the private letter I notes that in Comedy this idiom normally means ‘I'll do it’ (sc. moujcw), cf. Epitrep. 461 and Neil on Aristoph. Eg. 111: here it would have to be spoken by Gorgias, and so anticipate uncomfortably his commitment in 748.

741 7Aeo [ acceptable, the last trace ink on an isolated fibre.

742 .[, high ink on disordered fibres.

743 .[, upright, 7 acceptable.

744 Jur’ avtoc B: jv av, 0J¥8 adrovc edd. In 4019, 7 is represented by remains of two uprights on crumpled fibres; 6 is certain; the horizontal trace following does not suit a, and should perhaps be taken as an extension of the cross-bar, possibly joining an elision mark to the right.

745 JoAdenoc B: ote 7] Xeno edd.

746 add[c] acceptable (of a the top of an oblique descending from left to right; of A the left foot).

747 €k possible but not verifiable.

748 a, an oblique rising from left to right, 6 acceptable.

749 Taxicr[: similarly rayicr’ evpew B: rdyuc8’ edd. Did both papyri ignore the aspiration? or did 4019 have scriptio plena, taxict[a?

750 Traces, presumably of a nota personae, in the left margin.

P. J. PARSONS

4020. MENANDER, EPITREPONTES: Hypothesis a7 4020. Menanpver, Epitrepontes: Hypothesis

71/3(b) 4.7 X 11.5 cm Second century

On the back of this scrap, in a coarse hand, are line-ends from an account in drachmae. On the front (written parallel with the fibres), 7cm from the top, begins a text in a small, neat hand, rather featureless except for occasional serifs, to be assigned to the second century. Above stand five beginnings in a larger script; the first letter of 1, and all the letters of 3—5, written with a thick pen, or perhaps rather overwritten (note the thin final nu in 5), to give a heavy blotched effect. Since 2 reads emitpe[, and 3-5 contain Menander fr. 600.1 Kock, which is normally assigned to the opening scene of Epitrepontes (fr. 1S), and 6 ff. plainly refer to the content of the same play, we must be dealing with an hypothesis preceded by an ornamental heading. I have found no parallel to this lay-out (on the detail, see 1 ff. note). 1V 663, hypothesis to Cratinus’ Dionysalexandros, is superficially similar; but there it seems that the calli- graphic ‘heading’ represents the title of the play itself, around which the hypothesis text was later added (see E. W. Handley, B/CS 29 (1982) 114 and pl. 7).

There is no means of telling whether our papyrus belongs to a copy of the play, or to a collection of hypotheses, or represents (given the ineptness of the ornamental script; note also the enlarged initial in 2, documentary style) a short copying exercise. Dr Coles wondered whether the heading might be (at least in part) the work of the same hand as wrote the account.

The only well-preserved Menander hypotheses, X 1235 (Koerte I pp. 146-50), follow a pattern which we could reconstruct as (a) title (b) &v (or the like) dpyy (c) first line (d) didascalic comment (e) 7 8 tadBecic (f) plot summary (g) critical judgment. Of the other scraps, XX XI 2534 (CGFPR 202+ 111) contains the end of one hypothesis, with (f/f) and (g), and the beginning of another, with (a), probably (b) (restored in lacuna), (c) and probably (d) (ézro.76[ does not seem to suit a plot summary, though what follows might; see the discussion in CQ 59 (1965) 56 f.); PIFAO 337 (CGFPR 203 + 119) has the end of (g) and a beginning with (a), (b) and (c). 4020 follows an irregular pattern: (a), then (c), in the heading; (g) in the main text. But the form of (g), which discusses the characters of the play by function, not by name, has parallels in 1235 95 ff. and CGFPR 203.

I am much indebted to Adam Beresford and Mark Pobjoy for detailed and illuminating discussion of the text.

cat

EMITPE| Enurpé[movrec ovxot[ ovx 6 T, podi— pLocco| [0c CO,U, Tpoc

1 __{, first, 7 or possibly 17; second, foot of upright hooked to the right?

28 MENANDER

5 Bewv | beay \ A A > f. Todpapatwval| TO Opaya THv al pictwv mepuye ovev al Tepryeyovev al nOwv mtavtwr| HOV amavTwy | dvoTovpevcwd| dvo, TOV ev cwo| povwc 10 darcyuvopeva| om aicxuvopéevan| c JamernvKocuiw| ylaperny Kocuiw|c Jeratpavadedwc , [ eTaipav afpedadc _ [ |PiAapyvpovroyicu| diAdpyupov Aoyicu[ ]..amovrad,.. [ — OJepdrovra 8. [ 7 ye., lower part of upright, narrow band of stripped fibres above and continuing to the nght nals probably 7, the left-hand extension of the cross-bar shortened by damage; possibly y, but no certain example of this letter has the cross-bar projecting to the left of the upright 12 _[, upright with high cross- bar projecting a little to the left, 7 or possibly y (see on 7) 14 ]_., first, upper left-hand arc of circle, more ink at mid-height to right (would suit cross-bar of e) 6. __., first, parts of upright

1 ff. In 3-4, the quotation fixes the line-lengths; to judge from a traced reconstruction, 4 must have projected to the right of 3 by nearly two letters. If 2 contains the title em:tpe[zovtec, that line projected by nearly four letters beyond 3: i.e. the scribe could have made equal lines by writing oc at the end of 3 instead of the head of 4. I do not know how to account for these irregularities, except as ineptness or if 2 was abbreviated. One explanation would be this: 2 contained the full formula expected from other hypo- theses, emitpe[movtec wv apxyn, and what followed this long line was set out in two narrow parallel columns. But this seems otherwise most implausible. Why chop up the first verse in this way? What occupied the second column? Even if it projected downwards into the blank that divides heading from text, there would hardly be space for a list of characters, perhaps not even for a didascalic notice. (The elaborate arrangement of PAnt I 15, see CGFPR 240, is different.) I conclude that dv apy was omitted.

1 After «, perhaps 7, perhaps cr; then perhaps the foot of an upright just below line-level. The writing is substantially larger than in 2; and if that is to be reconstructed as the title, I do not see what to do with this. Possibilities: (1) the title copied twice—but then line 1 would have projected substantially to the right of line 2; (2) a number, see e.g. 1235 106 f. and CGFPR 111—but I see no ordinal that would fit; (3) an alternative title (such as existed for other plays, Gomme & Sandbach p. 130)—but I can think of nothing more plausible than Ez.[tpo7 7], which would at least make a line of the same length as 2.

3-5 Epitr. fr. 1, thus confirmed as the first line of the play. The small blank at the end of 4 suggests that the final Ovjcuwe was omitted (too long to be concluded in this line alone).

5-6 A blank of nearly 2 cm. There is no trace of the formula 7) 8’ d76é8ecuc, which in a typical hypothesis introduces the plot-summary; even if it had been centred above a double column, the beginning should show. But in any case our text omits (or postpones) the plot-summary.

6 ff. 1235 96 ff. reads ro de b[paua ra@v]|a[pictwy: Ex]er dé mp[ecBUTHY] |ed[dpynro]v «rA—the characters listed by function, not by name, asyndetically, with one or two adjectives to describe their personality. So here in g~12, but with adverbs (so that one or more participles must be supplied).

6-7 Two possibilities. (i) write t@v a[picrwv, and end the sentence there. (ii) Write tay a[AAwy (...) |epuyéyovev.

(i) 1s the formula that Wilamowitz restored at 1235 96 f. (cf. CGFPR 202.5). He will have had in mind the thumbnail verdicts known from various medieval hypotheses to plays of Sophocles, Euripides and Aristophanes (listed by Achelis, Philol. 27 (1914-6) 132 f., and summarised in P. T. Stevens’ edition of Euripides, Andromache (1971) p. 27). All follow the pattern 76 (...) dpaua trav ..., without verb. dpictwr is not actually attested, but suits the style (and the spacing in 1235).

4020. MENANDER, EPITREPONTES: Hypothesis 29

Since the formula suits the genre, we should adopt it. In that case repuyéyover, ‘it is superior’, belongs to a new clause; since this clause should explain the verdict, yd[p makes a tempting reading (the trace suits y, or perhaps 7). In that case, a[pécrwy ends its line, and that line was no longer than 4.

We have therefore to assume that the text, like the heading, was set out in a fairly narrow column. For example:

(3 \ / mepyeyovev yal[p defer nav aravtwv, [é€xov SovAouc dvo, Tov pev cwh[ pdovwe, Tov 10 & aicyuvopevw[c dixalovra, ylawerny Kocutwl[c épacav, « / > a , éraipav apeAdc, y|épovta / \ ” girapyupov Aoyicu| ov Eyovta,

Beparrovta dik. [

Thus the characters come in contrasted pairs, the title characters first; the fragment breaks off before we reach Charisios. But there remains a difficulty about the line-lengths. If we take 4 and 6 as standard, 9 and 13 would project two letters to the right, and 8 and 10 more. 8 is the crux: unless I have misunderstood the construction, we need both a word to govern the following accusatives, and a noun to go with 6évo, and I| do not see how to do with less space. It is true that, if we take 2 emitpe[movtec as the norm, even 8 would just about conform. But the result will still be a notably irregular right-hand margin. You may account for that by the scribe’s evident desire not to divide words over the line-end. But these irregularities are not cheering. The notes which follow assume the short line. But 8 may be a warning that all the lines were longer; even if 7 was self-contained, it could be restored at greater length (say, ra@v a[yav émuterevy- pevwr, cf. Arg. A4 to Aristoph., Pax, p. 3.27 Holwerda, etc.).

g-1o Syriskos and Daos. The suggested restoration leaves room for doubt. (i) g comes out a little longer than 6. (11) In its only attested use, DH 7.50, aicyvvouévwe is linked as a synonym with cwdpdvwe. But here we expect, if not the direct contrast which Smikrines’ verdict might justify, at least some differenti- ation. On the other hand, there seems no philological reason why, in the right context, the word should not mean ‘shamefully’.

11-12 Pamphile and Habrotonon. It is a question whether the adverbs refer (i) to their conduct, or (ii) to Charisios’ conduct towards them. If we fill the end of 10 with a participle, we have no room for veaviay or the like; and in any case there is no room for veaviav tHv prev, which looks like the minimum requirement. | therefore prefer (i). What should the participle be? I had thought of épa@cav; Professor Kassel suggests A€éyoucav, because 780c shows itself above all in speech.

12 adeAdc: ‘frankly’ (Epitr. 432)? or, if there is a contrast with 11, more ‘openly’ than befits a slave (Theogn. 1212)?

13 Smikrines. We expect his entry to begin with the noun designating him. This must come in 12. That leaves no room for a separate participle (say, 7[paccoucav) to go with ddedac. y[€povta seems likely, although the trace might equally suit 7[. I propose y[€povra ... Aoyscu[ov éxovra or the like (Adyicu[a sroowvta Kassel) because (i) y[épovroc ... Aoyecu[dv would spoil the run of personal accusatives; (11) y[épovra ... Aoyicu[ Ge sounds forced; (iii) éyovra or the like would continue the string of participles.

14 Onesimos. dcx is likely, then faint traces. One could think of d/cas[ov, to agree with the preceding noun, or with Aoy:cu[dv, if Smikrines and Onesimos are presented, like Pamphile and Habrotonon, as a pair (cf. Epitr. 1078 ff?). d:cas[oAoyodvra Kassel.

P. J. PARSONS

30 MENANDER 4021. MENANDER, Epitrepontes 150-164 etc. 87/331(a) Fr. 15.5 X 9.5 cm Third century

On one side, these scraps carry cursive writing parallel with the fibres: frr. 1 and 2 seem to belong to the same document, fr. 3 has line-ends and beginnings from two other documents, joined together in a synkollesimos. Fr. 1.15 preserves part of a date- clause referring to the sole reign of Caracalla (AD 213-217) or the reign of Alexander Severus (222-235).

On the other side (verso), upside-down and across the fibres, are remains of lines suggesting comedy. The slovenly script has strong cursive tendencies (a often as an open hook), with many ligatures; deep-bellied p, flat-based w. In itself, it could be assigned to the second or third century; the recto document shows that it cannot be earlier than the third. In fr. 1, change of speaker is indicated by paragraphos and space; dicolon perhaps in 157, not visible in 155 (but on damaged surface); at 160 the space is filled with an oblique stroke in paler ink. The text-hand added abbreviated names of speakers in the margin and above the line. No lectional signs survive, except for elision mark and (fr. 3.10) diaeresis. In general, this looks like an amateurish copy; note the itacism in 161, and the doubled paragraphoi in 155-6 and 160-1.

Fr. 1 twice offers the nomen personae aBpo’: if this is Habrotonon, the likely source is Menander’s Epitrepontes, for the name is rare (known otherwise only as a bit part in Pertkeiromene). A coincidence of text confirms this: fr. 1.10-15 evidently provide beginnings for the headless lines Epztrep. 159-164, fr. 2 odd letters from the latter part of these lines. For the lines as so far known we depend wholly on the Petersburg parchment (P). This leaf contains 127-148 on the recto and 159-177 on the verso; its lower part is missing, and the lacuna appears in the conventional numeration as ten lines, but may in fact have been substantially more (see below, fr. 3 note). 4021 frr. 1 +2 now add the beginnings of 150-158; confirm some but not all of the standard supplements in 159-64; and show that, as most editors have argued, Habrotonon, Smikrines and Chairestratos are all on stage at this point. Fr. 3, line-beginnings appar- ently from a monologue, remains unplaced. The sloppiness of the script, and surface damage in frr. 1 + 2, make the readings more than usually unreliable.

Ero Fr. 2 150 | haa

| Tpoce|

eee oe ee [

| amrodou , 6[

JaBpo’ tm, pe, [Jo

4021. MENANDER, EPITREPONTES 150-164 etc. oi

evcuyuoour | [

]

lew Tarra. [ |

|

]

Bret

150 ___[].[, second, long descender as of 1, p 153 v_[, scattered ink at line-level 154.7. .p, upright, junction at foot, perhaps v if the oblique trace at the top right of » forms the beginning; then perhaps 7, but unexplained oblique dash above the right-hand side; then p (or possible B) likely, but unexplained ink (parts of circle?) above 155 oy, ink above y to right, i.e. y’? , top of upright Ca probably a pin-headed p 157 .[, perhaps left-hand branch, and base, of uv ]., low ink, then high trace (right-hand end of horizontal?) 158 aA’, the elision mark is enormous and might have been taken for part of a letter; unexplained ink above v (breathing? variant?) —)_[, parts of left-hand arc of circle? #[ perhaps not excluded —_158-g thick paragraphos (double stroke?) 159 __ v, parts of circle? [, top of upright 160 _[, perhaps top left hook of v 160-1 thick paragraphos (three strokes) DOVE |) top left-hand arc? 162 _, tip of oblique sloping down from left to right [, parts of high horizontal 164, [, faint trace on darkened surface

[did 158 ]_, two points of ink ranged vertically on the edge (branches of x, ¢, x?) 161 |. €, second long horizontal at mid-height, short vertical trace below (7?)

The known text in 159 ff. fixes the relative positions of frr. 1 and 2. The surface is much damaged. In fr. 1.154~7 the last two traces of the line stand on a scrap which is attached only by straggling fibres (I have considered placing it a little lower, so that the traces assigned to 156 would attach to the character- name below, giving xa:; but there seems to be more ink than would be expected of a typical abbreviated form). Many traces on fr. 2 are too ambiguous to describe, and the traces on the extreme right are on a partly detached vertical strip which is too fragile to straighten.

152 I can make nothing of the remains. Perhaps , 7, [, the last trace only a point at line level; but uw might be Aa or another similar combination, 7 would be anomalously shaped. At the beginning, at least two overlapping horizontals, one or both of which might be a paragraphos, but rather low in the line; then a high curve that might represent the joined upper loop of €. «un, Kady equally unsatisfactory.

153 azrodovc 6[ possible.

154 ABpor(ovov). The name in fact stands a little higher than the verse itself, ranging with the paragraphos.

Perhaps r7v tpotka,[, but y would be cramped, and this reading does not account for ink above 7 and p (suprascript letters?) and between p and the putative o. As sense, it could combine with 153 and 155 (if rightly read), ‘The old man’ wants Charisios to ‘give back’ his daughter and her ‘dowry’: the dowry is already an issue in 134, and Smikrines will come to claim it in 1079.

155 oyepw, : this continues Habrotonon’s remark (there is no trace of a paragraphos between 154 and 155). 6 yépwy is tempting; if right, it shows that here at least Habrotonon and Chairestratos talk about

32 MENANDER

Smikrines, not to him. But the reading does not explain ink to the top right of y, which might suggest Oy épav.

Then Smikrines speaks: cw suits the suprascript letters, perhaps cf rather than cyt. épwral, epw Talv, épwrw[v (the first w vestigial and doubtful; the second substantial, not a)?

156 Ink in the left-hand margin, but I cannot reconcile it with any of the expected character-names. Smikrines begins a speech in 155, Chairestratos in 157; the speaker changed again in 156 (paragraphos). At simplest we could assume a speech of Habrotonon in between. But of course Smikrines may have resumed towards the end of 156.

dcxod might be read (the noun; the verb seems rare in the middle; cf. 127 ff. for wine-bibbing); or dyxood (not suited to the style?) or dyxou (but elsewhere amdyyxecBar).

~~ In ]_«a,ov, the first trace is very vague; after «a, the foot of an oblique descender, with more ink above, some apparently joining o at its top left—too much, it seems, for « or p, and wrongly spaced for r.

Professor Kassel notes that ddd’ od [8]écavov would be plausible in itself; but the first A, at least, looks impossible as a reading.

157 Thc vu[«]7dc? Charisios’ nocturnal dissipation (136)? After that, a point near line-level: remains of a dicolon? A new speech, headed X[aip(éctparoc) |, follows.

Fr. 2 provides |Bpo[; to judge from 159 ff., this should come about 20 letters from the beginning of the line. Since Habrotonon speaks next, it’s likely enough that we have a vocative A]Bpd[rovov, ending with the fifth or sixth foot.

158 ABpdr(ovov) seems certain from the f and the raised 7; the putative p is intersected by a rising oblique, as if aB/ had been the intention at one time.

