MASTRONARDE, DONALD J., Notes on Some Manuscripts of Euripides' "Phoenissae" , Greek, Roman and Byzantine Studies, 26:1 (1985:Spring) p.99 Notes on Some Manuscripts of Euripides' Phoenissae Donald J. Mastronarde 1. An almost-forgotten Thomano-Triklinian manuscript Together with Vaticanus graecus 1825, the manuscript Vaticanus grae­ cus 1824 contains a miscellany of poetic texts, some of which ap­ parently derive from a single scriptorium.1 Vat. gr. 1824 figures in A. Turyn's great work on Euripidean manuscripts2 in two places. On p.359 he lists the contents of folios 8Jr-87v, pages from a fourteenth­ century3 codex unrelated to other portions of the miscellany: Or. 1385-1557 and 1558-91 are extant with a single leaf of Phoen. (lines 802-42) bound among them (f.86r-v). For Phoen. this page is a worthless witness, showing no consistent affinities with any of the families identified by Mastronarde and Bremer, 4 and carrying no new readings of interest. Turyn (254 n.238) menti Dns the Aeschylean portion of the manuscript {ff.54r-80v = Fb)5 to record that Triklinios himself seems to have worked briefly with this codex. What is not reported in either of Turyn's studies is the fact that on ff.3Jr-53 r, on the same paper and written by some of the same hands as the Tho­ mano-Triklinian Aeschylus that follows, is a copy of Phoen. 296-673 and 937-1766 (between 37V and 38r five leaves containing 674-936 have been lost), which is also Thomano-Triklinian. This section of the manuscript was recorded by K. Ziegler in 18826 but not by others 1 See P. Canart, Codices Vaticani Graeci. Codices 1745-1962 I (Vatican City 1970) 240-50. 2 A. Turyn, The Byzantine Manuscript Tradition of the Tragedies of Euripides (= Illinois Studies in Language and Literature 43 [Urbana 1957]). 3 So Canart (supra n.D 245f, correcting Turyn's ascription of these pages to the fifteenth century. 4 D. 1. Mastronarde and 1. M. Bremer, The Textual Tradition of Euripides' Phoinissai (= University of California Publications: Classical Studies 27 [Berkeley 1983]), hence­ forth Text. Trad. 5 See also A. Turyn, The Manuscript Tradition of the Tragedies of Aeschylus (= Polish Institute Series 2 [New York 1943]) 71; O. L. Smith, Studies in the Scholia on Aeschylus I: The Recensions of Demetrius Triclinius (= Mnemosyne Suppl. 37 [Leiden 1975]) 22 n.49, and Scholia graeca in Aeschylum 11.2 (Leipzig 1982) vi. 6 NJbb 125 (1882) 826, where the contents are listed as 296-1766 and the separate Phoen. fragment on f.86 r- v is not noticed. 99 MASTRONARDE, DONALD J., Notes on Some Manuscripts of Euripides' "Phoenissae" , Greek, Roman and Byzantine Studies, 26:1 (1985:Spring) p.99 100 SOME MANUSCRIPTS OF EURIPIDES' PHOENISSAE (editors of Theocritus and scholia) who have discussed the manu­ script, until in 1970 Canart provided a full description of the manu­ script. As he notes, the Phoen. and the Aeschylean portion of Vat.gr. 1824, together with a section of Hesiod in Vat. gr. 1825 and pages of Theocritus divided between the two manuscripts, probably reflect the work of a single scriptorium. Canart also mentions that Turyn identi­ fied the text and scholia of Phoen. as "Thoman" in a private commu­ nication of 1 January 1959. The Phoen. section of Vat.gr. 1824 may suitably be given the siglum Zv. The paper used for Zv and for other sections of the re­ lated miscellany is western, with watermarks identified by Canart as known from the years 1297-1318. This span of years agrees with the presence of a few Triklinian corrections in the Aeschylean portion. There is one column of text per page, usually containing 24-27 lines. I designate as Z v 1 the hand (s) that wrote the text, some scholia in brownish black ink, and the personae notae in red; I designate Zv 2 the