individual all the vast numbers of variations which occurred among the millions of ancestors from the unicellular organism downwards can be recapitulated; and (2) as regards the highest animals, that it is obvious that the embryo in its various stages cannot present exact copies of its remote ancestors, for the simple reason, that in its various stages it is incapable of living apart from the parent. For instance, no animal that even approximately resembled the human embryo of two months could be capable of living outside the body of its mother. But these objections are met when we take into consideration what is undoubtedly true, viz. that variations occur, and that natural selection (reversed selection in this case), and particularly cessation of natural selection, act, not only at the end of the ontogeny, but during the entire period of development; natural selection during that period seizing upon as favourable and accumulating all such variations as tend to shorten and simplify the process of development, and cessation of natural selection tending to bring about the disappearance of all characters which during the phylogeny had been useful, but which to the embryo, living under different conditions, protected and sustained as it is within the body of its parent in the highest animals, or advantageously placed as regards the environment by the parent in the case of lower animals, are no longer useful.
One other point remains to be cleared up. It has been observed that the embryo of a high animal, e.g. man, resembles the embryos of lower animals, never the lower animals themselves; for instance, the embryo of a man in one stage of its ontogeny resembles the embryo of a frog, never the adult frog. The reason for this is obvious; man has not descended from the frog, but the frog and man have descended from a common ancestry, and therefore both animals as they recapitulate
