Seeing, then, that it is difficult to estimate the amount of purely individual increment or decrement, and that it is difficult, if not impossible, to exclude the disturbing effects of cessation of selection with economy of growth on the one hand, reducing the size of organs, and artificial selection on the other hand, increasing the size or efficiency of parts, it is clear that such cases cannot afford convincing evidence that the observed variations are the directly inherited results of use and disuse. Indeed, I am not aware of any experiments or direct observations on animals which are individually conclusive in favour of the hereditary summation of functionally produced modifications. It may, however, be said--Although no absolutely convincing experiments or observations are forthcoming (for, from the nature of the case, it is almost impossible logically to prove that this interpretation of the facts is alone possible), still there are cases which are much more readily explained on the hypothesis that the effects of use and disuse are inherited, than on any other hypothesis. But, so far as Professor Weismann and his followers are concerned, such an argument is wholly beside the question. They are ready to admit that inherited modifications of the body, if they could be proved, would render the explanation of many results of evolution much easier. It would, no doubt, they say, be easier to account for the shifting of the eye of a flat-fish from one side of the head to the other on the supposition that individual efforts were inherited, until, by an hereditary summation of effort, the eye at last came round. The question is--Are we justified in accepting the easier explanation if it be based on a mere assumption, at present unproved, the _modus operandi_ of which is inexplicable? Let us consider very briefly these two points--first, the "mere assumption;" secondly, "the inexplicable _modus operandi_." Is there any reason why we should not assume the inheritance of effects of use or disuse as a working hypothesis, if it is not in opposition to any known biological law, and if it does enable us to explain certain observed phenomena? I see no such reason. We do not know enough about the causes of variation to be rigidly bound by the law of parcimony. I am not aware of any biological law that would render the acceptance of this view as a provisional hypothesis unjustifiable.