MrShewring, C.Q. XXV. 14, gives statistics of the clausulae favoured in Aristotle's Ethics. I have applied them to test a few conjectural emendations that I happen to have published, with the following encouraging results: (a) Emendations that substitute a good clausula for a bad one:

96a 18 ὓστερον λέγομεν for ὓστερον ἐλέγομεν κb, ὓστερον ἔλεγον cet.

09b 5 ἑαυτοὺς ἀφέλκειν [δεῖν].

48a 14 ἀκόλαστον τίθεμεν [καὶ ἐγκρατῆ καὶ σώφρονα].

63b 13 τὴν φιλίαν [καθάπερ εἴρηται].

71a 35 αὐτῶν <ἡδονὴ εἶναι (the crasis gives – ∪ – –: some additional figures that Mr. Shewring has been good e nough to send me show that this is the second commonest clausula in normal Attic prose, and that in the Ethics its frequency is slightly above the normal).

(b) Emendations that substitute one good clausula for another:

97a 9 τὸ αὐτὸ τοῦτ for τὸ αὐτὸ τοῦτο ἀγαθόν κb, τὸ αὐτὸ or αὐτὸ τὸ ἀγαθόν cet.

25a 22 τῶν ἀγαθῶν [καὶ ἀγνοεῖν ἑαυτόν].

I may be pardoned for omitting my one or two alterations that substitute a bad clausula for a good one.

My text adopts a MS. variant, or a conjecture of another scholar, similarly supported by Mr. Shewring's test in the following places: 99b 6, 02a 20, 10b 31, 15a 8, 20a 22, 20b 2, 22a 28, 32a 11, 33a 1, 34b 33, 50b 34, 53a 1, 59a 3, 61a 2, 63a 3, 64a 23, 66a 34, 75a 6, 79a 22.