'E>'7JOOTWC] 1TEpt>'E>'7JOOTWC Re Such readings and the brief selection of scholia render this MS. of no value. Codices Oi Mf Se are shown to derive from a lost hyparchetype since all three MSS. contain the same scholia on Or. 1, and Oi Mf the same scholia on Orr. 3 and 4 as well as conjunctive and separative errors: 31.1-8 8ELvwe praehh. post 31.14 XP7JJLaTWV Oi Mf Se 33.5 lCT£ov-19 1TE1ToL~KcxeLv om. Oi MfSe 32.22 EV T. 8 ... . o{3o>.ove Oi Mf EV T. 8. transp. post o{3o>.ove Se 47.28 oi-48.2 avcxxwp~CWCLV (OJL.) Oi Se, om. Mf .28 Kcxt-48.1 OVTWC Se, om. Oi Scholia contained in these MSS. constitute a selection from Scholia Ulpiani and are found on folios following the text of Orr. 1, 3 and 4. Other orations in Oi Mf contain scant excerpts from scholia, which do not merit consideration (Or. 10) or scholia from Y (Or. 11). Oi has scholia from Y on Orr. 15 and 22; 13, 14, 16 and 17 lack scholia (Mf lacks these orations). Conjunctive and separative errors indicate that Vd and Ah derive scholia on Or. 1 from a common exemplar: 55.4 ciJc om. Vd Ah .7 ECTtV om. VdAh 57.1 JL~ om. VdAh 58.8 < KCXt TOVTO> T0 V d Ah 31.10 To-32.3 1TE1TELCJLCXL Vd, om. Ah 57.22 xpfje8cxL-58.2 Ah, om. V d Scholia on other orations appear in both MSS., but Vd derives these scholia from Vb (see n.7 and section III below) and Ah from Parisinus gr. 2940, T. Codex Ah is dated to the year 1525 and consists of selected scholia copied by Lazarus Bonamicus,12 who doubtless composed several scholia not known from other sources (e.g. 1.9 vvv ~, ''']''~' ~ '8 I " \ I ~,~ \ I OE KCXLPOC 7JKEL opcx OE 1TWC cx VJLLCXC CXVTOVC 1T1I7Jpwccxe oW TOV KCXTCXIIOYOV 12 See M. E. Cosenza, Biographical and Bibliographical Dictionary of the Italian Humanists . .. I (Boston 1962) 644-45. 60 MSS. OF SCHOLIA ULPIANION DEM. 1, 3 AND 4 ~ > \ \' "\ () , ~ \ ~ > ~., '\ 'TWV a7TO/\W/\O'TWV XWPLWV, 7Ta/\tV 7Tapa/LV 7]ca'TO OLa 'TOV EL7TE£V on 7Ta/\LV fI \, "') 7]KEL KaLpoc ov XELPWV EKELVWV • In sum, Oi Me Se and V d Ah contain a briefer selection from Seholia Ulpiani than Ca Fu MID., which in turn lack many scholia found in T Be. One notable feature of these eight MSS. is that they do have from codex F readings which preserve the truth when T Be are in error. 13 In addition, the samples of text collated for this article have brought to light accepted readings found neither in T Be nor in F: 56.12 a7To 2 Ca Fu Mm. Oi MeSe Vd Ah t17TO T Be 113.20 on 7TAE{OVC Eld Ca Fu MID. on Eld 7TAE{OVC T Be, LJ am. Oi Me 141.12 'TOV &'\'\ov Xp6vov Ca Fu MID. 'TWV &,\,\wv Xp6vwv (-vov Tao) TPO Be, E om. Oi Me .15 ayopEvELV Ca Fu MID. ayopEvEL T ayopEvwV Be, E am. Oi Me These involve minor corrections of the text of T Be and could be interpreted merely as felicitous conjectures. However obvious this conclusion may be, it is based on a collation of a portion of scholia on Orr. 1, 3 and 4 (see supra n.6), and further collation might well produce readings which derive from a source of equal or greater value than T Be. Moreover, since Ca (saee. 14) is nearly coeval with T Be (saee. 13) and since Ca contains a more complete text ofscholia than its cognate MSS., it seems prudent to collate all scholia in Ca and weigh the res1.llts. Meanwhile codices Fu Mm. Oi Me Se V d Ah can be dismissed from further consideration. II Conjunctive errors indicate that the following MSS. derive scholia on the Olynthiacs and first Philippic from a common exemplar, lost codex W:14 13 Note, for example the following readings cited above (pp. 57,59) 56.11,58.13,64.16, 142.13. 