to be too many of this sex or too few of that. And as the power of locomotion increased and the distribution of the sexes became more and more irregular, so the opportunity would be afforded for the development of any variation which would have tended to facilitate the process of pairing, and by so doing have conferred upon the individuals possessing it, some slight advantage over their fellows.
What would have been the most likely direction for variation to have taken? Any restriction upon the freedom of movement of both sexes would only have added to the difficulties of mating; but if restriction had been imposed upon one sex, whilst the other had been left free to wander, some order would have been introduced into the process. That the territory serves to restrict the movements of the males and to distribute them uniformly throughout all suitable localities, there can be no question; and since the instinctive behaviour in relation to it is timed to appear at a very early stage in the seasonal sexual process, the males are in a position to receive mates before the impulse to mate begins to assert itself in the female.
We will take the Ruff as an example. According to Mr Edmund Selous, pairing, in this species, is promiscuous—the Ruffs are polygamous, the Reeves polyandrous. Suppose, then, that upon this island of some few miles in circumference, whereon his investigations were made, the movements of neither Ruff
