"Science" works with several types of "law." The most frequent and useful type is that of the "elementary law,"--that of the composition of forces, that of gravitation, of refraction, and the like. Such laws declare no concrete facts to exist, and make no prophecy as to any actual future. They limit themselves to saying that if a certain character be found in any fact, another character will co-exist with it or follow it. The usefulness of these laws is proportionate to the extent to which the characters they treat of pervade the world, and to the accuracy with which they are definable. Statistical laws form another type, and positively declare something about the world of actuality. Although they tell us nothing of the elements of things, either abstract or concrete, they affirm that the resultant of their actions drifts preponderantly in a particular direction. Population tends toward cities; the working classes tend to grow discontented; the available energy of the universe is running down--such laws prophesy the real future _en gros_, but they never help us to predict any particular detail of it. Spencer's law of Evolution is of the statistical variety. It defines what evolution means, and what dissolution means, and asserts that, although both processes are always going on together, there is in the present phase of the world a drift in favor of evolution. In the first edition of "First Principles" an evolutive change in anything was described as the passage of it from a state of indefinite incoherent homogeneity to a definite coherent heterogeneity. The existence of a drift in this direction in everything Mr. Spencer proves, both by a survey of facts, and by deducing it from certain laws of the elementary type, which he severally names "the instability of the homogeneous," "the multiplication of effects," "segregation," and "equilibration." The two former insure the heterogeneity, while "segregation" brings about the definiteness and coherence, and "equilibration" arrests the process, and determines when dissolutive changes shall begin. The whole panorama is resplendent for variety and inclusiveness, and has aroused an admiration for philosophy in minds that never admired philosophy before. Like Descartes in earlier days, Spencer aims at a purely mechanical explanation of Nature. The knowable universe is nothing but matter and motion, and its history is nothing but the "redistribution" of these entities. The value of such an explanation for scientific purposes depends altogether on how consistent and exact it is. Every "thing" must be interpreted as a "configuration," every "event" as a change of configuration, every predicate ascribed must be of a geometrical sort. Measured by these requirements of mechanics Spencer's attempt has lamentably failed. His terms are vagueness and ambiguity incarnate, and he seems incapable of keeping the mechanical point of view in mind for five pages consecutively.