14 33.16 fLOVOV TBe CaMVb fLa).).ov VfWd Vs, EP om. Pv Ao 56.10 t/hA[TTTTOV F B T Be Ca MVb rpL>.tTT7TOV TTpaYfLaTa VfWd Vs Pv Ao 58.24 EXELV F B T Be Ca Vb EX€LV TOV >.oyov VfWd Pv Ao, ap. Vs E desunt post 58.5, Eom. M .25 cVfLq,£POVTL 7TpOC~I(EL F B T Be Ca Vb TTpOC~I(OVTL cVfLq,£P€L VfWd Pv Ao, E om. M MERVIN R. DILTS 61 Vf,15 Rome, Biblioteca Apostolica Vaticana, gr. 76, saec. 14 Wd, Vienna, Oesterreichische Nationalbibliothek, phil.gr. 105, saec. 14 Vs, Rome, Biblioteca Apostolica Vaticana, gr. 2207, saec. 14/15 Pv, Paris, Bibliotheque Nationale, gr. 2999, saec. 15 Ao, Milan, Biblioteca Ambrosiana, Q 12 sup., saec. 15 Vi/6 Rome, Biblioteca Apostolica Vaticana, gr. 1367, saec. 15. Apographs of codex w contain scholia not found in T Be: 29.5 ~lhKOv-ll 8€pa7T€V€' F B VfWd Vs Pv Ao, if. 3l.1-7 O€pa7T€VH T Be 156.19-21 F B W d V s Pv Ao, E om. Vi T Be .28-157.2 FBWdPvAo, Eom. ViTBc 143.26-144.2 P Wd Vs Pv Ao Vi, E om. F B T Be 144.7-9 PWdVsPvAoVi, Eom. FBTBc .10-12 P W d V s Pv Ao Vi, E om. F B T Be Like MSS. represented by Ca, readings and scholia in apographs of w point to contamination from F (B). In addition w contained scholia found neither in F (B) nor in T Be (e.g. the last three scholia listed above). These were first edited byMorelli in 1570, and he probably derived them from PV.17 Three MSS. (V s Pv Ao) derive still more scholia from recensions independent of T Be. Vs contains scholia which doubtless derive from codex A, Monacensis gr. 485, saec. 10: 156.8-10 AHVs, om. WdPvAoVi .11-14 AHR Vs, om. WdPvAoVi .21-24 A Vs, om. Wd Pv Ao Vi Similarly Pv Ao contain scholia found in the recenSIOn of F (B), which are lacking in TBcVfWdVs: 4l.11-12, 44.17-24, 26-45.5; 113.11 'lTEp1 'lTOt6T7JTOC 'lTpoccfmwv A H R Wd Vs(?) Pv Ao 'lTEp1 'lTOt6T7JTa 'lTpoccfmov T Be Ca Vb, E om. Af 142.2 TpufKOVTa 8150 F B T Be Ca Af Vb Ag' W d Pv Ao TpufKovTa T6 'g Vi, ~ om. V 5 15 Vf contains scholia only for Or. 1, ff. 5v-I3v • These folios are part of the first of two recensions of Demosthenic scholia in this codex; see Dilts, opp.citt. (supra n.4) 37-39 and (supra n.9) 207. 16 Vi contains scholia from this recension only for Or. 4; Orr. 1 and 3 have scholia from Be. 17 112.11 8ta TWII £PywII odKvva Wd Vs ap.a yap OELVOII xpijcOat 'lTp&.yp.aa TOVTOV OEtKVVCL Pv. Morelli attributes this reading to codex 71, the seventh of the MSS. he used (a-B); see further J. T. Voemel, ed. Demosthenis contiones quae circumferuntur (Halle 1857) 183. Scholia first edited by Morelli also appear in Af and Vb (see section III). 