recognised as the true reading, the suppressed object being not the audience but the rhetoricians.
Once more, in xxiv. 9. 1460a 23, where Ac gives the meaningless ἄλλου δέ Be, I read (as in the first edition) ἀλλ' οὐδέ, following the reviser of Ac. This reading, which was accepted long ago by Vettori, has been strangely set aside by the chief modern editors, who either adopt a variant ἄλλο or resort to conjecture, with the result that προσθεῖναι at the end of the sentence is forced into impossible meanings. A passage in the Rhetoric, i. 2. 1357a 17 ff., appears to me to determine the question conclusively in favour of ἀλλ' οὐδέ . . . ἀνάγκη . . . προσθεῖναι. The passage runs thus: ἐὰν γὰρ ᾖ τι τούτων γνώριμον, οὐδέ δεῖ λέγειν• αὐτὸς γὰρ τοῦτο προστίθησιν ὁ ἀκροατής, οἷον ὅτι Δωριεὺς στεφανίτην ἀγῶρα νενίκηκεν, ἱκανὸν εἰπεῖν ὅτι Ὀλύμπια γὰρ νενίκηκεν, τὸ δ' ὅτι στεφανίτης τὰ Ὀλύμπια, οὐδέ δεῖ προσθεῖναι• γιγνώσκουσι γὰρ πάντες. The general idea is closely parallel to our passage of the Poetics, and the expression of it is similar, even the word οὐδέ (where the bare οὐ might have been expected) in the duplicated phrase οὐδὲ δεῖ λέγειν, οὐδὲ δεῖ προσθεῖναι. One difficulty still remains. The subject to εἶναι ἢ γενέσθαι, is omitted. To supply it in thought is not, perhaps, impossible, but it is exceedingly harsh, and I have accordingly in this edition accepted Professor Tucker's conjecture, ἀνάγκη <κἀκεῖνο> εἶναι ἢ γενέσθαι.
