I easily comprehend how careful we ought to be to employ in Anthropology the term _species_. It can scarcely be used with certainty until science has clearly circumscribed the limits of each species of men. This moment is not come yet, and may, perhaps, never arrive, for, in the midst of constant changes produced by crossing, migrations, and conquests, and with the certainty that several races, or a great number of them, have disappeared within historical time,[15] it seems impossible to appreciate the degree of purity of certain races, to discover their origin, to know whether they are autochthonic or exotic, whether they belonged originally to this or that Fauna, and re-establish the Ethnology of our planet as it was in the beginning. To fix the number of primitive species of men, or even the number of actual species, is an insoluble problem to us, and probably to our successors. The attempts of Desmoulins et Bory de Saint Vincent have only produced imperfect sketches, which have led to contradictory classifications, where the number of arbitrary divisions is nearly equal to more natural divisions. The term species has, in classical language, an absolute sense, implying both the idea of a special conformation and special origin, and if some races--the Australians, for instance--unite these conditions in a sufficient degree, to constitute a clearly marked species, many other pure or mixed races escape, in this respect, a rigorous appreciation. It is for these reasons that many polygenists, after having proclaimed the multiplicity of the origins of humanity, and having recognised the impossibility of determining the number and the characters of the primitive stocks, have justly avoided methodically to divide the human genus into species. Many among them, however, who thought that they were, nevertheless, bound to establish divisions, have committed the error to accept the basis of the classification of the monogenists, and, like them, to establish five chief human families, and, like them, to admit that the individuals of each family are issued from a common trunk, with this difference, that, whilst the monogenists assume that the five primary trunks have proceeded from the same stock, and have the same roots, the pentagenists (if we may use this term) assume five distinct and independent stocks. Logically speaking, it would have been requisite to term the five fundamental races of the monogenists _species_, but it is easy to perceive that, for many reasons, the term species cannot be employed here in an absolute sense. The pentagenists have felt this, and, for want of a better term, use the word _race_, which has thus been diverted from its real acceptation.