62 MSS. OF SCHOLIA ULPIANION DEM. 1, 3 AND 4 135.4-10, 15-18; 138.5-6, 8-11; 144.3-5, 14-18; 147.8-18.18 These scholia appear to derive from a comparison of F and B, since Pv Ao sometimes agree with F but not B (e.g. 134.16 K. T. C. B T. C. K. F Pv Ao) and viee versa (e.g. 154.13-17 B Pv Ao, om. F), and at other times Pv Ao combine readings in F and B: 136.11 wc €:rravop(}WTtKOV F lrravop(}WTtKWC B we ETTavop(}wTLKwe PvAo 137.5 TTAEtova EV F TTAELOV Ta B TTAELOV Ta EV Pv (TTAEtW Ao) Ao The following separative errors indicate that codex Vf is an inde- pendent copy of lost codex w: 32.12 Ei1TELV Wd Vs Pv Ao Ei1TELV Sta Vf 54.26 yap WdVsPvAo yap 0 P~TWP vr 56.4 jJ-EAAEt Wd Vs Pv Ao TO jJ-EAAOV vr Three readings for scholia on Or. 1 aver a common source for Wd and Vs: 32.26 0 1T. 0 1Tp. CPo T Be 0 CPo 0 1Tp. 1T. Wd Vs 0 1Tp. CPo 1T. Vf (32.24-33.19 om. Pv Ao) 55 3 ~ , T B Wd V ~, ~ \' \ \ ~ • OEtKVVet C S oEtKVVct jJ-tjJ-HTat yap aEt T7JV TWV 1Tpay- jJ-aTwv ¢>VctV VfPv Ao 57.3 1Tap' aAAOV Kat avaYKa'OjJ-EVOC F B Kat avay. 1Tap' aAAOV T Be Wd Vs Kat avay. {m' aAAWV VfPv Ao and the following separative errors indicate that each of these MSS. is independent of the other: E post 58.5 TOK
Vb 142.5 cay. 'TW(Xc MT Be Ca Wd nvac caVa Vb Thus far we have seen that MVb are independent witnesses, and at this point it is appropriate to determine whether or not these MSS. are of value in establishing the text of Scholia Ulpiani. While MVb lack some scholia contained in T Be, these two MSS., like Ca Wd, sometimes preserve correct readings when T Be are in error: 31.7 a. av'T. aV'Totc €V(}vc Y MVb a. av'T. €V(}vc aV'Totc Wd v~op JLofiv'Ta T Be v~oPJLwv'Ta Ca 56.12 a7T()2 MVb Ca Wd tJ7TO T Be 56.12 orov-57.! 'TO LOV'TW V om. FBHRS 66.9 ~AAOLC hab. F B MVb Wd, om. T Be Ca 113.2 JL€'TaC'T~CaL A Vb JL€'TaC'T~VaL T Be Ca Wd, E om. M 118.18 7TpO(}VJLt~ MVb f107](}€t~ T Be C7TOVa€t~ Wd (ut apparet), E om. Ca In three instances (31.7, 66.9, 113.2) MVb or Vb preserves correct readings variously found in F Y A. Since these readings do occur in independent recensions, they do not alone justify full collation, but in other cases MVb preserve the truth for Scholia Ulpiani not found MERVIN R. DILTS 65 in independent recensions. At 56.12 MVb have the correct reading along with Ca Wd. This could derive from independent conjecture in all four MSS. or from a double reading in the archetype. The case for archetypal doublets becomes more cogent in light of 118.18, where MVb alone preserve the accepted reading and T Be Wd variants. In addition, the following variants are most readily explained as double readings in the archetype: 66.15-16 7TfXV'TfXxo8EV • •• 'T. CPo Ca M 'T. CPo 7TfXV'TfXxo8EIJ T Be (l7To8. 7T(xv'TfXxo8EIJ 'T. CPo Wd Vb 139.1 ZYj'TOVCL EL Wd ZYj'TOVCL yap T ZYj'TOvcl TtIJEC Ca ZYj'TOVCL 8l TtVEC OTt Be, E om. MVb 141.23 OV'TW Il-EIJ OV'TOC T Be OV'TOC Il-EV OV'TW F B W d OVIJ Il-EIJ OV'TW CaM 142.2 7ToAAfj F B T Be Vb 7TOAV Ca Wd M MVb date at least to the beginning of the fourteenth century,20 and they are therefore, after T Be, the oldest witnesses to the text of the archetype. This fact, as well as the correct readings found in M or Vb or both, justify collation of these MSS. * * * * * Contrary to the closed tradition established for Scholia Ulpiani on Demosthenis in Timocratem (Or. 24),21 MSS. of Scholia Ulpiani on the Olynthiacs and first Philippic contain readings and scholia from inde pendent recensions. Thus of the eleven MSS. containing Scholia Ulpiani on Or. 24, none contains a text contaminated from indepen dent recensions, but an average of 50.6 per cent of the MSS. containing Scholia Ulpiani on Orr. 1, 3 and 4 show signs of contamination. 