GREEN'S METAPHYSICS OF KNOWLEDGE. 79 which he gives of Intellectual Idealism, this side of his doc- trine must not be lost sight of. Considering how fundamental a portion of his system this doctrine is, it certainly seems to me to have been treated by the author with less fulness than its importance, and with less clearness than its difficulty, renders desirable. The official proof, which will be found in chap, i., 26-9 of the Prolegomena, may be summarised as follows. It being granted that the world of experience consists in " a single, all inclusive system of relations," the question remains " what is the condition f its possibility ? " " What is implied in there being such a world ? :> Now " relation involves the existence of many in one ". One thing is only one thing because it involves many relations, one relation is only one relation in so far as it is between manifold things. "But a plurality of things cannot of themselves unite in one relation, nor can a single thing of itself bring itself into a multitude of relations." ..." There must, then, be something other than the manifold things themselves which com- bine them without effacing their severally. . . . With such a com- bining agency we are familiar as our intelligence." A relation of succes- sions, for instance, between two sensations is only possible if on the one hand they are united, and if on the other hand the distinct being of each is maintained. " But if it were not for the action of something which is not either of them or both together, there would be no alternative between their separateness and their fusion," and " the same or an analogous action is necessary to account for any relation whatever ". Hence the world of relations, if it is to be real for us, must be the " product of our combining intelligence " ; and " if it is to be real otherwise than merely for us," it must be " the product of some unifying principle analogous to that of our understanding ". Xow whether Intellectual Idealism be true or not there seems to be the gravest objection to this mode of establishing it. Related objects must be, it appears, at the same time both many and one. Left to themselves, however, and without the assistance of some " combining agency," they would be condemned either to be united in a " featureless identity," or separated in a rigid isolation. The one never would become many, nor the many one. If we are to avoid such a consummation a " combining agency " of some kind has therefore to be* found, and as none presents itself adequate to the task except intelligence, by intelligence it must be accomplished. There certainly appears to me to be something eminently unsatisfactory in this method of procedure. The philosopher is made to look about for some- thing which is involved in the existence of a connected system of nature, much as a geologist looks about for what is involved in the existence of an ancient moraine. And as the geologist assumes a glacier as the only thing which renders a moraine possible, so the philosopher assumes a
