¹ Storia della Tradizione e Critica del Testo 247.

² On the MSS. of the Laws and Spuria much light has been thrown by Post, L. A., The Vatican Plato and its Relations (1934)Google Scholar; and in his Budé edition of the Laws (Parts i and ii, 1951) desPlaces has set a new standard of precision in presenting the manuscript evidence.

³ Though a long list of Burnet's errors in reporting W in the Phaedo was published by Klos, and Minio-Paluello, in CQ 43 (1949), 126.CrossRefGoogle Scholar

⁴ CR 16 (1902), 98 ff.; 17 (1903), 12 ff.

⁵ De Platonis … F memoria (diss. Göttingen, 1922).

⁶ Published in the series Editiones Helveticae (Francke, Bern, n.d.).

⁷ See the articles referred to in note 4.

⁸ Adam, , CR 16 (1902), 215Google Scholar; Immisch, , Philologische Studien zu Platon II, 84, n. 1Google Scholar; Theiler, op. cit. 138.

⁹ In his Ecloga Vocum Atticarum Thomas condemns the forms ὀψοποιητικός at Gorg. 465d6 and αἰσχυντηλόζ at 487b 1, both of which are found in F; he also omits μή with F at 511a6. But it is no doubt possible that he found the text so quoted in the indirect tradition on which he drew.

¹⁰ Cf. Dain, A., Les manuscrits 135 f.Google Scholar

¹¹ Cf. Irigoin, J., Histoire du texte de Pindare 107Google Scholar, on the difficulties experienced by these late transcribers in transliterating uncials.

¹² Pages, not columns. Had the exemplar been written in two columns, as F itself is, the standard series of intervals between lacunae would have been 22–66–22 and not 22–22–22.

¹³ The Descent of Manuscripts, 414 ff.

¹⁴ Roberts, C. H., ‘The Codex’, Proc. Brit. Acad. 1954, 195.Google Scholar Examples of third-century papyrus codices of Attic authors having similar dimensions are P. Rylands 549 (Xenophon) with an average of 39 letters to a line and 32 to 35 lines to a page, and P. Oxy. 459 (Demosthenes) with about 42 letters to a line and 32 to 34 lines to a page.

¹⁵ CR 16 (1902), 391. Immisch had already spoken in similar terms of F, op. cit. II, 15.

¹⁶ Deneke put forward the opposite contention, that in the Gorgias (though not elsewhere) the F tradition shows traces of having been revised by an Atticist. But he produced as evidence only two words, one of which, ἑπτέτη at 471C2, turns out not to be in F, while the other, ἀρτοποιὸς for ἀρτοκόπος at 518b6, has no claim to be called an Atticism.

¹⁷ Y is a ‘Mischcodex’ whose contents are drawn from various sources, and as Alline observed (Histoire du texte de Platon 235), its value varies widely in different dialogues. In the Gorgias, and also in the Meno (for which Mr. Bluck has kindly shown me the results of his collation), I doubt its claim to primary status. In both dialogues Y appears to me to descend from W through a MS. which was corrected in placesfrom F; to this mixture it adds a good many false guesses, as well as accidental corruptions of all sorts. In the Meno it seems to contribute nothing; in the Gorgias very little, and nothing that exceeds the range of easy conjecture.

¹⁸ Platon II, 334. Ritter had made the same point in a review of Burnet's, text, Bursians Jahresbericht 161 (1913), 64 f.Google Scholar

¹⁹ op. cit., 134 f.

²⁰ Über d. Platocodex in Venedig, 68 f.

²¹ Flor has on the flyleaf a note referring to events of that year which was almost certainly made at the time of their occurrence; it is not in the scribe's hand. Immisch, overlooking this, assigned the MS. to the fifteenth century; Rostagno made it late thirteenth.

²² I am indebted for these particulars to my pupil Father H. D. Saffrey, O.P., who kindly examined V for me. The fact that Immisch and Post have considered V a primary authority for the Axiochus has thus no bearing on its value in the Gorgias.

²³ M belonged to Dr. Giovanni Marco da Rimini, who left it at his death to the library of the Franciscan convent at Cesena, which formed the nucleus of the Biblioteca Malatestiana.

²⁴ Lewis Campbell described M in J. Phil, 11 (1882), 195–200, and collated it for his edition of the Republic; but so far as I know it has not been collated for any other dialogue. For tetralogies I–VII and Spuria collation would probably in fact be labour wasted, but its remaining contents should be examined.

²⁵ Platocodex, 56 ff. and 104. Post has since shown that M derives from Par in the Spuria also (Vatican Plato, 53 f.). It seems to be a direct copy, while Flor is an indirect derivative. Parisinus 1809 (Bekker's C) appears to have (as Schanz thought) the same origin, but I have not personally examined it.

²⁶ e.g. 491b8 [sic] T Par M Flor. Schanz, , Platocodex, 47 ff.Google Scholar, cited instances where Par omits a complete line of T, and others where Par is corrupted through misunderstanding T's corrections.

²⁷ Notably at 507c8, T: B W F Par Oxy. Stob. Here T's false order could not have been corrected by conjecture.

²⁸ The sole exception which I have noticed is at 523d7, where Par and its derivatives have νῦν μὲν with Plutarch (νῦν B T W F).

²⁹ M has a few small and obvious corrections which I cannot trace in Par as it now stands and which Stallbaum has not noted in Flor: for at 454b9 (also in E and Y); for at 456b6 (also in J); μοι s.l. for με at 486d7 (also in E, Y and V); ἴδια for ἰδίᾳ at 514C2. J's only independent contribution would seem to be οὔτοι (which is not in F) for οὔ τι at 450e4. It is a hybrid MS.: its text hasbeen systematically contaminated from F as far as 472d, and perhaps sporadically elsewhere. On f see above, p. 25.

³⁰ The collation of Par, M and f is my own, and I have personally checked some though not all ofthe readings cited from Stallbaum's collation of Flor and Bekker's of V. For the unimportant J, I am entirely dependent on Bekker.

³¹ Cf. Maas, Paul, Byz. Zeitschr. 1935, 299 ff.Google Scholar, 1936, 27ff., and Gnomon 25 (1953), 441 f.; also Turyn, A., ‘The Sophocles Recension of Manuel Moschopoulos’, T.A.P.A. 1949Google Scholar, who shows that the Byzantine recensions of Sophocles reach well back into the thirteenth century. As Sandbach, F. H. has recently observed, ‘there is a danger of underestimating the powers of the late Byzantine scholars, and so, through unnecessarily crediting them with access to unknown traditions, of according unwarranted honour to their conjectures’ (CR 68, 1954, 251).Google Scholar

³² While the main basis of V appears to be Flor, it has readings apparently derived from F (e.g. at 481d6) and others characteristic of the Y group (e.g. at 508e5). Theiler himself expresses uncertainty ‘coniecturaene debeantur bonae lectiones unius codicis V’.