hand that used a lighter brown ink to make corrections in the text and to add most of the scholia. Zv should be of interest to editors of the "Thoman" scholia, since it shares with Zrn (Milan, Ambros. I 47 sup.; middle [or early?]7 14th cent.) or with ZrnZu (Uppsala, Gr. 15; first half of 14th cent.) ele­ ments absent from Z (Cambridge, No. 3.14) and sometimes from the published "Thoman" Gu-scholia in Dindorf. ZrnZv have a fuller set of scholia than Zu, and Zv contains some items not in Zrn or any other source that I know of. I record a few examples to illustrate the relationships: 409 (marginal genealogy of Adrastus): ZrnZv (not in ZZuT) [=Dindorf 137. 15f] 409 () XP"f1u"w~ 1nrO MvauEov OV'TCI)() avaYE:ypa'Tr'Tat ... av'T"I} it in'TOpia O'1rtUOEV: ZZrnZvT (not in Zu) [=Dindorf 138.12-139.3 and app. ad 139.4] (likewise, e.g., schol. 1185 [=Dindorf 317.1-91, 1188 [=Dindorf 317.15-25]) 410 gloss O'TrEP Et'Tra~: ZrnZuZv (not in Z T) [=Dindorf 138.10] 415 gloss El.~ 'Ttl 'Trp0mJAata: ZZrnZuZvT; add. am) ,dpo~ 'TO 'TraV 71YOVV 'Ttl~ mJAa~ ZrnZuZv (om. Z T) [=Dindorf 140. 16f] 441 gloss avapiO,.,,"f1'Tov: Zv (not in ZZrnZuT) 448 'ToVro 'TO Kat 'TrpO~ 'TO 'TrOAtV UVVaTrTE'Tat (ZrnZv)/uVVaTrTE (Zu): Zrn ZuZv (not in Z T) 448 AOXO~ KVp~ it EVEapa, Ka'Taxp'TIU'TtK~ aE Kat it 'Ta~t~' () ae AOXO~ aw 'TrE,WV avapwv yiVETat: Zv (not in ZZrnZuT) [Dindorf 144.19f] 7 Smith, Scholia (supra n.S) vii, in describing the Aeschylean portion of this codex ( = Fe): "ineunte s. XIV scriptus." MASTRONARDE, DONALD J., Notes on Some Manuscripts of Euripides' "Phoenissae" , Greek, Roman and Byzantine Studies, 26:1 (1985:Spring) p.99 DONALDJ.MASTRONARDE 101 1228 '1T'pOUTaKTtKov: ZrnZuZv; add. 0 'Yap a7Ta'YopEVTtKO~ /-1/1, Oll J-Wvov 7TPO~ mro1'aK'TLKO"V £lAAll Kat 7TPO~ 7TPOG"TaK'TLKO"V Zv (not in Z T) The colon-divisions of Z v also point to a close relationship with Zm Zu, and particularly with Zm. The following divisions that are unique to Zm Zu among the manuscripts studied in Text. Trad. are shared by Zv: 1351 KpaTa T€/, 1502 Ta8€/, 1528 'TT'poc;/, 1535 J.W­ KP07rVOvv/, 1577 V€KPWV/, 1578 B€!, 1580 BOJ.WUFLV/, 1717 BT,'Ta/ (Zm only, Zu not extant). In 296-354 Zv's division is like that of Zm (and other MSS.) and unlike that of Zu (and other MSS.): e.g., 298 'TOVSE/ ZmZv+ vs J,Ui'TEp/ Zu+. ZmZv also share a false di­ vision of trimeters at 95lf ( ... E'TEPOV/ ... 'TT'OAw/). The readings of Zv also show that it is very closely related to Zm, so close that a superficial inspection might lead to the conclusion that one is a copy of the other. First, among conjunctive errors of ZmZuZv we may mention 440 avfJpdmouFLv, 581 vvJl4>as, 622 K'Ta­ vwv (sic), 953 TavT', 1047 BE, 1113 wytyva, 1234 vVUEUfJE,8 1337 Bt1''TWS', 1363 aAKT, (Zv 1pc), 1580 aJUX'TEpOtUt.9 Then, among conjunc­ tive errors of ZmZv we may mention 305 I-'VPUxtUtV (1" om.) (Zv 1 in rasura), 348 AOV'TPO'Tpocpov, 368 EV'Tpaqn1V Zv2 EV-Tpacfn1v ZmuV, 376 J.WV, 1020 Exvi&xS' a.c., 1038 E'TT'E'T'TO'TV~E, 1158 aJ.La{-, 1558 yp. cpotVUxtS', 1697 E'TEOKA-, 1715 afJAice (also GXsXas), 1755 accents on u'TOAtUaJ,LEVa 'TT'O'TE. Both Zv and Zm contain corrections by the first hand (s) and by a second hand, and all sorts of combinations of agree­ ment can be found: Zv ac - Zv1PCZm: 305 (above), 369 Ct7T'EA.aufJEls Zv ac , 426 SEVP' E'TT'E­ ufJat UOt u'Tpa'Tov (Zv1 in ras., Zm), and about 20 other passages Zv1 - Zv2Zm: 368 (above), 382 aVTap Zv 1, 407 Bvvatl-" av Zv1, and 8 other passages ZvZmac - ZmPC: 349 EtUOSOt ZmPC, 405 1-'7] "'XEtV Zm2, 408 EUXEfJES' Zm2, 950 51-'J.Lautv Zm 2, and 14 other passages Zv ZmPC - Zm ac: 312 'TT'av Zm ac, 322 AEvKoxpwa Zm ac, 500 acpi­ AEK'TOS' Zml, 1002 mhpi"s- Zm ac, 1299 'TaAawES' Zml, 1628 7rEPUW Zm 1, and 3 other passages 8 In JJV~Hr(J€., Zm