I take it, then, that the burden of proof lies with Professor Lloyd Morgan to show why the withdrawal of selection is _not_ sufficient to account for degeneration any further than the mediocrity-level in the former presence of selection. Why does "the strong tendency[146] to deterioration demand an explanation," further than the fact that when all variations below the average in every generation are allowed to survive, they must gradually lower the average itself through a series of generations? To answer that any such tendency "would have been bred out of the race" by the previous action of selection, is to suppose that the function of selection is at an end when once it has built up a structure to the highest point of working efficiency,--that the presence of selection is no longer required to _maintain_ the structure at that point. But it is enough to ask in reply--Why, under the cessation of selection, does _complexity_ of structure degenerate so much more rapidly than _size_ of structure? Why is it, for instance, that "the eyes of crustacea" in dark caves have entirely disappeared, while their foot-stalks (when originally present) still remain? Can it be maintained that "for hundreds of generations" natural selection was more intent on developing the foot-stalks than the eyes which were mounted upon them--so that while the latter were left by selection with "a strong tendency to deterioration," the former have had this tendency "bred out in the race"[147]? [146] No one has supposed that the tendency need be "strong": it has only to be persistent.