This distinction governs the theory of Kant. It is more easy to point out the embarassments and inconsistencies into which it leads him, than to get rid of the distinction itself. Ordinary criticism of Kant, indeed, has not taken much heed of the distinction or of its perplexing results. It has been too busy in refuting his doctrine that laws of nature are derived from the understanding, to enquire closely into his view of the relation between nature, in his sense of the term, and things-in-themselves. It has been gaining apparent triumphs, due to a misunderstanding of the question at issue, over the strongest part of his system, while it has left the weakest unassailed. There have been abundant proofs of what was not in dispute, that our knowledge of laws of nature is the result of experience; but the question whether phenomena could be so related as to constitute the nature which is the object of our experience without the unifying action of understanding is seldom even touched. Given an experience of phenomena related to each other in one system—so related that, whatever an object is really, or according to the fulness of its relations, it is unalterably—it is easy to show that our knowledge of laws of nature is derived from it. Such experience in its most elementary form is already implicitly a knowledge that there are laws of nature, and only needs to be reflected on in order to become so explicitly. When it has become so explicitly, the development of the experience—through cognisance of relations of which there has previously been no experience, or of which the experience has not been reflected on—becomes a growing knowledge of what the laws of nature in particular are. (§ 31 ¶ 1)

But the derivation of knowledge from an experience of unalterably related phenomena is its derivation from objcts unalterably related in consciousness. If the relation of the objects were not a relation of them in consciousness, there would be no experience of it. The question then arises how a succession of feelings becomes such a relation of objects in consciousness. If a relation of objects existed or could be known to exist otherwise than for consciousness, this would not help to account for what has to be accounted for, which is wholly a process of consciousness. The feelings which succeed each other are no doubt due to certain related conditions, which are not feelings. But granting for the moment that these conditions and their relation exist independently of consciousness, in accounting for a multitude of feelings they do not account for the experience of related objects. Of two objects which form the terms of a relation one cannot exist as so related without the other, and therefore cannot exist before or after the other. For this reason the objects between which a relation subsists, even relation of succession, are, just so far as related, not successive. In other words, a succession always implies something else than the terms of the succession, and that a something else which can simultaneously present to itself objects as existing not simultaneously but one before the other. (§ 31 ¶ 2)