has a very angular ligature of upsilon and acute accent that I wrongly treated as i in the collation of Text. Trad. Zm is written in tiny letters in a faint-colored ink, and it is often hard to read the smallest details on microfilm. Com­ parison with Zv led me to re-examine the film of Zm and to discover a few details missed earlier: e.g., the second accent in 622; the rough breathing on iTEOKA- in 443, 1223, 1390, 1407; yp. 7Tpo8po/-W" in 296; 368 EV-Tpa.!frrJV ut vid.; 636 -vEiKTjV in ras.; 982 fS oil8a" Zm 1 (E" eras. Zm 2); 1006 qX,vwv. At 583 Zu has 'KEivwv, like PZmZv; at 612 ~7j is in Zu, not Zm; at 1012 read Zu for Zn. 9 There are a few conjunctive errors of ZmZu not shared by Zv: 1367 E~V; 1530 OTOTTOL; also the errors at 1060, 1259, and 1466 cited infra. MASTRONARDE, DONALD J., Notes on Some Manuscripts of Euripides' "Phoenissae" , Greek, Roman and Byzantine Studies, 26:1 (1985:Spring) p.99 102 SOME MANUSCRIPTS OF EURIPIDES' PHOENISSAE Zv 8C Zm - ZvPC: 608 J.t'YIKVlIaIS Zv 8C Zm, 966 av'Tov ZvlZm, 977 m; Zv l Zm, 1063 AdJo{30AOll Zv l Zm, and 5 other passages Zm 8C Zv 8C - ZmpcZvPC: 331 avjjgE a.c., ch"ryi'gE p.C., 449 7TOAtV om. a.c., s.l. add. Zm, in ras. Zv2~ 569 EiO'" a.c., Ei~ 0'" p.C., and 6 other passages Zm 1 Zv2 - Zm 2 Zv 1 : 1228 a7TEJ.t7ToAE'i'TE Zm 1 Zv 2, -aTE Zm 2 Zv 1 These instances present too complicated a pattern to allow the hypoth­ esis that one codex might have been copied from the other at a time when some, but not all, corrections had been made in the Vorlage. Zv is obviously not a copy of Zm since Zv1 has written horizontal strokes (copied from strokes written from one end of an erasure to another in its Vorlage) in the line at 445, 959, and 1583, where Zm has no era­ sure, no line, and no empty space; and in some passages Zv ante rasuram clearly had a longer text than what we find in Zv post rasu­ ram = Zm (e.g., Zv 8C had a longer exclamation in 1530). The pas­ sages in which Zv is correct and Zm in error are not very significant (e.g., Zm has 1060 "tEviI/M(J', 1259 ~, 14667Tp0J.tv(J-, 1698 &~v, 1724 E~ a.c.), but it is even harder to find significant uncorrected errors in Zv where Zm is correct (1065 E7TEO"O"V'TO Zv hardly counts). Never­ theless, at 1284 Zm has / __ ai ai, reflecting an erasure of part of ai ai ai ai in its Vorlage, which here cannot be Z v, which has simply / ai ai (c! 1590: Zm has 'TEtPEO"ta~ ** OV J.tiJ7TO'TE, Zv 'TEl.PEO"I.aS ov J.tiJ- 7TOTE).lO Zm may be later than Zv, and it is impossible to prove that it does not descend from Zv via a corrected/corrupted intermediary. But for all practical purposes, an editor may assume that ZmZv are gemelli and be content with citing Zm alone (the more complete witness), if a generous citation of Thomano-Triklinian manuscripts is desired. 2. A new member of the family Ab R Mn S Vr W Copenhagen, Gr. 417, known under the symbol Hn (Hauniensis), was first used by A. Matthiae in 1814, and its readings were made known to the scholarly world in Matthiae's volumes of critical notes to his edition published from 1821 on. Turyn regarded the Phoen. and Hipp. portions of H n has an apograph of V r (Vatican, Pal. gr. 343), made when V r was still complete (V r has lost the argumenta to Phoen. and most of Hipp.).ll Diggle has recently shown that for Hipp. 10 At 938 Zv has !3po-nwlJ, Zm has either !3po TEWIJ (space) or !3po*nwlJ (erasure). At 610 Zv has ~, Zm has JE. 11 Turyn (supra n.2) 329-33. See also K. Matthiessen, Studien zur Textiiberlie/erung der Hekabe des Euripides (=Bibliothek der kl. Altertumswissenschaften N.F. 11.52 [Heidelberg 1974]) 42 (where first use of the manuscript is wrongly ascribed to Kirch­ hom. On Vr see Text. Trad. 14 and (for its relation to the Aldine edition) 20. MASTRONARDE, DONALD