A COMPENDIOUS CLASSIFICATION OF THE SCIENCES. 33 This is partly but not fully admitted by Mr. Spencer in relation to his own scheme when he says that a true classification of the sciences ought to be figured in three dimensions, and not on a surface. For not only do his tables, as he himself notes, exclude subjective psychology, which he regards as coextensive with all the objective sciences and antithetical to them ; but, more than this, the use of a model in three dimensions would not enable him to bring it in. The present adaptation of Comte's scheme to a more metaphysical doctrine and indeed the original scheme itself does not seem to be necessarily in rivalry with Mr. Spencer's. When it is recognised that every diagrammatic representation must be inadequate, the two classifications may very well be taken as expressions of different points of view. For philo- sophical use, Comte's point of view has this advantage. It brings out clearly that the sciences, in their ideal order, form a single organism of knowledge to which each is subservient. Mr. Spencer's scheme, on its side, brings out what is also a perfectly real aspect of science ; namely, its tendency to branch into divergent specialties, which arrange themselves, like groups of organisms at the termination of a process of biological evolution. This, however, is a less important aspect for the philosopher. And to keep it primarily in view seems less conducive to the reception of science into the system of general culture. When the sciences are thought of as organically related to a whole, the advantages of the circular arrangement are easy to see. For this by no means indicates a definitively closed system. On the contrary, it might have served as the least inadequate representation from the time when cosmic science or philosophy first began vaguely to differen- tiate into particular sciences. New sciences would thus be: seen introducing themselves in accordance with that process of " intussusception" by which a biological organism grows, and which Kant regarded as the true process of develop- ment for an architectonic system of knowledge. This, and not the direct historical succession of the sciences in agree- ment with their logical order, has been the real course of intellectual history. The supposition that the logical order of the sciences and the historical order in which they become "positive" are one and the same, is a defect in Comte's classification as it stands ; though, as may now be seen, it is unessential to the use of it. There is no diffi- culty indeed in fixing arbitrarily the time when a science is positively constituted, and thus making the two orders 3
