only defined and distinguished by its coalescence with a memory-image, which we send forth to meet it. Only thus is attention secured, and without attention there is but a passive juxtapositing of sensations, accompanied by a mechanical reaction. But, on the other hand, as we shall show later, the memory-image itself, if it remained pure memory, would be ineffectual. Virtual, this memory can only become actual by means of the perception which attracts it. Powerless, it borrows life and strength from the present sensation in which it is materialized. Does not this amount to saying that distinct perception is brought about by two opposite currents, of which the one, centripetal, comes from the external object, and the other, centrifugal, has for its point of departure that which we term 'pure memory'? The first current, alone, would only give a passive perception with the mechanical reactions which accompany it. The second, left to itself, tends to give a recollection that is actualized—more and more actual as the current becomes more marked. Together, these two currents make up, at their point of confluence, the perception that is distinct and recognized.
This is the witness of introspection. But we have no right to stop there. Undoubtedly there is considerable risk in venturing, without sufficient evidence, into the obscure problems of cerebral localization. But we have said that to separate from one another the completed per-