J., Notes on Some Manuscripts of Euripides' "Phoenissae" , Greek, Roman and Byzantine Studies, 26:1 (1985:Spring) p.99 DONALD 1. MASTRONARDE 103 Ho is a gemellus, not a copy of Vr.12 I shall now show that Hn and Vr are gemelli for Phoen. as well, and that Hn is accordingly an addi­ tional member of the family Ab R Mn S V r W, which will appear in my forthcoming Teubner edition as e (in Text. Trad. the siglum P2 was used). Hn is dated by Turyn to "around 1475," a date that already makes it unlikely that it is a copy of a manuscript written ca 1500. It is written on western paper, with one column of 25 lines on each page. There are very few corrections and no glosses or scholia on the same pages as the text. Folios 91L 92 r contain various items of prefatory material of Phoen. (arg. 1-5, 12, and 9: see Text. Trad. 78-88); 92v-124v contain the text of the entire play; 125 r-126 r contain fur­ ther argumenta (13, 10, 14-17) followed by old scholia (which con­ tinue to f.139 v). Though divided into two sections in Hn, the prefa­ tory items are the same as those found in the subfamily Mn S, and their order is identical to that in S (where, after argo 9 [= dramatis personae], the text is begun; but then abandoned so that the remain­ ing items can be added). The vast majority of Ho's errors are conjunctive with e or with several codices in the family and not just with Vr. For example, if we confine ourselves to omissions and additions, we may cite: 5 Y7JV (Jewv MnSHn [Vr]; 20 aOI) om. AbMnSHn [Vr]; 198 (Jr(AEtWV YEVOI) [cpv MnSVrHn; 408 8' om. B, 9Hn; 413 'TT,V om. 9Hn; 478 a~(JLI) om. FSa, MnSVrHn: 778 8' om. RVrHn; 1196 o~v om. MnSVrHn; 1277 8e om. eHn; 1317-1318 -aTEAAWV add., YEpwvet ypaLav om. MnSVrHn; 1500 alterum OJ om. RMnSVrHn; 1622 y' om. RSVrHn+ [Mnl; 1626 a' om. SVrHn+ [Mnl; 1706 1TOV om. Mn S Vr Hn. Conjunctive errors uniting Ho and Vr alone (or alone among e) are very few, but one is striking: versus 74, 73, 75 hoc ordine, sed numeris {3, a, y adscriptis, habent Vr Hn; 407 8vvap.,' av VrHn, CRw; 460 acpW VrHn (acPW P); 556 8' om. VrHn, L. Fur­ thermore, the colon-divisions of Hn are identical to those of Vr (which are often like those of other members of 9, which are not uniform among themselves). In particular we may cite 1060 JJ-/8e, 1061 eV'Te-/ KVOt, 1062 8paKOV-hol) in Vr Hn only; also 209 1Teptp-/ pv­ 'TWV, 673 gvv7JtjJe/ in Ab Vr Hn only; 1510 1Tpo1TapOt-/(J' in R Vr Hn only; 249 /-lOt! and 686 1TaV'TWv/ avaaaa in MnSVrHn only. Nevertheless, there is abundant evidence that Hn is not a copy of Vr, nor Vr a copy of Hn. Errors in Vr not found in Hn include 156 1TOV 1TOV Vrpe (ZbZrnZu); 199 ei 9 (Tiv SpeHn); 253 oiae'Tat 12 J. Diggle, "Five Late Manuscripts of Euripides, Hippolytus," CQ 33 (1983) 34-43. MASTRONARDE, DONALD J., Notes on Some Manuscripts of Euripides' "Phoenissae" , Greek, Roman and Byzantine Studies, 26:1 (1985:Spring) p.99 104 SOME MANUSCRIPTS OF EURIPIDES' PHOENISSAE AbMnSVr+; 285 1TOAEL RVr; 399 i}BEl.aV 8EOV Vr (Mn'YPS'YPSa'YP); 432 (rrpaTEVOV(J'(. Vr; 444 ilKEL Vr, AaPRfRw (yp-variant in rest of @); 464 1.J:·"'OO~c; AbMn Vr+; 512 av om. Vr; 979 uE om. Vr; 1033 versum om. S Vr; 1185 Kat om. Vr.l3 Errors in Hn not found in Vr include 88 8aAAoc; Hn +; 99 TOl,uB' Eyxpi1TTETaL Hn +; 114 xaAKoBET' E/-'f30Aa OHn; 127 at at MnSHn+; 234 EAl.uUWV Hn+; 246 cpoi­ VLuua xwpa MnSHn+; 262 e1JETEiaC; Hn (EVE1TEiaC; fere MnS); 267 T4J~~ TwBE XEI,pa (transp.) Hn; 428 EUT' EI-'Oi MnS Hn; 432 E1Ti yav MnSHn; 439 TlX ante n~TaTa add. Hn; 452 T7JV xapLv EXEL Hn; 478 WUT' aVT71c; Hn; 479 Of30v Hn; 502 TWV B' EPYwv Mn S Hn; 1185 KWAEC; EC; KVKAW/-" MnHn; 1223 1TpOV'TT'ijpgE RSHn, VRfRv. In sum, the readings of Hn are closest to those of the subfamily MnSVr within @, and in several