The world which seems to be spread out before us in space and time is, he tells us, a world of things _as they are revealed to our senses and our intelligence_; it is a world of manifestations, of phenomena. What things-in-themselves are like we have no means of knowing; we know only things as they appear to us. We may, to be sure, talk of a something distinct from phenomena, a something not revealed to the senses, but thought of, a _noumenon_; but we should not forget that this is a negative conception; there is nothing in our experience that can give it a filling, for our experience is only of phenomena. The reader will find an unmistakable echo of this doctrine in Herbert Spencer's doctrine of the "Unknowable" and its "manifestations." Now, Berkeley had called all the things we immediately perceive _ideas_. As we have seen, he distinguished between "ideas of sense" and "ideas of memory and imagination." Hume preferred to give to these two classes different names--he called the first _impressions_ and the second _ideas_. The associations of the word "impression" are not to be mistaken. Locke had taught that between ideas in the memory and genuine sensations there is the difference that the latter are due to the "brisk acting" of objects without us. Objects impress us, and we have sensations or impressions. To be sure, Hume, after employing the word "impression," goes on to argue that we have no evidence that there are external objects, which cause impressions. But he retains the word "impression," nevertheless, and his use of it perceptibly colors his thought. In Kant's distinction between phenomena and noumena we have the lineal descendant of the old distinction between the circle of our ideas and the something outside of them that causes them and of which they are supposed to give information. Hume said we have no reason to believe such a thing exists, but are impelled by our nature to believe in it. Kant is not so much concerned to prove the nonexistence of noumena, things-in-themselves, as he is to prove that the very conception is an empty one. His reasonings seem to result in the conclusion that we can make no intelligible statement about things so cut off from our experience as noumena are supposed to be; and one would imagine that he would have felt impelled to go on to the frank declaration that we have no reason to believe in noumena at all, and had better throw away altogether so meaningless and useless a notion. But he was a conservative creature, and he did not go quite so far. So far there is little choice between Kant and Hume. Certainly the former does not appear to have rehabilitated the external world which had suffered from the assaults of his predecessors. What important difference is there between his doctrine and that of the man whose skeptical tendencies he wished to combat?