I know full well that a wish has been felt to go further back. Under the more or less perceptible influence of transmutationist doctrines, terms of comparison in the estimation of these differences have too often been sought for among animals, and especially among apes. Eminent men, without even adopting these doctrines, frequently use the expressions, _simian character_, _animal character_. Why forget the embryo or the human fœtus? Why not remember even the infant? Question their history. It furnishes all the elements of a _human evolution theory_, certainly much more precise and true than the _simian theory_. This is again a result which will be made clear, I hope, by the facts which I shall have to mention. But whether or not I may be able to explain the appearance of the features which distinguish races from each other, and whatever origin may be attributed to them, I shall only take the word _character_ in the sense which is given to it in botany and zoology. III. An animal species is not characterised solely by the peculiarities manifested by its physical organism. No history of bees or ants omits to speak of their instincts, or to show how these differ in different species. With much stronger reason ought we to point out in the history of human races the characteristic points in their intellectual, moral, and religious manifestations. Of course, when approaching this order of facts, the anthropologist ought none the less to remain exclusively a naturalist. This very simple consideration is sufficient to determine the relative value which ought to be attributed in anthropology to characters of different orders. Here, as in botany and in zoology, the first place ought to be given to the most persistent characters. Now, a man, tribe, or an entire population can in a certain number of years change its social state, its language, religion, etc. They do not on that account modify their external or anatomical physical characters. It is therefore to the latter that the anthropologist will attach most importance, contrary to what the linguist, the philosopher, or the theologian would certainly do. Nevertheless we shall see that, in some very rare cases, the linguistic characters preponderate over the physical characters, in the sense that they furnish more striking indications on the subject of certain ethnical affinities. Considered from a physical point of view, man exhibits characters which may be divided into four distinct categories, namely: external characters, anatomical characters, physiological characters, and pathological characters.