passages Hn maintains the error of MnS that the tradition of Vr seems to have removed by collation with another tradition. In addition, Hn has some curiosities of its own (262, 267, 478, 479). Whereas Vr has some interest for its rela­ tion to the Aldine edition (see supra n.l 1), Hn makes no further useful contribution to our knowledge of @ and may safely be dis­ pensed with even in a generous apparatus criticus. For the prefatory material it may be cited as an additional witness for the items peculiar to (or almost peculiar to) Mn S. 3. T and its descendants One of the few readings of any substance in which descendants of T differ from T itself in Phoen. is 659 valJ-aTa T', where the transmitted text valJ-aT' was written in T by Triklinios himself. Au­ topsy confirms that T contains no correction or supralinear addi­ tion, only the gloss T1YOVV T7JV BipK'YIv, which also appears in the copies.14 valJ-aTa T', if not an unbelievably lucky accident, must be viewed as an emendation that creates responsion between 640 ~uxoc; a&iIJ-aUTOv 1TEUTJIJ-a (- ~ ~ ~ - - ~ - -, or, as Triklinios de­ scribes it, hypercatalectic ionic a maiore dimeter composed of first paeon + third epitrite + syllable) and 659 valJ-aTa T' EvvBpa Kat pEE8pa (- ~ ~ ~ ~ - ~ - ~ ) .15 As such, the emendation ought to be Triklinios', and his failure to write valJ-aTlI T' in T must be due to an oversight Gust as in 250 Triklinios himself wrote 1TOMV even though he analyzed 239=250 as a lecythion, which required 1TTOALV). It is 13 This list is based on collation of over 500 lines of Hn and selective checking of omissions in the remainder of the play. 14 I examined T twice during a visit to Rome in October 1984. 15 In modern editions, a&i~.utTov in 640 (Elmsley on Soph. OT 196) convincingly solves the same problem. MASTRONARDE, DONALD J., Notes on Some Manuscripts of Euripides' "Phoenissae" , Greek, Roman and Byzantine Studies, 26:1 (1985:Spring) p.99 DONALD 1. MASTRONARDE 105 hard to believe that the scribe of, say, Ta carefully compared Trikli­ nios' metrical scholia with the text and made adjustments: no such adjustment was made at 250, and Triklinios' intentions were frus­ trated by an unmetrical aA.A.ouTL in the trochaic tetrameter 1338 (aA.­ A.ov; T) and by unmetrical xeLpa in 1711 (opeye XEpa c/JiA.alJ T = iambic penthemimer in Triklinios' analysis). The metrical emendation lJa,uaTa T' is also found in four other manuscripts containing the Triklinian Phoenissae: London, Arundel 522, Vat.gr. 2241, 897, Pal.gr. 223.16 Of these, Turyn argued that Vat. gr. 2241 was copied from T directly rather than via Ta, but sug- gested that the others descended via TaP Turyn's hypothesis would compound the problem of explaining lJa,uaTa T', since we would have to posit two scribes reading the scholia with care and arriving at the same solution. Since this is so improbable, we must ask instead whether there was not an intermediary between T and Ta that Tri­ klinios himself had revised in minor details. This intermediary would have served as a conduit to Ta and to Vat. gr. 2241, and even manuscripts identified as copies of Ta may have descended from it rather than from Ta, since there are a few other agreements of these descendants with T rather than with Ta (see list below: 23, 228, 725, 1155). T is a hybrid production, contain­ ing pages of different appearance and sometimes cramped scholia. It would not be surprising if, after completion of his work on the Eurip­ idean triad, Triklinios had a more presentable copy made by another scribe and then acted as diorthotes, making a very few alterations or additions. I list here the most significant cases where it might be appropriate to posit such alterations or unnoticed errors: 23 qxh.."v TZ, Arundel 522, Vat. 897, Vat. 2241 ante corr.: cpaTLV Ta, Vat. 2241 post corr., Pal. 223 228 /3aKXELWv Tt, Arundel 522, Pal. 223, Vat. 897: /3aKXEiwv Ta: /3aKXEWJv Vat. 2241 475 EcJ>8E.