except when they directly affect the reproductive cells, has not been proved. On the other hand, as we shall presently show, there is much reason for believing that such acquired characters are in their nature non-heritable.
Variation and Selection Overpower the Effects of Use and Disuse.
But there is another objection to this theory arising from the very nature of the effects produced. In each generation the effects of use or disuse, or of effort, will certainly be very small, while of this small effect it is not maintained that the whole will be always inherited by the next generation. How small the effect is we have no means of determining, except in the case of disuse, which Mr. Darwin investigated carefully. He found that in twelve fancy breeds of pigeons, which are often kept in aviaries, or if free fly but little, the sternum had been reduced by about one-seventh or one-eighth of its entire length, and that of the scapula about one-ninth. In domestic ducks the weight of the wing-bones in proportion to that of the whole skeleton had decreased about one-tenth. In domestic rabbits the bones of the legs were found to have increased in weight in due proportion to the increased weight of the body, but those of the hind legs were rather less in proportion to those of the fore legs than in the wild animal, a difference which may be imputed to their being less used in rapid motion. The pigeons, therefore, afford the greatest amount of reduction by disuse—one-seventh of the length of the sternum. But the pigeon has certainly been domesticated four or five thousand years; and if the reduction of the wings by disuse has only been going on for the last thousand years, the amount of reduction in each generation would be absolutely imperceptible, and quite within the limits of the reduction due to the absence of selection, as already explained. But, as we have seen in Chapter III, the fortuitous variation of every part or organ usually amounts to one-tenth, and often to one-sixth of the average dimensions—that is, the fortuitous variation in one generation among a limited number of the individuals of a species is as great as the cumulative effects of disuse in a thousand generations! If we assume that the effects of use or of effort in the individual are equal to the effects of disuse, or even ten or a hundred times greater, they
