In neither of these passages, however, have I altered the MSS. reading.
(c) Passages on which the Arabic throws no light:—
i. 9. 1447b 22. The only point of interest that emerges is that in the Arabic rendering ('of all the metres we ought to call him poet') there is no trace of καί, which is found alike in Ac and the 'apographa.'
x. 3. 1452a 20. The words γίγνεσθαι ταῦτα are simply omitted in the Arabic.
xxv. 18. 1461b 18, ὥστε καὶ αὐτὸν MSS. The line containing these words is not represented in the Arabic.
XXV. 19. 1461b 19, ὄταν μὴ ἀνάγκης οὔσης μηδὲν . . . The words in the Arabic are partly obliterated, partly corrupt.
In conclusion, I desire to acknowledge my obligations to friends, such as Mr. B. Bosanquet (whose History of Aesthetic ought to be in the hands of all students of the subject), Dr. A. W. Verrall, Mr. W. J. Courthope, Mr. A. O. Prickard, and Rev. Dr. Lock, who have written me notes on particular points, and to many reviewers by whose criticism I have profited. In a special sense I am indebted to Professor Susemihl for his review of my first edition in the Berl. Phil. Wochenschr., 28th September 1895, as well as for the instruction derived from his numerous articles on the Poetics, extending over many years in Bursian's Jahresbericht and else-