~aT' Tz: EcJ>8E.y~aT' Ta, rell. 659 valJ,aT' P: valJ,aTa T' Ta, rell. 725 UcJ>aAEIS P, Vat. 2241 s.l., Vat. 897: UcJ>aAE/s TzTa, Arundel 522: ucJ>a­ Aill) voluit Ta S (T/ s.l. scr.), Vat. 897 s.l., Vat. 2241 (ijl) p.c.; ij a.c.): ucJ>a­ AijEtl) (sic) Pal. 223. 759 EyXE.yyvOV TZ: EXE.yytJOV Ta, rell. 1155 KaTafTKatjlwv TZ, Vat. 897, Taac: -uKat/Jw Tape, Vat. 2241, Pal. 223 1388 aAAotl) Tt: aAAotUt Ta, Vat. 2241, Pal. 223: aAAouTtv Vat. 897 16 The other Triklinian manuscripts of Phoenissae, which I have not seen, are Milan, Ambros. A 104 sup. and A 115 sup.; Paris 2812; Salamanca 243. 17 Turyn (supra n.2) 194-202. MASTRONARDE, DONALD J., Notes on Some Manuscripts of Euripides' "Phoenissae" , Greek, Roman and Byzantine Studies, 26:1 (1985:Spring) p.99 106 SOME MANUSCRIPTS OF EURIPIDES' PHOENISSAE 1364 E~ TI: el~ Ta, rell. 1403 a7TEO"TEprIlUVOI..V TPC: a7TEO"TEP'TIJJivoLv Ta, rell. 1473 #-/..vptwv TZ, in linea Ta, rell.: #-/..vpwv s.1. Ta, rell. 1711 XEpa T': XE'ipa Ta, rell. I take this opportunity to report a few further details that are vis­ ible in T in the original but not legible on microfilm (and so not contained in the collation of Text. Trad. ). T Z is the other, original scribe of the iambic portions of the play; Tt is Triklinios himself, writing the lyric portions or correcting the iambic portions; T is used when (on pages written originally by Tz) it is impossible to determine which hand made a change. 10 KAi1'O~/' T in rasura; 98 EV(JEV& TZ, EV(JEVS' T'; 310 ~AL~ PI; 378 TATII'.wVO~ TIs (coni. Markland); 407 EO"TL TZ, EO"Ttv Tt (so too a final nu is added by Tt at 542 S..wPtO"E, 733 &.P~o"l.., 905 'o"TL, 950 o#-/..~o"t, 1115 OW ~O"t, 1166 aiTWAUTt, 1174 EKo#-/..7TaO"E, 1194 a{oo"L, 1650 KVO"i); 424 o~v EliTVXE'i~ transp. Tzac, corr. TZ; 438 o~v om. TZac, corr. TZ; 472 ailTC;J Tac, aVT4' TPC; 504 UVaTOA.a~ Tzac, UVT- T post rasuram; 522 7Ti#-/..7TAaO"(J' Tz ante rasuram; 548 r4'S' a7TOvE#LEI..V P'YP; 559 AO'YO~ TS; 591 (J"I(T,7TTPOV Tt; 596 OV P'YP; 603 ¢'rIJ,£i TZ, eJr/,#-/..' P; 606 f>W~(J' P in rasura; 623 TEKV' T'PC; 720 EO"W TZ, EWW Tt; 724 7TpO{3&.Aot#LEV Tz, 7TPOO"{3&.Aot#-/..' av Tt; 734 7TOAE~f.,(J"t &J TZ, -w/,~ f>WO"w Tt; 739 ,,",Aat~ TZ, ,,",Aa~ P; 748 EA(JWV ... 7TOMV TZ, (EA(J. om) ... 7TOA/,V #LOAWV Tt; 755 EA(JE'iV TZ, EAE'iV P; 792 aU' &.P~o"L in rasura Tt; 851 Ka7To~ TZ, Ka7To~ Tt; 865 KA'TItua~ TZ, sed s.l. (J"Vv~'TI(J"L~ ser. Tt (idem s.1. ad 945 T,t'(JEO~, 1041 7TOAE~); 881 7TEpi Tt'YP et VEKPO~ Tts; 884 7TOAEL TPC, 7TOA/'~ Tzac; 902 #-/..' aUo &'i Tt'YP; 924 -7Ti7TTEL~ TZ, -7T/,TvE'i~ Tt; 931 (JaA&#LOI..~ TZ, -~I..~ Tt; 933 cpOvwv TZ, ~ivwv Tt; 939 TJ"uv TZ, v"uv Tt; 964 7Tpg6T,vat TZ, -(Javal.. Tt; 980 SEA~~ usque ad 983 ST,T' rescr. in ras. P; 982 6E(J"7TPWTWV Ttac, -TOV Ttpc; 983 ;'Yv~ fort. omiserat TZ; 983 ;pv~ Tt; 985 7TapE{w Tt'YP; 986 vvv TZ, vvv Tt; 1084 AEl}O"O"EI.. (non AEVO"EL) TUv; 1100 -AEi7T- T&eTs, -M7T- TPC; 1120 a(J"7Ti~ fort. TZ, TIS, U(J"7TiSL Ttuv; 1124 ap'TI TZ, ap'TIv Tt; 1140 7TPO(J"CpEPOVTL Tt in rasura; 1188 Ka6T,O"Ev Tts; 1218 #-/"'TIvvO"at T in rasura; 1352 TOV {3iov Tt in rasura; 1404 cip7TaO"avTE~ Tac, -aVTE TIIC; 1425 OWi7TO~ oO"ov T in rasura; 1428 EAI..7TET'TIV Tzac, -AEl..7T- TIIC; 1431 TETpWJJiVW Tzac, -JJivo~ TIIC; 1464 7TEAOt TZT'S, 7TEAEI.. Tt; 1506 {vVETO~ pS; 1603 a(JAwv TZ, a6ALaV Tt; 1689 WAf3W'EV EV TZ ante rasuram; 1701 (aOAI..' om.) a(JAwv 7TaTpo~ TEKva fort. etiam TZ ante rasuram; 1707 &lJ~(J' Tal:, &J~ 6' TIIC. 4. Notes on other manuscripts V (Vat.gr. 909) In the collation of Text. Trad. I was unwilling in many passages to assert on the basis of microfilm