AX’ odk éxad [? (éxado[ looks possible, but not é«aAe[) But this reading does not explain the extra ink above v (unless it is a breathing).

159-164 survive headless in P. Joining the two texts, and adding the traces on fr. 2 (often very uncertain, especially in 162, and those on the far right uncertainly placed) we have:

LA > / / \ / ovTwe ayaboy , Te cor yév oLTO, fL.7) A€ye 160 aero, (Xaup.) od, K ec KOpak jac; olww, €er paxpa. (Cw.) eicKe>yue 0° ody €, icw cab@ jc TE TV, Gopevoc

ov A A He / amavta T[ad |, ta tHe Ovy ,atpd,c, Bovredcopar

acw €v aUT@ TOD TOV HK, ovT’ evOdde; Pacw| ey, 4 {TOV HK.

This confirms the supplements proposed for 159 and 164, but corrects those in 160-2; in 163 not enough can be read to confirm or disconfirm 6vrwa.

159-60 There is no paragraphos below 159; that implies that the beginning of 160 continues the speech from 159. I can make nothing of it. (i) The ink. aero, _, the penultimate trace has the shape of a small nu, but there is more ink above it, touching the right-hand upright; then the tp at line-level of an oblique rising from left to right, which apparently continued over a small gap as a thick rising stroke; below this, in slightly lighter ink, is a thinner rising stroke, which I have taken to denote change of speaker (see 4022 introd.). (11) The context. What do we expect from py Aéye? It could be followed by an accusative (fr. 612.1; Men. et Phil. 1.217) or a clause, tic AcBa (Men. et Phil. 1.259), 67u (3.27); it might mean simply ‘shut up’ (Aristoph., Vesp. 37, Pax 648). On the face of it, odrwe «7A attaches to it as a formula of request; but we could also take the formula as object, ‘Don’t say “Bless you ...’’’. Elsewhere in the play, the phrase is put in the mouth of Syriskos (264 f.); and see fr. 3.7—8 note.

Since @ edov is excluded by prosody (and in any case leaves the last trace unexplained), we seem reduced to de¢é. But what then? My only idea was ore; but then (i) 7 does not suit either the main trace or the suprascript; (11) the phrase (see editors on Dysk. 426) seems generally to belong in general statements, not commands— unless here it begins a new sentence, which Smikrines breaks off in order to get down to business (6° odv).

160-1 The papyrus gives the curse to Chairestratos, presumably addressed to Smikrines. Then, clearly, Smikrines speaks 161-2; presumably his name is concealed in the left margin of 161. If we assume that Chairestratos would not address Smikrines with the same freedom that Smikrines uses towards a slave

4021. MENANDER, EPITREPONTES 150-164 etc. 33

(1068), his curse here must have been an aside; and that raises the question whether, even at the end of the scene, Chairestratos and Habrotonon converse with Smikrines, or simply comment on his monologue.

Frag bol eh Feph te...sal J.ov.a. nl }. parpeBou| 7 |aparpiBop| 5 _|wpacava, | Joyapr[ Joa. [ ].aAda.n...[ Mpalleverss [ a |yaba yévoulro _Jabevd’avac [| K|abevd’ avac. [ 10 Jicacunveyw, [ icacw nv exw, [ Powe Ren partes | |] ayabovyevo[ ayabov yévole | eAarevdewor | eAdArer depor | | avrovede € [ avtov eUéde | oO 15 ] vpacer] |yAel dpac evoyAe| | ,vée deou, [ ovbev Seon. [ }.@7_ xv [ jovece:.,f ]. varx[ 20 |poarro[ | daxo[ I €., upright n{, vy (7) plus left-hand are not excluded? 2 ca, perhaps za not excluded 3 |., ink at line-level a, loop of ¢ or lower loop of B? 4 |., a acceptable 7 a., horizontal joining at mid-height _..[, lower arc of circle; foot of upright with serif to left; lower part of upright on the edge? 8 _[, spot of ink at half-height g c.[, unexplained trace to top right of ¢ (possibly diastole?), not certainly ink; then 7? or left-hand part of 7? 10 Of the first iota, the top of the upright and the left-hand dot of the diaeresis _ny, if right, in ligature {, high spot on the edge riggs se left-hand end of horizontal just above line-level 13 po, unexplained ink above [, trace on edge (left-hand end of horizontal or descending oblique?) just below letter-tops 14 €., oblique feet as of A [, perhaps top of upright ligatured to e 15 |, ], blotted letter; alterations also to following y

(darker ink) 16 _v, left-hand arc 18 [, 7? or part of 7?

34 MENANDER

Some suggestions of dialogue. But no paragraphoi can be seen in 12~15, where the line-beginnings survive, or in 3-11, where the right-hand end at least would be expected to show. Of course, we may have a continuous speech which quotes a conversation. One character begs (8, 12), another dismisses him (7, 9)?

Fr. 3 was copied by the same hand as frr. 1-2. The chances are that it came from the same play. It may even have come from the same scene, if the repeated dyaba (—Oov) yévorro is significant (see fr. 1.159 note). But in fact the phrase recurs in other scenes (below, 7-8 note).

In the immediate context, there are physical arguments. (i) 4021. Fr. 1 has the foot of a column. Fr. 3 cannot belong to the same column, since the documentary texts on the backs are different. At closest, it could belong to the preceding column, or the following; if the former, we need to allow for a lacuna, i.e. the lines which originally stood above fr. 1; how large a lacuna, we have no means of determining, but presumably not less than 15 verses, since at this narrow spacing the column must have contained at least 30 lines. (ii) Membrana Petropolitana. The lower part of the leaf is lost. Early editors guessed a page of 30-35 lines, and therefore estimated the lacuna at c. 10 lines (149-158, 178 ff.); this is still implied by the conventional line-numbering. But Turner produced parallels for a codex-format, in which the page would hold c. 50 lines (GRBS 10 (1969) 311 f.; Typology of the Early Codex (1977) 112 no. 227(a)). If we assume that, the gap between 148 and 159 amounted to c. 28 lines. Of these, fr. 1 provides beginnings of the last g. If the 21 lines of fr. 3 immediately preceded, they might just fit in the lacuna, or just overlap the headless lines 145-8. But the argument under (i) suggests that they cannot have come immediately before; and if we do allow for (say) another 15 lines between, the top of fr. 3 should substantially overlap lines 141 ff. Since I see no way of splicing the two, I conclude that either the page in P was even larger than Turner suggested or fr. 3 does not beloug here.

Alternative placings would be: (a) in the lacuna before 127 or (b) in the lacuna after 177. With (a) we might visualise Smikrines relating a conversation with Charisios (how else did he know so many details, cf. 138?); but that goes against the normal assumption that at 127 Sm. ‘can hardly have been long on the stage’ (Gomme & Sandbach 294). As to (b), the yévouro formula might connect this speech with Daos (264 f.); but again it is normally assumed that 218 is the first line, or nearly, of the scene. That leaves the opening of the act, a monologue (it seems), perhaps by Onesimos (or by Smikrines?); or of course a scene totally lost in the gap between the end of P and the beginning of C.

3 |, ovca possible.

4 tJaparpBoyu[ seems inevitable (the spacing would allow the line to begin with Ja; but clearly dpa is excluded by the metre, and dpa as a first word).

5 xl]@pac (or ¢]wpac) would be one division; then perhaps avaz[ or avar[ (but the traces might also allow avay[, even avav[; space and trace tell against avacr{, cf. 9).

6 r]o yap m[€]pac [ could be considered, cf. 76 d€ mépac Dysk. 117 and elsewhere. But the final trace is rather angular for sigma.

7-8 a ]|raddayn could be read, and indeed 4 JraAAdyn. A brisk dismissal (Austin on Asp. 246)? Then

[© otrw modAd cou (or por) /dyaba yévorro. For the plural in this formula, cf. Dysk. 300 f. More usually dyaov, see 12 and fr. 1.159, Misoum. 433. Such phrases may reinforce a request (Epitr. 264 f., Syriskos) or an asseveration (Epitr. 1071, Smikrines). Here presumably it goes with the imperative preceding (if rightly read) or following in 9.

8 The height of the final trace suggests yevo.8[ rather than yevort[o.

g Pk. 469 xdBevd’ dreAfdy (Pataikos tries to get rid of the drunken Polemon, in a scene in which Habrotonon plays the flute). Here dvacz[de could be read (or dvacz[?), but ‘get up and sleep’ rings oddly: therefore dvact[n@ Kassel.

12 ovTw tT cou (or prot) |/ayabov yevo[ito, see on 7-8.

13 éAdAer Epitr. 886, Sik. 213. Probably 6€ jou, but the trace allows S€ wory[.

15 It seems that evoyAe [ was the final intention (o written above a blotted a, parts of y in darker ink?); the last trace suitable to iota. jude may be the subject of this verb, or perhaps the object (we expect a dative on the model of Dysk. 693; but the accusative is adequately attested, e.g. Diod. Com.2.18, even if Misoum. 189 is a special case).

16 Pk. 180 od6€ €v/dedpevoc, fr. 89.2 oddev dedpevov mpoikdc.

17 Space and trace allow od] r@ tvxov7[e.

19 ]yvvarx[ could be considered. This reading would not explain a horizontal trace between v and »; but the apparent ink may be delusory.

4021. MENANDER, EPITREPONTES 150-164 etc. 35

20 mpoamo—. Many compounds are available (note Dysk. 391 mpoaroAei 1.’ (edd.: mpoc— pap.); mpoam0A@ CGFPR 272.5).

21 If daxw[v, see Austin on Sam. 356.

4022. MENANDER, Epitrepontes 290-301, 338-345, 376-400, 421-447 88/157(b) c. 8x 12cm Second century

These tattered fragments represent two consecutive leaves of a papyrus codex. The first (fr. 1) had about 47 lines to the page, the second (fr. 2) about 45 lines. On both, ‘verso’ (the side on which the writing crosses the fibres) precedes ‘recto’; Turner observed that this arrangement, | on the right-hand page, and unlike facing unlike, was common in earlier codices (Typology 67). The written surface can be estimated, very approximately, at 6.5 X 21 cm.

The graceless and informal script, ornamented with occasional serifs, could be compared with the hands of 841 (Roberts, GLH 14), Pindar, Paeans, and assigned to the second century. If this dating is right, 4022 should be added to the short list of the earliest codices (Roberts & Skeat, Bzrth of the Codex 71).

In fr. 1], where alone line-beginnings survive, paragraphoi indicate change of speaker; within the line, this is indicated by blank space or by blank with punctuation (stop, 393?, 394; high oblique dash, 382, 396, 435; stop without spacing, 395, 4367). The stops, a rough breathing (391) and the notae personarum (marginal and suprascript), are or may be by the same hand as the text.

Though the ink is sufficiently clear, the papyrus presents great difficulties, because it is much broken and in some places the traces stand in straggling fibres or pieces which are connected only loosely to the main massif. The text confirms some supple- ments, and presents a number of variants. The apparent character-names in fr. 2| are a particular problem.

Freer Fri

Peele L now llth:

36

300

Fr. 375

380

385

age

21

MENANDER

eal Jev[ 345

aS oo

lel oynwle Kar Badile, cv de [] |. |. [ AaBouca mpoc Tov Tp odmoy ,€ jv, ad XalpecTpaTov vujy yap MEV OV, meV eic auptov 6 em ep you e€opuync, opev Tnhv atopopay ; amodovtec aAA, a TavTa LoL Tmpwt amap.l juncar Kab ev ex, ELC KOLTLOG TWA ovnc[ | Badd exc To mpoKo ,Amov | wayerpo,v Bpadutepov ovdeic EopaKke , THVLKAUT Exec 7, adaL c]up

].[ Jouv ovroce pev | adeKT puw,v, Tic Kat wa, A,a cTLd| polc[ Tout O,€ 61a ,Abov Te meXEKU, C OV jT\ OCL Te TAU To Lu ymoxpucoc daxT, v[A]to jc, Tuc ovTOc. auToc c Lonpouc yAupp,a T,aupoc N , TPayoc ovK av, dtayvounc KAEo, cTp jatoc Se | TLC ec|7w 0 70 ncac we Ae, yer TA yp ape, ATA

[

]..0 ] ty ev (6 ec tee ovr joc €¢, Te Tuc

esc!

o Olax, TVA \10,¢, © Trot, oc ov yap pav Ja, vw

[

Tov | decTTOoTOU TOULLO, U Xapccov |" x0, Aatc

al

ov] amwAecey * Tov , daxtuAov , Dec , abAr je

4022. MENANDER, EPITREPONTES 290-301, 338-345, 376-400, 421-447

400

420

ge) 2

440

Ree

onl

to|v nwetepov: co,t Ow mobev 6 avtov, AaBw, v

Re [

ex lec ! amroAA o,v Kae Beor Sewou Kak jou }. ov eco[ | [ m]atdoc: 0 mp, oceADwv

tlov daxrvAL, ov

Tee sceloal sal

i

wv eyyuc n|d]n Kat mpoc, avTw,t TavTEAwe avadvopat Kal TW, TPOTEPW) Vv LoL wETapeEAeL pnvupatewy Aeyer, yap emer, KwWC TUKVA we Tov Ppacavta TavjTA LoL Ka, KOV KaK[we o Cevc amrodeca jt wn pe Oy KaTal[ Tpoc THY yuva tka Tov PpacavT,a Tav[ta Kau cuverdoot a davicy[ | AaBwv ka, Awe [ ETEPOV TL 7 POC TOUTO\L\C KUKAaY Oo. [| Kavtavia , KaKov evectw emEe, iKwe [peya ETE [JL LKE )TEUW CE KL LN LOL K, aKa TAPEXET Ea jUTHV wc, corey aOA[ta | [ AeAnba yAeval jouca ,€ ;pacbat mpoc[edloxw|v fevov de juicer, fru, c joc avOpwrro,c, we Te. [ OUKETL fk € a, t yap, ov, d,€ KaTaxkercBar Tara, v

ovnc map avto vy adda xa, pic, | avd atroéw 7, adw

Al

map ov tapeAaBov apt two atom,o\v | [ OUTOC TL TO COU, TOV apyupLoO jv amo, AAvEL ETE TO Y ETL TOV, TwL THC Dea jc Pep, EL yV KQV 0, UY E404 Y’ 0, LoV TE VUV €\cTW TaAav ay vn \yapjwv yap pac 7 wepav Tpirnv 70) Kk, aOnp jar 7, we av ov,v mpoc Tw Dewy

a0

38 MENANDER

mojcayv |_| | zeal

|TEprepxo, evdov ouToc |. wyabe

Tov dak, TvALov y devéov we EAA EL OTE o|vnec 445 Kpwwped eADew Seu pe trou T OLvouTovEL ect To Tpayp avOpwre Tov ev dectroT, ov

EcT 010 akpiBwe ouTOCL Yapt ;cLov [

293 Paragraphos above and below; abbreviated marginal notes to the left of 293 and 294, most likely notae personarum. At 293 apparently | 1% vdz: the beginning can easily be read as c]ux(puvyc) or cuJerK (pwc); for the rest, Dr Austin suggests (8/(a), ‘aside’, ‘to himself’. At 294 confused traces with a raised letter at the end; this too could be read as «, but also (if part of the ink belongs to the final iota above) as e.g. w. Allowing for blots, cue“ might be possible (but who could have thought he spoke this line?), or sdvw(c); I cannot read what is expected, 1.e. cupick(oc) or avOpax(evc).

C too has paragraphos above and below, and divides speeches thus: eypyKkev: ouxnKoucaceipy Kev: KaAWwC cup. Modern editors generally follow the dicola by giving the first word, as well as the last, to Syriskos. 4022 might be interpreted as giving the first word to the character named first in the margin, i.e. Smikrines: an arrangement already proposed by Lefebvre and Sudhaus.

376 de: the following traces may perhaps allow ra[uzv.

378 e[vlov[uer, the traces (on straggling fibres) suit ov very well, but high ink (suprascript?) remains unexplained.

382 ovnc is a good reading of the suprascript letters (of o the left-hand arc; the right-hand upright of v fused with the left-hand of y, as at 444); and in 384 clup is clear. These then are the expected notae personarum in what editors take to be a two-character scene. oy[n¢e can be recognised again in 395. But I cannot recognise either name in the suprascript letters of 392, 393, 394 and 396. See 396 note.

384 Jouv: emvov:ourocyrevewarpawerar C. Presumably 4022 had a variant ovx«|odv ottoci pev dlaiverar (in that case, the nota personae stands a little to the night of the beginning of his speech, as perhaps in 396). Syriskos has a similar od«odv in 294.

385 crid[: crpidvoc C. Editors have already been tempted to restore ctidpdc, which Moeris, and XV 1803 quoting Men. fr. 389, declare to be the Attic form.

386 The scribe left a space after 71. Punctuation? There are no suprascript letters to suggest that he intended change of speaker.

387 The scribe left a space after ra. Any suprascript nomen personae would be lost in the lacuna.

389 drayvornc: diayvorny C. Editors seem not to have doubted the first person. But the second person seems more idiomatic (and more easily corrupted to the other).

391 qv: above the following letters traces which look more like ov[ than cup’[: i.e. téc 8 ef c¥ was assigned to Onesimos?

The rough breathing is clearly written.

392 Three or four letters written above zo:[, the first two (vo?) apparently deleted with horizontal strokes. I cannot read either of the expected character-names. See on 396.

393 Above yo[ damaged remains, apparently not cv, perhaps 7, [. See on 396. Before it, a high trace of ink, perhaps the tip of a letter, perhaps a substantial high stop (not dicolon, for one would expect to see the lower point as well).

394-7 The line-ends stand on a very damaged and confused scrap, placed to the right here on the evidence of the text on the back.

394 Above tov damaged letters: see on 382.

395-7 Doubtful traces on straggling fibres near line-end, perhaps 395 avt]ov[, 396 dewolv[.

4022. MENANDER, EPITREPONTES 290-301, 338-345, 376-400, 421-447 39

395 Heavy stop after nwerepov; then co[« with o, [ above, the trace vestigial, presumably ov[nc. It seems that 4022 attached rév jérepov to the speech of Syriskos in the preceding line.

