Now, what would be the result of this alternation of good times and hard times? During good times varieties, which would be otherwise unable to hold their own, might arise and have time to establish themselves. In an expanding area migration would take place, local segregation in the colonial areas would be rendered possible, differential elimination in the different migration-areas would produce divergence. There would be diminished elimination of neutral variations, thus affording opportunities for experimental combinations. In general, good times would favour variation and divergence. Intermediate between good times and hard times would come, in logical order, the times in which there is neither an expansion nor a contraction of the life-area. One may suppose that these are times of relatively little change. There is neither the divergence rendered possible by the expansion of life-area, nor the heightened elimination enforced by the contraction of life-area.[BA] Elimination is steadily in progress, for the law of increase must still hold good. Divergence is still taking place, for the law of variation still obtains. But neither is at its maximum. These are the good old-fashioned times of slow and steady conservative progress. They are, perhaps, well exemplified by the fauna of the Carboniferous period, and it is not at all improbable that we are ourselves living in such a quiet, conservative period. On the other hand, hard times would mean increased elimination. During the exhibitions at South Kensington there were good times for rats. But when the show was over, there followed times that were cruelly hard. The keenest competition for the scanty food arose, and the poor animals were forced to prey upon each other. "Their cravings for food," we read in _Nature_, "culminated in a fierce onslaught on one another, which was evidenced by the piteous cries of those being devoured. The method of seizing their victims was to suddenly make a raid upon one weaker or smaller than themselves, and, after overpowering it by numbers, to tear it in pieces." Elimination by competition, passing in this way into elimination by battle, would, during hard times, be increased. None but the best organized and best adapted could hope to escape. There would be no room for neutral variations, which, in the keenness of the struggle, would be relatively disadvantageous. Slightly divergent varieties, before kept apart through local segregation, would be brought into competition. The weakest would in some cases be eliminated. In other cases, the best-adapted individuals of each variety might survive. If their experiments in intercrossing, should such occur, gave rise to fertile offspring, more vigorous and better adapted than either parent-race, these would survive, and the parent-forms would be eliminated. But if such experiments in intercrossing gave rise to infertile, weakly offspring, these would be eliminated. Thus sterility between species would become fixed.