Another strange circumstance is that the wing-bones have diminished _in length only_. The shortening is about 6 per cent. more than in the shortened legs, and it amounts to 11 per cent. as compared with the weight of the skeleton. Such a shortening should represent a reduction of 29 per cent. in weight, whereas the actual reduction in the weight of the wing-bones relatively to the weight of the skeleton is only 9 per cent. even in the breeds that never fly. Independently of shortening, the disused wing-bones have actually thickened or increased in weight. In the Aylesbury duck the disproportion caused by these conflicting changes is so great that the wing-bones are 47 per cent. heavier than they should be if their weight had varied proportionally with their length.[25] The reduction in weight on which Darwin relies seems to be entirely due to the shortening, and this shortening appears to be irrelevant to disuse, since the wings of the Call duck are similarly shortened in their proportions by 12 per cent., although this bird habitually flies to such an extent that Darwin partly attributes the greatly increased weight of its wing-bones to increased use under domestication. We find that _all_ the changes are in the direction of shorter and thicker bones--a tendency which must be largely dependent upon the suspension of the rigorous elimination which keeps the bones of the wild duck _long and light_. The used leg-bones and the disused wing-bones have alike been shortened and thickened, though in different proportions. Natural or artificial selection might easily thicken legs without lengthening them, or shorten wings without eliminating strong heavy bones, but it can hardly be contended that use-inheritance has acted in such conflicting ways. The thickening of the wing-bones has actually more than kept pace with any increase of weight in the skeleton, in spite of the effect of individual disuse and of the alleged cumulative effect of ancestral disuse for hundreds of generations. The case of the duck deserves special attention as a crucial one, if only from the fact that in this instance, and in this instance only, has Darwin given the weights of the skeletons, thus furnishing the means for a closer examination of his details than is usually possible. If we ignore such factors as selection, panmixia, correlation, and the effects of use and disuse during lifetime, and still regard the case of the domestic duck as a valid proof of the inheritance of the effects of use and disuse, we must also accept it as an equally valid proof that the effects of use and disuse are _not_ inherited. Nay, we may even have to admit that, in two points out of four, the _inherited_ effect of use and disuse on successive generations is exactly opposite to the immediate effect on the individual.