2. The races of men are deficient in some of the essential characters of species. It is true that they are reproduced with considerable permanency; though a great many cases of spontaneous change, of atavism, or return to the character of progenitors, and of slow variation under changed conditions, have been recorded. But the most manifest deficiency in true specific characters is in the invariable shading-off of one race into another, and in the entire failure of those who maintain the distinction of species in the attempt accurately to define their number and limits. The characters run into each other in such a manner that no natural arrangement based on the whole can apparently be arrived at; and when one particular ground is taken, as color, or shape of skull, the so-called species have still no distinct limits; and all the arrangements formed differ from each other, and from the deductions of philology and history. Thus, from the division of Virey into two species, on the entirely arbitrary ground of facial angle, to that of Bory de St. Vincent into fifteen, we have a great number and variety of distinctions, all incapable of zoological definition; or, if capable of definition, eminently unnatural. There are, in short, no missing links between the varieties of men corresponding to that which obtains between man and lower animals. 3. The races of men differ in those points in which the higher animals usually vary with the greatest facility. The physical characters chiefly relied on have been color, character of hair, and form of skull, together with diversities in stature and general proportion. These are precisely the points in which our domestic races are most prone to vary. The manner in which these characters differ in the races of men may be aptly illustrated by a few examples of the arrangements to which they lead.