Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Four Manuscripts of Unattached Scholia on Oppian's Halieutica by Andreas Darmarios Leverenz, Lynn Greek, Roman and Byzantine Studies; Spring 1995; 36, 1; ProQuest pg. 101 Four Manuscripts of Unattached Scholia on Oppian's Halieutica by Andreas Darmarios Lynn Leverenz KANUSCRIPT TRADITION of scholia on appian's Halieutica, separate from the text of the poem, appears to have de- - vel oped during the second or early third quarter of the sixteenth century in Spain. 1 Andreas Darmarios, the disrepu­ table buttreeminent scribe and manuscript dealer of his day, . discovere such a Ms. late in the 1580s at Madrid and exploited the relatively novel nature of its content; he is responsible for six of the eight extant Mss. of unattached Halieutica scholia. After he had emigrated from Greece to settle in Venice as a young man during the 1560s, Darmarios also spent a good deal of time in Spain, particularly in the seventies and eighties, selling Mss. to and copying from the libraries of prominent Spanish humanists and collectors. In fact, all of the surviving copies of Darmarios' Oppianic scholia were written in either Madrid or Salamanca, derived from exemplars that have not survived. Of the four such Mss. to be considered here, the earliest copy, Salamanca 2730 (previously Palacio gr. 39), is dated in Dar­ marios' colophon to 24 July 1577.2 We may view this as the time of his introduction to this corpus of scholia. It was evidently a fruitful discovery, for Darmarios returned to the source at least three times to make virtually identical copies of the text. At the termination of his second oldest copy, Escorial gr. 569, dated in t For the fullest available exposition of scholia on Opp. Hal. see U. C. Busse­ maker, Scholia et paraphrases in Nicandrum et Oppianum (Paris 1849). 2 In translation the colophon reads, "The end has been reached with the help of God by Andreas Darmarios the Epidaurian on July 24, 1577 in Ma­ drid." For a full description of this codex see C. Graux, Notices sommaires des manuscrits Crees d'Espagne et de Portugal (Paris 1892). 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 102 DARMARIOS' SCHOLIA his colophon to 21 October 1578,3 Darmarios identifies the location of his source as the library of the Cardinal of Bourgos, Francisco de Mendoza. As I shall demonstrate, all four of Darmarios' Mss. discussed here are derived from the same source, and we may thus assume that the source of all four was housed in that library in Madrid. The four Mss., Salamanca 2730 (S), Escorial gr. 569 ( E), Brussels 85 (Be), and Beinecke 269 (B), each described in turn below, were written between 1577 and 1579. Brussels 85 and Beinecke 269 (formerly Saragossa Pilar 2097) were subsequently taken to Salamanca in 1580, where Darmarios augmented them with scholia from a second source. In these two codices converge the two lines of descent in this Ms. tradition of independent Hal. scholia, one from Madrid, the other from Salamanca. Darmarios' second source, as we shall demonstrate, was a lost sibling of Salamanca M 31, or z 1 (described below), a codex in which the scholia accompany the text of the Hal. 4 z 1 is closely allied not only to the second layer of scholia in Br and B, but also to Darmarios' two other copies of unattached scholia that contain the entire Z corpus (see below). These two Mss., Monacesis 134 (Mon) and Londin. Royal 16 D XII (R), together with the later of sections of Br and B constitute the Salamancan branch of the tradition, which is separate from that of the four Darmarios works originating in Madrid. The Salamancan branch of the scholia tradition, central to a future discussion, will be treated here only as it impinges on the Madrid Mss. One of the Salamanca group, M onacensis 134, bears mention at this point, as it represents a unique conflation of the two camps in this tradition. M onacensis 134 