^{page 57 note 1} See Iriarte, Joannes, Regiae Bibliothecae Matritensis Codices Graeci MSS. (Madrid, 1769), pp. 86–120Google Scholar; Sikes, and Allen, , Homeric Hymns, p. XIIGoogle Scholar.; Ludwich, A., Eudocia …, p. 123Google Scholar For the Homeric Hymns it is T; for Proclus, G.

^{page 58 note 1} See Puntoni, V., Indice de' codici greci della biblioteca Estense di Modena, in Studi Italiani di Filologia Classica, Florence, IV. (1896), pp. 487–8Google Scholar; Nigra, , l.c., pp. 216Google Scholarsqq.; and Hollander, in Jahrbb. f. Philol., 1892 p. 544Google Scholar; Ludwich, , op, cit., p. 122Google Scholar.

^{page 58 note 2} Formerly III. E. 11 and 164. The Librarian of the Bibl. Estense kindly informed me of the change in designation. For Homeric Hymns it is E; for Callimachus Wilamowitz calls it q (Schneider did not know it); for Proclus it is B.

^{page 58 note 3} Nigra, , I.e., says it was written towards the end of cent. XVGoogle Scholar.

^{page 58 note 4} See Ludwich, l.c.

^{page 58 note 5} Propertius II. 1, 40.

^{page 59 note 1} Cf. II, F, and At.

^{page 59 note 2} -s at the end of a line occurs only thrice in the text, once in a suprascript correction.

^{page 59 note 3} See Martini, E. and Bassi, D., Catalogus Codium Graccorum Bibliothecas Ambrosianae (Milan, 1906)Google Scholar; Nigra, , l.c., pp. 204–5Google Scholar; Schneider, , praef., XXXIGoogle Scholar. for Proclus it is D.

^{page 59 note 4} Ff. 58, 82^r, 83^r, 106^r, 110^r, 111^r are partly blank.

^{page 59 note 5} So the Ambrosian Catalogue, dating it 1509 (from f. III^r); Nigra, , i.c., p. 204Google Scholar, ascribes it to the end of cent. XV.

^{page 60 note 1} See Ludwich, , op. cit., p. 122Google Scholar.

^{page 60 note 2} Catalogue, Ambrosian, with a reference to Bassi in Bollettino di Filol. Class. IV., pp. 32–4Google Scholar.

^{page 60 note 3} Of Nauplia, who lived about 1475 (Nigra, , l.c., p. 204Google Scholar, who says that q has all the faults characteristics of the books which issued from Suliardo's workshop). He was an Argive ( έκXώραѕ тὦ⋯ ⋯рУίων, subcription to Mut. III. C. 6).

^{page 60 note 4} Cf e. He wrote Ambrosian 629 (P 78 sup.; formerly S 556), described thus in the Catalougs; ‘“‹Ex› ercitam ‹en› ta utriusque linguae ‹sc. gr. lat.› Vincentii Pinelli sed praescrtim Graecae.” Passim sim sunt uero notuale uarii generis, nullius tamen momenti. Inter alios adferuntur loci Aristophanis, Callimachi, Euripidis, Homeri, et Thucydidis.’

^{page 61 note 1} I. 5 ⋯μϕlλοyo⋯ (q): 14 oÌ ⋯pXɑîol (q): 25 θŋplα (q omits): 28 ʏṔ Ṕɛlŋ Q (SQq have ŋpŋ for Ṕɛlŋ in text: q omits the variant):36 μpdXɛpo⋯oς roÛ ĸɛ⋯Ƭαuṕ S, μpα Xlpo⋯ roÛ κɛ⋯Ƭαup (clipped on margin) Q (μp⋯ Xlpo⋯ roÛ κɛ⋯ƬαÚ po⋯ q). S has all these in the text ink and Lascari's hand. The scholia on I. 5, 14, 36 are from the common stock; at 25 cett. have a longer gloss.

