But the matter is by no means so simple as may appear from this statement. In the first place, it is a matter of common observation that all the individuals of a species are never injured in the same part of the body at the same time. In those cases in which it is known that a special part is often injured, an examination has shown that there are not more than ten per cent of individuals that are injured at any one time, and in the case of the vast majority of animals this estimate is much too great. Thus there will be very little chance for competition of the injured individuals in each generation with each other, and the effects that are imagined to be gained as a result would be entirely lost by crossing with the uninjured individuals. But it is not necessary to consider this possibility, since there is another fact that shows at once that the power to regenerate could not have been gained through selection. The number of uninjured individuals in each generation will be much greater than the injured ones, and these will have so great an advantage over the injured individuals that, if competition approached the degree assumed by the selectionists, the injured individuals should be exterminated. A slight advantage gained through better powers of regeneration would be of little avail in competition, as compared with the competition with the uninjured individuals. Since selection is powerless to accomplish its end without competition, and since with competition all the injured individuals would be eliminated, it is clear that an appeal cannot be made to selection to explain the power of regeneration. In many cases the power of regeneration could not have been slowly acquired through selection, since the intermediate steps would be of no use. Unless, for example, a limb regenerated from the beginning almost completely, the result would be of no use to the animal. If the limb did regenerate completely the first time it was injured, then the selection hypothesis becomes superfluous. There are also a few cases known in which a process of regeneration takes place that is of no use to the animal. If, for instance, the earthworm (_Allolobophora fœtida_) be cut in two in the middle, the posterior piece regenerates at its anterior cut end, not a head, but a tail. Not by the widest stretch of the imagination can such a result be accounted for on the selection theory. Again, we find the reverse case, as it were, in certain planarians. If the head of _Planaria lugubris_ is cut off just behind the eyes, there develops at the cut surface of this head-piece another head turned in the opposite direction. Here again we have the regeneration of a perfect structure, but one that is entirely useless to the individual. The development of an antenna in place of an eye in the shrimp, when the eye stalk is cut off near its base, is another instance of the occurrence of a perfectly constant process, but one that is of no use to the organism.