nervous system consists of separate masses; Articulates have a hard external skeleton and jointed limbs, and their nervous system consists of two long ventral cords; Radiates have ill-defined nervous and muscular systems, and in their lowest forms possess the animal functions without the animal organs. This well-rounded classification of animal forms is in a sense the crown of Cuvier's work, for the principle of the subordination of characters, in the interpretation which he gives to it, is a direct application of his principle of functional correlation. Each of the great groups is built upon one plan. The idea of the unity of plan has become for Cuvier a commonplace of his thought, and it is tacitly recognised in all his anatomical work. But he never takes it as a hard-and-fast principle which must at all costs be imposed upon the facts. Cuvier has become known as the greatest champion of the fixity of species, but it is not often recognised that his attitude to this problem is at least as scientific as that of the evolutionists of his own and later times. No doubt he became dogmatic in his rejection of evolution-theory, but he was on sure ground in maintaining that the evolutionists of his day went beyond their facts. He considered that certain forms (species) have reproduced themselves from the origin of things without exceeding the limits of variation. His definition of a species was, "the individuals descended from one another or from common parents, together with those that resemble them as much as they resemble one another."[62] "These forms are neither produced nor do they change of themselves; life presupposes their existence, for it cannot arise save in organisations ready prepared for it."[63] He based his rejection of all theories of descent upon the absence of definite evidence for evolution. If species have gradually changed, he argued, one ought to find traces of these gradual modifications.[64] Palaeontology does not furnish such traces. Again, the limits of variation, even under domestication, are narrow, and the most extreme variation does not fundamentally alter the specific type. Thus the dog has varied perhaps most of all, in size, in shape, in colour. "But throughout all these variations the relations of the bones remain the same, and the form of the teeth never changes to an appreciable extent; at most there are some individuals in which an additional false molar develops on one side or the other."[65] This second objection is the objection of the morphologist. It would be an interesting study to compare Cuvier's views on variation with those of Darwin, who was essentially a systematist.