We do not think solely by concepts or images; we think, first of all, according to Mr Bergson's expression, by dynamic schemes. What is a dynamic scheme? It is motive rather than representative, inexpressible in itself, but a source of language containing not so much the images or concepts in which it will develop as the indication of the path to be followed in order to obtain them. It is not so much system as movement, progress, genesis; it does not mark the gaze directed upon the various points of one plane of deliberate contemplation so much as an effort to pass through successive planes of thought in a direction leading from intuition to analysis. We might define it by its function of calling up images and concepts, representations which, for one and the same scheme, are neither strictly determined nor anything in particular in themselves, concurrent representations which have in common one and the same logical power. The representations called up form a body to the scheme, and the relation of the scheme to the concepts and images which it calls up resembles, mutatis mutandis, the relation pointed out by Mr Bergson between an idea and its basis in the brain. In short, it is the very act of creative thought which the dynamic scheme interprets, the act not yet fixed in "results." Nothing is easier than to illustrate the existence of this scheme. Let us merely remark a few facts of current observation. Recall, for example, the suggestive anxiety we experience when we seek to remember a name; the precise syllables of the name still escape us, but we feel them approaching, and already we possess something of them, since we immediately reject those which do not answer to a certain direction of expectancy; and by endeavouring to secure a more intimate feeling of this direction we suddenly arouse the desired recollection. In the same way, what does it mean to have the sense of a complex situation in active life, if not that we perceive it, not as a static group of explicit details, but as a meeting of powers allied or hostile, convergent or divergent, directed towards this or that, of which the aggregate whole tends of itself to awaken in us the initial reactions which analyse it? In the same way again, how do we learn, how can we assimilate a vast system of conceits or images? Our task is not to concentrate an enumerative attention on each individual factor; we should never get away from them, the weight would be too heavy. What we entrust to memory is really a dynamic scheme permitting us to "regain" what we should not have succeeded in "retaining." In reality our only "knowledge" is through such a scheme, which contains in the state of potential implication an inexhaustible multiplicity ready to be developed in actual representations. How, finally, is any discovery made? Finding is solving a problem; and to solve a problem we must always begin by supposing it solved. But of what does such a hypothesis consist?