We need not discuss the various suggestions that have been made to explain this difficulty, none of which, as Pearson points out, have been satisfactory. He himself believes that a process of segregation of like individuals must occur, during the incipient stages at least, in the formation of species. Afterwards a correlation may exist between the new organs and the germ-cells, of such a sort that a relative or an absolute sterility between the incipient species is attained. After this condition has been reached the two new species may freely intermix without a return to the primitive type, since they are no longer fertile _inter se_. It seems to me, also, that this would be an essential requisite if we assume that species are slowly formed out of races from individual differences, as Pearson supposes to be the case. There are, however, other possibilities that Pearson does not take into account, namely, that from the very beginning the change may be so great that the new form is not fertile with the original one; and there is also another possibility as well, that, although the new and the old forms are fertile, the hybrids may be like one or the other parent, as in several cases to be given later. Not that I mean to say that in either of these two ways can we really offer a solution of the question of infertility, for, from the evidence that we possess, it appears improbable that the infertility of species _inter se_ has been the outcome of either of these causes. In support of his main thesis Pearson gives certain data in respect to preferential mating in the human race. By this is meant that selection of certain types of individuals is more likely to take place, and also that the fertility of certain types of individuals is greater than that of other types. The calculations are based on stature, color of hair, and of eyes. The results appear to show in all cases examined that there is a slight tendency to form new races as the result of the more frequent selection of certain kinds of individuals. But even if this is the case, what more do the results show than that local races may be formed,—races having a certain mode for height, for color of eyes or of hair? That changes of this kind can be brought about we knew already without any elaborate measurements, yet we should not conclude from this that new species will be formed by a continuation of the process.