334 CHAELES MEECIER: sideration. But those who do admit the application of evolution to this region of being are committed beforehand to an approval of the basis of my classification. That the details or even the grosser structure are correctly worked out is not claimed, is perhaps not even probable ; but it is claimed that evolutionists are logically bound to accept the principle of the classification, and that the burden of disproof lies with those who reject it. In the classification of animals and plants, the primary divisions are marked out by differences in some fundamental attribute by the mode of germination in the one case, and by the presence or absence of organs of profound impor- tance, as limbs, bloodvessels, and nervous system, in the other. The division of the primary into secondary groups follows the variation of some attribute of the primary group that is not only of secondary importance, but is, or may be wholly wanting in the other primary groups. The phanero- gamic plants, for instance, are divided according as their seed possesses one or more cotyledons organs that have no existence outside this group. The dicotyledonous plants are classified according to the number of their petals organs that no other kind of plants possess. The sub-orders of the leguminosae are determined by variations in the characters of the pod a form of fruit that is confined to this order. And the classification of animals proceeds in the same way. The inference is obvious. If the feelings have come into existence by the same process of evolution to which plants and animals owe their origin, and if the classification of the latter, following the lines of this process, exhibits certain characters, then it may be expected that the classification of the former, following the same process, will exhibit similar characters. When it is found that the mode of classifying the feelings that is here proposed does exhibit a well- marked similarity to the accepted mode of classifying animals and plants, it may be claimed as an additional indication that the principle on which it is founded is correct. It may be objected that we have now got two principles of classification, one according to the way the interaction is begun, the other according to its position in the evolutionary scale, and that these two principles are so widely different that the lines of division that they regulate cannot possibly coin- cide with or even approximate to each other ; and this must be admitted, but it does not invalidate the classification. A classification which follows the course of evolution is often and very aptly compared to the structure of a tree. It may be said that a tree forms a solid diagram of such a classifica-
