text of this edition differs from that of the last:—
vii. 6. 1451a 9. Here I keep the reading of the MSS., ὥσπερ ποτὲ καὶ ἄλλοτέ φασιν. Schmidt's correction εἰώθασιν for φασίν seemed at first sight to be confirmed by the Arabic, but, as Vahlen argues (Hermeneutische Bemerkungen zu Aristoteles' Poetik, 1897), this is doubtful, and a more fundamental objection the question arises whether the correction can, after all, convey the sense intended. Can the words as emended refer to a known practice in present time, 'as is the custom on certain other occasions also,' i.e. in certain other contests, the ἀγῶνες of the law-courts being thus suggested? As to this I have always had misgivings. Further observation has convinced me that ποτὲ καὶ ἄλλοτε can only mean 'at some other time also,' in an indefinite past or future. With φασίν (sc. ἀγωνίσασθαι) the reference must be to the past. This lands us in a serious difficulty, for the use of the κλεψύδρα in regulating dramatic representations is otherwise unheard of. Still it is conceivable that a report of some such old local custom had reached the ears of Aristotle, and that he introduces it in a parenthesis with the φασίν of mere hearsay.
ix. 7. 1451b 21. I accept Welcker's Ἀνθεῖ for ἄνθει. Professor Bywater is, I think, the first editor who has admitted this conjecture into the text.
xvii. 5. 1455b 22. I restore the MSS. reading ἀναγνωρίσας τινάς, which has been given up by almost all editors, even the most conservative. Hitherto a parallel was wanting for the required