is a composite: its core layer, not writ­ ten by Darmarios, is the precise complement (down to the word in mid-entry) of the truncated assemblage of scholia in Matritensis 4715 (described below). The fragmented state of Matritensis 4715 (M) will prove influential in providing evi- 3 Darmarios' colophon in E reads very much like the one in S: -The end has been reached with the help of God by Andreas Darmari'ls the Epidaurian on October 21, in the year 1578 in Madrid, Spain." For a full description of Escorial gr. 569 see G. de Andres, Catalogo de los Codices Griegos de Ia Real Biblioteca de El Escorial III (Madrid 1967); also E. Miller, Catalogue des manuscrits grecs de Ia bibliotheque de l'Escurial (Amsterdam 1966). 4 For a description of z1 see A. Tovar, Catalogus codicum Graecorum Uni­ 'lJersitatis Salamantinae I (Salamanca 1963); for a description of the M 55. com­ prising the voluminous and convoluted Hal. tradition see D. Robin, -The Manuscript Tradition of Oppian's Halieutica," BollClass SER. 3 2 (1981). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LYNN LEVERENZ 103 dence for that Ms.'s relationship to the Madrid group of Damar­ ios' Mss. M and its complement, the anterior sections of M ona­ censis 134, are the only non-Darmarios copies of unaffixed Hal. scholia we possess. The importance of M is twofold: it is the earliest witness in this tradition of unattached scholia and the closest extant relative to Darmarios' Madrid quartet. Rather than bearing a close relationship to z 1 like the Salamanca branch of the tradition, M and the Darmarios group are by comparison only distantly connected to it. z 1 bears the same uncular relationship to M as M does to the Darmarios foursome, i.e., z1 was a sibling of M's (lost) exemplar, as M was a sibling of the (lost) exemplar for the Darmarios group. Each of the two branches of Mss.-those closely affiliated to z 1, and those related more distantly to it-has features that distinguish it from the other. The Salamanca line includes a Life of Oppian and a brief discussion of heroic meter as preface to the scholia, and a paraphrase of Opp. Cyn. follows the scholia. These are z 1 properties. On the other side, the Madrid line has its idiosyncracies too, such as the epigram to Plato (omitted in Br), the Orphic hymn, and the closing hexameter verses that are appended to the end of the text. And, most distinctively, all the Madrid Mss. possess an identically abridged collection of the scholia. Hybrid members of both branches of the tradition (Br and B in this raper, also Monacen­ sis 134), inherited features of each. And al members of the tra­ dition, of either line of descent, written by Darmarios or not, share one essential characteristic: they are all comprised of Z scholia, and they all exhibit Z-family features, such as the two epigrams between Books 1 and 2 (d~ apXlo'tpa'tll'Yov and d~ aYYEAov), and the cosmological diagrams that occur in the middle of Book 1. Laurentianus 31.3 (Z) is one of the oldest archetypes in the Hal. tradition, as well as one of the most scholiated. Written by Manuel Spheneas in 1291, it has proved to be a bountiful exemplar, with seven surviving descendants and seven other Mss. bearing its scholia independently. Of the former seven, z 1, written by Johannes Calliandros in June, 1326, is by far the closest to its archetype-originating only thirty-five years after Z-and it reproduces the entire corpus of Z scholia. Any Ms. that has a Salamancan derivation has also inherited the full set, for the source of its exemplar was also z1's source. The scholia reveal what the very faithfully rendered poetic text in z 1 does not, that z 1 is not an apogram of Z. The first two sets of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 104 DARMARIOS' SCHOLIA readings below imply an intermediary between Z and two sibling descendants, z 1 and the putative parent of M, grandparent of the Darmarios group. All six Mss.