munity a considerable infusion of English or French or Spanish blood. And though, in such a mingling, the blood of one race and the language of another may preponderate, yet even this fact is not perplexing. Apart from history, the speech alone, rightly studied, will indicate with sufficient clearness the origin and the circumstances of the mixture.
A striking and indeed crucial test of the decisive value of language in ethnological classification is found in the case of Madagascar. In seeking the origin of its inhabitants we should naturally turn first to Africa; and there, in fact, we find, among the Nubians and the Hamitic tribes of the eastern coast, people bearing sufficient resemblance in shape, features, complexion, and hair, to the natives of Madagascar, to warrant the opinion of their relationship, in the absence of any stronger evidence to the contrary. Remembering, however, that the Arabians in early times had much intercourse with the great African island, we turn to their country and find in the tribes of Yemen a similar resemblance. We then, perhaps, consider how readily the swarthy and curly-haired Dravidians of Southern Hindostan might have found their way to Madagascar, with the help of the northeast monsoon. To decide from which of these probable sources the ancestors of the Madagascar natives were derived, we have recourse to their language; and we ascertain, beyond the shadow of a doubt, that they were neither Africans nor Arabians nor Dravidians, but Malays. To reach their new abode they had to cross the entire width of the Indian Ocean, a distance of three thousand miles. This origin is a fact which no ethnologist now thinks of questioning; and the only decisive evidence to establish it is the language of the islanders. It is true, that when we have ascertained this fact by the linguistic evidence, we find ample material in the character and customs of the people to confirm it; but without the positive test of language this subsidiary evidence would be altogether insufficient as proof of such derivation. No one who considers the case of Madagascar can reasonably doubt that in language, and language only, resides the true distinction of races.
From the great number and the marked peculiarities of the linguistic stocks of this continent, America may be considered to offer by far the best field for the study of scientific ethnology. This fact was early apparent to that remarkable group of philologists, among whom
- Against this comparison of the linguistic stocks with the chemical elements (which is offered, of course, merely as an illustration, and not as an exact parallel), it may be objected that, according to the latest theory, these elements are all merely allotropic forms of a single substance. But, in fact, if the truth of this theory should be established, it will only serve to make the force of the illustration still more striking. In a "vice-presidential address," on the "Origin of Languages," delivered before the section of Anthropology in the American Association for the Advancement of Science (and published in the "Proceedings" of the Association for 1886), I endeavored to show in what manner all the linguistic stocks have probably originated from a single primitive language. Both theories, it may be added, simply exemplify the tendency of science to trace back all varieties to unity.