of climatic variation in the present discussion is excluded. In particular, so far as any question of heredity is concerned, all these facts are as assuredly as they are cogently relevant. It is perfectly certain that there is "a large proportional number" of named species--particularly of plants--which further investigation would resolve into climatic varieties. With the advance of knowledge, "bad species" are always increasing at the expense of "good species," so that we are now justified in concluding with Kerner, Häckel, and other naturalists best qualified to speak on this subject, that if we could know as much about the past history and present relations of the remaining good species as we do about the bad, all the former, without exception, would become resolved into the latter. In point of fact, and apart altogether from the inductive experience on which this conclusion is based, the conclusion follows "as a necessary deduction" from the general theory of descent. For this theory essentially consists in supposing either the past or the present existence of intermediate varietal forms in all cases, with the consequence that "good species" serve merely to mark _lacunae_ in our knowledge of what is everywhere a finely graduated process of transmutation. Hence, if we place this unquestionably "necessary deduction" from the general theory of descent side by side with the alleged "necessary deduction" from the theory of natural selection, we cannot avoid the following absurdity--Whether or not a given form is to be regarded as necessarily due to natural selection, and all its characters necessarily utilitarian, is to be determined, and determined solely, by the mere accident of our having found, or not having found, either in a living or in a fossil state, its varietal ancestry.