396 I cannot read the suprascript letters as either of the expected character-names (382 note). marc might suit, but leaves the last two traces unexplained (in any case, it would not distinguish Syriskos from Onesimos). Dr Rea suggests |7r0« ||cup[: zrou(znv) would have resulted from a confusion between Daos, who was indeed a shepherd, and Syriskos. On that basis, one could try (very doubtfully) to read the parallel suprascripts thus: 392 [70 |ic[up, 393 a[, 394 [roe I<Lup-

397 |. .ov, first, right-hand are of small circle? then two traces at mid-height, perhaps to be combined (since the fibres are distorted) as part of a single upright. ], []cwcae C: ofov] 1d c@car Koerte, ofov a |z[o]c@car Lefebvre. 4022 certainly allows Jovov (though there might be room for another, narrow, letter before 0). But what follows is clearly ye or re, and more likely te, to judge from the position of the upright. That is, ofév re cHcae as a question (Austin)?

The scrap on the right may preserve a high trace from the line-end.

400 The surviving letter tops allow 7o]u Seczoz[.

421-4 Unassignable traces on a loose vertical fibre to the left.

422 mporepw[v: omega, though damaged, is certain: mpotepov C.

425 Kata[AAayeic: diadAay[erc C.

426 Kadwe[ C.

428 Kuxav o, [: kuxav[ C (some have seen further traces to the right; the photograph is indecisive). In 4022 the traces are (1) sigma, or left side of omicron; (2) a point of ink on a high projecting fibre. This excludes the normal supplements (azecyouny, 08 BovAopar, apeopar, Puddéopar, drorpéepopar, ovK ect’ e.dv). KaAwe [roel .. dc aly BAniw? Kadwde [dpa ... dco[y taxoc; (Austin: a question, as at 435 f., Onesimos thinks how to create a diversion)?

430-6 Dr Rea observes that yu in 430, and the ranging letters in the lines below, stand on a narrow strip of vertical fibres—a patch on the surface?

431-3 Unassigned traces on a loose vertical fibre to the left, in 432 perhaps yxA,e,v.

432 epacba:{ C. 4022 confirms the supplement zpoceddxcwyv (Capps).

436 Confused traces from the earlier part of the line, possibly ]eAaBov[. Apparently a stop before atoz[, but no sign of a speaker-name; after |v, probably a stop and traces (the first a long high horizontal) reconcilable with ta{Aac, and above them suprascript remains which might be read aB[por.

438 ee: to y €]: ink from some letters on a lone projecting fibre. Beale: Beov C.

439 €|ctw tadav: ect w tadav C.

440 Croiset’s supplement confirmed.

441 7, wc, 7 represents an uncertain trace on twisted fibres, perhaps to be discounted (we might expect a blank, to mark change of speaker).

442 Unassigned traces to the right of rw]c a[v. roucr[ | .@veyw C: 4022 might be read |ytwveya, with |], |cupe suprascript; eyw altered to ayw with a suprascript alpha.

443 |, wyale: Jwyafe C. azrodo]c possible.

P. J. PARSONS

4023. MENANDER, Epitrepontes 655-65 etc 63 6B.63/B(1—2)b 3.5 X 7.5 cm Third/fourth century

A scrap from a parchment codex, written in a sloping Severe Style assignable to the third/fourth century. Another hand (blacker ink) was responsible for (all?) the lectional signs. Upper margin preserved to 1.5 cm.

The flesh side preserves text identifiable as Epztrepontes 655-65. The text on the hair side must clearly belong to the same play, but no line of it has been identified in what otherwise survives of the context. We have no means of telling which side came

40 MENANDER

first, or how many lines intervened. Ten lines of this small script occupy 5 cm; we might therefore have to reckon with a column (page) of 50 lines or more.

A transcript of this piece, with some notes, was found among Sir Eric Turner’s papers; he had printed a text of 655-63 in CE 54(1979) 120 n. 3 (the same in Actes du VII’ Gongrés de la FIEC I (Budapest, 1983) 254 n. 24).

Flesh side Hair side

ll ecmpalues | (Re es jor xav | 5 eevee) 660 ] .-péctpar |. eco. ‘repipa, [ Lee low we Pere |

Flesh side

In the top margin, doubtful traces. Turner read them as 7a, a page number (less likely a quire number, if the Hair Side carries a similar numeral). But nothing reliable can now be seen.

655-65 This part of the text survives otherwise only in C, on a lacunose page (H. Riad & A. el-K. Selim, edd., The Cazro Codex of Menander [London, 1978] pl. xvii) consisting of two separate fragments juxtaposed by Lefebvre; if there were any doubt (which there is not, since a quotation overlaps the two in 692), our parchment proves the placing correct in 658.

655 |, ,dud[, second probably v, the accent certain; |, _[, shadowy traces partly concealed by dirt, apparently including a diagonal that descends from left to right. rouro[ c. 13 |y-adXicwceyw C. | rodTo,v a tov A.d[vuco]y [ suits the space, and the final trace, in 4023. There would then be a question of syntax, which in turn depends on the division of speakers. Some have seen in C a dicolon at the end of 654 (and a paragraphos below its beginning). If that is right, and if the oath looks forward as at Sam. 139 (as emended), todro[y must stand alone (interrogative); if it is wrong, the oath may also look backward, to a clause of which rodro[v is the last word. It would in principle be possible to join todro[y with 4id[vuco]y; but the word-order and the practicalities discourage that (we find no evidence for statues of Dionysos on stage, as there were of Apollo, see Handley on Dysk. 659). Alternatively, we might seek a different reading Of 705m:

656 | .v@[, two high traces with damage above, well-suited to 0; ] [, dirt, high oblique above (grave accent?). modumpayy[ c. 9. |, TeTwveuwvrpartw C, restored and corrected as oAumpayp[ova mel ]w te mpdtTw Tav epav. 4023 had modvmpayy jova[. Then spacing would allow [Aew 7é], only the accent showing; but we cannot parallel this use of the gravis.

4023. MENANDER, EPITREPONTES 655-65 etc. 41

657 Jo°[, the trace is an oblique so high and so dark that it is likely to be a grave accent, not part of a letter. catadoyovefoval[mev]arryvOvyatepa C. In 4023 xara Aoyov e€ ov { would suit the general spacing; but the accent stands well to the right of its vowel.

658 Jul... |.¢.[, before < the feet of two uprights, perhaps the end ofa mid-height horizontal crossing the second (i.e. 7); after c the left-hand are of a circle. AaBovra touto pe[v mloncw Kae cyedov C (on not verifiable on the photograph).- In 4023 space and traces should allow AaBovta tovto, p,€[v mo ,ncw, 5 possibly an acute accent on », but it is difficult to be sure in the general darkening.

659 Jou... .vav [, feet of three uprights, perhaps more ink to the right touching the left-hand top of x; after v, upright on edge, perhaps trace of cross-bar at mid-height and rightward hook at foot (i.e. €). dedoypevov por Tvyx aver waprupopar C (nothing after ov verifiable on the photograph). In 4023 dedoypevov joe Tvyxave,e would suit.

660 | «pécrpar[, first trace perhaps an oblique descending from left to right. iwacd’opo[ C (the elision mark, and oo, not verifiable on the photograph). ‘The favourite supplement has been 6wo[Aoyeiv (Sudhaus). But 4023 offers X]a:péctpat[, where the last syllable of the name must (as the accent shows) be short by nature; c. 10-12 letters preceding. That suggests tude 8’ ouol,,, , Xalupéctpar’ (the vocative to take up budc); or, if both Chairestratos and Simias are present, X]Jatpéctpat[oc Kal Cystac. To judge from a traced reconstruction, duo[d (and even dualc, if w[ could be read in C) would be short for the gap; dpue[cae (or 6u0[cac Kassel) would fit.

661 “|euya_[, at the end a very short trace, level with the letter-tops, sloping down from left to right; so isolated that it may be a stop, rather than the tip of a damaged oblique. peAwvc{ C (nothing visible after wy in the photograph). In combination: pel’? dv [ c. 6 €|repipa?

662 |_. .av_[, third probably 8, first and second suitable to ay; after v, a point of ink on the edge. @vyarepa[C. In combination: Avyarepa[ c. 6 Alay Bav [. Thus bvyarépa [viv aoAlayBav [| would fit, but not Ouyatép’ a[zoA] (too short) or Avyarépa [rv éunv dod] (too long).

663 Confused traces distorted by a fold.

664 ]’..»_[, perhaps ] ’ywv [, the last y or 7. E.g. Aléywv.

Hair side

Traces in the upper margin, the lower apparently a long thick horizontal: that is, a page number was set off by under- (and over-) lining?

1 ff. All readings are very uncertain; the upper part is badly warped.

1 ].e’c_[, the accent is clear, the letter below (which should thus be a vowel) possibly a or 0; of c the back and lower curve; then tall upright, more ink to the right.

3 ]/ [, indistinct; if the trace below the accent is really ink, perhaps the sloping back of a.

4 ]., possibly o, but the trace may be delusory. | [, upright on the edge? but more likely a phantom produced by folding and staining.

5 6.. .[, of 8 only the base and the apex; then short horizontal or narrow arc at line-level; then perhaps c; then perhaps an upright trace. E.g. ]ev deca[or—? (But metre excludes an overlap with 629.). evdeec, evdeec(tep— less likely (the third epsilon could be read only if its cross-bar were lost in damage; but the parchment surface looks relatively intact). In any case, not 682.

6 |. ac, first perhaps 4, second dispersed points of ink, « not indicated but perhaps not excluded (C:]uiac). 6. [, first perhaps c (or @, ifa trace to the right represents the projecting cross-bar); then shadowy upright on the edge, most likely not ink.

7 Much damaged traces; |, vy Aoyov [ not excluded.

8 ¢, or perhaps c.

11 |’. _f[, for the first a 7 « w seem possible, o perhaps most likely; then probably 7.

+E. G. TURNER—P. J. PARSONS

42 MENANDER 4024. MrNANDER, Leukadia? 103/210(b) 10X7cm First century

On one side of this morsel, cursive traces run parallel with the fibres and at right angles to a sheet-join. On the verso, iambic trimeters punctuated by paragraphos and double point. The scribe wrote a graceless informal hand to be compared with XXV 2435 (GMAW 57) and assigned to the first century.

Line 5 coincides with Menander fr. 686 KT, quoted by the Etymologicum Genuinum, without specifying the play, for the rare word Caxopoc. The same entry quotes the same word from fr. 112, Dis Exapaton, and fr. 257, Leukadia. Given the references to 7érpa (2, 8, 102), and perhaps to a temple of Apollo (1), there is a good chance that our fragment comes from Leukadia.

On that play, see most recently K. Gaiser, Menanders Hydria (1977) 445 ff.; E. W. Handley, BICS 26 (1979) 85 ff. A few quotations survive, frr. 255-62 KT; and some fragments of Turpilius’ version (ed. L. Rychlewska, Turpilit Comici Fragmenta, Teubner 1971). The Mytilene mosaics include a single scene from the play, with no indication of Act; the central figure, female to judge from its clothing, wears a crown and carries a palm branch, and has therefore been interpreted as the priestess of Apollo (S. Charitonidis, L. Kahil, R. Ginouves, Les Mosaiques de la Maison du Ménandre a Mytiléne (1970) 53-7; T. B. L. Webster, rev. J. R. Green and Axel Seeburg, MNC* XZ 12; XZ 13-14 are possibly related scenes, but without identifying title).

Handley l.c. publishes another piece of comedy from Oxyrhynchus (inv. 50 4B.30/H(5)a, fr. 1); and, since it mentions ‘the great rock’, suggests that it might come from Leukadia. (See further K. Gaiser, PE 39 (1980) 99 ff., who thinks of Synaristosaz; H. J. Mette, Lustrum 25 (1983) 29 f.) This too is a verso text, in a first century hand. It would be tempting to recognise 4024 as part of the same roll, especially if both fragments are attributed to the beginning of the play (see Handley l.c.; and below). But the hands seem different; and in relation to the literary text, the recto document stands right way up in 4024, upside-down in Handley’s papyrus. Nonetheless, it is a notable coincidence, and one must bear in mind the possibility that a single roll was made up from old documents, some one way up and some the other, and written on by more than one scribe—or by one scribe whose writing varied from place to place.

4024 carries no character-names. But the dialogue seems to involve two persons: an older woman (7), the Zakoros (5); and a ‘child’, presumably a girl (3 madiov, 5 téxvov). I speculate below that 4024 belongs to an early scene of the play, and rep- resents the first meeting of the heroine with the Priestess (7) Caxopoc to be identified with the crowned figure of the mosaic).

4024. MENANDER, LEUKADIA? es

top JroAAov[. Jee L. | Karl] , xecOncro[

|ravrarertpa. | | adadarrectivk|

ToAXa |SewdoBepa, [, ]c: yatpemracd.iov : [ ate: STE

mel SC Sicterertled Tore: nricemme [ 5 ] nlaxop cyxocuovcar, . vewly ][ | ehvdmpBadilerc : varye: route, _[

tack Japa: pnteppidra |] |

] .moumre pac weurewor | ] ,vewatouc: evOade [ 10 LL. . ]. vpn dnvreyec[

A|moAdov, etic [ofo|y katwxicOync 76[ ov.

a |ravra métpa[ | Kal dAarr’ écrw K[

(|det doBepa,[. ]c. (ZA.) yatpe woAAa, rardiov.

(IIAI.) vy cat cv y’ Aree ef 708’. (ZA.) Aric ety’ ey[o; CaKkopoc 7) Kocmovca TOV vew, | TEKVOV.

ITAI.) ef’ vdwp Badilerc; (ZA.) vary’. (ITAI.) route, [ a Jawa. (IIAI.) prep diAraz[y]

a ys |. 70d métpa’ ctw, etmé por

oO Sa

a

| ,vewarouc. ( ) evOadd [ 10 JL. . ] vpnAny Aeyere[

1 ]z,or].7 e.[, apparently c (not 7) ]., ambiguous []., cross-bar (oblique? horizontal?) joining upright to right 2 a_[, shadowy trace on edge? _]_,, low oblique foot 3 a_[, traces on edge at upper and lower level? or delusory? 4 ],., first, upright with descending oblique joining from left; then, after space, strongly curved right-hand side qjoTe’, OY ToTeE:? [, traces on edge 6 _.[, perhaps left-hand part of 7, then high tip of oblique 7 |...[, first €« (or possibly 6)? then perhaps top of oblique descending from left to right; further trace to right (high horizontal), but so high that it may be simply discoloration? [, left-hand end of high horizontal (7, possibly 7?) Cmlpeatop ofupright 7e_, high horizontal touching pc _, corrected letter (not just r)? you [, dicolon? mi) |] [Is high trace, perhaps tip of tall upright ]_, p?

1 A]zoAAov. The narrow space following has a damaged surface; it might have contained a stop, or dicolon; it may serve by itself as punctuation.

etc, the trace seems to point to sigma, not e.g. ein[é.

[ofo]y. The badly damaged trace seems to suit v better than other case endings; the space is not large (thus [ofo]v rather than [zroio]v).

44 MENANDER

xatwxicOnc. The trace (and the sense as guessed) suit this better than KaryKicOyc; the spacing suits katw rather than cata.

2 mérpa[] «ai. Afier pa, shadowy traces on the edge, but I cannot be sure that they are ink; then, to judge from the normal size of « a short blank (punctuation?) in the lacuna. Dr Rea points out that 7érpa[«] (or 7érpax) would be equally possible.

In either case, 7érp— represents a tragic prosody (contrast 9; compare fr. 258.3, anapaests from Leukadia); cf. Handley on Dysk. 414. I have not found a close tragic parallel for the line as a whole (Soph., Ph. 902 dmavra Sucxépeva is somewhat similar). Virg., Aen. 3.193 caelum undique et undique pontus, Ov., Tr. 1.2.23 quocunque aspicio, nihil est nist pontus et aer.

Turpilius perhaps had this line in mind, Lewcad. XI Rych. miseram terrent me omnia|maris scopult, sonitus, solitudo, sanctitudo Apollinis. (The text is so transmitted by Nonius 174.4; in view of the Greek, and the rhetoric, would one not expect a nominative in place of maris?)

At the end, perhaps «[arw.

1-2 How many speakers? No paragraphos shows below line 1; but, to judge from the paragraphoi below 3, 4 and 6, it might be completely lost in the initial lacuna (whereas a paragraphos below line 2 would be expected to show its right-hand end). So far as the sense goes, it depends who is addressed in line 1. (i) Two speeches: X addresses the Child, and the Child replies; AzoAdov is, as usual, an exclamation. (ii) One speech: the Child addresses Apollo. If (i), we have to assume a third character, since the Priestess does not appear until 3; and the general description in 2 follows oddly on the aorist in 1. (ii), proposed by Dr Rea, is clearly much preferable: the Leukadia, to which this fragment can be assigned on other grounds, took place near the temple of Apollo Leukatas (fr. 258).

3 (det seems inevitable, and fits well with ¢oBepa,[ (Aesch., Pers. 27 etc—another tragic touch?). After ¢oBepa, shadowy traces on the edge, quite likely not ink at all. Then either (i) PoBép’ a_[. Jc or (ii) doBepa [, ]c. With (i) we might look for an emphatic adverb, but a[év@ Jc (Kassel), a[7A@ Jc, a[xpwc, all look too long. This approach creates a split anapaest (or divided tribrach). (ii) requires a monosyllable. mwe and dic seem feeble: Ofc Rea: possibly 7éc— (interrupted question).

xaipe: an older woman, carrying a water-jar, interrupts. Professor Kassel points to a similar scene, with roles reversed, in Plautus’ Rudens: 263 iubemus te salvere, mater.—salvete puellae (the heroine Palaestra greets the priestess Ptolemocratia); 285 ego huius fani sacerdos clueo (Ptolemocratia); 430 the priestess sends Ampelisca to fetch water. Plautus too, and perhaps his Greek original (Diphilus?), set his play near a temple and the sea-shore.

ma.diov: ‘child’, not ‘slave’, cf. 5 téxvov. Nothing in the fragment determines the sex; but the general cast of the scene, and perhaps the fearful tragic tone of 2 f., suggest a girl.