-zl, M, and the Darmarios tetrad-share the following set of readings against Z, indicating a common ancestor, written sometime between 1291 and 1326, that is closer to Z. 1.29 ordo partium mutatus: (a) post Ilovro hab. on ... O'UAAaIl~avO'\)ow (b) post O'uAAull~avOU(nv hab. 0 IlEA.ACJ)V ... UUA(CE'tUl. zl, M, S, E, Br, B 1.47 post OIlIlU'tul om. ~OAat ... OIlIlU'tu2 zl, M, S, E, Br, B 1.51 1touv/ vouv zl, M, S, E, Br, B 1.68 'taxos! 'taxou~ zl, M, S, E, Br, B 1.70 post 1tUpU add. 'to zl, M, S, E, Br, B The following separative readings eliminate z 1 as a potential source for M and the Darmarios group. As z 1 precedes the others by two-and-a-quarter to two-and-a-half centuries, none of the other five Mss. could have served as exemplar for z 1. 1.1 UhaCouomnhaCouO'ul. z 1 1.1 post 1tOVO~ om. 1tov'to~ zl 1.17 post ~uAoi~ om. lCut 'toi~ ICM~O\~ zl 1.18 post 'tp(u om. 'ti}v lCohT\v z 1 1.43 post 1tU1t'tu(vCJ) om. cetera zl 1.417 post 01 om. UTtP zl 1.424 post O''tOVOEV'tU om. MYEl. 'tu z 1 Scores of conjunctive readings shared by M and the Dar­ marios group against z 1 confirm a common source for those Mss. The following list is but a representative sample. 1.2 'tW1./ 'tU M, S, E, Br, B 1.4 ~Xl./liXpl. M, S, E, Br, B 1.5 Al.1tU(VCJ)/ At1tCJ) M, S, E, Br, B 1.17 O'UV~EV~Po\s! O'UVE~POl.~ M, S, E, Br, B 1.20 post ~E om. 'to M, S, E, Br, B 1.29 post YEvvroO'l. om. ICUt EV 'to UEPl. M, S, E, Br, B 1.425 UAl.Ut£'tUl.! UA,l.Ut£'tov M, S, E, Br, B 1.793 CT\vot! Cilv o~ M, S, E, Br, B We are thus compelled to posit a second lost exemplar, a sibling of z 1, as the source of the M matrix of Mss. Mark Sosower has suggested Escorial desp. 286 as that source. S This 5 A brief description of the lost Escorial gr. 286 appears in G. de Andres, Catalogo de los Codices Griegos Desaparecidos de la Real Biblioteca de EI Escorial (Escorial 1968). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LYNN LEVERENZ 10! codex contained scholia on all five books of the Hal., scholia on Aesch. PV, Sept. and Pers., Eust. ad Dionys. Per., and scholia on an assortment of epigrams. The proposal is cogent for a number of reasons. First, Escorial 286 contained scholia on the same Byzantine school trilogy of Aeschylean tragedies as those whose scholia also appear in M.6 Secondly, it was associated with the Mendoza library in Madrid, as were M and Darmarios' I Mss. Thirdly, its scholia on the Hal. were attributed to the Byzantine polymath John Tzetzes, as they were by Camillus Gianetos in M in 1552 and Darmarios a quarter-century later in his copies of the scholia.7 And finally, from its description as containing scholia without the Hal. text, it is an attractive consid­ eration that Escorial desp. 286 could represent the inception of this tradition of independent scholia that culminated with Darmarios (Sosower 299). Whatever the identity of this lost codex, it produced three abridged offspring: Matritensis 4715, its complement, the ante­ cedent layer of Monacensis 134, and the lost exemplar for Darmarios' four Mss. M would seem to be the obvious candi­ date for Darmarios' source. 8 The codex comprises scholia on the Anth. plan. (also found in S), scholia on the same Aeschy­ lean triad contained in the lost Escorial gr. 286, scholia on the Hal, commentary and scholia by Proclus Diadochus on various works of Plato, and an anonymous paraphrase of Arist. Eth. Nic. Camillus Gianetos wrote the greater portion of M (in­ cluding the Hal. scholia) in 1552 according to a subscription in 6 Sosower has pointed out the possibility of a relationship beween the M­ family and the lost Grimani 230, owned at one time by Domenico Grimani, Cardinal of Venice. The codex disappeared in a conflagration that destroyed Grimani's library in 1687. Sosower notes that in addition to the scholiated Hal, Grimani 230 also featured the three plays of Aeschylus whose scholia were featured in des. Escorial 286 and appear in M as