perception and memory. Our perceptions are undoubtedly interlaced with memories, and inversely, a memory, as we shall show later, only becomes actual by borrowing the body of some perception into which it slips. These two acts, perception and recollection, always interpenetrate each other, are always exchanging something of their substance as by a process of endosmosis. The proper office of psychologists would be to dissociate them, to give back to each its natural purity; in this way many difficulties raised by psychology, and perhaps also by metaphysics, might be lessened. But they will have it that these mixed states, compounded, in unequal proportions, of pure perception and pure memory, are simple. And so we are condemned to an ignorance alike of pure memory and of pure perception; to knowing only a single kind of phenomenon which will be called now memory and now perception, according to the predominance in it of one or other of the two aspects; and, consequently, to finding between perception and memory only a difference in degree and not in kind. The first effect of this error, as we shall see in detail, is to vitiate profoundly the theory of memory; for if we make recollection merely a weakened perception we misunderstand the essential difference between the past and the present, we abandon all hope of understanding the phenomena of recognition, and, more generally, the mechanism of the unconscious. But, in-