^{page 61 note 2} Valla is the writer of a manuscript containing Plato's Critias, Timaeus, Minos, and Definitions, which is now in the Regia Biblioteca Estense; see Allen, T. W., Notes upon Greek Manucripts in Italian Libraries (1890), p. 11Google Scholar, No. 89.

^{page 62 note 1} In S these are in the text hand but in a lighter ink. They are I. 17 καθάρμαϒα: 17: ⋯πολοαιτο: 18 and 26 ποταμòς ⋯:ρκαδίας 25 κινώπεδα: 34: σπήλαιον: 35 έθρεψε At 35 έθρεψε was corrected later to αν in S: e had probably arisen from a misreaóing of the sign for α At I. 42 Ѕ has πόλιѕ κα⋯ ⋯λσοѕ in the text ink, Q in lighter ink but from the text pen, which suggests that Valla overlooked it in copying the text-ink series from Ѕ.

^{page 62 note 2} These are I. I ε⋯η: 22 δρ⋯:25 κཱαταδὑεις:25 κινώπεδα altered to ετα, and τ⋯ έν τ⋯ π⋯δω κινώπεδα θηρ⋯αadded before the change: 33 ν⋯α (on: ⋯να 57 παῖς ν⋯οѕ: 66 θεόν: 72 μικροῖς: 74 γεωργòѕ: 76 Ủμνοûμεν: 77 κνγήϒ(sic): 84 ƥεÛσιν ⋯Φένον, repeated in margin, |where stands also ƥνηΦονίην ƥ Ûσιν ϕόνον: 90 ⋯νέδην ϕ+(?). II. 19 κα⋯ λνκωρε +(?): 20 θρηνεÎ: 21 πρόσΦθεγμα κατα Φωνούτοѕ: 23 (for 24) γ⋯ѕ γιόβηѕ: 32 περόνη: 42 βοηθόѕ: 45 ψ⋯ѕ νιόβης: 32 περόνη: 50 blurred: 70 μ⋯ντικον: 89 blurred (? - + ΦνǶσον): 92 ή κνρίνη ή τοÛ 94 ἅÖσα. III. 3 παίξειν: 45 όπαδούς έκ τοÚ ἆμα òρούεσθαι: 62 ëτλησαν: 112 έλάΦοις: 114 λαίλαψ: 155(for 154)νεβρούς. IV.12 λ⋯ροις: 36 ήν (rather a correction of δι' ὧσοι then a scholium). V. 76 μελανίξων. Of these, except in so far as they form part of longer scholia, Q has only ϒεωρϒός (I. 74); and there and at II. 92 (where he has only ή κνρήνη he agrees, with Ee, from some kinsman of which, as we shall see later, he derives most of his scholia and also some corrections of his text. Of S's marginal reminders he reproduces only ώαρίων (III. 265). These reminders and the thirty-five scholia, though all are undoubtedly from Lascaris' hand, appear in inks of various hues; and their lettering, while in all cases paralleled in some part or parts of the text itself, is also varied, so that Valla may well have formed a hasty judgment that they were not authentic enough for admission to his carefully select text (for the working of this principle in the matter of textual additions, see below, p. 70).

^{page 62 note 3} But one (ίερ⋯ τριήρηѕ, IV. 314) may be meant as a schlium.

^{page 62 note 4} Which is held, at least as regards the Homeric Hymns, by Allen (who makes them brothers; see J.H.S. XV. 1895, p. 163) and Hollander, H. (in Hermes, 1891, pp. 170Google Scholar sqq.).

^{page 62 note 5} With two qualifications; for we shall find a few changes in S that were perhaps made by Valla, and a few other corrections of both S and Q (e.g. at V. 17) have reached each independently.

^{page 63 note 1} These all concern quite trivial points, and cannot independently counter my view that S does not come from Q. The remaining reading both S and Q got by correction –III. 201 σϰίνος (S from σϰό, Q from σϰύ-).