well. As the scholia to all these works accompanied the poetic texts in Grimani 230, that Ms. was not a member of the independent scholia tradition with which we are concerned here, but it may have formed a link in the chain of Mss. preceding it. It is not, however, a necessary component of the stemma here, and without readings the evidence is too sparse to assign a definite affiliation. See M. SOSOWER, • A Forger Revisited: Andreas Darmarios and Beinecke 269," JD BG 43 (1993: hereafter 'Sosower') 298. For a descripiton of Grimani 230 see G. Tomasinus, Bibliotheca Venetae manuscriptae (Utini 1650), where it is catalogued as Pluto 21.6. 7 The question of Tzetzes' authorship of Hal. scholia is complex and contro­ versial and will be treated in a future discussion. • For a description of Matritensis 4715 see G. de Andres, Catalog de los Codices Griegos de la Biblioteca Nacional (Madrid 1987). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 106 DARMARIOS' SCHOLIA another hand. 9 It belonged to Francisco de Mendoza's library, from which Darmarios would find his exemplar for Hal. scholia twenty-five years later. Nonetheless, two sets of evidence dem­ onstrate that M could not have been Darmarios' source. First, there are broad discrepancies between the constitutions of M and the Darmarios group. The latter possesses a sequence of scholia that is peculiar to and shared by all four Mss. It is ordered as follows: 1.1-73, 406-31, 784-97 (end); 2.6-688 (all); 3.1-647 (all); 4.1-203; 5.1-621 (in entirety, excert in E, where 5.460-601 are omitted). Over three-quarters 0 the Z scholia from Book 1 and two-thirds from Book 4· are absent. This assemblage of scholia diverges in order and content from that of M, whose material is arranged as follows: 1.1-237, 265-471, 564-785, 471 (continued)-563, 237-61, 779-end; 2.6-167, 333 (in part)-416, 634-end; Book 3 omitted entirely; 4.58 (in part)-end; Book 5 omitted entirely. One sees that M comprises a very different set of scholia from that of the other four Mss. In fact, the two sets overlap only in the portions of Book 1 that are contained in the Darmarios group, in the sections of Book 2 that are found in M, and in 4.58-203, amounting to roughly one fifth of the Z corpus. There is no evidence that material was removed from M after its original composition; rather, M was copied from an exemplar from which parts had been extracted, specifically, those sections of the scholia that surface as the nucleus around which Darmarios wrote Monacensis 134. The disparity in contents would in itself eliminate M as Darmarios' source; the separative readings that follow corroborate its ineligibility. M here deviates from the standard readings, while the other four Mss. (except at 1.42) concur with Z and z 1. 1.7 IJ.! IJ.Tt M 1.42 0.1 (J M, B 1.73 (Juvaepo(~rol (Juvavaepo(~ro M 1. end (0 'ta~uipXTt<;) pa1/ EAacppal S, E, Br, B 1.28££ post eXAl£U'tllCi1V am. 'tEXVTJ S, E, Br, B 1.417 post iO"'tEOV add. BE S, E, Br, B 1.417 OJ.LOlOlhal/oJ.Lolouv-ral S, E, Br, B 1.425 £X£l/ £XElV S, E, Br, B 1.784 ante £lCa'tov am. 'teX S, E, Br, B 2.396 a?JAor;j IXAA' au'to~ S, E, Br, B 4.185 post ro~3 am. 'to S, E, Br, B 107 " Another piece of evidence from which a common source for the Darmarios group can be inferred lies in a recurrent variant reading of 1t£pi for 1tapa found in Darmarios' four Mss. At these loci M shows the standard Z reading 1tapa, but in tachygraphical form. The scribe of M used the same abbrev­ iation, but with a pair of dots added beside the superscript epsilon, to designate the preposition 1t£Pi. M's exemplar prob­ ably had the same abbreviated forms of these words, which Gianetos then repeated in M. But the copyist of Darmarios' exemplar evidently wrote these prepositions without differen­ tiating the two abbreviations and transmitted them all as 1t£pi, which is how they appear in the four Darmarios copies. In­ stances of this syncretism