4 vy Kal cv y’: the reply as at Sam. 128; CGFPR 257.77. Both these continue with a vocative. But here the speaker of 3° (addressing zadéov) identifies herself in 5 (addressing réxvov). That gives good reason to reconstruct a question in 4”, j7uc edu’ éy[&;, with change of speaker before it (the papyrus shows a high point, damaged surface below: high stop and dicolon are equally possible readings). Then what follows cv scattered ink on straggling fibres, do not exclude this reconstruction, but do not confirm it. The objection would be that 7[ is narrow (and if the first trace there is taken as a large elision mark, the remaining trace is too curved to begin an eta).

Written above | zore is | 7e,-: the first trace seems to be the right-hand arc of a circle; after « a narrow patch of damage; the stop at the end is faint. This should represent a correction or variant; it stands too far left to be a nota personae. My only idea is m]o7 ex, a variant on ef zo7’. But in that case the text before and after must have been different too.

5 Men. fr. 686 KT. The sources are: (i) EtGen (I am grateful to Professor K. Alpers for allowing me to print his collation) Cdé«opoc: vewxdpoc, Hyovv 7 Siaxovodca epi To fepdv. Mévar8poc Alc eLanaravte 0d MeyaBuloc jv dcric yévouro Cdxopoc [fr. 112]. kal madw map’ adt@ Cdxopoc % Kocuodca Tov vadv, TéKvov [fr. 686]. Kat 6 umnpéryc. Aevxadia: éribec 7 mip 7) Lakopoc odtwa Kadac [fr. 257]. 7.6 tepede [so A: B has danpérnc 7) Cdxopoc in place of the underlined words] 6 rév vadv capdv «TA. (ii) EtMagn 407.23 same, but om. Aic— map av7@. Other versions omit fr. 686: (iii) Photius p. 244 Naber same, but om. xal 7éAw—réxvov. (iv) Suda Z g same, but om. jyouv—tepov and Kal médkw—réxvov.

Professor Alpers observes that the subscriptio in EtGen, confirmed by the parallel tradition in Photius and the Suda, states the source of these glosses as the ExAoyai (probably deriving ultimately from Seleucus) and the pyropixov.

4024. MENANDER, LEUKADIA? 45

Meineke, Men. et Phil. p. 107, and after him Chr. Dedoussi, @[ATPA: Tyntixos topos &. I. Kayswpeévov, Thessalonike 1975, 21~3 (I owe the reference to Dr Austin), took Caxopoc alone as quotation, 7 Kocuodca tov vady as gloss, the final réxvov as dittography. 4024 refutes this in every detail (réxvov is lost at the line- end, but implied by waSiov above), and confirms that Sylburg was right to recognise a trimeter. Note (1) the papyrus supplies the initial 4, which the quotation lacks (suppl. Sylburg); (2) the papyrus gives vew (originally vewy; the final v apparently blotted out, and an expunging dot above), the quotation vadv; no doubt the papyrus is right (the word does not occur elsewhere in extant Menander).

Gaiser p. 463 assigned fr. 686 to Leukadia, and this papyrus tends to confirm the assignment. But on his view, the line addresses the Caxopoc as réxvov, whom he identifies as a girl serving the temple under the orders of a senior priestess. Our fragment shows that this is not so; see 7 note. The fragments contain nothing about a priestess other than the Caxopoc; that Caxopoc is here addressed as ujrep, and could therefore be identified with the central figure of the mosaic.

6 Clearly é¢’ vdwp begins a new utterance, but there is no sign of a paragraphos above (perhaps we should assume a dicolon at the end of 5).

Here and below I have assumed a simple dialogue between Priestess and Child. But it remains possible that a third person intervenes.

At the end, the likeliest reading seems to be rout: 7, [, the last an oblique crest as of a § A. Assuming a simple dialogue, and that the dicolon in 7 is correct, we need to allow for a short utterance of the Child, and another of the Zakoros, before the Child resumes with prep: presumably question and answer, and making the transition from the water to the rock. Perhaps (ITAI.) rout! 7A[ncéov; | (Za.) év [7H mérpa 7d (or pec Austin) v]aua (even ev 7[y, if the third trace is really ink).

7 At the end, the papyrus is broken away just above line-level. On the whole it seems likely that, if there had been writing after ¢iAraz[7], it would have left traces. I therefore assume that the verse ended there. But the assumption is not secure.

Jaya: one possibility, in the context of fetching water, would be v]éua. Again a solemn word, used by Menander for comic effect at Dysk. 947.

unrep: before this, a possible or likely dicolon (an initial paragraphos would not be visible?). It would be simplest to correlate this with madéov and réxvov, which, since the Priestess needs to identify herself, must be terms of benevolence, not of family. Gow on Theoc. 15.60 found no example earlier of u#rnp as a term of respect; but see now Dysk. 495!

It would be easy to guess that the Child will indeed turn out to be the priestess’ child; so that the terms of respect have a particular irony. Fr. 258 gives scene-setting anapaests, normally thought to be spoken by the priestess, and normally identified with the irregularly grouped verses which the scholia to Hephaestion found as the eicBoAy of Leukadia (the beginning? or near the beginning?: [Longinus] 38.2 «vOvc ev 77) etcBoAn corresponds to §8 of the speech!). Was this soliloquy followed by the scene in our papyrus? If so, the structure shows a clear likeness with Euripides’ Jon, both dramaturgically (the scenic solo, the sacred place, the fetching of water) and in plot (parent and child, one a new arrival, one serving the temple— Leukadia reversing the age-roles).

8 |., the top of an upright; in zw, tau does not explain all the ink (a correction? or an exceptionally long descender from ¢ above?).

The letters could be articulated in more than one way. The text printed, wot 7érpa ‘ctw, assumes a reference to a particular rock. But, as Dr Rea points out, | ou (mov, mov?) mérpac: tiv’, eté ror would be equally possible.

After jor, dim traces before the papyrus breaks off. This is likely, from its position, to be the verse- end (only 5, as reconstructed, would be longer); I therefore incline to take the traces as a dicolon, although the lower point is higher in the line than elsewhere.

On the argument made above, this line should continue a question by the Child. If one rock is in question, and has been mentioned before, we have to account for the absence of article. One pattern would be 7 tod Pawvoc (but |c cannot be read); another, continuing the theme of 6—7, ag’ Fc ddpeve]v?

g Ifa dicolon is rightly read at the end of 8, the surviving text, unless the answer was very brief, must belong to the Priestess’ reply; a pity, since prima facie it would be tempting to take évOad/ as answering zrov. But in any case I can do nothing with the letters. | _vcwwatouc seems unavoidable (not azouc); the first trace the top of an oblique that slopes down from left to right, rather close to the next letter, perhaps the upper right of an angular loop (i.e. € 8 0 c; p), but a6 A w could not be excluded. —Jovci, | dvcw (direction),

46 MENANDER

] pocw (water), 7e]pucurd could be considered: but how to continue? Perhaps we should assume corruption: a<c>rove or a<v>Tove Austin.

10 |_[, if rightly seen, the top of a tall upright (so tall as to suggest ¢ or the like?). |v, apparently the underside of a small loop, with spots of ink from a descender to the left: i.e. p? tynAnjv (preceded by yd |p? Austin); ] v[7e]pupnAnv not suggested, since the high trace suggests neither v nor a diaeresis.

éyetc shows that this is the Child speaking: a comment on the Priestess’ description? or more likely a further question, “You mean the very tall rock?’.

P. J. PARSONS

4025. MENANDER, Misoumenos? A 9B4/3(1) 1 3.5 X 3.7 cm First century

A scrap (back blank) with parts of seven lines. The script, ornamented with hooks, serifs and half-serifs, has a gawky early-Roman look; a in the capital form. Compare GMAW 37 (assigned i BC/i AD) or XX XI 2555 (datable to the later 1 AD). No lectional signs.

What little can be seen of matter and metre would support an attribution to New Comedy; and in 2—4 may be recognised parts of three proper names, Krateia Demeas and Kleinias, which occur together in Menander’s Misoumenos.

ieee

|revaveéay [ Kpa |revav e€aye|

] Kpvedny, [ dlaxpver Ane

JvovyysrKpaka | [ |v obdxt pixpa KAe[wve 5 ].. .vrovrad, [

] a, devpobvy[

eral

1 ],..[, first perhaps foot of oblique descending from left to right, second foot of upright Pelle e or 0 3 ]., ink at line-level, perhaps foot of oblique descending from left to right [, upright trace on edge, slightly convex to the left ihe lcs Ors ( 5 ]..., second apparently cross-bar and lower curve of e or 0, third foot of oblique descending from right to left, then top and foot of oblique descending from left to right [, foot of oblique rising from left to right 6 ]., right-hand end of high horizontal, lower part of upright below, more ink at line-level to left a_, point high in line, hole below

2 Kpa|revav. ctpa|retav would have attractions, since é&dyew occurs often in military contexts (cf. fr. 555.2). The case for Krateia is the coincidence of names in 3 and 4; 6 @vy[ gives some support (she is Demeas’ daughter).

5 |. €avrov looks likely (of a the left foot and parts of the right-hand oblique, A also possible), or perhaps | ceaurov: the first trace consists of scattered ink, some below the line, which might form parts of

4025. MENANDER, MISOUMENOS? a7

a circle, but I am not certain that c would account for all of it. I have tried other possible readings and divisions (~¢ adrév with elision, Avrov, —v rdv), but without finding anything more plausible.

In rad [, the trace might represent a or A (not ray[). If (c)eaurdy is right, we can exclude e.g. parts of radaimwpoc and raAavrov; rd\A[a could be considered, or a part of raAac.

Dr Austin notes a possible overlap of 5—6 with Misoum. 354~—5, lines poorly preserved in XX XIII 2656 (latest text, CGFPR 151): there it might be possible to read the end of 354 as ov (rather than o 0) rad df, the end of 355 as Jo [@v]yarpio[v. To confirm the placing, 2~4 should be identifiable in 351-3; but the degraded remains in 2656 are indecisive (I am grateful to Dr Coles for his advice).

Even if the placing is right, the problem of reconstructing 5 remains: assuming (c)eautov, what can be made of rad _8[?

6 |. a: the first may be +, with the rightward extension of the cross-bar lost in damage (then a trace at line-level to the left belongs to the preceding letter) or possibly 7, a trace of the left-hand upright, the cross-bar projecting beyond the right-hand cross-bar. After a, the high trace, if not delusory, might be taken as the top of a narrow letter, i.e. « (—ra1, va?); better as a stop, i.e. high point or the upper part of a dicolon. Both metrical schemes outlined below require a short syllable here, which would exclude a. In that case, dedpo must begin a new clause, or a new speech, as imperative rather than simple adverb (cf. Sik. 146); which in turn might recommend the vocative Ovy[atep or Ovy[drpiov (cf. Misoum. 355, see above on 5).

2-7 If (c)eavrov rad | is right in 5, and assuming that these lines are trimeters (but they may not be), we would consider two possible arrangements:

(1) line-beginnings x Kpa |revav ~~ dSjaxpver dy obdx? ~] (c)eautov

MM mM OM

~ | a dedpo

(i) line-ends Kpda |revav eEdye[w djaxpter Anpué[ac ovx! puxpa, Kre[iwia (c)eaurov tad. [ dedpo Ouy[atépa

P. J. PARSONS

4026. MENANDER (?), PROGAMON 38 3B.82/E(1)a 6.5 x 10.5 cm Third century

A scrap of thick, coarse papyrus. On one side, along the fibres, a few line-ends in a good practised cursive, perhaps from a land-register; line 1 mentions érnvt(Anpéry). On the other side, and upside-down, stands the title of a literary work, written across the fibres; the script, a not very well executed example of the Severe Style, could be assigned to the third century. The top edge is so straight that it may be original; the papyrus is broken on the other three sides. To the right, the vertical fibres have been stripped, except for one narrow, isolated patch about 2.5 cm from the top.

In principle, the piece may be (i) a szllybos or (11) a colophon or (111) an independ- ent writing-exercise. In favour of (ili) we could point to the reused papyrus (assuming this side to be the verso), and to the misspelling 7; but the nature of the text tells against it. As to (i), the other examples (see Turner, GMAW nos. 6-8; Dorandi, SC'8 (1984) 185 ff.) are narrower (lesser dimension 2—3.5cm), and most, though not

48 MENANDER

PAnt I 21, have their inscription written parallel to the greater dimension; 4026, as it survives, carries writing parallel with the narrower edge, with a blank of at least 5.5 cm above. That leaves (ii): the text was copied on the back of the land-register, and this colophon was added to the right of the last column; the author’s name may have followed further down. But an apparent trace of writing above, to the extreme right, remains unexplained (4 note).

The format certainly suggests a pair of alternative titles, even though the scribe failed to centre the linking 7, and wrote a rough breathing on it. Lucian’s "Ove:poc 7 AXextpuwy provides a parallel; but I have not found a placing in this genre. Comedy would be the most obvious source (for alternative titles see Gomme & Sandbach, Menander 129 f.). To judge from the lists in Kock, Koerte’s Menander and Austin’s CGFPR, the possible authors are not many. Diocles wrote an "Ovecpou (Test. 1 K.-A.), Menander a J] poyau@v (Koerte II p. 128, who refers to a comedy of the same name by an unknown author, IG II? 2323.136). No doubt Menander is the best bet.

The meaning of Menander’s title has been disputed. ‘rpoyapety significat concu- bitum ante nuptias exercere’, says Koerte. The alternative title provided by 4026 suggests that the marriage was anticipated, carnally or not, in a prophetic dream.

]

|] overpoc [

iy [ mpoyanwv — [

3 ..[, shadowy traces, perhaps just stray ink. 4 ].[, clear traces, on the isolated patch of vertical fibres: an upright with a high horizontal joining from the left just below the top. The ink is quite thick: this might be the same pen as in the main text.

P. J. PARSONS

IV. AESCHINES

The following twenty-nine items represent all the remaining unpublished papyri of Aeschines that have been identified among the Society’s papyri from Oxyrhynchus. They have been collated principally with the most recent edition, the revised 1978 Teubner text of Blass-Schindel. Among other editions use has been made particularly of the Bude text (3rd edition, 1962) and of Schultz’ edition (1865).

These papyri nearly treble the total of known papyri of Aeschines. To those listed in the introduction to the 1978 Teubner edition may be added P. Oxy. Hels. 1, P. Koln VI 254 and P. Duke inv. G44 (ed. W. H. Wills, GRBM 10 (1984) 311-4). P. Colon. inv. 5927 (wrongly cited in the Teubner edition, pp. xxi and xlv) has been republished as P. Koln II 65. P. Mil. Vogl II 41 has been re-edited by J. Lenaerts, Miscellanea Papyrologica (= Pap. Flor. XIX) II 335-340. The Rainer papyrus (‘P1’ on p. xxi in the Teubner introd.) has been re-edited by U. Schindel, ZPE 46 (1982) 1-31.

It is very difficult to obtain precise and reliable information about mediaeval MS—readings from the app. crit. in either the Teubner or the Budé edition. The apparatus in Schultz’ edition is much fuller. Just how much more extensive the mediaeval evidence is may readily be seen from the list of Aeschines MSS published by R. Roncali, Ann. Fac. Lett. e Fil. Univ. Bari 14 (1969) 381-390. We have tried to avoid citations of MS ‘groups’, given how frequently individual MSS within so-called ‘groups’ provide variant readings.

For convenience we append a conspectus of passages where our papyri exhibit readings different from the text as presented by Blass-Schindel. We add where we can details of the mediaeval MSS that support the papyrus reading or Blass-Schindel’s text, as well as other variants that there may be in these passages. We also include selected indications of the readings adopted in various other editions.

4027. ArscHINEs Jn Tim. 3

38 3B.84/H(3)b 10.2 X 6.5 cm Second or third century

The papyrus preserves part of one column, broken above and below, which was perhaps the first column of the roll. Written along the fibres in a hand of severe style type. There are no accents or lectional signs. Iota adscript occurs in ro but is not used in 8. Elision is regularly employed (before rough breathings in 7, 8 and g). The back is blank.

