It must not be believed that all human races can with equal facility be ranged in either of these divisions; nor must we believe that the characteristic traits of one group are equally marked in all the dependent races; nor even that they are found combined in any of these races; nor, finally, that in the centre of each group we find a typical race in which all the characters have their maximum of development. This might be the case if all known races had descended from five primitive stocks, as admitted by several polygenists, or if, as many monogenists think, humanity, _one_ in the beginning, had soon afterwards been divided into five principal trunks, from which issued, as so many accessory branches, the numerous sub-divisions which constitute the secondary races. But there is no race which can pretend to personify within itself the type to which it belongs. This type is fictitious; the description is an ideal one, like the forms of the Apollo de Belvedere. Human types, like all other types, are merely abstractions, and in proportion as we attach more importance to this or that character, we obtain a more or less considerable number of types. Thus, Blumenbach had five, Cuvier only three, and Bérard describes fifteen types. This is also proved by the fact, that whilst many races attach themselves directly and evidently to a fixed type, there are others belonging to two very dissimilar types. Thus the Abyssinians are Caucasian in form and Ethiopian by colour. The description of the principal types is thus merely a methodical process, fit to facilitate, by the formation of a certain number of groups, the comparison of human races, and to simplify the partial description of each. This division has, moreover, the advantage of establishing for the greater part of the races, their degree of relative affinity or divergence. It even accords to a certain point with their primitive repartition upon the surface of the globe, which has permitted, without doing any violence to the facts, to distinguish the types by denominations borrowed from geography.[14]