occur in all four Mss. at 1.12, 17 (bis), 26, 45, 73. Where 1tapa is spelled out in M, and thus presumab­ ly in its immediate source (also the source of Darmarios' exemplar), Darmarios' Mss. do not have the 1t£pi error. The discrepancy is well illustrated at 1.17, a scholion that includes three 1tapa's. The first two are abbreviated in M; accordingly . they appear as 1t£pi in Darmarios' works. The third 1tapa is spelled out in M, therefore the correct reading was transmitted into the Darmarios grour as well. The most economica model for our hypothetical stemma would posit one member of the quartet as the source of the other three. Each of the Mss., however, possesses variant read- .~ that disqualify it as a potential exemplar for the others. manca 2730 is, according to Darmarios' colophon on f. 300v, the oldest of the four, completed in Madrid on 24 July 1577. "The codex, 448 folios in length, comprises scholia on Maximus Planudes' Anthologia, Oppian's Halieutica, Julius Africanus' Kes­ to; and a letter to Philo. As the first of Darmarios' copies, S would be the most logical candidate for internal intermediary; the readings that follow, however, reveal that S could not have served as the source for the other three Mss. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 108 DARMARIOS' SCHOLIA 1.9 ante 9ava'tou om. dval S 1.18 EMaivrol EMairo S 1.28££ a:YPEuouaal aYPEuaaaa S 1.42 post 'tou'to 2 add. Se S 1.793 ante 'tlK't6J.1£vo~ om. 'to S 5.20 £AKUa'tfjpal £A.KUaTItPla S Escorial gr. 569, ninety-three folios in length, comprises only scholia on the Hal. with the addition of the epigram to Plato, Orphic hymn, and closing gnome at the end of the text that is one of the earmarks of Darmarios' Madrid productions. E is the second oldest of these copies, written in Madrid, according to Darmarios' colophon on f. 92r, on 21 October 1578 and ob­ tained from the library of the Cardinal of Bourgos (f. 93v). The following separative readings demonstrate that E could not have been the exemplar for any of the other three. 1.41 arol oro E, M 1.42 1tAaVT\ 1/ 1tMvT\ E, M 2.238 i;EcpavT\aavl i;EcpavT\aal E 3.8 aM'ta9 aSuva'tOl~ E aM'to~ z12, S, Brl, Bl aSu'tou zl, Br2, B2 5.460-601 om. The third Madrid copy, Brussels 85, is a codex consisting of 176 folios containing scholia on the Hal., preceded by a Life of the poet, and a paraphrase of Opp. Cyn. 10 The Hal. scholia were written in two stages; the first layer, comprising the Madrid 'core' of scholia, was completed in 1579, as Darmarios' colo­ phon on f. 64v states. The following readings reveal that Br could not have been the prototype for any of the other Darmarios copies. 1.2 post 'tTl 10m. 9t,AUKWV ... 'tT\2 Br 1.47 ao~apw9 ~Aa~Ep~ Br, E 5.6-9 post avacpuval add. 'tou~ Br 5.21 ante KOW'I> add. 't'l> Br 5.203 h:a'tEp9Evl £Ka't£pro9Ev Br 5.259 om. ~plap.q ytvu~.q 'tou alS,;pou Brl Finally, Beinecke 269 is a codex of 176 folios with the same composition as Br. A Life of Oppian precedes and a paraphrase of his Cyn. follows the Hal. scholia-features that characterize 10 For a description of Brussels 85 (11329-31) see H. Omont, Catalogue des manuscrits grecs de la bibliotheque royale de Bruxelles (Gand 1885). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LYNN LEVERENZ 109 Mss. closely related to z 1.11 Like Br, B was written in two stages. The first phase of composition, which presumbably took place in Madrid, resulted in the same corpus of Hal. scholia found in Darmarios' other three copies. Darmarios apparently appended this Ms. to another codex and paginated it accordingly with the numbers 205-63. We do not have a date for this part of the Ms. because the end of the text of the scholia (f. 118r), where Dar­ marios' colophon, together wi th the standard Madrid features of epigram, hymn, and hexameters, normally appears, was prob­ ably altered by Darmarios in the process of integrating the material from his two sources (see below), and is now missing in B. But it seems reasonable to surmise that the first stratum of B was written in 1579 along with Br, as the second phase in the composition of the two Mss. is so closely allied. Moreover, Sosower has observed (296) that the watermarks in a large por­ tion of the antecedent layer of B (ff. 80-117) match those from Monacensis 162, a Ms. that Darmarios wrote in Madrid in 1579. The following separative readings eliminate B as a potential exemplar for any of the other Darmarios works. 