{502} _Origin_, Ed. i. p. 137, vi. p. 170. In as far then as it is admitted as probable that the effects of disuse (together with occasional true and sudden abortions during the embryonic period) would cause a part to be less developed, and finally to become abortive and useless; then during the infinitely numerous changes of habits in the many descendants from a common stock, we might fairly have expected that cases of organs becom<ing> abortive would have been numerous. The preservation of the stump of the tail, as usually happens when an animal is born tailless, we can only explain by the strength of the hereditary principle and by the period in embryo when affected{503}: but on the theory of disuse gradually obliterating a part, we can see, according to the principles explained in the last chapter (viz. of hereditariness at corresponding periods of life{504}, together with the use and disuse of the part in question not being brought into play in early or embryonic life), that organs or parts would tend not to be utterly obliterated, but to be reduced to that state in which they existed in early embryonic life. Owen often speaks of a part in a full-grown animal being in an "embryonic condition." Moreover we can thus see why abortive organs are most developed at an early period of life. Again, by gradual selection, we can see how an organ rendered abortive in its primary use might be converted to other purposes; a duck's wing might come to serve for a fin, as does that of the penguin; an abortive bone might come to serve, by the slow increment and change of place in the muscular fibres, as a fulcrum for a new series of muscles; the pistil{505} of the marigold might become abortive as a reproductive part, but be continued in its function of sweeping the pollen out of the anthers; for if in this latter respect the abortion had not been checked by selection, the species must have become extinct from the pollen remaining enclosed in the capsules of the anthers. {503} These words seem to have been inserted as an afterthought. {504} _Origin_, Ed. i. p. 444, vi. p. 611. {505} This and similar cases occur in the _Origin_, Ed. i. p. 452, vi. p. 621.