applies to birds, for Darwin was informed by the Rev. W.D. Fox that his flocks of white and Chinese geese kept distinct.[61] This constant preference of animals for their like, even in the case of slightly different varieties of the same species, is evidently a fact of great importance in considering the origin of species by natural selection, since it shows us that, so soon as a slight differentiation of form or colour has been effected, isolation will at once arise by the selective association of the animals themselves; and thus the great stumbling-block of "the swamping effects of intercrossing," which has been so prominently brought forward by many naturalists, will be completely obviated. If now we combine with this fact the correlation of colour with important constitutional peculiarities, and, in some cases, with infertility; and consider, further, the curious parallelism that has been shown to exist between the effects of changed conditions and the intercrossing of varieties in producing either an increase or a decrease of fertility, we shall have obtained, at all events, a starting-point for the production of that infertility which is so characteristic a feature of distinct species when intercrossed. All we need, now, is some means of increasing or accumulating this initial tendency; and to a discussion of this problem we will therefore address ourselves. _The Influence of Natural Selection upon Sterility and Fertility._ It will occur to many persons that, as the infertility or sterility of incipient species would be useful to them when occupying the same or adjacent areas, by neutralising the effects of intercrossing, this infertility might have been increased by the action of natural selection; and this will be thought the more probable if we admit, as we have seen reason to do, that variations in fertility occur, perhaps as frequently as other variations. Mr. Darwin tells us that, at one time, this appeared to him probable, but he found the problem to be one of extreme complexity; and he was also influenced against the view by many considerations which seemed to render such an origin of the sterility or infertility of species when intercrossed very improbable. The fact that species which occupy distinct areas, and which nowhere come in contact with each other, are often sterile when crossed, is one of the difficulties; but this may perhaps be overcome by the consideration that, though now isolated, they may, and often must, have been in contact at their origination. More important is the objection that natural selection could not possibly have produced the difference that often occurs between reciprocal crosses, one of these being sometimes fertile, while the other is sterile. The extremely different amounts of infertility or sterility between different species of the same genus, the infertility often bearing no proportion to the difference between the species crossed, is also an important objection. But none of these