3.21 0 lCUpOAE~(U/ lCUpLOAE~(U S, E, Br lCUpOAE~(U B 3.291 post OUlAEIC'tOV 2 add. 1tPOOICEl'tUl 11 jll] EV oup8oyy,? oux 'to liou<; B 3.308 a1tof3oAU 2/ a1tOAU B 5.6-9 post avucpuvul add. 'touc; S, E, Br om. B 5.146 a'tEAEo'tOl<;/ a'tEAEl)o'tOlC; B Whatever the identity of Darmarios' exemplar, it was a considerably reduced rendering of the Z corpus as presented in Escorial desp. 286 or its equivalent, a Ms. that evidently under­ went dissection before it disappeared altogether. Although in addition to falsifying dates and authors Darmarios has been in­ dicted for abridging bodies of commentary, it seems unlikely that he did so here wi th the Hal. scholia, for the truncation is identical in four Mss. written over a span of two to three years. Darmarios apparently did not work from a master copy of his own but from a prototype in the Mendoza library in Madrid, and it is unlikely that he would have curtailed the text so uniformly in all four copies. Moreover, Darmarios seems to have been eager to reproduce whatever scholia on the Hal. he 1\ For a description of Reinecke 269 see B. Shailor, Catalogue of Medieval Imd Renaissance Manuscripts in the Reinecke Rare Rook and Manuscript Library (Binghampton [NY] 1984). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 110 DARMARIOS' SCHOLIA could find, an inference afforded by the second chapter in the genesis of both Br and B. In 1580 Darmarios took these two Mss. with him on a trip to Salamanca. Whereas B had been annexed to another codex, Darmarios evidently left Br unattached to any other codex. While in Salamanca Darmarios discovered a Ms. closely related to z 1, with that codex's replete corpus of Z scholia, large por­ tions of which the Mss. he carried with him already had. He copied the balance of the scholia from the z 1 sibling and simply appended it to the end of B r. Because Darmarios did not integrate the scholia from the two sources in their natural se­ quence, the order of presentation is skewed. The arrangement of material in Br is: 1.1-73 (with a note by Darmarios at the end of this section informing the reader that he will find additional scholia, beginning with aUXTlv, at the end of this portion of the codex), 406-31 (with another marginal note on the last folio announcing the continuation of this scholia at the end of the Ms., starting with XEAlOcOV), 784-end; 26-end; 3.1-end; 4.1-203 (with a third note added at the end to the same effect as the previous two, this time commencing with the scholion on 't iYPTl<;); 5.1-end. After the usual epigram, hymn, and gnome that conclude the Madrid Mss. the second series of scholia, the Salamanca repertory, ensues, consisting of 1.66-406, 428-783 and 4.204-end. Darmarios also copied from the z 1 relative the Life of Oppian that precedes the scholia and the Cyn. para­ phrase thatfollows it in z 1-related Mss. In addition, he emended marginally and interlineally many of the original readings from his Madrid exemplar with z1 readings. These emendations occur most frequently in Book 1 and more sporadically in the books that follow. Darmarios obviously regarded this Ms. as the higher authority of his two sources, justifiably in view of its greater volume and probably its greater antiquity. The following set of tandem readings taken from Madrid portions of B r demonstrate the alternation of sources in operation. The other Madrid Mss. share the first tier of readings (Brl); the second, emended set (Br2) corresponds to z 1 readings. 