physiological peculiarity. Thus, incipient species which were infertile together would have an advantage over other incipient species which were fertile, and, whenever the struggle for existence became severe, would prevail over them and take their place. Such infertility, being correlated with constitutional or structural differences, would probably, as already suggested, go on increasing as these differences increased; and thus, by the time the new species became fully differentiated from its parent form (or brother variety) the infertility might have become as well marked as we usually find it to be between distinct species.
This discussion has led us to some conclusions of the greatest importance as bearing on the difficult problem of the cause of the sterility of the hybrids between distinct species. Accepting, as highly probable, the fact of variations in fertility occurring in correlation with variations in habits, colour, or structure, we see, that so long as such variations occurred only sporadically, and affected but a small proportion of the individuals in any area, the infertility could not be increased by natural selection, but would tend to die out almost as fast as it was produced. If, however, it was so closely correlated with physical variations or diverse modes of life as to affect, even in a small degree, a considerable proportion of the individuals of the two forms in definite areas, it would be preserved by natural selection, and the portion of the varying species thus affected would increase at the expense of those portions which were more fertile when crossed. Each further variation towards infertility between the two forms would be again preserved, and thus the incipient infertility of the hybrid offspring might be increased till it became so great as almost to amount to sterility. Yet further, we have seen that if several competing species in the same area were being simultaneously modified, those between whose varieties infertility arose would have an advantage over those whose varieties remained fertile inter se, and would ultimately supplant them.
The preceding argument, it will be seen, depends entirely upon the assumption that some amount of infertility characterises the distinct varieties which are in process of differentiation into species; and it may be objected that of
