But while scientific anthropologists have long given up the idea that, if we want to know the condition of primitive man, we must study it among the Fuegians or Eskimos, the subject has lost none of its charms. It is, no doubt, a very amusing occupation to run through the books of modern and ancient travellers, traders, or missionaries, to mark with pencil a strange legend here, and an odd custom there, to point out a similarity between a Shâman and an Archbishop, between a Hottentot and Homer. This kind of work can be done in the intervals of more serious studies, and if it is done with the facile pen of a journalist or the epigrammatic eloquence of a young lawyer, nothing can be more delightful. But it is dangerous work—so dangerous that the prejudice that has lately arisen among scientific anthropologists against Agriology seems justified, at least to a certain extent. There are truly scholarlike works on savages. I say scholarlike intentionally, because they are based on a scholarlike study of the languages spoken by the races whose mental organisation has to be analysed. The works of Bishops Callaway and Caldwell, of Brinton and Horatio Hale, of Gill, Bleek, and Hahn, the more general compilations of Waitz, Tiele, Lubbock, Tylor, and Reville, the clever contributions of A. Lang, John Fiske, and others, are but the first that occur to my mind as specimens of really useful work that may be done in this line. But the loose and superficial appeals to savages as the representatives of a brand-new humanity, fresh from the hands of the potter, the ignorant attempts at explaining classical myths from Melanesian tattle, the wild comparisons of Hebrew customs with the outrages of modern cannibals, have at last met with their well-merited reward, and the very name of savage is gradually disappearing from the best works on anthropology and philosophy. And yet there are savages, only we must distinguish. There are, as I pointed out long ago, two classes of savages, to say nothing of minor subdivisions—namely, _progressive_ and _retrogressive_ savages. There is a hopeful and a hopeless barbarism, there is a growing and a decaying civilisation. We owe a great deal to the Duke of Argyll, particularly in his last great work, _The Unity of Nature_, for having laid so much stress on the fact that of all works of nature man is the one most liable to two kinds of evolution, one ascending and the other descending. Like the individual, a whole family, tribe, or race of men may, within a very short time, rise to the highest pitch of virtue and culture, and in the next generation sink to the lowest level of vice and brutality.