1.1 ante y€vou<; add. 'tou BrI om. zl, del. Br2 1.2 YEVVW/ YEV£W Brl y£vvw Br2 1.5 'Atnalvro/ Alnro Brl 'Atnalvro Br2 1.17 post 'tonot<; et post u'Ao<; om. KVcOoa'Aa BrI add. KVcOSaS'Aa his Br2 1.29 au'AlC£'tat/ u'AloKE'tat Brl au'AlCE'tat Br2 3.1 u'At€a/ ~aOt'A€a BrI u'At€a Br2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LYNN LEVERENZ 111 3.7£ E1tEJ.I.\j1CXV/ 1tEJ.I.\j1CXl Brl E1tEJ.I.\j1E Z I, Br2 3.16 ilcxV091tcxV'toC; Z I, Br2 Darmarios sold the completed Ms. to Andreas Schott on 10 November 1580. About this time he separated B from its adop­ , ted codex and manufactured for it another copy of complemen­ " tary scholia from z 1's sister Ms., together with the Life and I Cyn. paraphrase. But this time Darmarios interwove the scholia from the two sources to agree with the sequence that is found in z1 and its lost sibling, inserting the supplementary segments from his second source within the antecedent layer in three places and paginating the entire corpus according to the new order.12 As a result, the older sections have two sets of pagina­ tion, while the later segments are singly paginated. With few exceptions, he did not, however, bother to emend the original . readings in the first layer of the Ms. with z1readings. Darmarios probably did not wish to take the time to do so, for he was , taking the Ms., now represented as a unified work, away from Salamanca, and potential buyers elsewhere would not be ex­ pected to discern any such omissions. Darmarios' colophon appears at the end of the Cyn. on f. 176r; it is dated 17 Novem­ ber 1580 at Salamanca, one week after Darmarios sold Br to Andreas Schott. Contrary to initial impressions, the second stratum of Br and B are not derived directly from z 1, although, as the foregoing and following sets of readings imply, they are closely related. Each of the errors below is shared by z1, Br, B, and the other two Salamancan branch Mss., Monacensis 134 (Mon) and Royal ms. 16 D XII (R). These readings are pervasive throughout the pertinent sections of the Mss. (1.74-405, 432-783,4.204-692), and can only be fractionally represented here. In no instance does M agree with these variants. In fact M, the sole representative of the Madrid branch for these segments of the scholia (obviously, Sand E do not contain them), is peppered with singular readings throughout these parts, underscoring its dis­ junction from Br and B here. 1.185 post cxu'toc; om. EK£l ... dOl z1, Br, B, Mon, R 1.208 (hcxv/ O'tl z1, Br, B, Mon, R 12 For a more detailed description of the composition of Beinecke 269 see D. F. Jackson and L. Leverenz, "The Sources of Beinecke Manuscript 269," , Re'CIHist Textes 22 (1992). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 112 1.215 1.238 1.287 1.296 1.360 1.367 1.438 1.450 1.461 1.475 DARMARIOS' SCHOLIA post Kul add. EV zI, Br, B, Mon, R 'ta:x.oq 'tu 'np{CilVl il~ zl, Br, B, Mon, R post PAocrupo~l om. ~h' EVO~ cr zl, Br, B, Mon, R post Ku8oAlKro~ add. vuv; post ix8uCilV om. vuv zl, Br, B,Mon,R post \jIUjljlOV om. ajlucr8lXlzl, Br, B, Mon, R post cr11jluivEl om. 1(l)piCil~ zl, Br, B, Mon, R post EPlPpuX11v om. 0 ... jlquAoPpuX11v zl, Br, B, Mon, R Perhaps even more pervasive than the foregoing class of error, however, are readings shared by Br, B, Mon, R, to the ex­ clusion of z 1. These variants, together with the separative zl readings that follow them, demonstrate that z 1 was not the exemplar of the others. This situation forces us to posit yet another anonymous Ms. in the stemma and confirms a growing suspicion that there were, as recently as the late sixteenth century, more copies of this material at large than one might have supposed. 1.180 post hu'ipo~2 add. 0 Br, B, Mon, R 1.228 nA£'i/ nA£'iv Br, B, Mon, R 1.249 post 'til~ om. u')'til~ Br, B, Mon, R 1.252 OU\jIlAroq OU\jIlArov Br, B, Mon, R 1.296 ro~4/ rocr't£ Br, B, Mon, R 1.349 E1tl