The two conjectures of my own above mentioned are based on or corroborated by the Arabic. I ought to add, that in the Text and Critical Notes generally I have made a freer use than before of the Arabic version (concerning which see p. 4). But it must be remembered that only detached passages, literally rendered into Latin in Professor Margoliouth's Analecta Orientalia (D. Nutt, 1887), are as yet accessible to those like myself who are not Arabic scholars; and that even if the whole were before us in a literal translation, it could not safely be used by any one unfamiliar with Syriac and Arabic save with the utmost caution and subject to the advice of experts. Of the precise value of this version for the criticism of the text, no final estimate can yet be made. But it seems clear that in several passages it carries us back to a Greek original earlier than any of our existing MSS. Two striking instances may here be noted:—
(1) i. 6-7. 1447a 29 ff., where the Arabic confirms Ueberweg's excision of ἐποποιία and the insertion of ἀνωνυμος before τυγχάνουσα, according to the brilliant conjecture of Bernays (see Margoliouth, Analecta Orientalia, p. 47).
(2) xxi. 1. 1457a 36, where for μεγαλιωτῶν of the MSS. Diels has, by the aid of the Arabic, restored the word Μασσαλιωτῶν, and added a most ingenious and convincing explanation of Ἑρμοκαϊκόξανθος
