GERM-PLASM.--While Weismann regarded germ-plasm as absolutely stable, the only mode by which congenital variation could be brought about was that of _amphimixis_, or intermingling of individuals in the process of generation. As modified, however, by his latest work, "The Germ-plasm, a Theory of Heredity," published in 1892, his theory now allows the plasm to be capable of modification, and he ascribes that variation to the direct effects of external influences on the biophors and determinants of the germ-plasm. The instability of this substance is so slight, however, that congenital variations cannot be acted on and perpetuated by natural selection, and the influence of amphimixis is thus required for the purpose. Mr. Herbert Spencer, however, in criticising Weismann's theory, declares that "functionally produced modifications of structure are transmissible," and he refers in support of his contention to the remarkable effect of arrested nutrition on the structure and habits of wasps and bees. It especially affects the reproductive organs, and hence there is no occasion to call in the aid of amphimixis to perpetuate the variations produced, its office being the blending of the elements on which the characteristics of offspring depend. If it be asked how modifications are actually transmitted, we may say that it can be only by an affection of the germ-cell. This probably takes place by deviations in the structure of what Weismann calls determinants, or of groups of determinants, through rearrangement of their primary units. The modification would be preceded, however, by a corresponding change in the nerve centers concerned in the use or disuse of the organs affected. Mr. Spencer shows that under certain conditions changes take place in the conduct of certain insects, and that "the maternal activities and instincts undergo analogous changes,"[77:A] facts which point to a loss of nervous energy and to an intimate connection between the nervous system and the reproductive function. Use or disuse first increases or diminishes the activity of certain nerve centers, and this leads to a modification of the corresponding germ-cells. If so, the determinants, instead of being first affected, as proposed by Weismann, and thus determining the variations, are in reality modified as the result of the functional changes, and are thus capable of transmitting these changes to succeeding generations.