The text is faulty; repeated pév odv (0d pev odv dAov, 3 init.; wept wev odv TOUTUY, 3 ad fin.) has led the papyrus copyist to jump back to the former, so that from the end of line 6 the papyrus will represent a repetition (only lines 11-12 are actually dupli-

apng q 6&6 ‘udiaq “ustiy ZNYIS lh 4 6 PHP Q aur-yng ‘q 01-6 1 A 620F >[e]a2é[ay]oy 611 | qqv qieg apyodri py 1.493 IQUPIAM awyodorpy 212 1123 ZyNYIS q 686 ‘ugg ‘usyiy b y DL WwYoArI PAT 1433 apng DL 1433 awyodrr py [zor 5 ‘Aur “-ynq ‘dl La A 620¥ 1439 >| oy [odary7] giv I apng “2xsloy d madodlinlig rela ZyNYIS b—£ 1 A 6Z0F aadodunle[ k]r Cron y apng ‘zaynyos ra9nadodAnup 9° Z-1 I A 620F 1asdo0dvuv azTl o g1‘oF I ZNYIS rmnaodoyogidyy avrV ayuely vtavuv roraodoyogidyp avy b mpnaodoyogidwy avy DLaDu apng mnaodoyogrd»p qqy 1 queg j rnaodoyogidwy vLapu GLY LADLD ADIV 1-08 Ul Y 6ZOF ronladdoyog|1d] 10 pil[a]Duv Fi—-S1ov | 2png qqv qreg bd yjp art lil A art at lint ZNYIC nt av lil 6z 1 Y 6ZOF 3i3Qlinl €1:0F | al oXdor]i3 7, apng ‘ziyNYyIS aoXdom 7 1Lp110> OL aoXdorh 7 <mL>1L011m> OL L—gz UY 6Z0F ML i[ zorlm> mL} Z1OV I apng ‘zijNyIs 4331393 aoddaor oLav 9—M 61 1 Y 6ZOF “WO o1—6'0P [| I 2a0L0D apng gi-Li 0 apEge qe apLgD znyos 1 A 6Z0F COE Gof | apng 1-11 (thgpr2) zaynyog ougr ae HY 6ZOP = c9432999'04? 10 0199990»? Lo¥ } apng ‘zi[~nNYyIS 13D NOLAD W3 “qieg 20LaD us 13D Lo Il Y 6ZOF 10 |24[ av 43] Mm >m]}130 G-6§ I sniuodurgq a mda apng ‘zynysg = ss —-E “Wl + EGZOF ahoda V6 | aaodak apng ‘ziyNYyIS pinla3godu aaodad hranioLr o=15 l YW 6ZOF kjiavi[o]s [171>290du o—1 66 [| 9png doy qed P dV J b zynyog ‘ayuely 6% 1 Y 6Z0F i z'6E | d ‘a “wIO apng ‘Zang Lo 1 Y 6Z0F [pdaxn] 2-66 | apng “Zanyos >yDu &% 1 Y 620F [p]io# 20200 7z'6& | apng ‘zaynyog pesnos Xd rh J 1 qieg p qqv J zz 1 Y 6ZOF mXdovha 7 zz'6E | ney “wo apng “Zanyos 79 811 Y 6Z0F one 13°6E | apng ‘zijnya¢g MmL3Va0g d0LaD IM 73.13 ZI-11 I Y 6ZOF 131 |3 ivL32ya0g 3[oLan 3] 61-91 “gE if 2prd ZYNYOS 10 agpdXoyD 4 bud yqueg pqqvwy 9-5! UY 620% >09|d]X2[1 dod m100] L1-g1 “gf | znyos “png hakg, tie) iby qrgqpqqwy &-2 1 Y 620F a|]orla[aa|s50/g7¢ git J MOT[OF OYUM “Pps JIYIO “surpess [s40}1po “‘ppe iayi0 Aq pamoTjoy sty) Ajddns ulapowl 2] ‘paul SSW Jo/pue “paul SSjY 194}0 Jo ssuIpeas JURLIEA, 1ey “paul SSW JopuryIG/sse[g Ul 1x9], sgouaprag Sunsoddng ‘you ‘deg x0} snakdeg PY

b

x qqy queg mney yy p ‘ZyNYIS “Iq b

Ine'T

Biel

‘zynyog “Weig “Iz ‘aysiay

d pid

qqv qed 45 P ab

png “susg “17 JIOM “H

qqv ney A OF

y 3

ZNYIS ‘Qyuely souprem “qqy bys p queg

mgov>0Ad1339 mgrv2DAd>33 1.493 M100

Lap u> v . € ©

“WO aoyyoT!

1220NOMD D10L amra

airla

doLpndpdu a0L r01LA0

20LA0 s01071 Addu aoL

Muo0d1143 L120 / cer

muodsusdou 2400

aj20L00

aoLan 1d3u 2mg~m13 13X32 ‘ ? % ‘ « “ aoLan 1dau amX>

sitet plat F

aoLan 1d3u 2Mg9MI> amX3 Vues es ‘ « x

Z\NYoS ‘oyuesy opng ‘Ziynyog

‘aqv “qeg UF P

apng “zanyog png ‘wp sqe

Ine ‘qay ‘qieg 9

‘aqv J °P

png “Zaynyog png ‘zijnyosg epng ‘Zaynyos apng “zinyog png “Zaynyos ‘qieg

JA ‘Qyuely ‘addneg-sa11eg ‘gpng “Zjnyosg epng “ouplo ‘qieg ‘qqvibyp opng

ZVNYIS

AD aral

Zt S

Asie

ANAS

Heatel

2666

‘udaq ‘usily Aer | ZINYIS ob

ee acd

aA IDLADLD ant dod amd1aX 1L9

[199297 p4d252|

MLav 199 - © € “ MLaAO if an1IXndQ aoyyon dL DANQAIMD ; ioe

1DINOHD

a2 aia

(ainjoafuos) A 9m soL1vndndu aor

d3u10»

1N21310L 313373Q0UD IDL

azaodsd 1071

“wo 1030

1030 a

MmA> “UO 2, « 20L0D “WO

[muodiu> dow 3100 | f Fe, i

10LN0 ‘ ?

AaoLap 1d3.u a13X> 2M9M)3 rae} ‘ K \ ‘

IDAIL 13D ak

ZyNYIS “opng opng

qqv qieg byjp png ‘zinyog I8qqawe

apng

qqv qeg bys p 2png

zynyps “A , Iq ‘A4Stay *FTOMA “Tuos

apng ‘zi[nysg

‘

I—O qqv qieg b yy p

qqv buyp

queg qqv buy p eyurig Soddneg-sajieg Joyyog ‘ZiNyos

uw qe

ZNYIS

1% Il O€OF £1 Wl O€OF Z1 I O€OF

ZI—I1 I OOF

Ul O€OF 82 Ul O€OF ZZ Ul O€OF 01-6 Ut O€OF 6-8 1 O€OF

8 1 O£0F ££ 1 O€OF 6—z 1 O€OF Lz 1 O€OF 1Z—OZ | OOF 1 1 O€OF

cOl | O€OF

g-St 1 A 6Z0F

Ol A 6204

11-o1'bh 5 ‘Aut ynq ‘gq gi-Li i A 6Z0F

/1. A 620

2v1a[vuv| (é) 273 d]n[A amd13aX A> 109239 DAd3 33

Miav 19199

2MLA0

(c) aoyyo]7 avn [Xn]de 3Q va[aga]ino

19[a0a]6»[

a3|71 ara

(¢) mA2 2M

C-9 201]a0 soLvriAnd | u ida 10»

212313Q0LD IDL ]1Dg[ 21310

101 asaodcd

éaomamadrda j[ 030

mA]3 iL6

[d0]idd[m] r0La0 muodiiu3 J)ou ri[ao aoLa[p]

d01[a0]

413 |X[>

>| mgm13[aoLa|o rd [>]

b 2png peytoodsun “ssw

zZNYyIS

IOUPIAM

JUPIAM

imeyT Aidowy]3

1 IIUPIOM

mmgr3ard

2myl1>30du 10713 jon 20» litradpu

aoaonl uid 1D

amAzy 220LAD MF}3QIL? XYP

‘T?P a1Lr2

imLAoA1Dp

2OLND LO

p2003423d0u[ L001]

zNnYs ‘png

ZNYI “UIIg 2Prgd

ZNYIS “gpng Syelte|

SO |

I=o (8)] aneq qed Paved

apng “Znyog

apng “anys

apng *ZaNYsg wp[sqe 2eNd

(teh

& PMP) zINYIS apng ‘Za[NYoS png ‘Zzanypg apng “ajnyps wi, sqe apng ‘Zaynyosg png *ZzaNYyIg ayueiy

q ‘e

apng 1b

zany “png zynys “gpng

pidkssliy

DLADL *

DLADL *** tant “" pLave

mprradrd

rp Aranda

20Xdoh 7

‘yoas [pols smml1>3d011 | aoaonl lil amAay

<2g)20400[ 312912 1D]

“wo

1071 12vDm aarl aoLmdu

Beer Ue RS

2314097 AaoLavd Zins Gienaa ti

aao s

A0LIOMDIDLLIL

“wo 2Q niadapodu

AMYYD AML

1aD> OLAO IOLAD - > 2 hie

D20NIMI3SdDL MLQ0L

ZyNYIG ‘ayurly]

(3) | ney 1 qieg pj

4 qqV

eee

queg byyp

qqv qieg byyp

qqv qaeg bygp

q

‘qqv qed 5 byp opng ‘Zany

ZyNYIS IQUPIOM,

qqv qieg bd yp qqv qieg bysp

o1—-6 ¢¢oF

[eer]

‘z1 ‘ol ‘6 EEF © €€0F

S 1 ZE0F

£ 10+

o—! [EOF 6& A OOF

1 A O€OF pei2]9P ‘1-06 A Q€0F

z—1& Al O€OF

6—gz Al O€OF Lo At O€OF 1Z Al O€OF

61 At OOF QI Al O€OF

o1—6 At O€OF S At O€OF Z Al O€OF

Lo 1 OOF gz Wt O€OF

peTfaoueo gp Oi [40] [rdlirrly yh ba p[idhssky

ZW DLaADL

I ‘Ul OLAOL Fay adi [10 ]g23acd 1 ‘un [10]9>>094

¢a13 [Azazrdan]

aoXdonh 7

[>] [lsd] on ionl[> aoaonl ao

“WO [aug[7]syx0dg [ap 4 70» ||

a3|7 ao1[m]|du [1]o7 13y0»

i

1D» 3Q [7p 3[190]o i aii[>]5

AaoLaAD 33 [+] aog|1

lug

[a]or.04vdn>234

10413 23LA000|Q

40 10413 a1900H0|Q 1D |>% 213d [apodu] amyl |yyo

(¢) MLaD> d0LAad

m[1La0L p2DaIwI>IdDL

a

‘ppe 1ayi0 Aq paMoTjoy Jo/pue “paur Sg Jeujo Jo ssurpead JUeLIeA

MoT[oy OYM “Ppa

LYIO ‘Burpess

sty Ajddns

1ey) "paul SSW

fepuryos/ssefg Ul x9],

[s1011pa

uslgpoul 29] “paul SSP ssouapiaa Sunsoddng

you ‘deg

1x9} snuddeg

ore

Ss

2 o$ bowT ld

boa

oS boyy] | ineyq [gy aN [BY Me

1OpJ

CHASE HL gh 2) [2

A

i

2pngd

iney

y

d

(saouasayyp

Joyo ym) 9 [PH “d

amLaD> idasdorn1g amLapridaidonvigodz amaznomrmaidvandow amasnowrmaidvdnw

“WO

1o1datdor lyor>1u> pidasdor

ADILADAILA

ZyNYIS

39S ‘s}UBLIVA J9YIO [eIIAIg

(7D™ *WO) pasnodAayi Lads vw ‘ “

snidodluand 3A and aoidoAluand ardu

(210L00104

(2202) Jo goed ut) r02a02

AMLNOIOL 270L

Sata Joyewepy ‘fuoo

ZINYIS “png

zyNYIS “png (d ‘ur ‘t pappe “7) Iney

znyog ‘gpng peps0934

“SW jueueA ON oSboy]

J neq neq py

guoddns ‘sp ou ‘fu0s s/g

zNYyIg

S

2png

oS boyy

J ineqy (j22L0poy13)

IneT [4

ZyNYIS apng y ine'y

ZNYIS Zanyyg ‘apng z® qqV bry3syq zanypg ‘apng ney eA dusqe

(9 18H ‘d] znyys ‘png

amLanodatdonnigs0du

amaznonrmadidvavdou

luyorr12> mmdardon

1gmadvan yD» 4 « \

DILADAILA

pasloAsyiLapL 10» ; ook

spridodp raLoply>

“WO DAIL

soidodlaar rut adn

210L0010L JI0L

& 2

papnyour 2202

znyos oyuely oS boyy] J aney ineq [gy

[pia zyNyS ayuely

oS boyJ J aneyT ineyT [yf

oS boy] J sneq aneyq [4

I

OPE

‘Doyo “TWgag “oystoy oS bow] | aneyq [y

st by quegyp

oped SSVedae dee Saeed

qieg iney A wp ,e

ZI-11 LEO

g-L LE0r

v_-& 160%

S 9¢04

11—-O1 1 EOF

8 1 CEOF

ams] avsidasdorinr9 [pod

ama ]3rlo%>ma13[ Avavdvu |

vidat[ dor | hyoi[rra]

(32 10) [439 1]gmadvan

a] o1Lapas| wa

pDaaT0 {4] 3YI1LAD DL 1D

saL0pliyra avridogp

aoeds Jo suoseas 10} “pout [varL]

anidoAlans|ali[s ardu ‘20] pv 210U 39g é[210LA010L] a10jaq 201 “WO

cnr 2Q] 43 @ “Ul “ppe $1 ‘UI "WIO D104

11°SE1 JJ

o1-6'SEr J]

g Per TT

o1°Sg J]

8'¥9 II

bev IT]

&a'bF TT

6°EP T]

g—LS6r |

G61 |

L-z61 |

1161 |

TQUPlOM Iayyog I ‘9

CFLS -ineT

snueydaais

ZyNYIS

0S ‘boy] ‘F:09 ‘Sh'LS “neq A UU Re

A Uwe Pp‘

ZyNYyIS oS boyy] J ney ineq jy ¥

9L0dL

apng “Znyos

ZNYIS epng

paisjap aseryd amayou rt amt 3

2oraLnyou Dt > ‘ aN

Z[NYIS oS boy ane'yT

aodagpap

Fog ine'y ZNYIS ‘apng 2Png rorpaligAy @ ¥ D0 ZNYIS ‘apng

opng

oS ‘boy

‘F-09 “Sh-LS neq Sz9I

“uo £20 on

apng ‘Zznyag

Spee

‘JIUPRM “SxSOY

addneg

-ioueg ‘gpng

‘opurd ‘wW1g

“‘rayyeg ‘ZNYIS

“wo aod» zA 10,4 ds, x0 ag bt opng “Zys IIUPIOM

rat)

zynyps “opng

q

zynypg “gpng

o§ bow] sney [y zynyssg “opng ZINYIS 9Pug

oS boyy] J ineq epng “eqop

1BYOL lit I

arpr0du

‘(uo »oLaopoddvan

aonlig aor ova 4 - \ >

2 “UO

aodangnrsad

AML AIQQ0 aogpy ,4Q0 140

romvalig 3adQa

3. *8 2S 63

19NIVIM

md>1ara OL 1Laodapriad OL

Cape

aodmin 19>139dk azglirl

an

alaaoXdoua .Q slaludaa

pis ELEY

amdlim> 13Vou it wo»

>| apng *Za[NYIS

zZynYyoS syueLy 9png

cZ9I

opnd

(gi1G1 ysuaziayY) Ug 142 SZ9T

qoig “xery “Utd

Jouplam ‘zipnyag Aq ‘qqoy {(6Sg1) ayuery fuos

IOUPIO\ al?

ZINYIS

Jajasuag ‘ayursy ‘oystay

png ‘zynypg ‘oyuely uwse

mey Aj 3 e (aq)

J uneqy

IneT [4

01-6 OFOF

61 6£0F

C1-¥1 6€0F VI 6€0F

[8] 6€0F

St gE0r

61-21 QEOr 8 8E0F

8-L gcor

él SE0F

3p20[du oviopoddAsdvan

jaonke aot OLA “WO if ws ‘ ?

am [syou abt »]> jaoa[agasuaan| am[4 aogvoy> 4ao|

1oywalig Ee 23dQan @ “wo

éa[or0]

1INIVIM “WwOo 7

1ta[ odagrlad 1md |aLaria 194

ap> “UO

19 [p19db aod» |a71 arpgur

[ éapL0|

aiana

alaa0Xdnua sald] 19

pia 10»

a[32430Qa ]3 é[pvo]z kt afon|d[ usm 10>]

L166 [JT 116 [II

oz—61°€ JI]

L£°S% TIT 61°22 TI] glee TIT

o1-6'91 JI] 8°91 III

oz'St JI]

ons Ty

61-4 TIT

EE, TEU Lee Til

¥1°9 III

freee Til

gi—-S1'2L1 J] Visi T]

Gri TJ]

Dienst [if

a

‘pps ayo Aq pamoToy 1o/pue ‘paul SCJ JoyIO Jo ssutpess jUeLIeA

MOTO} OYM “ppa IIyIO “surpeal sty Atddns

1eyI “pow SSW

jopuryos/sseyg Ur XA],

[s10j1pa

usspour 2] “pou SSW sgouapiAa sunsoddng

Homecery

xo} snuddeg

gol

b38 w

“qieg dary

zyNYyIS

eles

zanyog ‘9png

qi, 1

NSPE EL

doastiopad4 ; aoasrioundL

doasriog3d 14> ‘

alurlmaddas amna dpu aonayou a14>3 aonsyou ; 7 7

aoLn Aan 1133 aorlsyou 7 ae ?

az udaaar

POLIILD

‘ yD AOLLIVIG,

70m nord graar aombyy7, a13X2D131 AOLLIVIG,

a0>013 son. 2341.10

IDLIDH1Q DNOL / N

apng

‘ane'y Se ZNYIS ‘png Td Sine]

ZNYIS

png ‘zanYg

apng ‘zaynyog IOUPIOAA,

ZyNYIG “oyuRly

png ‘zanyosg epng “Z2ajnyosg epng ‘Zaynyos

apng

14? zynypg ‘apng zany ‘gpng zynyog ‘apng

ZNYIS ‘gpng | ‘9

znyg ‘epng zNyS ‘gpng

ZNYIS

‘Jojasueg ‘IyxuRLy ‘addneg/saiieg

dn0d313 2Q 4 cay 2n01¥1X>1Q

iLaoppodd

doastound4

pouodjsod spu0d4

AMLIYDLD 4 AMIDAIMIDLDM

amApaad

suurleaddas amna dow ao4Dm 1499 aoinidan aorisyou oy 3 $

A139DL DLIIXIID Nyou kL Z ‘ Vas Ad IL lugaaar

« 4

JOLNOIOL

20d3L0101D aoda1ana OL

aononle OL 11L3 939LAD1909 7 Neon 7

aivd4ovig,

‘ yom | 20rd Qaaar gomalyy Gy, aaXr01371| ‘AouLvig 02019 2oMIAa 23a14 19 a eS K aod1D»

[279190419 ri0L]

(949g ‘pap) [rug G2]

[seiqoq ‘fuoo] 217132g07I0a

ZyNYOS

qua “do ut Japuryog-ssetg

[ti

2png jppurys-sselgq Aq parts “sseQ od’ JO

y 2oasnowads ans)

IPE OS -ssetg Aq paito sjayfesed

Aresoyy JO nq ESSW ON

apng

znyos ‘gpng 149 IIUPIOM

14?

