which he had although απο ‛υπαρχοντος (proceeding from an actually existent thing) was not κατα το ‛υπαρχον, i.e. did not truly represent that existent thing. Aug. _Cont. Acad._ II. 11 quotes Cicero's definition and condenses it thus; _his signis verum posse comprehendi quae signa non potest habere quod falsum est_. _Iudicium_: κριτηριον, a test to distinguish between the unknown and the known. _Eo, quo minime volt_: several things are clear, (1) that Philo headed a reaction towards dogmatism, (2) that he based the possibility of knowledge on a ground quite different from the καταληπτικη φαντασια, which he pronounced impossible, (3) that he distorted the views of Carneades to suit his own. As to (1) all ancient testimony is clear, cf. 11, Sextus _Pyrr. Hyp._ I. 235, who tells us that while the Carneadeans believed all things to be ακαταληπτα, Philo held them to be καταληπτα, and Numenius in Euseb. _Praep. Ev._ XIV. 8, p. 739, who treats him throughout his notice as a renegade. (2) is evident from the _Academica_ and from Sextus as quoted above. The foundation for knowledge which he substituted is more difficult to comprehend. Sextus indeed tells us that he held things to be _in their own nature_ καταληπτα (‛οσον δε επι τη φυσει των πραγματων αυτων καταλ.). But Arcesilas and Carneades would not have attempted to disprove this; they never tried to show that things _in themselves_ were incognisable, _but_ that human faculties do not avail to give information about them. Unless therefore Philo deluded himself with words, there was nothing new to him about such a doctrine. The Stoics by their καταληπτικη φαντασια professed to be able to get at _the thing in itself_, in its real being, if then Philo did away with the καταλ. φαντ. and substituted no other mode of curing the defects alleged by Arcesilas and Carneades to reside in sense, he was fairly open to the retort of Antiochus given in the text. Numenius treats his polemic against the καταλ. φαντ. as a mere feint intended to cover his retreat towards dogmatism. A glimpse of his position is afforded in 112 of this book, where we may suppose Cic. to be expressing the views of Philo, and not those of Clitomachus as he usually does. It would seem from that passage that he defined the cognisable to be "_quod impressum esset e vero_" (φαντασια απο ‛υπαρχοντος εναπομεμαγμενη), refusing to add "_quo modo imprimi non posset a falso_ (‛οια ουκ αν γενοιτο απο μη ‛υπαρχοντος), cf. my n. on the passage. Thus defined, he most likely tried to show that the cognisable was equivalent to the δηλον or πιθανον of Carneades, hence he eagerly pressed the doubtful statement of the latter that the wise man would "opine," that is, would pronounce definite judgments on phenomena. (See 78 of this book.) The scarcity of references to Philo in ancient authorities does not allow of a more exact view of his doctrine. Modern inquiry has been able to add little or nothing to the