This is certainly ingeniously and cleverly thought out, and it reads even better and more smoothly in the original than in my brief summary, but it will hardly be regarded as a refutation of my position; the hypotheses are all too daring for that. We have no knowledge that particular modifications in form can be produced and conditioned by particular kinds of food, and, indeed, the contrary has been proved, namely, that the two or three different castes of polymorphic species have precisely similar diet. I need only recall the six forms of female in _Papilio merope_, of which at least three have been obtained from the same set of eggs, and by feeding with the same plant. It is true that there are ants which lay in stores of nourishment, but these consist, for the most part, of one kind of seeds, or of honey, not of different substances, and we have no knowledge that the different persons use different food, or even that there is any diversity in the mode of feeding the helpless larvæ. The feeding in some species takes place from mouth to mouth, and therefore cannot be precisely investigated, and we can only suppose from analogy with bees that the larvæ of the males and females frequently receive not only more abundant, but qualitatively different food. They are fed from the crop unless the food consists of the pith of a tree in which the larvæ are imbedded, as Dahl informs us is the case with some tropical ants of the Bismarck Archipelago. But even if we assume that the soldiers take different food from the ordinary workers, and different again from that of the sexual animals, is it by virtue of the quality of their food that they have become what they are? Have our breeds of pigeons or hens been produced by different diet, or do we know anything in the whole range of animal life of such a parallelism between food and bodily structure as Zehnder here assumes? And if, in reality, let us say, the breeds of pigeon had arisen through specific dieting, and we were to feed one pair with the specific food-stuffs of three different breeds, would the descendants of this pair exhibit the form of these three breeds? Or would they exhibit them in precisely the proportion in which the food-stuffs had been mixed? It seems to me that Zehnder's assumptions diverge so far from what we are accustomed to regard as solid ground in biology that they hardly require consideration, and yet he not only uses them for the explanation of the case of the ants, but bases upon them the whole of his theory of the inheritance of acquired characters.