318 INTEODUCTORY REMARKS. can hardly be doubted by any reader who is acquainted witb Aris- totle's style and method. For example, it is obvious that the grammatical details in chapters XX. and xxi. are not in the style of Aristotle, and with regard to the former, where eight parts of speech are enumerated, we have the express statement of Diony- sius of Halicarnassus {de Comioositione Verhorum, c. 2, -init.; de Free- stantia Demosthenis^ p. 1101), and of Quintilian {Inst. Oral. 1.4, § 18), that Aristotle and Theodectes reckoned only three parts of speech. In the following translation I have indicated by brackets those passages which Ritter regards as interpolations, but I do not think that there is in every case an equally good reason for the ejection of the clause. J. W. D.
