unworthy of discussion. Rejecting this and other similar theories, I think that those who maintain the transmissibility of acquired variations are practically reduced to the choice of one of the three following hypotheses. The first of these supposes that an acquired variation produces, through the blood, such changes in the nutrition of the germ cell as bring about after fertilization a similar variation in the organism of which its remote cell-descendants form the component parts. To take the same example as before, it is supposed by this theory that the blacksmith, by exercising the muscles of his arm, produces such a change in the nutritive fluids which bathe his germ cells as causes them, after conjugation with germ cells from another individual, to proliferate in such a manner as to produce new organisms with muscles similarly enlarged. To my mind this is as wild a theory as that which I have already dismissed. Even were it possible, and I think it is impossible, that such transient changes as are produced in the nutritive fluids by exercising the arm muscles could so permanently and profoundly affect the constitution of the germ cell as to cause it to develop, after long separation from the parent organism, into an individual structurally different from the individual which would otherwise have resulted, even, I say, were this possible, how is it possible that these nutritive changes can cause in the germ cell such particular structural alterations as result, after its proliferation, in an individual who has a variation similar to the particular variation in the parent organism, a variation which resulted from exercising the arm muscles? To put it in fewer words: how can the change which any acquired variation produces in the nutritive fluids that bathe the germ cell affect that cell in such a manner as to cause it, after fertilization and separation from the parent
