pletely dispose of Mr. Spencer's objections. The co-adaption of animal parts is clearly due to, and maintained by, the inborn power the parts possess of developing in response to stimulation, so that the development or atrophy of any part, owing to increase or decrease of stimulation, is followed, owing to alterations in the strain put upon them, and therefore also to increase or decrease of stimulation, by the development or atrophy of all co-ordinated parts. It must be remembered that Mr. Spencer, while admitting that inborn variations are transmissible, attributes co-adaptive evolution to the accumulated effects of use or disuse, that is, he attributes it to the accumulation of those variations which are acquired in response to stimulation. But in the offspring these variations are not reproduced except in response to fresh stimulation, therefore, as regards any structure, that which is inherited is not the variation, but only the power to acquire the variation afresh in response to stimulation. Now there is no tittle of evidence tending to prove that stimulation increases the power of a structure to vary. The power to vary is inherent; it is called into activity by stimulation, but there is nothing to show that the inherent ability to vary is increased by stimulation. On the contrary, as animals grow older, their structures become less and less responsive to stimulation; the power of responding to stimulation in their long-stimulated structures actually declines, and the structures atrophy. Therefore as the power of varying is not increased or (presumably) decreased by stimulation or the lack of it, by use or disuse, it is clear, since variations of this power cannot be acquired, that acquired variations in it cannot be transmitted, and therefore if we hold that acquired variations are transmissible, we must not attribute the reproduction in the child of the parent's acquired variations to transmitted increase or decrease of this power.