apng ‘zijnyog apng “Znyog

9png ‘1949S

9 L¥OF V L¥OF

g—S OSIDA OFOF C_P 2 GcoF

g—Lo p ccor

L—9z P SCOP Sz Pp ¢coF 1I—OI 9 CCQP 8 ° ScOF

1 9 CO

S—-¥ q ¢coF 2 q SCOP

@ OIL OFOF [gf] | chor

L—of 1 GEOFF VE | CbOF

1& | ChOF

b1—ZI FOF

S—2z | CFO L1—QI 1 EOF [1] ' GEO Ol | FEOF

6 €FOF

[9] £404

dno[d]a1> Q

saordaring é[tavprodA] (iLavp|odd

20a |arlouadi

[pvuods 2001 | ayn

gamui[yoiomAs am}Da3z»>[ Dd |ou

am|dvdnaas

amna dou{ sir

me 1 heal aa g 1433[aorlsyou] you Gf

43 Q lugsand ‘wo Anuaiedde >01a0102

doda101|>1uD a” ees 23[Lap1305|

isrvdwovigy

é[7» aordesaa>| aomaky [vq

a13X>0 |i371 10» a[ounr|yig 10» ] ao[1dQ>]aar

aoma[ yy 213X>]v1971 ey oa }

1 .aoil| LIVI7g

dora aprm13 >3[ art 13 ‘wo Apuoiedde aod» 21DL9041Q >[10L onlg[ms

[ ¢20139 ]or/6a

61°g6 TI] g1°g6 II]

€°16 TI]

8°06 TIT

g—L-06 TI]

L-06 [II 2°68 TIT 1°68 III

61°68 IIT

S1—¥1°98 III vi—61°98 TIT

818 III 6°6S TT]

6—g°6S JIT £-gS III

98S III

&—2'gS II] 2o'gS TI] gi°LS JI]

o1-6-LS JIT oz°6& TT]

61°6& TI]

IPN 7p “h oup g4gi UPI

‘22140 “14? “@' 94P zynyos ‘Sgi ‘quiey “g

by ‘p yaqory ‘yap aseayd joy TUPIM “2p “P'@ Ou”

(Ppuryos

-sse[g 0) 908) [| 4 2 (TPPUTYIS

-ssefg ‘apng 0} ‘90R) ¥ ayueigq *Z[NYIS ‘addneg-sai1eg ‘p 4

[ neq] neq ‘oj uw se UN

s1sy10 2 9png “ZNYsg Mey 10, J (gpng 90") ¥

opng

lojyyj uur se

ZYNYIS J neq iney ¥ 2

ZyNYyIS

apng ‘jp

(221g ‘T193) PpuUryos-ssrTg (znyos <9an2Aoyrua) 4 png

yPUd

Tye

ZNYIS “apng ane’y “1 2png

Z\NYIS

“aIne'T]

royaogaond¢g slyng oup

orao9r oy

apng ‘Znyog

aotoe? lL rodgan u

“

ORS |

2040191 dodgap (zyNYIS aoLong? 0} 099%) [¥y a aarughog J mney po Pbavel

PP i)

2073La071 at 10» amndo AML 10 apdp rkt suave ah >073La071 at 10» amydo amL 104 DID rhs rhsav1 shaarloasd

2013LaD71 abt 10» amxdo

amt 1D¥ 2DdD aut surave

apng “Znyos JOUPIOM, apng

pLoidp1

roynogarndg [>kyag exp]

oLapndoyrua ¢ ’ amLaogy2Lpy Aad

La slyag oun

roLargaazodnuL doLmdu OL

anda1ara

dod nyou

noummen cern e

ialglog

[puasrioaad| 2073La071 ah 10» amydo

amt 1D» apdp sks suave

doAg> lu 21019) G2 one

pindp DL “ \

‘queg ‘9 {Sg ‘quiey] “q ZyNYIS ‘oayues bi

IIUPlLOM

a Z144 39

Jlopurq, “twg4g “19yJaq 1442

Io¥Hg d

ZNYIS ep leh Pesca tal [ ney IneyJuuUuse

apng ‘Zaynyps

01-6 ZGOF

9 7S0F £-% TOF

6-8 1S0F

9-S 1S0F

9 OSOF

511 660%

SIT 6F0F

61 640F

rhyag ou|v soynogasv[d@

oraojzidoy[ sua slyag o [uo

mLan | amLaog [y2L04da>

doLtmdu or doLAsgaazodn [u

avd [a1 jan

you dod

“=

a{o]zmg7 & s0dga[p

azglog

aor dauja

am[xdo amt 10%

avdo rt 10[ rlaarioaa Ad anatapM dls sutave (okie:

i] 3&[1]>vAag [lk roa] Qa amt 1071[ Le

p1>|1d0

€°S6r JI]

z°S61 [II

$z°G61 [I]

o1—6'191 [I]

6191 [II

13°9S1 [JI

Gi-vri [I]

€1-z1-V11 [J]

g vit TI]

oz—61°€11 JI]

OZ'OIT JI] OI'101 [IJ OI‘I01 [IJ

ee ee

Jo/pue “paul SS Joyo Jo ssurpeos juRUeA

MOT[O} OUM “ppa JI1Q “Surperosi sty Ajddns

yey “pau SSW

“‘ppe say30 Aq paMmoroy

jopuryos/ssefg Ut axa],

[s1011pa

ulgpow 2] ‘pou SSW :gouspiaa Sunsoddng

‘you ‘deg

3x9} snuddeg

got

JouplaM, ZipNYIS “uO

apng A>AD9>LD le JOUPIO\\ palajep 144? r0plupt 70 diay

Z ‘1109 ‘1O],J UW YY SF pode (S/q 01 ‘998) 25

mLhunyodou ?

alt DLD» “WO

J ineyT po

y aramrmar Arner

Jouptaa, (1qoq) ‘jap aseryd

col eee avLaogpodd aonoandou 2n0L

opie JQUPIA\

‘ping ‘so0diepy

apng “anys png *z1]NYy9¢g

apng “zinyog

ZyNYIS ADE at

ZINYIG Ages | Z\NYIS ADE ZNYIS PAPAS |

AQ

ZYNYIS 7Pud

apng ‘zinyog

Lasanopae sal

>opplip

[e190 r0plpp 70] DL13u>

Fs

mMg2DANg nap no liolion

r0Llasdan

(jurod sty ye ¢/gq ut yuasaid you 9» si9}a'T)

mmLhuisyodou

23Qlinl Da 4 a 1013

st 213Q03Hf a0 DLD™ “WO aanrmadrdnig

IL

piaoppdd dnorloa dn0L vdow 4 4 4 ‘ ‘

oug ‘ping ‘soodiezy

ZNYIS

‘ping ‘swo0diepy png ‘zanyog

ZNYDS 2png

Z\NYIS a) Mb!

Hopurq, “saxpjag 1y4e Ggl “quiey ‘d + JIO9 SB

$1 ¢coF

[€] 3 ¢coF

Z—1 1 COP

z f ¢coF

! $SOF

o1—6 1 ¢¢oF

SI 1 ECOF 6-8 | ECOF 9 | SOF

11 €cOr

©1-Z1 JOOP

azap |g>2D 1»

[caoaor]

1|mLav 10g[ lp 10]

DLI>

1992 |daag irl [>]13224[9 [@p>hiow|

@n0Lkasdp “WO

|

1|dLlhuisyoL[ Do»

213|Qlir! 3Q[ 0a 192322 Lang 313Q03| fh] ak. vLv¥ aarmardnig aaonoa

3[aox| vdnu (14 viaopodd

0} “1109) 14 amdodd

1°2Sz JI]

1°aSz JI]

1a°3Sz JI] Vabz TI] Gab TI] 61-zbz TI]

grSzz TI]

€1-z1'9'bzz [I]

F1-glz JI]

1-giz TI] 1a'F1z [J] 61-g1°€1z TI] Lr°G1% TT]

Gr-€1z JI]

S-¥-S61 [I]

58 AESCHINES

cated on the papyrus as it stands, repeating lines 1-2). There is a subtle change in the script after line 6, suggesting an interval in the copying. See also 13 n.

The column width of 8 cm is wider than is often found for oratorical texts, see E. G. Turner, GMAW? p. 7. The text lost before the fragment would occupy approxi- mately 31 lines, which would give a column height of at least 44 lines.

xpelc] BleBoxore py dnuyyo| pew emi[taypa we ye On eyw Kpt| vw ov xad[emov emutagavtec aA] Aa Kat tlavu pardvov ewe 6 €| 5 Env av[twe ev ecwh|p[o]v[ec| un cuxo[pavrew rept jzJev ovv odov {a} aywvoc dalvy|cetar ovd y roduc aitia ovca Tiwapyw ov8 ox vo[ por] ov8 vpeic o[ut] eyw add avto[c ov] 10 Toc avtwe ot [pev] yap vou[or]

mpoevto[v avtwe aic]x[p]we [BeBu] wore [ c. 15 letters |

10 avtw pap.; atr@ a b t Laur; atr@ fo p q Vat; €avr@ omnes edd. 13 A further error appears to have been introduced here, cf. introd. The traces seemingly read

have written all along. Because of this, we refrain from supplementing line 12.

LUCIANA SABINI

4028. ArscHINEs Jn Tim. 14-15, 17-18

100/73(a) 4.4 x 8.4 cm Second century

The fragment comes from a roll and preserves parts of two columns with the intercolumnium and lower margin, in a semi-cursive hand. The back is blank.

Between the end of the first column and the top of the second column as preserved on the papyrus, most of sections 15-17 have been lost. We can calculate from the average line length that 24 lines (assuming omission of the text of the véuor) have been lost from the top of col. ii, giving a column of 35 lines. The height of the roll may be calculated as 23-24 cm, including 3 cm for the preserved lower margin and

4028. AESCHINES, IN TIM. 14-15, 17-18 59

allowing 2—2.5 cm for the lost upper margin. The column height was c. 18-19 cm, the width c. 7 cm, with an intercolumnium of roughly 1.5 cm. We may futher calculate that the full text of the oration would have required approximately 90 columns which would have occupied a roll of c. 7.5 metres.

There are no accents. The scribe punctuated with a middle point (i 6, 9 and 10). Paragraphi occur below ii 2 and 11. Tota adscript occurs in i 6 and 9.

Col. 1 ; ect 4églettersea |] ct. 5 4 [eOnxe dudaka Tw v|weTepw [7avdwv tov tyHc| mpoayw [yerac Ta peyic|ta emiTipia

5 [ eAevbep lov maida [n yuvaika tpoayw|lyeunu Kar [zrovov adAov tov tT]ne vBpewe [oc eve Kefadaanr |ravta Ta [rovavta cvAAaBwv | exe ev we

10 [Scappydynv yeyplanrac:

[eav tic vBprlne evc] mada

Col. 11

o[vK« emtndevov nyncarTo| e[war cuptroArtevecbat| x[axewvo Se joe cuvor] a[uvnuwovevcate w avdpec Abn]

5 va[voe ote evtavd o vopobe]| Ty[c oumw diadeyetar avtwr| Tw[t cwpate Tov mraLdoc| aA[Aa rou rept Tov mada] ma[tpe adeAgat emitpoTrwr|

10 d1d[ackadoic Kar oAwe Tou]

Kup| vouc emreLoav é eyypagyt|

Col. i

5 Ifthe line is restored on the basis of the transmitted text (émvypdibac, édv tic €AevOepor), its length of 29 letters would seem too long compared with the 22 letters of the adjacent lines; we suspect therefore that something different and shorter may have been written.

6 The middle point signals the end of section 14. It is likely to have been combined with a paragraphus in the left margin, cf. col. 11.

60 AESCHINES

8 mavta Blass-Schindel, Budé; adzvavra a b 1 mo V Laur f Schultz, Franke. We print the Blass-Schindel text, although it is possible that the papyrus had a}zavra.

Col. ii 2 The paragraphus signals the break between sections 17~—18; it was probably accompanied by a middle point, cf. i 6 n.

LUCIANA SABINI

4029. Arscuines Jn Tim. 38-43 101/221(a) Approx. 17.5 X 22 cm Third century?

A much-damaged leaf of a papyrus codex, reassembled from sixteen fragments. There are two columns each side, with page numbers 13 (—) and 14 (|) at the head (in a different hand). The text on the leaf is lacunose and presents numerous variants. From the page numbers we may calculate that no other work is likely to have preceded the /n Timarchum in the codex, and that this speech would have occupied around 64 pages (=16 double leaves). The margins are preserved in part (upper 2 cm, lower 3 cm, inner side margin I cm, outer 2—2.5 cm). The intercolumnium each side meas- ures approximately 2 cm. Column areas are roughly 6 x 17 cm; there are 35 lines in the first column and 36 in each of the other three, with an average 13—15 letters per line (the outside limits are 11 and 19). The resulting leaf size of not less than 17.5 cm broad by 22 cm high (the margins may have been greater than what survives) would place it among the aberrants of Group 5 in E. G. Turner, The Typology of the Early Codex 18 and 24.

The script is a rather broad and heavy severe style, with (across the fibres) a notable contrast between thick and thin strokes. Along the fibres, this contrast is much less pronounced. There are no accents. There are occasional stops as punctuation, in at least one instance a later insertion. A paragraphus signals the start of §40. Diaeresis occurs twice. Elision is frequent but not universal, and scriptio plena is preferred at the ends of lines. Iota adscript does not occur. Three times a supralinear horizontal bar represents v at the end of a line.

The format of two columns per page in a papyrus codex implies an attempt at a prestige production, see Turner op. cit. 35-7. This codicological level, however, is in contrast with the textual quality. The papyrus presents inversions, additions and omissions which are often unjustifiable and not all of which are represented in the mediaeval tradition. There are very many differences from the text in the Blass- Schindel edition, but a conjecture in that edition is confirmed (recto ii 26). In general, the papyrus does not coincide with any particular MS or group of MSS.

4029 is the sole papyrus witness for §§38-40 and 42-3. is recorded in part also by P. Duk. inv. G 44 ed. W. H. Willis, Studies presented to Sterling Dow (=GRBM to (1984) ) 311-4 with plate 21. The first part of §41 (here verso i 7-24) is also cited by

4029. AESCHINES, IN TIM. 36-43 61

Athenaeus, Deipn. 339 b—c. At verso 117-18 4029 shares the word order of P. Duk. (lines 10-11), unique at the time of Willis’ publication, and at verso 1 7 it attests the wording cited by Athenaeus and proposed for P. Duk.

There is no evidence of any manufacturer’s kollesis on the leaf.

Recto Col. 1

(m.2) [Jy (m.1) [uncarte ev| TL cadwe

“abo pe dea upalc BovAopevoc] [aA]Aa z[oAv paddov]

: t[ov]tw [ovrw yap ar] cx[plwe [tvyxaver] BeBiwx[we were Tov] Ta TOUTW TETpA yueva dveEvovta

10 aduvatov Eewvat wie avto|c BovAerau e[urew elav uy Te [Ka Twv] TovouTwv [pbeyE|yrar pnua

15 [tTwv ev|AaBycouat 8[ avro] movew we a[v duv]wpor wa Nie eee ae yar w avdpec Abnvar

20 ou we peTpic jer Aw mpochepecBau Teiwapxw eyw yap o[cla pev ovtoc trar[c] [w]v exc To cwpa 7[o]

25 [ea]utov nyap|7nKev] [a}den[ue «Jax [ecre] [ra]ura [axupa werep|

[ra ela Tw[v Tpraxov]

62 AESCHINES

[ra] kar tla mpo Ev]

30 [kAewd Jou 7 [er tec addy] [aw |motle mpobecuia| t[o]rav7|n yeyove a 3] ¥ dy dplovwv Kar pecpaki[ov] w[v] Kau

35 touc vo[uoluc [e|mctla]

Col. i [{Levoc TOUC TNC 770 [Ae]we dvatrempakrat. TEpl TOUTWY Eyw ye TAC KaTHYOpLac

5 TOLNCOMAL KAL UV pac a€i||we]w [er av] t[ouc c]z[ovda€]e[uw]

ov[ro]e |y]ap zav

tLov] we[v] tpwrov e[ze|cdy [alm a JAAa [y]y ex m[ar]owv exa

[oJ ]cer, [ev Teipat

emt Tou [FE ]v@véicou

[e}

[u]atprov mpodacer

5 pev [THC] TEXvYC pladnz|nc [7] de a[Anbeva] mw[Alew [avrov pon |pnue [voc ocou ev] ovr

20 [tTwv eutropw|y 7 Tw(v) ent | [adAwy Eevw|v 7 [twv todTwv Twv] [muete]pwl[v Kart €| [xewo]uc Tove xpo

25 [vouc ex] pncavro

[tw cwpat|t tw Tex

30

35

Verso Col. i

(m.1)

on

20

4029. AESCHINES, IN TIM. 38-43

[wapxo]v exwy Kau To[vt]|ouc umepBnco wale clva pnderc ex

7 [we a]pa araly|ra ax[pr]Bodoyoupar

wv [d ely tT[lar]c ouKe auc [yeyove klaTau

cx [pve To cwula [T]o ea[utou KQL TYYV lo

Aw [prcbapvwr €|

(m.2) 1d T QUTW TOUTW O Mev amrayopevet 0 vol poc | un mparrew 7 u[n] Snunyopew rept]

TOUTWV TOL1)CO

[ar tT]ouc Aoyouc [Mucyo]Aa[c e]cte Nav Kpa[touc w] av[dplec Aénvai[o.] Ko[Au]rz ile adda KaAd[oc] Kay[aboc] kat ov[da]un av

tic av[ tov pleura

To: Tepe d[€ T]o pa ypa Touvt[o] dao viwe [e|co[vdax]w[c] Kat Tle|pe alvtov] ex wOw[c elxlew «Ba] [p]wdol[uc n KBapr| ctac: tT[aute be Aeyw]

ov Tou [doptixov |

63

64 AESCHINES

[evexa add wa yvw |

[puecnre alvu[rov oc]

[ruc ectu]v* ov[toc ac] 25a wv [evexa tac}

2

[<]Popevoc d[carpr]

Oo

Bac erove[tro Tyrap | xoc ovto[ce em Tov] Latpevov apyu[prov] 7] mpoavadweac a[v]e

30 = etn |cev avrov [K]au ECXYEV TAP EAUTW evcapk|olv ov[ra] Kau veo[v Klar [Bdle Avpov [Kat emurn|

