(see Critical Notes). This emendation is introduced for the first time into the present edition. Professor Margoliouth tells me that Diels' restoration of ἑπευξάμενος in this passage is confirmed by the fact that the same word is employed in the Arabic of Aristotle's Rhetoric to render εὔχεσθαι.
Another result of great importance has been established. In some fifty instances where the Arabic points to a Greek original diverging from the text of Ac, it confirms the reading found in one or other of the 'apographa,' or conjectures made either at the time of the Renaissance or in a more recent period. It would be too long to enumerate the passages here; they will be found noted as they occur. In most of these examples the reading attested by the Arabic commands our undoubting assent. It is, therefore, no longer possible to concede to Ac the unique authority claimed for it by Vahlen.
I have consulted by the side of Professor Margoliouth's book various criticisms of it, e.g. by Susemihl in Berl. Phil. Wochenschr. 1891, p. 1546, and by Diels in Sitzungsber. der Berl. Akad. 1888, p. 49. But I have also enjoyed the special benefit of private communication with Professor Margoliouth himself upon a number of difficulties not dealt with in his Analecta Orientalia. He has most generously put his learning at my disposal,
