hand, deals with things to be sought and avoided, with human life and with the end-in-chief. I They reject dialectic as superfluous ; holding that in their inquiries the physicists should be content to employ the ordinary terms for things a Now in The Canon Epicurus affirms that our sensations and preconceptions and our feelings are the standards of truth ; the Epicureans generally make perceptions of mental presentations b to be also standards. His own statements are also to be found in the Summary addressed to Herodotus and in the Sovran Maxims. I Every sensation, he says, is devoid of reason and incapable of memory ; for neither is it self-caused nor, regarded as having an external cause, can it add anything thereto or take anything therefrom.' Nor is there anything which can refute sensations or convict them of error : one sensation cannot convict another and kindred sensation, for they are equally valid ; nor can one sensation refute another which is not kindred but heterogeneous, for the objects which the two senses judge are not the same c ; nor again can reason refute them, for reason is wholly dependent on sensation ; nor can one sense refute another, since we pay equal heed to all. And the reality of separate perceptions guarantees d the truth of our senses. But seeing and hearing are just as real as feeling pain. Hence it is from plain facts that we must start when we draw inferences about the unknown. e For all our notions are derived from e More precisely d57?\oi/ = that which does not come within the range of sense. Compare e.g. § 38 to irpoaixevov Kai to a07)\oi>, and the way in which the conception of void is obtained in § 40. In § 62 it is called to Trpoo-do^a^bjxevov irepl tov dopaTov. vol.ii 2o 56