35 de[vov TpOoc TO mpa |

y[wa o mponper]

Col. ii to exe[woc ev mpla[T] [te] o[vroc de zJa [cyew Kat Tavta] ovK w [kvycev adlA vtec

5 [7m Tiwapx]oc [ov]roc [ovdevo|c wv Tw(v) [wetprwv] evdenc: [a ]oAAnv yap mavu KateA[u]zev o traTnp

10 [a]urov ovela]v y[v olv Toc KaTednd[oKe|y we eyw mpotov|toc €]7 devEw Tov Aloyou aA] A erpagev [tavTa]

5 dovAevwy [Tac aicxe]

cTaic ndo[vaic os |o

dayral [ic] ] xa [modvr]e yo

4029. AESCHINES, IN TIM. 38-43 65 Pathe: ioe dee <b

TvwV Kat avdAnTtpice(v)

20 Kat ete] ‘ac’ parc Kar KvBorc Kat Touc addAoic VP wv ovdevoc xpn Kparet [c]@ac rov yeveov Kau [eA]evBepov Kar ov[x 7]

25 [exuv 8 9 rapoc o[v]

[toc ex ]Aim[w]v pev

[7Hv matpway [o]e[Ke] [av drartwulevo[c d¢ 7a] [pa MicyoAla outle arp]

30 [kw ovte dilAw [ov8 nrc [kiwtn ov|re wa[p em] [tpomw alAAa z[ap addo] [7]ele]@ [xa] z[pecBurepw] [eJavrov [x]a[¢ wap] ax[o]Aa

35 [c]rw mlep|e t[av]ra avroc

[«w]pasloc] ew[v mo] Aa pey

Recto Col. i 12 1. TTeupacet 14 l. larpetou 20 Tw 26. 1. Tiwdpyou Verso Col. ii 4 u? 6 Tw 17-18 |. wodutedeta 18-19 1. detrvwy 19 avdAnrpice

23 |. yervaiov

Recto, col. i 2-3 ddJacxe[w] vale with f Abb d Barb h q r; dude om. Blass-Schindel, Budé, Schultz. 5-6 [ovrw yap at|cx[p]we with f Abb d Barb h and pr. q, and Budé; e atcypac ot7w Blass-Schindel, Schultz. II-12 wlc avto|c BovAerar e[umew. elmeiv we avtoc BovAetau Blass-Schindel, Budé, Schultz. 18 yap. dé Blass-Schindel, Budé, Schultz; om. Laur. 22 Teapyw (1. Tysdpyw) with f Abb d Barb t; Tysdpyw tovrwi Blass-Schindel, Bude, Schultz. 23 ovtoc trat[c. maic Blass-Schindel, Budé, Schultz. 27 dxupa restored because of the space, with Budé and Schultz; om. Blass-Schindel with V, p. 29 «ai Franke, Schultz; 7 f Abb d Barb h o p Blass-Schindel, Bude. 31-2 mpobecyia] t[o]ravt[n yeyove. roratTy yéyove mpobecuta Blass-Schindel, Budé, Schultz.

Col. ii 3-4 éywye Bude, Schultz; éyw re Blass-Schindel with Emperius. 6-7 at we jw [em av]|t[ouc. aim én’ adtoic Barb. én’ adroic aéi@ Blass-Schindel, Budé, Schultz. 7 Paragraphus below this line inserted by a different hand. 10 The correction has been made by a different hand.

66 AESCHINES

11-12 exa[,, ],er,. Perhaps éxa[6/]fero or exa[Oé |Lero, against é«d4y70 in Blass-Schindel, Budé and (€xaOyr’) Schultz.

12 ITje:pai. Final iota and diaeresis added by a different hand.

14 [c]atpiov. Read larp: ‘ov; the same error occurs in cod. o.

17-18 éavrdv Schultz; airov a b, Blass-Schindel, Budé; adrév 0, abrouc |.

19 After zponpyuévoc the papyrus must have omitted dc adr todpyov éde.éev, an important clause but one which is not indispensable syntactically. Its omission might then have been intentional. I would exclude an error due to jumping from the final —oc of mponpypeévoc to the initial —oc of 6cor, because there is space in the lacuna for oc oc. The traces in the margin level with 20 may relate to the omission but I have been unable to read them as part of the omitted text.

20 Marginal note by a different hand.

26-7 [rw cwuart|e tw Tei[uapyo]v (1. Tysdpxyov). Blass-Schindel added <7), against 7@ caHpare Tiwapyov in Schultz and Bude.

29 pydetc. wy pe Tic Blass-Schindel and Schultz, wy te tue V, wy tue d fh p q Barb Abb and Bude. The papyrus reading appears to be novel.

30-1 amal[v]ra ax[pe|Bodoyoupar. mavra axpiBodoyodpac f Barb t Abb; Acav anavt’ axpiBodroyovpat Blass- Schindel, Budé, with V p; wdvra Atav axpiPoAoyodpar q; Atay axpiBoAoyobpat Grravra Franke; Alay axpiBodoyod- prac Schultz.

Verso, col. i

1-2 0 pev amayopevet. 6 amayopever Blass-Schindel, Bude, Schultz.

3 At the end of the line ,u7) (so Schultz) is more likely than y7dé (p, Reiske, Blass-Schindel, Bude) for reasons of space.

7 [Micyo]Aa[c €]er. This word order with omission of tic was proposed for P. Duk. inv. G 44, line 2, and is found in Athen. Dezpn. 339 b.

g-10 The last traces of 9 suggest 7 rather than v; thus Kodurrevc (P. Duk. inv. G 44, Athen, Dezpn. 339 b) is more likely than KoAdutevc (Blass-Schindel, Budé, Schultz).

14 High point a later addition.

17 After cai, omission of ded rwac (present in P. Duk. inv. G 44 and Blass-Schindel, Budé and Schultz).

17-18 a[e]pe alutov] evwOw[c e]x[ew. Similarly in P. Duk. efwac exew epi atrov Athen., Blass- Schindel, Bude; éyew etwOwc wep! abrov Franke, Schultz; éywv epi adtov d fh q Abb, Weidner; éxwy etwhwe mepi avtov Barb.

25a This interlinear insertion may be by the original hand, although in a lighter ink and more informal.

31 Final v of écyev has been wrongly added before a consonant.

Col. 11

2 Above of, a rough mark in a lighter ink, possibly intended as a high stop.

3 Unexplained ink above and before ouk.

5 [ov]roc. odroci Blass-Schindel, Bude, Schultz; odrociy g, h.

10 [a]urov. adt@ Blass-Schindel, Bude, Schultz.

14 empagev. Cf. verso 1 31 n.

17 The ink traces are confused, but correction from dative plural to singular (—¢ayral[ [cc] ]) must I think be preferable to correction from singular to plural (—dayia[:]S) because of the avoidance of iota adscript in this MS and the too wide gap which would follow uncorrected —¢ayia. Thus the high ink trace will be from the erasure, not a supralinear sigma. Schultz retains the plural; Blass-Schindel and Budé prefer the singular, with f Abb d Barb.

18 After Ava, a cancelled letter (<?) with possibly w (cancelled?) after that, but it is hard to discern p[ in the traces following that. I cannot absolutely exclude —Aalic «[ac(?); the repeated plural might be expected, and together with dodayiaic is retained by Schultz. The traces transcribed after u[ or «[ are remains of the erasing strokes only. Either way, there is no other trace in the text tradition of other words between modvtedela and defmvwy (dumvwy pap.) except in f Abb Barb which give xat deimvw. The fragmentary marginal note (in a different hand) does not help to reconstruct the text.

20 etepaic by the original hand, an error for ératpaic, was corrected by a second hand which wrote supralinear az above the second e. The correction was taken beyond this, however, with the addition of

4029. AESCHINES, IN TIM. 38-43 67

supralinear € above a following, perhaps a failed attempt (iota omitted) to write éra:pedarc, which would be a variant unattested elsewhere. 31-2 ov|re malp emtpomw. Similarly Schultz. tapemirpémw fo v Laur Abb. map’ del. H. Wolf, Fr., Bens., Budé. The whole phrase was deleted by Weidner, and is printed but bracketed by Blass-Schindel. 35-6 avroc [w]pa:[oc] w[v. There is no other trace of adrdc in the text tradition.

GIOVANNA MENCI

4030. Arscuines /n Tim. 43-52

47 5B.47/B(7-8)a Cao x23 cm Late second century + 46 5B.51/C(3-4)a

Five consecutive columns contain §§43—52 of Aeschines’ Jn Timarchum on the back of a second century register. There are two fragments, each reassembled from many smaller pieces; the gap between them comes down the middle of col. ii. This is the first papyrus witness for this section of the oration. Cf. the addenda in the edition of Blass-Schindel, pp. xxili-xxvi. The papyri of the Jn Timarchum listed there, to which P. Duk. inv. G 44 (see 4029 introd.) and also 4027-9 and 4031-4 should be added, all come from MSS different from 4030.

The upper half of the last column has been lost. The height of the roll was approx. 23cm, with an upper margin of 2cm and a lower one of 2.5cm. There are two (this of course is on the side of the roll with vertical fibres), giving a sheet width of 21 cm. The papyrus was a working copy rather than a prestige production; the height and width of the columns are uneven, the number of lines per column is not constant (33 in col. i, 37 in col. iv), and there are many writing errors corrected by the scribe himself in the course of making this copy. The same scribe has also inserted some punctuation and lectional signs at the same time. However, most of the corrections, cursive supralinear additions, accents and punctuation have been effected in a paler ink at a revision stage, although they may still be the work of the original scribe. Diaeresis occurs over initial « and v and in ii 8 and iv 33 over epideictic v. Elision is marked in iii 4, and iota adscript occurs in ii 26. Final v at the end of a line is occasionally represented by a supralinear bar. Written in a semi-cursive hand.

Compared with the Teubner edition of Blass-Schindel (1978), the papyrus pre- sents frequent changes in word order, and in many places the reconstruction of text lost in lacuna is uncertain. From an overall view of the table above, pp. 51-2, the papyrus tends to align itself (but not consistently) with the readings of d fh q Barb Abb.

68 Col. i

20

25

30

AESCHINES

CAMLEVOU de | TOU Micyora K[ ac]

tov Paidpov] tou Eevorc: Ka[e|

TO dEecuwrT|npLov' OTL MELpaKLOV eAevbepov dilehbe|plav. PoBybev

TEC OC €VOL €| VVYOV | TEC WYOVTO [ Eevor pe[vyov] XovTé

[ [ | keAevovTwr | non akoAovbew ELC [ [

[kataAimovtec Ta 7 lapeckevacue [va kat TavO]| ote e[yw alAnOn rAeyw. [mavrec (?)ou K]at EKewvouc TouC [ypovouc (?)oco]e eyeyvwekol] pe |v [Micyorav car] Tysapyxov icace [8] 4 87 [kau mavu xaip|w ore yeyove Lou [y dx mpoc]| avOpwrov ovk 7 [yvonuevo|y ud vuwv ovd a

[7 addAov ytyvwckope|vov o[v de [voc 1 am avtovu Tov em|Tndev [watoc Tepe ov Kat THY] Wydov [weAreTe hepew trepi wlev yap [Twv ayvoovpevwv ca]deic

[ucwe mpocynKer trovetc |ar

[tac amoderéerc Tov Kal|TyHyo

[pov zrepu de Twv opolAoyou [wevwv ov Arav eywy |e pe

[ya epyov ewar vol|uilw to [katynyopew alvayrvyncar yap [wovov mpocyk jet To[v]c axovov [Tac eyw Towv|y Kau rept ogo [Aoyoupevou 7] payyatoc ov

[toc €.5 emjedn ev duka [cTnpww ecuev| yeypada pap [ruptav rw Mic]yoda adnby

[ev ovk amrat]devtov de [wc]

[(?)we eyw ewav]rov mewn:

($43)

844

Col. ii

20

25

30

33

4030. AESCHINES, IN TIM. 43-52

[avro] pev yalp rovvoy|a

[rov e]pyo[v] o ex[parte mpo]c [Tour]ov- ove elyypadw] |... | [ovd] addo yey| pada ov|dev- [.. . Jemefn[prov €]erw

[ex tT]wv vou[wr tT]w Ta A[n@]y waptu[pycart] a

8 ect[e]v vu pwlev a]xo[vov] ct [y]vwpima axi[vdu]va

de Tw waptupo[vrti] Kat

un alic]xpa tLavra yleypa

ga: e[av ple[v ovr €]beAn §46 cn o M[ucyoAac ¢.5 Tadly On wlaprupew ta duxarla troun|[cer eav de po]

aipnTta[e exkAnrevOnvar | paddro[v 7 tadnbn paptv]

pew. vpulerc To oAov pa]

yua [cu]ydlere ev yap o]

pev [7 ]pacalc aucyuvertar|

Kat T[polarpync[erar yiAc]

ac dp[ax]uac p[aArov a] tot[eccar] tw [Snuociw we]

te [yn] devé[ar to mpocw]

mov To eau[Tou vpww o de] tretrovOw[c dnunyopycet|

coor [,."] voploflerne 0] touc outwe B[deAupouc €€]

etp[y|wv azo t[ov Bnuaroc]

ea[v 6] apa umaxk|[oucn pev Tpa] 847 mnt lau de em [To avacde |

cra[Tolv em zo] [e€opvucbar]

tac [aA]nbevac: w[c Tywapxyw |

t[ev] xapirac a[mrod.douc]

[erepolic de em[wWeréw |

69

70

AESCHINES

Col. ii

20

25

30

TOLOUJLEVOC WC €, ETTLCTA

TAL TA TOLAUTA CUVKpUTTEL(V)

TPWTOV [LEV ELC EAUTOV

efapapTncetau emeiT’ ov

dev ectat avtTw TAeov' €

teplav ylap ey|w yeypa]pa pap

tu[pralv tou[c ec]doc[c] Tuwap

xov tovtovi KataXit|[ov|ra

THY TaTpwav oiKiav Kalu] du

aitw|plevov mapa Micyoda:

Tpaypa ouwar xaderrov €€

epyalecBar eyxerpwv" ov

[re yla[p ewe Sleu trouc ewavrov

[drAouc apt |upac mapa|c]ye

cba [out]e toluc] routwy exOpouc:

OUTE TOUC LNdETEPOUC NUw(V) [.]

yryv[w]cKkovtac: adda Touc

tovtwyv gdiArov[c] av 6 apa Kau §48

TOUTOUC TELCWCL LN) LAPTU

pew’ we ovk [orlouau ec de wy

add ovy [ama]vrac exes[vo] ye

[ CELO vt jac

[adeAecBar tyv adnfer|av:

[ovde tHv ev TH TroAEL] TEpu

[Tiwapxou dnuny nv ov]x ¢

[yw mapeckevaca TouT]we

[aAA avtoc eavtw ovt|w yap

[xpn Kkabapov e|iar tov Be

[ov c.3 Tov clwd[polvoc av

[Spoc were nd erd[e]xe

[Bar] do€av aitiac movnpac:

[BovAo]war d€ KaKewvo po §49

[evrer]v- eav apa vmaKkoucy

35

4030. AESCHINES, IN TIM. 43-52 [o Micy]oAac roic vopou[c] Kae

[uv ev]cu ducerc avOpwrw(v)

Col. iv

20

215

30

ToAv diad[epoucat o| POnvar adrdA[nAwv ra}

Tepe THY y[AcKcav eve]

ot [wev yap v[eor ovrec] [pode ]peic K[ae mpecBv] [repo]. dawo|vrat etepor| [de w]oAvuv api[Auov xpo] [vo]u yeyovot[ec mavta| [w]ace veo d[oxovew ec] [vac Tov|twv [6 ecte Tw] [avdp|wv o M[ucyodac tuy] [xavlec wev y[ap nArkww | [tn ]¢ wy exol[c Kar clur[e] [é]Bol[c] Kas [ect] nuw TouTt meut[Toly Kau Teccapakocto|v] eToc: eye plev TOCauTac TroALac €

xw oca® ipurc opate adX ov K exewoc: dla TL ON TAU

ta [mpoAleyw [i]va py €€ aup[vyc t]dov[7]ec avtoy favp[acn|re [Kar t]ovov tov tT[u Tn dtjavo[ta] v70Aa Byte & [nplaxAler|c aAA ov roc {n} ye tout|ouc, |{z}ou fe Av diagepe[t] apa pev

yap e[c]rw 4 pluci}e rovav Tn Tov avdpo[c aya] de

Kal pipakiw [avT]w ov

tu emAncial[e|v wa de §50

71

Ve:

35

Col. v

20

25

30

35

Col. i

AESCHINES

pn dcatper[Bw] Kader pol] mp|w|rov plev toluc idorac Tipapxov t[ovro]vi diac [Twp | Je[voc] [ [ev tly Micyo [Aa ox lia’ em[ec|]ta tyv Par [dpo|v papry| pilav ava [veyv|ewle]* reAe[v} ray

{1-15 lost]

ul

[

Wa sler<|

voc avtTw plot] doxer toy[Tw] evoyoc €[t|var eay de v

pac avammncac ETL

derEw vU7epBawwv

toucde Touc ayp|ujovs Kyndw vidnv Kat AvtokAedy(v)

kat Oepcavdpov |[kar y A[v]

ena te ]dnv | WV €V TALC

In hetere [a]verAnpper|oc| yeyove ov jovov Tapa

tw Micyoda pepicbap XNKOTA aUTOV ETL TH,

cwpatt adda Kat Tap ETE

(§51)

§52

4 The intercolumnium before col. ii is preserved from here to the foot of the column, and measures

approx.

1.5 cm wide.

8 drt éyw: so a b m and Bk., Turr., Fr., and Schultz, cf. §65; Blass-Schindel, Weidner and Budé om. eyw following d fq t Barb Abb.

4030. AESCHINES, IN TIM. 43-52 73

Q mavrec of d fh q Barb Abb. mdvrec bcou Blass-Schindel, Budé. davrec dcou Bk., Turr., Fr., Schultz.