22 This is all the more remarkable since in both Orr. 24 and 1, 3 and 4 Scholia Ulpiani are often cognate with independent recensions (e.g. A Y F B). Clearly the MSS. considered here reflect a stage in the text tradition of Demosthenic scholia in which interest in comparing Vorlagen for scholia on orations at the beginning of the corpus was greater than for subsequent orations. This is borne out by the fact 20 On Vb, see]. Irigoin, "Les filigranes de Fabriani ... ," Scriptorium 12 (1958) 46,47. 21 See Dilts, op.cit. (supra n.4) 35-50. 22 For Or. I, 48 % (IS of 31 MSS.) are part of the contaminated tradition; for Or. 3, 50 % (12 of24 MSS.); for Or. 4, 54 % (13 of24 MSS.). 66 MSS. OF SCHOLIA ULPIANI ON DEM. 1, 3 AND 4 that contaminated MSS. contain Scholia Ulpiani only for the beginning of the corpus, and other orations (13, 14, 16, 17, 24) lack scholia or contain scholia from other MSS., most notably Y. The terminus ante quem for such philological activity is the end of the thirteenth or the beginning of the fourteenth century, since M is dated saee. 13/14, Vb has been dated to the beginning of the four teenth century, and the lost exemplars ofCa and Wd (both saee. 14) were doubtless as old. In other words, these MSS. date from a period coeval with or slightly later than T Bc. Fifteen of the seventeen MSS. discussed in this article have been shown to derive from two hyparchetypes, which can be cited adequately from Ca and Wd. MVb together with Ca Wd represent a branch of the MS. tradition which is independent ofT Bc, as can be seen from the following simplified stemma for Seholia Ulpiani on Orations 1-4: FBYAS , ,./'/ , df\ Ca Wd Af Vb /\ T Be Prior to Dindorf, editions of Seholia Ulpiani were based primarily on the Aldine edition of 1503 with accretions of readings from MSS. in Paris collected by Morelli (1570). Frequently the Aldine edition departs from the paradosis, and one of Dindorf's contributions was to eliminate some of these aberrations; but many remain in his text, regrettably without any account of the paradosis. Moreover Dindorf used an imperfect collation of only one primary MS. (T). 23 A new edition is clearly in order, and for Scholia Ulpiani on the first four orations of Demosthenes it should be based on Bc Ca Wd AfVb as well as T. Making use of these six primary MSS. will result in a text 23 He also used two apographs Paris. gr. 2944 (D) and Paris.gr. 2946 (C). Dindorf (viii) did recognize that D is a copy of T, but he reports to excess readings from C, which preserves an interpolated version of Sclwlia Ulpiani; see Dilts, op.cit. (supra n.4) 42-45. MERVIN R. DILTS 67 which conforms to modern critical standards and contains a full account of Scholia Ulpiani not found in T24 and variant readings. 25 NEW YORK UNIVERSITY October, 1978 24 These include several scholia in Bc, which have not been edited as well as those edited by Morelli (see supra nn.5 and 17). 25 A grant from the American Philosophical Society has facilitated purchase of micro film copies of Demosthenic MSS., and travel grants from the American Council of Learned Societies (1973 and 1977) and the Arts and Science Research Fund of New York Uni versity (1975) have made possible inspection ofMSS.