The possession by every cell of primitive sexuality on which the secretion of the sexual glands has little effect might be shown further by consideration of the effects of grafting male genital glands on female organisms. For such an experiment to be accurate it would be necessary that the animal from which the testis was to be transplanted should be as near akin as possible to the female on which the testis was to be grafted, as, for instance, in the case of a brother and sister; the idioplasm of the two should be as alike as possible. In this experiment much would depend on limiting the conditions of the experiment as much as possible so that the results would not be confused by conflicting factors. Experiments made in Vienna have shown that when an exchange of the ovaries has been made between unrelated female animals (chosen at random) the atrophy of the ovaries follows, but that there is no failure of the secondary sexual characters (_e.g._, degeneration of the mammæ). Moreover, when the genital glands of an animal are removed from their natural position and grafted in a new position in the same animal (so that it still retains its own tissues) the full development of the secondary sexual characters goes on precisely as if there had been no interference, at least in cases where the operation is successful. The failure of the transplantation of ovaries from one animal to another may be due to the absence of family relationship between the tissues; the influence of the idioplasm probably is of primary importance. These experiments closely resemble those made in the transfusion of alien blood. It is a practical rule with surgeons that when a dangerous loss of blood has to be made good, the blood required for transfusion must be obtained from an individual not only of the same species and family, but also of the same sex as that of the patient. The parallel between transfusion and transplantation is at once evident. If I am correct in my views, when surgeons seek to transfuse blood, instead of being content with injections of normal salt solution they must take the blood not merely from one of the same species, family and sex, but of a similar degree of masculinity or femininity. Experiments on transfusion not only lend support to my belief in the existence of sex characters in the blood corpuscles, but they furnish additional explanations of the failure of experiments in grafting ovaries or testis on individuals of the opposite sex. The internal secretions of the genital glands are operative only in their appropriate environment of arrhenoplasm or thelyplasm.