whether the correction was made by the original scribe V· or by the later scribe V2 (whom Murray re- MASTRONARDE, DONALD J., Notes on Some Manuscripts of Euripides' "Phoenissae" , Greek, Roman and Byzantine Studies, 26:1 (1985:Spring) p.99 DONALDJ.MASTRONARDE 107 ferred to as v, Wecklein as b). The distinction is an important one in the case of this manuscript, because VI (whatever the date of V) seems to me to be a relatively accurate transmitter of a Vorlage which probably antedates 1200, whereas V2 seems to reflect Palaeo­ logan scholarship to the extent that he draws readings from other sources, some of which may be the product of contemporary con­ jecture.IS I was able to examine the original manuscript in October 1984 and would now like to report the results in as brief a fashion as possible. I record first a number of items by line number only. By referring to the published collation, the reader will find a single entry under each number that involves vac and VPC (or only one of the two when it is implied that the other agrees with the lemma). In the following lines Vac is in fact VI and VPC is V2: 98, 223, 224, 244, 276 (?), 277, 412, 451, 463 (?), 578, 596 (ov), 632, 658, 702, 763, 844, 902, 1246, 1415, 1632, 1643. In the following lines VPC is VIPC: 90, 169, 190, 327,417, 629, 1018, 1460, 1490, 1530, 1687, 1689, 1721. At 738 and 1095 corrections were made by a hand that Wecklein's collator called manus recentissima: this hand uses a thin stroke like V2'S, but a darker black ink. This hand I shall henceforth term V3; V3'S correc­ tions are earlier than the copy of V made in the fourteenth century, Vat. Pal. gr. 98 (Va).19 In the following lines the corrector was either V2 or V3: 489, 571, 606, 713. The rubricator (vr) is responsible for VPC at 101, 618, 687. Some addenda (marked +) and corrigenda to the published colla­ tion based on autopsy of V may be recorded here:20 35 T' VI in rasura; 82 Bovpo~ (not BEpO~) VI; 103f delete entry;21 145 TaU in rasura VI; 209 1TEptPP1JTC!!lJ VI (i.e., the spacing of the letters and the appear­ ance of the ink suggest to me that V I wrote the whole word thus all at once and did not first write 1TEptPPVTqJ and then change to 1TEpptPVTWlJ by adding a nu [without deleting the subscript]); +245 E1TT- VI, E1TT- V2; +2711Tou~ VI, 1TO~ V2; 295 c;, (with subscript) Vac, OJ VPC; 299 (misprint) V1TCxppo(Ja VI; 18 Text. Trad. 33, 113. It was the connection of V2 with Palaeologan scholarship that prompted the use of the derogatory lower-case italic siglum in Wecklein and Murray. 19 On Va see Turyn (supra n.2) 91f; Matthiessen (supra n.ll) 45f. 20 In several places a Jortasse can be removed (e.g. 74, 981, 1216). 21 The scribe wrote the text through line 102 (in darker, fatter script), then wrote the scholia (in lighter, thinner script) and, finding that the scholia did not fill the lower margin, then added 103f as the last lines of the page, but without changing the style of his writing. Hence, these two lines appear lighter, but were not really omitted; nor were the other lines rewritten, only written in a heavier style by the same scribe. The same occurred in the MS. Vr at lines 73-78, which were not omitted, but added when extra space was found available after the writing of the scholia. MASTRONARDE, DONALD J., Notes on Some Manuscripts of Euripides' "Phoenissae" , Greek, Roman and Byzantine Studies, 26:1 (1985:Spring) p.99 108 SOME MANUSCRIPTS OF EURIPIDES' PHOENISSAE 529 r,f.L11'- yI, r, Ef.L11'- y2; +539 aiEl, in rasura yl (£lEi yac?); +584 £l,."a8Lat y2s; +585 Tav8' yI, Tav8" yl; +625 OVKE8' in rasura y2 (fort. OVK yI); 683 ai vel at yI, a ypc (rasura facta), at yl'YP; 715 01'8' yI, oi 8' y2; 778 corr. y2s (not yls); 787 A.OTOV y2s (not yls); 795 'YEvva V., 'YEVV~ y2; 811 aA.A..cp yac:, corr. yls, aA.A.7] y2s; 870 corr. y2s (not yls); 885 TIS habet y; 927 delete entry KaTaKTE VOL~; + 1098 nullum punctum Y I, punctum post £lA.K7} y2; 1113 delete "(et fortasse ... y)"; +1114 EUT7]X' yl in rasura (EUTELX' a.c.); + 1299 aVTiK' (et aif.