Proclus: Commentary on Plato's Timaeus, Vol. III

VII
PREFACE

The two lacunae of codex D (p. 18 ff. and 228 ff., cf. p. IV), which are of the same size, would have filled about 70 lines of this edition.

From p. 320, 27 to p. 356, in both Paris books, at equal intervals, gaps of about sixty-five or seventy lines of this printed edition appear, becoming larger and larger; yet so that the damage in codex D is both more frequent and always greater from the beginning than the gaps in codex Q: whence you might rightly conjecture that the lowest pages of the leaves of the exemplar were fragmented rather than the first. The smaller gaps that appear in the upper parts of the commentaries, now in D, now in Q (p. 231, 6 s 234, 30. 255, 13. 270, 9. 274, 7. 276, 3. 302, 17 s 304, 30 ff.), can be circumscribed by approximately the same range of pages, with the final leaves corrupted. Indeed, the lacunae of codex D mentioned above, which we contend arose twice because a page of the largest format of the exemplar was omitted by the scribe by mistake, would have filled individual pages of the same size.

The fragment Melissēs eis ton en Politeia logon tōn Mousōn On the speech of the Muses in the Republic (cf. Proclus in remp. ed. Kroll II 1 ff.), which codex D exhibits on p. 685 ff., having long been recognized by Usener and myself (cf. Mus. Rhen. vol. 54, 196 ff.), is divided into two parts by three or four empty lines in the books, of which the former, which is larger, would encompass along with the lacuna more than 60 lines of this edition. It follows that the leaf which we established to have been appended to the exemplar of the D, Chisian, and Escorial codices was added to protect the last leaf of the archetype of books Q and D (with the German ones), on the verso of which only a few things were written (p. 356, 26—28).

But, so that we may not be silent about anything, the lightest traces of those gaps do not seem to be absent even from the common recension. For gaps of one or two words have appeared (p. 320, 27. 334, 25. 352, 1) either in $\varsigma$ or in one of the other two books, and the common text is corrupted in those places where empty spaces occur in Q D (p. 326, 23. 349, 17. 21), a matter which, given the nature of the $\varsigma$ recension, it is not permitted to press too far.