^{page 63 note 2} Under (f) come also IV. 167, 300, both discussed below; and III. 99, where, though προμολѕՋ१ᓦ⋯ S has μ with a long stem, followed by an undeniable o, Q gives προκαλ⋯σ, and later sets μ above κ in lighter ink and finer pen.

^{page 63 note 3} A later hand Produces γον (sic). Probably Valla made the change όΦθαλμώς in both manuscripts.

^{page 64 note 1} μακηδόνικοι ρα νέεσθαι II.; cett. divide correctly.

^{page 64 note 2} Perhaps we should add his ⋯δέ(⋯δε S cett., VI. 98). His ⋯πολΪηѕ (so q: ⋯ποϔηѕ S cett., III. 230) is a more doubtful case.

^{page 64 note 3} If so, this stroke is probably Valla's work. βιάξεθαι was due to the influence of μαντεύεσθαι (just following).

^{page 64 note 4} Lascaris or his source several times confounds β, κ, and μ. At VI. 97 τοοιαûτ corrected to τοΊα may have been a gloss in the source; and at VI. 101 ύύπ' ⋯ύ' π⋯π'όλλωνοѕ looks like an ancestral doublet ύύπ'

⋯ύ'

 πόλωνοѕ

^{page 64 note 5} δνσμάς occurs here only in S and (in the text) in GIBr (δημ H). In S it perhaps comes from a later hand; certainly we have nowhere else any proof of interaction between the z manuscripts. on the one side, and cett. on the other.

^{page 65 note 1} — marks the word as a personal name Why ην should be repeated is not clear, for the text is quite legible. The topmost mark seems to be ώληs, and the margin has ώληś in Lascaris' hand.

^{page 66 note 1} The additions to S from Politian (see below) must stand or fall together; and, since Valla received part or all of the text of four of them from S (pp.70–72 below), we may fairly assume that all were in S before Q was copied. The added ε of S in VI. 7 is the same semi-supine letter as in the Lascarean Φαμένα(V. 131); and ονѕ at V. 5, in almost or actually contextual ink, is clearly Lascaris'.

^{page 66 note 2} More probably he belived the ink to be non-contextual; for in six passeges (see p. 64 above) he accepts doublets that are undoubtedly contextual in S.

^{page 66 note 3} For δνσμάЅ (VI. 10) see above, p.64, II. 5.

^{page 66 note 4} Here δέ (II Politian Hq) probably underlies the blotched κí in S (see p. 71 below). κα⋯ is also found in FIBr and ed. pr.; but S shows elsewhere no trace of indebtedness to any manuscripts of the z group, if we put out of reckoning δνσμάς (VI. 10).

^{page 66 note 5} But έûσαν (V. 69) coincides with ed. pr. in th position of the breathing. Politian's text lack breathings.

^{page 66 note 6} For the details of C. Lascaris' life see Legrand, E., Bibliographic Hellénique (Paris, 1885), I. pp. LXXIGoogle Scholar. sqq.; Sandys, J. E., History of Classical Scholarship, II. pp. 76–8Google Scholar.

^{page 67 note 1} Their close kinship will appear when the lacunae of ABCK are treated.

^{page 68 note 1} Groups (c) and (h) on pp. 61–2 above.

^{page 68 note 2} Q's rejection of S's is, ⋯ς which it is probable that Lascaris had previously added there from Politian. follows his general rule of disregarding the non-contextual elements of S.

^{page 68 note 3} see below, p. 75.

^{page 68 note 4} q seems to have thought the solitary letter not worth recording; cf. his attitude at V. 136 and 139 (below, p. 75).

^{page 68 note 5} Also, at V. 18 the dark dot in έριν produces the reading (έρι) of ABCK alone.

^{page 68 note 6} It is not quite clear from the photograph whether the suprascript in Q's öστιѕ οι (VI. 116), written with the text-pen, is also in the text-ink or in this lighter shade; but the former seems more probable, q disregards the correction; H has оöτ,

οι

I őττοι, E őѕ τοι, cett. őѕ Q's correction is not necessarily drawn from E as the original slip would have been easy to make in copying from 5, and may have been corrected at once from S, or later from C or K.

