Page:Mind (New Series) Volume 6.djvu/174

 158 L. T. HOBHOUSE : we have just said it will be clear that our two forms of conception serve in practice as the basis of two very different types of classification. Take a system M in which a com- mon element A works out under modifications a, yS, y, etc., into the forms A a = a, A^ = b, and so on. Here we have, or may have, a hierarchy of classes in which as we descend the scale every new species is formed by a fresh modification of the common element. The classification is so far natural that it exhibits the real structure of the species composing it. And it wants but a step to pass into a genuine explana- tion. If A as unmodified is a reality, and if a, &, etc., can be regarded as real processes modifying A, the classification is in itself an explanation of the genesis of the diverse species. A would then figure in two senses, as a concrete being at the beginning of a process and as an abstract character running through the evolving forms. We are not likely to find an instance of so simple a type. It is perhaps most nearly realised in the sphere of human purposes. But normally the most primitive form of a content has a definite character of its own, something of which it loses in passing to a richer and fuller phase of development. Primitive society loses something as it develops into a higher form. The protozoon is generically, I suppose, the unicellular modification of the organic type. Still the classification once made it is relatively easy to place the species in a possible genetic order, and by verifying the historical charac- ter of the genesis to translate the classificatory system into an explanation of development. Further, such a system may throw light on genesis even where its hierarchical arrangement bears no relation to temporal development. In such a system to know the factors in the formation of any one member is virtually to know them for all cases. Thus we may arrange forms of art or types of society in an order of genera and species, and apart from positive evidence of the development of one form from another light will be thrown on their origin, since the arrangement will show of itself how the conditions issuing in one type will, mutatis mutandis, be responsible for others. In contrast with all this stands the disjunctive classi- fication for which the abstract concept is responsible. It must needs be that disjunctive classifications be made. They serve our purposes, prevent confusion, and provide ideally a complete enumeration of types. They belong to the pre- liminaries of science, the arrangement of the material and the marking out of the subject-matter. But as science progresses they tend to pass into the other type. Thus I