This triple division has been very popular in France, and to some extent in other countries. It is not, and it was not in its inception, a scientific deduction from observed facts, but was a sort of _a priori_ hypothesis based on the physiological theories of Bichat, and at a later day derived support from the philosophic dreams of Auguste Comte. Bichat, for instance, had recognized three fundamental physiological systems in man—the vegetative or visceral, the osso-muscular, and the cerebro-spinal. The anthropologists, in turn, considered it a happy thought to divide the human species into three races, each of which should show the predominance of one or other of these systems. Thus the black race was to show the predominance of the vegetative system; the yellow race, the osso-muscular system; the white race, the nervous system.[36] As Bichat had not discovered any more physiological systems, so there could be no more human races on the earth: and thus the sacred triplets of the Comtian philosophy could be vindicated. How little value attaches to any such generalizations you will readily perceive, and you will be prepared, with me, to dismiss them all, and to turn to the facts of the case, inquiring whether there are any traits of the red race which justify their being called “Mongolian” or “Mongoloid.” Such affinities have been asserted to exist in language, in culture, and in physical peculiarities, and I shall take these up _seriatim_ for examination. First, as to language. The great Mongolian stock is divided into the southern branch, speaking monosyllabic, isolating languages, and the northern branch, whose dialects are polysyllabic and agglutinating. The latter are sometimes called Turanian or Ural-Altaic; and as they are geographically contiguous to the Eskimo, and almost to the Athabascans, we might reasonably expect the linguistic kinship, if any exists, to be shown in this branch of Mongol speech. Is such the case? Not in the least. To prove it, I think it enough to quote the positive statement of the best European authority on the Ural-Altaic languages, Dr. Heinrich Winkler. He emphatically says, that, in the present state of linguistic science, not only is there no connection apparent between any Ural-Altaic and any American language, but that such connection is shown to be highly improbable. The evidence is all the other way.[37] I need not, therefore, delay over this part of my subject, but will proceed to inquire whether there are any American affinities to the monosyllabic, isolating languages of Asia.