Psychology has long since dispelled the scientific illusion of idealism. We do not wish to recall the pitiful failure of all the attempts formerly made, and even in our own times, to deduce _a priori_ the laws of the physical world. Everywhere, in this domain, the method of observation has superseded the deductive method. The reason of it is simple. An idea, however lofty, can only give out what it contains, _i.e._ other ideas. We know very well that our ideas are in our mind, but they are only in it in the state of ideas. How do we know that the objects which they represent exist outside ourselves? Only by logic can we pass from the idea of a thing to the external reality of that thing. Experience is necessary. Without it our ideas are empty forms. One may conjure with them for ever without ever reaching anything objective. They are shells without kernels. Pure idealism, so far from furnishing a solid theory of knowledge, ends in scepticism, _i.e._ in the negation of knowledge. The excesses and failures of idealist theories of knowledge have always given rise in history to the opposite theory of sensualist nominalism, according to which our ideas are simply transformed sensations. Unhappily, sensualism, in laying down this axiom, never explained the nature and still less the cause of that marvellous transformation. "There is nothing in the understanding," said Locke, "that was not previously in the senses." To which Leibniz rightly replied: "Except the understanding itself;" that is to say, the force which from sensation draws knowledge. By suppressing this ideal principle, you remove from science all element of necessity--that is to say, all general worth. With Hume, the sensualist theory, so far from giving an account of knowledge, ended in pure phenomenalism, _i.e._ once more, in scepticism. It is, in fact, with isolated sensation as with pure idea; you may press it as much as you will, you will never get out of it anything but what it contains--that is to say, contingencies without any connection between each other. Materialism is still more embarrassed to furnish any theory whatever of knowledge, for it does not even succeed in explaining sensation. Between a mechanical movement and a phenomenon of consciousness there is an impassable abyss. One of the most evident marks of the inferiority of the philosophy of French positivism is that it has not even approached this problem of knowledge, and that it has been able to constitute itself without any other than the popular psychology. 2. _The Kantian Theory of Knowledge_