4. Ehrlich’s work on trypanosomes seems to indicate a remarkable power of adaptation on the part of organisms to certain poisons. If the writer understands these experiments correctly they consisted in infecting a mouse with a certain strain of trypanosomes, and treating it with a certain arsenic compound, which inhibited somewhat the propagation of the parasites but did not kill them all. Four or five days later trypanosomes from this mouse were transmitted to another mouse and after twenty-four hours this mouse was treated with a stronger dose of the same arsenic compound; and this process was repeated. After the third transmission or later, the trypanosomes can resist considerably higher doses of the same poison than at first and this resistance is retained for years. Ehrlich seems to have taken it for granted that he had succeeded in transforming the surviving trypanosomes into a type which is permanently more resistant to the arsenic compound than was the original strain. The writer is not entirely convinced that in these experiments a possibility was sufficiently considered which is suggested by Johannsen’s experiments on the importance of pure lines in work on heredity. According to this author a strain of trypanosomes taken at random should, in all likelihood, contain a population consisting of strains with different degrees of resistance. If a high but not the maximal concentration of an arsenic compound is repeatedly injected into the infected mice the weaker populations of trypanosomes are killed and only the more resistant survive. These of course continue to retain their resistance if transplanted to hosts of the same species. According to this interpretation the arsenic-fast strain may possibly have existed before the experiments were made, and Ehrlich’s treatment consisted only in eliminating the less resistant strains. On the other hand, it has been shown that if an arsenic-fast strain of trypanosomes is carried through a tsetse fly it loses its arsenic-fastness. This fact may possibly eliminate the applicability of the pure line theory to a discussion of the nature of the arsenic-fastness, but it seems that further experiments are desirable.