istic hypothesis of an epiphenomenal consciousness. If they are only its occasion, I thereby suppose that they do not resemble it in any way, and so, depriving matter of all the qualities which I conferred upon it in my representation, I come back to idealism. Idealism and materialism are then the two poles between which this kind of dualism will always oscillate; and when, in order to maintain the duality of substances, it decides to make them both of equal rank, it will be led to regard them as two translations of one and the same original, two parallel and predetermined developments of a single principle, and thus to deny their reciprocal influence, and, by an inevitable consequence, to sacrifice freedom.
Now, if we look beneath these three hypotheses, we find that they have a common basis:The mistake is due to our believing that perception and memory are pure knowledge, whereas they point to action. all three regard the elementary operations of the mind, perception and memory, as operations of pure knowledge. What they place at the origin of consciousness is either the useless duplicate of an external reality or the inert material of an intellectual construction entirely disinterested: but they always neglect the relation of perception with action and of memory with conduct. Now, it is no doubt possible to conceive, as an ideal limit, a memory and a perception that are disinterested; but, in fact, it is towards action that memory and perception are turned; it is action that the body pre-