10 Calculation of the lacuna suggests that the papyrus had the reading dco]: with d fh q Abb.

11 icace[8}757, with a rough breathing over first 7 as well as the accent. icacidy was first written. A second hand revised the text by deleting 8 with a diagonal stroke through it and a dot above, and adding dy and the breathing and accents, to give # 81 (so a b h Vat.)

12 yeyove or. wou yéyovev edd.

20-1 moveicBar rac dmrodetterc with d fh q Barb Abb. rac amrodeiEeic moveicOar edd.

26 Je. very dubious. The ink is very confused, and there are traces of 2-3 letters above the line.

27 Kat wept. So o=r. xaimep Blass-Schindel, Schultz, Bude.

28-9 m]payparoc ov[roc. tod mpdyparoc dvroc q & ovroc mpdypyaroc d fh Barb Abb.; 10d mpdyparoc, om. évtoc edd.

29 The text tradition gives no indication of what might have filled out the lacuna.

33 The initial restoration is uncertain but is suggested by the deletion of we in 32, and fits the space and accords with the later manuscript evidence. Presumably we in 32 was deleted later, when it was seen that the word had been written twice.

mrevBa- Possibly mecAw||e]-?

Col. ii

4-5 v and high stop added at the end of 4 by a different hand. The expected text continues 6 émLruov, which is too short for the lacuna at the start of 5. If 4 originally ended ovSe, did 5 begin v (i.e. odderi, so l) @&

8-9 See the table above, p. 51.

10 de. re Blass-Schindel.

13 For the lacuna of approx. 5 letters, the text tradition offers Sedpo mapeAOwv. Did the papyrus contain dedpo and omit zapeABawv?

14 On the left, a heavy short diagonal stroke in the intercolumnium. It is misplaced if it was intended to draw attention to an omission in 13, see n. above.

u[aprupew. We print the text as in the Teubner edition, but either this (with d fh q Barb Abb, and so Bude) or u[aprupycar (so Schultz) would fit the space.

17-18 7 taAn67 paptupeiv bracketed by Blass-Schindel.

20 € of wey corr. [7]paca[c: 1. mpdgéac. For c in place of £, see Mayser i?.1.184.

21 A heavy dot of ink in the intercolumnium to the left.

22 yu[. Scanty traces only from the upper part of a letter. The text that preceded suggests that the papyrus had y[aAdov, with d f h q Barb Abb. paAdov om. Laur. xiAlac wadAov Spayydc Blass-Schindel, Bude, Schultz.

32 Correction by a different hand; w heavily stroked out and o added above the line.

Col. ii

1 It does not seem as if the traces will readily adapt to ed; was e: written?

2 CUVKPUTITEL =

11-12 See the table above on p. 51.

13 The space is not decisive, but éué (Budé) would probably fit better than pe (Blass-Schindel, Schultz).

14 A supralinear bar (representing final vy) appears to have been written over « at the end of the line.

16 nyw-.

17 eee first v, at least one letter, subsequently cancelled. First a of add corr. from o.

20 A spot of ink in the intercolumnium to the left. Mark of ink after 7 accidental?

21 The papyrus does not support the Teubner text’s insertion of ye following admavrtac.

22 There are scanty traces of ink within the indicated lacuna, but they are too small for identification as particular letters, and their lateral location is too uncertain for them to be usefully represented by dots. The space in the lacuna is not decisive for a choice between the readings transmitted or conjectured. It is less confusing if the words are divided. odédérore a b g 1 m p; oddeurmore d f h q Barb Abb; ov pymore conjectured by Emperius, foll. Fr.', Weidner, Schultz, Blass-Schindel. p inserts pe. d0vwrtar d f Barb Abb, Weidner; duvyjcovrat h q ab gl mp, Bk., Turr., Fr., Bude, Schultz, Blass-Schindel. Further conjectures are given in Schultz’s apparatus. The line length, including transcribed v7]ac, should be approximately 20 letters.

74 AESCHINES

26 mapeckevaca Tovt]w. The iota adscript is exceptional in this MS. This order of words is given by d fh q Barb Abb. TOUTW TrapEecKEvaca Blass-Schindel; rov7w del. Weidner.

27 The space before ovr]w yap suggests GAN’ adrdc éavt@, with d fh p q Barb Abb and Weidner. éavr@ omitted by t. add’ adroc odtoc éavt@ Blass-Schindel, Bude, Schultz.

29 The initial lacuna could contain some three letters more than the transmitted text; there are no variants which will guide us to a solution.

35 avOpwrw-.

Col. iv

2 add[nAwv. So Schultz, Weidner, Fr.; r@v adAAwv Blass-Schindel and Bude with q t.

5 «[ae. dé Blass-Schindel.

g-10 veor S{oxovew ewar. So dh q. S[oxovvrec (f Barb Abb) would also fit the space. Blass-Schindel, Budé and Schultz omit Sox. efvac with a b. véor daivovrar g 1 mo pr Vat Laur.

16 reccapaxocro[v]. So b (rerr— Blass-Schindel, Budé, Schultz). See F. T. Gignac, Grammar I 146. Punctuation and «ad at the end of the line added by a different hand.

17 Tocavrac. So p. tocavract Blass-Schindel, Budé, Schultz.

18 Marks of ink in the intercolumnium on the left. If not accidental, they may relate to the scribal errors in this line. The supralinear corrections have been added by a different hand.

19 5y. Sod fh q Barb Abb. odv Blass-Schindel, Budé, Schultz. Omitted (i.e. 61a 7/ rab7a) by Weidner.

20 Diaeresis over the iota visible.

21 w|dov[r]ec avroy. Inversion of the word order in Blass-Schindel, Bude and Schultz.

23 Ink traces above end of line probably accidental.

27 €[c]rw 7 d[vca}c. Sod fh q Barb Abb. 7 ducic écré Blass-Schindel, Budé, Schultz.

28 Ink traces in the intercolumnium on the left, probably accidental.

28-9 apa] de cat. dua dé 75y Blass-Schindel, Budé, Schultz (aua 8 767).

29 I, peipakiw.

29-30 [uvt]w ovr. dvte adt®@ Blass-Schindel, Budé, Schultz.

31-2 Kddev pou mpw@rov pev is the order also in d f h q Barb. mp@rov pev «ddAer ou Blass-Schindel, Budé, Schultz.

32 |. €iddrac.

Col. v

1-15 The upper part of the column has been lost, containing c. 15 lines of text. The loss from col. iv 37-v 19=§50.4—-§51.3, which may be calculated to have been distributed in lines with an average of 19 letters each.

18 Reconstruction on the basis of the scanty traces is difficult, bearing in mind also the various possibilities offered by the text tradition. a[v7]ov? But the lacuna calculable before 19 zmit. would not then be sufficient to contain the text as given by Blass-Schindel, §51.3. Perhaps, as often, there was a change in word order.

21-2 of ... mpatrovrec an error for 6 ... mpattwv. The scribe does not maintain the plural.

26 |. dvapvycac.

28 The correction is by a different hand.

29 avtoKAedn-.

30-31 [Kal y’ AvdpoxAedSyv |. An inclusion in (apparently) only this MS, but deleted in antiquity.

31 After the personal names, our text continues with dv év raic (so Bude, following Greg. Corinth. in Walz, Rhet. Graec. vii 1185). There is no trace of add’ émide(Ew adrodc Aéywy or its variants in the mediaeval tradition, for which cf. the app. crit. in Blass-Schindel and Schultz.

33 ov povor is apparently not attested elsewhere. kat 47 pdvov MSS unspecified; 7 pdvov Blass-Schindel, Bude, Schultz.

34-5 peucOapynKora. |. pepicbapynKora.

35 Curious vertical trace at the end of the line, in a dark ink like that of the original scribe. It is not an extension of final p above, nor, I think, is it an iota adscript.

ISABELLA ANDORLINI

4031. AESCHINES, IN TIM. 79 75 4031. Arscuines /n Tim. 79 g 1B.182/F(a) 7.4x6cm Second or third century

The ends of ten lines, broken above and below, written in a careful severe style, with extended or compressed letters to achieve a justified right margin. The back is blank.

There are no accents or breathings or iota adscript, but the scribe makes liberal use of punctuation marks.

This is the first papyrus to attest this section of the oration.

[ c.gletters €]uor 7a [pecty]kw[c] exnpwra [vuac] To ex Tov vomou Knpuypwa: twv Undwv 5 [n tet]puTern o

[tw doxer] memopvev [c]@ar Tywwapyxov: n de [7An] pyc oTw un: TL [av edn |dicacbe: axpu

10 [Bwe od olte Kateyyw

1-2 €uol mapectynKwe, attested by d f Abb Barb h t Laur. I(g), has not been accepted by any modern editor. See the table above, p. 52.

5 Cf. Schol. in Aeschin. I 79 apud Schultz, p. 268 (now ed. M. R. Dilts, Leipzig 1992, n. 174 a—d). Aeschines is cited by Harpocration, Lex. I, s.v. rerpumnuévyn, with II p. 436; and by Bachmann, Anecd. Gr. II p. 333.23, 373.8. For the technical expression retpurnuévn Yndoc see Arist. Ath. Pol. 68.4, 69.1 (P. J. Rhodes, A Commentary on the Aristotelian Athenaion Polttera (1981) 730 ff.); Schol. Gr. in Ar. Vesp. 987; Phot. Lex. s.v.; Poll. Onom. VIII 123; Suid. s.v. (T 417); Bekker, Anecd. Gr. III p. 307.18. For an illustration see Daremberg-Saglio, Dict. s.v. Aicacrai.

4 High point added, possibly by a different hand.

7 Tiwapyov. The accusative is accepted by Schultz and Franke. Tiwapyoc (d f Abb Barb Laur. I(g)) is accepted by Blass-Schindel and Bude. See the table above, p. 52.

LUCIANA SABINI

4032. Arscutnes Jn Tim. 131-2, 134 82/88 (a) gx 15.8cm Second century

The badly damaged remains of parts of two columns from a roll, with the lower margin. The first column has parts of §§131~—2, the second column part of §134. The number of lines per column may be calculated as 32, with 24-27 letters per line.

76 AESCHINES There is one inserted high point (i17) and a decorative line filler (i 4 end).

Written in a rather thick and blobby upright hand with some crude serifs. The back is blank.

Col. 1

c. 12 letters colu Ta k[opiba rav| Ta xAavickia mep|teAou|[evoc Kat] touc padakouc xu|Twri|cKouc e|v oc Touc Kata Twr| dilwy Aoyouc> Grid letters ] ew doin, exc crizletters’’ jtarc|) Cc oletters || c. 12 letters | oale av avtouc]

[xeupncer duacu]pew tyv oA[ nr]

[evctacw tov ay|wvoc: ov Kpuciw

[ [ [ [ [ [ [ [ev Tuc uN TpoE|iTwV TOUTO TOLH [ [ [ [ [ [

[eEeupnkevar ule dackwy ad [Aa dewyc amralidevciac ap[x]nv

20 [7apadepwv 7|pwrov pwev Touc

Col. i

[r]ouc [6 nbn yeyovotac ed ouc] mpoc[nKer ceuvuvecbar Tv] moA[v eav KaAAe Kat wpa] dueve[yKovtec exTrAnEwce|

5 twac K[au TEepysaynror €& €]

PwToc YE [vwvrac TouTouc |

4032. AESCHINES, IN TIM. 131-2, 134 ri

5 ].ew suggests an infinitive, perhaps mepueveyxeiv. tepievéyxac Blass-Schindel; no variant appears to be recorded here in the mediaeval tradition. The implications of an ink trace between don and eic remain uncertain.

5-7 The reconstruction of these lines is difficult. From Socy in 5 till owafe in 7, the text tradition offers us a total of 32 letters, to be distributed over a space sufficient for 50. This exercise is complicated by the fact that apparent aic in 6 will not fit with any word in that tradition.

g [av]. The deletion (indicated by a supralinear bar) is by the original scribe, who presumably caught himself writing avdpdc too soon.

ROBERTA BARBIS

4033. Arscuines /n Tim. 190-192 g 1B.170/J(a) 3.5 X 13.5 cm Second or third century

A tall narrow strip with part of the top margin and the middle letters of a narrow column. Distribution of missing text between lines is conjectural.

Lines 6—7 have been written closer together and in a smaller script, but by the same hand; 7, moreover, must have projected into the right margin (although not as much as the transcript might suggest, since the letters are smaller). Perhaps the writer left a blank line, because his exemplar was faulty in some respect, intending to fill it in later, but he then found that the space he had left was insufficient.

There are no accents. A high point is used in 7, 12, 18 and 24, and a double point in 23. At least those in 7, 12 and 23 are insertions. In 18, the heavy high point after dyJov is followed by a long horizontal line.

Written in a distinctive upright hand with occasional slight serifs and some ligatur- ing, with very few projections above or below the main line of writing. Cf. 4034. The back is blank.

[ ]vas 7” (vac.) [oux am alvOpwrw|v aceA] [yerac] yevecO[ar] [unde toluc nceB[nKorac] 5 [xaba|mep ev tatc Tl payworarc | [ITowa]c eXavvew kale Kodalew]

[Saciv] nupevaic: ad[A au mpomreterc Tov cwuaTtoc ndovar] 8

[kat T]o undev [txavor]

[rour]o mAnpot tla An]

10 [ernpi]a- Touto e.[c Tov]

[erax|tpoxeA[nra eur]

48 AESCHINES

a

[Bale |e: tovTo [ectw exa| ctw| [Town t[avta zapa| KeAleverat [char]

Tew | TOUC 770 [Acrac]

rolt|c U7Tr7) Pp plerew Tu| pavvoic|

cuyK jataAve|w Tov| dnp lov-——— [ov yap]

[er

[ke

[

ef [Covrac] add’ mA[ouc Kar | [

[

20

ovv w av|dpec A[Anvacor} tac Tovauta|c guc[ec Kat] Ta Twv vew|y Cy[Awua] Ta em apet|nv [po]

tpepecbe] ev [

1 The remains in the upper margin may be of an omitted word, for insertion in the text, perhaps at line 8 (see n.). Jvae is in the original hand, but on the small scale of 6-7. ‘7’is in a more cursive script.

3 Above the first € of yevécOar, the letters yy have been added perhaps by a different hand, with the effect of transforming an aorist infinitive into a present infinitive. The variant yevécfar is apparently unattes- ted in the mediaeval tradition.

8 There is no MS or editorial support for the supralinear correction of wydév to under.

At the end of the line, an infinitive is wanting after [c«avov] (if that restoration is correct), but there is no space for it on the papyrus. The overlooked infinitive might have been 7yeicHa (so Blass-Schindel, Budeé, Schultz) or efvar (cf. Ammonius s.v. «éAnc); in favour of the second might be the presence of }vac in the upper margin, see I n.

g Here and in to and 12 (and perhaps also 13, now lost) todto has been converted to tadra by the insertion of two supralinear alphas. As regards 10, cf. Schol. Patm. p. 154 (apud Blass-Schindel).

g-10 AyctHpra first of all was changed to the singular by the supralinear insertion, and then the insertion was itself cancelled with a line through it. We may suppose that t[a in 9 underwent a similar process.

16 The supralinear insertion of the article before tupdvvoic is probably due to the original scribe. The wording of his first version, without the article, is found in Codex Laurentianus conv. soppr. 84 (= Abb).

18 After v, a heavy high stop. After that, a long horizontal line, by the original scribe. Are we at the end of a later insertion as conjectured for 6-7, see introd., only here the space that had been left proved more than adequate?

19 The line seems unexpectedly short.

21 The modification may be due to the original hand, adding the apostrophe and supralinear e¢, with a dot at least above 7. It is not certain how aAdyA[ might have continued; we supplement the line so as to cause the minimum disturbance. é¢’ is bracketed by Blass-Schindel (i.e. aA’ ofc).

4033. AESCHINES, IN TIM. 190-192 79

28-9 mportpeecbe in the Teubner text is from a conjecture by Cobet; the papyrus may have had mpotpepacHe along with (it seems) the majority of the mediaeval MSS. P. Hal. 6 offers mpotpéare, see Blass-Schindel, p. xxv.

29 The traces are scanty but best support the reading év d¢ ed (fh q) against ed 5é of a b g m p Vat Laur with Blass-Schindel, Budé, Schultz.

ROBERTA BARBIS

4034. Arscuines /n Tim. 194-6 26 3B.53/G(3-5)a 6.2 x 8.4 cm Second or third century

Parts of the tops of the two final columns of the oration, with a fine coronis. The final column has three lines only, with a large blank area below. There is no trace of a colophon.

Line length varies between 14~18 letters. Normal column height will have been 41-43 lines. In col.i 1 elision is marked with an apostrophe. There are examples of diaeresis (1 4), a high stop (113), a double point (i 16) and a forked line filler at the end of 16. Another double point at the end of i 14 is wrongly placed for punctuation and perhaps serves a different textual purpose. In col. i the text shows traces of faint horizontal ink ruling lines, especially noticeable below i 2, 7 and 10. There is no trace of vertical ruling. The ruling was done a column at a time; the lines do not reach into the blank foot of col. ii, and no ruling is visible below any of the three lines of script there. No such rulings are present in 4033 (see below).

The script seems closely similar to that of 4033, and it is possible that they belong to the same roll despite the divergent inventory numbers. The amount of text lost between them would accord with one completely missing column. The back of 4033 was blank; here on the back are the remains of six cursive lines, the lower four very faint and possibly deliberately washed out.

Col. i [ctwew etepor| 54 Ex [twv axoAractw |v Kar [twv Tovoutoc KEe]xpy [pevwv apbovjwe t

5 [va tarc BonBevar|c av

[Tw muctevov | TEC >

[ paov TLVEC e€a|ap [Tavwew WV Trp T|nv [

cuvnyopiav ak jou

80 AESCHINES

cal TOUC Buouc ava |e

IO [ [pvncKkecbe Kat Tlouc [prev evc Ta cwpal|Ta [nuaprnKkorac py] v [puv evoxrew adr|Aa:

15 [mavcacbar dnuny|o [pouvtac KeAever |e:

Col. ii Tac veTlEpaic yw | jraic 1 mpag[ic Kata] Newer au’

(coronis)

Col. i

1 A tiny trace of ink above the beginning of this line, of uncertain significance.