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

VI
PREFACE

However, it appears that more good readings of the common text can return to the tradition preserved more integrally in $\varsigma$ than in the gaps of Q D, which appear in Q D toward the end of the fifth book—and they are wider, as one would expect, in D—whereas the context of the $\varsigma$ codices is generally not only continuous but also sound in those places, such that there the commentaries rely solely on the common recension.1 For this reason, we are all the more offended that at the beginning of the fifth book, all the codices Q D $\varsigma$ are lacerated in the same way at certain intervals. It is obvious that a great difficulty arises for one constructing the history of the commentaries. For codex D, which we recognized as occupying a place in some way between Q and $\varsigma$, acts not only in good readings, but also in bad ones, and in transmitting lacunae and gaps, sometimes with Q against $\varsigma$, sometimes with $\varsigma$ against Q, and with Q $\varsigma$ against M P.

For the rest of the fates of the context, which is slipping away more and more, you would expect the gaps of codex Q, which we see enlarged in D, to have become very large in $\varsigma$. But the exact opposite happened: we found them to be closed. Were the gaps perhaps filled from some very integral codex that is now completely lost?

This could happen all the more, because it is agreed that the common recension exceeds in age all the Proclian books that still exist (cf. p. 13). Nevertheless, it will appear below that that explanation also has features that give us offense.

But first, let us examine the origin and nature of the lacunae and gaps.

1. We consulted for those places the final part of the Marcian codex 195, transcribed in the 15th century (M²; cf. I p. VII). But we saw that greater authority must be held by the Munich and Oxford books, since the Marcian context seemed to be lost here and there with inept errors and conjectures, of which we consider it sufficient to have brought one example:

p. 345, 6s and 8s epi tou antiprosōpou autō hestōtos and tou ... horōntos Q: epi tou hen ti prosōpou (prosōpon) autō hestōtos and to ... horōntos $\varsigma$: epi tou eph' hen prosōpon hestōtos and tō ... horan M²

But wherever the reading of codex M², however it arose, seemed to be better, we either received it into the context or into the annotation (p. 326, 23. 330, 18. 334, 25. 352, 1).