ARISTOTLE 



(these include S and 0*5, and another ms. called /5 

 which contains only a very small part of the begin- 

 ning of G.A.) and the remaining two (one of which is 

 P) were collated by L. Dittmeyer from photographs. 

 Five others were collated (apparently from photo- 

 graphs) by Bitterauf sufficiently to establish their 

 character ; of the remaining eight he gives no report 

 on the character of their text. The upshot of 

 Bitterauf 's work is to show that Bekker was right in 

 basing the text upon PSYZ, and that although the 

 most faithful witness to the original text is Z, with 

 P a good second, no ms. has a monopoly of the truth, 

 since their common descent gives them all a fair 

 chance of preserving a good reading, just as it has un- 

 doubtedly ensured, as I mentioned above, that they 

 have all failed to preserve the text in certain passages. 

 With regard to the defective nature of Bekker 's 

 apparatus, the corrections which Bitterauf gives are 

 of value primarily in determining the comparative 

 trustworthiness of the mss. rather than in yielding 

 substantial improvements of the text " ; but there 

 are a good many places where they do make an 

 improvement possible, and all the suggestions which 

 Bitterauf makes for so doing I have carefully con- 

 sidered, and many I have adopted. ** When the 

 changes indicated are of a minor character, for 



" Examples are: 718 a 36, Bekker's app. aurais Z, 

 actually avrats SZ ; 719 a 31, Bekker's app. evros, to 8' e'/crds 

 Y, but actually PZ. Bitterauf had access to Bekker's own 

 copj'^ of the Basel Aristotle (1550), and shows that some of 

 Bekker's errors are due to his having used one set of symbols 

 for the MSS. in his collation and another set in his apparatus. 



* It should be remembered that Bitterauf's pamphlets are 

 merely " foretastes " of his projected edition, and therefore 

 the list of passages dealt with by him cannot be treated as 

 exhaustive. 