^{page 69 note 1} See below, p. 74, n. 2.

^{page 69 note 2} Similarly unavailing to guide us here is Q's link at III. 254 ĩνα ϰιώνησ (ȉνα ϰιώνης ȉνα ϰιώνης qz: ⋯να ϰιώνης cett., ed. pr.).

^{page 70 note 1} I infer this from two considerations—(a) that the ink used in these four passages is that employed for the C (or K) additions; (b) that, if suited before the readings in the four passages were settled, it would, while unavailing at V. 128 and the end of 136, have given Valla the whole of δλολνγαîς at 131 and ώς Φαμένα(131) woul have appeared in Q in the lighter ink used for class (2). However, the Ee additions may have preceded both classes (1) and (3); for we find them strictly limited to corrections of words already present in Q's original text, though (e.g.) Valla could probably have completed VI. 22 from this source (Ee have it in full), and Ee's ψεν (V. 136 ad init.), if their kinsman had it, might as readily have been used by Valla as CK's π (VI. 23 ad init.).

^{page 72 note 1} Sandys, J. E., History of Classical Scholarship, II. p. 77Google Scholar; Heiberg, J. L., Beiträge zur Geschichte Georg Valla's und seiner Bibliotheh (Beiheft zum Centralblatt für Bibliothekswesen, XVI. Leipzig, 1896), p. 7Google Scholar.

^{page 72 note 2} Heiberg, op. cit. p. 7.

^{page 72 note 3} Heiberg, op. cit. p. 61, Ep. 3; p. 88, Appendix, No. I.

^{page 72 note 4} Heiberg, op. cit. p. 7. Further, I think that some Greek MSS., owned by Valla and now in the Biblioteca Nacionale at Madrid (e.g. No. 113), must have been in Lascaris' possession at his death and accompanied his own collection to Naples and ultimately to Madrid.

^{page 72 note 5} Heiberg, op. cit. p. 88, Ep. 36.

^{page 73 note 1} Cited by Heiberg (op. cit. p. 61, Ep. 2), who idendifies the Hymns with Mut. III. E. II(Q).

^{page 73 note 2} See Schneider, , Praef. XXXIGoogle Scholar.

^{page 73 note 3} Seemingly a conjecture by Q, as is clear αι is clear in S.

^{page 73 note 4} Such subscription of an omitted letter όππόϒαν ούκ cett. rightly here) is found in S also at IV. 8 (ὃτιѕ)

ς,

319 (ὃτι)

τ,

V. 83 (έκόλασαν)

λ.

^{page 74 note 1} A converse case of confusion by q's scribe between η and ε, so often closely alike in Q, occurs at II. 10, where q reads Q's dubious ⋯δη (S has ⋯δῃ) as íδε íδε μέϒας, with íδἕ later in the line, is attractive in view of the similar Juxtapositions in I. 55, II. 25 (=80=97=103), and IV. 204; cf. II. 3–5, 109–110, Theocritus, VI. 19, etc. (see Schneider, vol. i. pp. 152–4). For a short final vowel lengthened before μ cf. III. 55, 61, 150, and the Homeric precedents (Monro, H.G., §§ 371–2).

^{page 74 note 2} Similarly III. 99, προκαλᚨἦσ μ Q, προκαλης μ q. A curious case is VI. 77, καλνξώ

π

Q, παλνξώ q. where seemingly q accepted the correction, but thought κ worth recording (? as a variant); cf, Br from I at IV. 42.

^{page 74 note 3} Cf. VI. 77, ἦνθε q. Q cannot be a copy of q; for it has III. 33, which q omits owing to homoeographon with 34.

^{page 75 note 1} In more than half of these q improves on Q. Except these and four or five successful emendations, noticed below, all his variations from Q are for the worse.

^{page 76 note 1} I call Schneider's citations Q¹ here to avoid confusion.

^{page 76 note 3} E. g. Q¹ omits II. 27 with SQqx.