Recently, there has been a revival of Lamarckism hitherto regarded as defunct. Guyer, Kammerer, and Pawlow profess to find factual justification for it, and Bouvier adopts it in his “La vie psychique des insectes” (1918), to account for the origin of instinct. Of the alleged facts of Kammerer and Guyer, we have spoken in a previous chapter. Here we shall content ourselves with few remarks on the experiments of Ivan Pawlow, as being especially relevant to the subject under consideration. The Russian physiologist has experimented on white mice, and claims that the mice of the fifth generation learned to answer a dinner bell in the space of five lessons, whereas their ancestors of the first generation had required a hundred lessons to answer the same signal. Hence he concludes: “The latest experiments ... show that conditioned reflexes, _i.e._, the highest nervous activity, are inherited.” (_Science_, Nov. 9, 1923, p. 360.) His results, however, do not tally with those recently obtained by E. C. MacDowell of the Carnegie Institution, by H. G. Bragg, and by E. M. Vicari of Columbia. MacDowell found that white rats trained in a circular maze did not improve in their susceptibility to training from generation to generation. “Children from trained parents,” he says, “or from trained parents and grandparents, take as long to learn the maze habit as the first generation used.” (_Science_, March 28, 1924, p. 303.) Having cited the similar results of Bragg, who experimented with white mice, he concludes: “The results are in full accord with those given above; they indicate that the training of the ancestors did not facilitate the learning of the descendants.” (_Ibidem._) E. M. Vicari, using a simple maze and white rats, obtained the same results. “It seems clear,” she says, “that the latter generations have not been aided by the training of their ancestors.” (_Ibidem._) Bouvier’s conception, then, that the automatisms of instinct originate as automatisms of acquired habit, the latter being appropriated by inheritance, still stands in need of reliable experimental confirmation. Moreover, a theory of this sort could never account, as Weismann points out, for such phenomena as the specific instincts of worker bees, which are _excluded from propagation_. Nor can the theory explain, as originating in acquired _habit_, those instinctive operations of enormous complexity, like the complicated method of emergence employed by the larva of the emperor moth, which only occur _once in a lifetime_, and could not, therefore, fasten themselves on the organism as a _habit_.