£-) yl, aVTix' (et aif.£-) ylpc; + 1391 i1T1To8pof.LOv ut vid. yI, Vrro8- y2;22 + 1396 Kav yI, corr. y2; 1496 Kpav8Ei~/u' yl (ut vid.), Kpav8Ew' yl, delete "(et voluit ... delevit)"; 1505 Ta~ yac, Tar; ylpc; 1515 Ti~ habet y, nulla correctione facta; + 1534 £lEPOUKOTOV yac, corr. ylpc; + 1564 Of.L~TWV y2s; 1575 f/>otVLaV y2s (not yls); 1649 ~v yI, Tiv y2; 1649 (add subscript to lemma 11) i; yI, ~ (sic) y2; 1687 8avwv y2s (not yls); 1697 uw,."a Yls (not y2s); +1725 0 TA.iX~ y2s. Sa (Vat.gr. 1345) and R (Vat.gr. 1135) Several points of uncertainty in these two interesting manuscripts have been cleared up by autopsy, and corrected readings will be apparent from the Teubner apparatus. Here I mention only that in Sa, line 1 does in fact contain 6J, written by the rubricator in an ink that is invisible on copies (likewise the heading lnro8. cpo"v. is present in argo 1: Text. Trad. 393), and ~(J'1Tap8EVOf) is the reading in 1023; and in R Aoxa),ov was omitted from l3lf by R 1, but added in the margin by R 2. Rv (Vat.gr. 1332) In Text. Trad. 9 I mention Livadaras' report of a date on f.18 v of this manuscript. When I inspected the manuscript itself, I found f.18 v to be blank. The alleged date may be somewhere else in this many­ page manuscript, but I did not find it in a brief perusal.23 Additional 'Moschopoulean' manuscripts Inspection of Vat. gr. 56, Vat.gr. 50, and Pal.gr. 42 confirms that all are faithful carriers of the x-tradition (cf. Text. Trad. 169). The latter two have the Moschopoulean version of the epitome (arg. 1), but 22 A. Tuilier meant to record Vac, not Lac, in his Etude comparee du texte et des scho­ lies d'Euripide (= Etudes et commentaires 77 [Paris 1972]) 86. What Tuiler reports as a sigma above the line written by Vl, implying (he thinks) a variant im08poJ.LO(j, is in fact a cursive alpha, which, together with the beta over rxvo~ and the gamma over KW)o.OV, is a guide to the prosaic order of the words. Similar schoolmasterly guidance is given in the same way in other lines. 23 I was able to clarify many obscure readings of this damaged manuscript, but the points are all so minor that I refrain from reporting any here. The collation in Text. Trod. will not mislead. Likewise, inspection of Zb (Vat.gr. 50 revealed that the cor­ rectors are even more numerous than I supposed on the basis of microfilm, but it is not worthwhile to publish here the details of this relatively unimportant witness. MASTRONARDE, DONALD J., Notes on Some Manuscripts of Euripides' "Phoenissae" , Greek, Roman and Byzantine Studies, 26:1 (1985:Spring) p.99 DONALDJ.MASTRONARDE 109 other prefatory items as well (arg. 4-6 plus the iambic trimeter epit­ ome from the tradition of Sophocles' Oedipus Tyrannus are found in . Pal. gr. 42; argo 2-6 in Vat. gr. 50) .24 THE UNIVERSITY OF CALIFORNIA, BERKELEY F"bruary. 1 ()95 24 I take this opportunity to correct a few oversights and errors in Text. Trad. 11 (and 186): Z contains Phoen. 1-333 and 377-1766, not the whole play. 18 and 425 Oine 1175): P. Here. 1609 I (127, p.48 in Gomperz, Herku/anisehe Studien II: Phi/odem Uber die Frommigkeit [Leipzig 1866]) should be added as carrying a testimonium to line 1175; ef Philippson, Hermes 55 (1920) 259f. 38f (last line and first line): read P2 = AbMnRSVr and P3 = AaPRfRvRw. Collation, line 229: add S as a witness of Ka(J­ TjJ,LEPW/I; line 297f: the space in the papyrus is probably ea 18-20 letters (not ea 15); line 444: add S'YP as a witness of T/KEt.