The incompatibility of blood is essentially a phenomenon which distinguishes different _species_ of animals, since in no case can the blood of one species circulate unaltered in the blood-vessels of another kind of animal. This serological specificity may be in some way related to the sterility of one kind of animal with another, though not actually causing it, and so be merely an incidental phenomenon. It cannot be in any sense protective, since it never happens in the course of nature that blood is transferred from one animal to another. In the same way it is difficult to see how there can be any biological “purpose” in similar differences between individuals of the same species, and, so far as is at present known, the possession of a particular group does not confer upon its owner any advantage over the individuals of other groups, such as a relatively greater immunity from disease, longevity, or fertility. It is quite clear that there is no connexion between incompatible blood groups and sterility between individuals. An investigation of a possible relation between blood groups and disease has been begun by W. Alexander at St. Andrews University. In a preliminary communication concerning the blood groups found among fifty patients suffering from “malignant disease” of all forms, including leukæmia, he has found that there is a considerably higher proportion of Groups I and III than among healthy people. On the other hand, the groups are found in the normal proportions among people suffering from tuberculosis, syphilis, and tetanus. It would, however, be premature to assume that individuals of Groups I and III are more liable to suffer from “malignant disease” than other people, as the numbers tested are, at present, too small for definite conclusions to be formulated. Also it remains to be proved that the presence of malignant disease does not produce an alteration in the agglutinating reactions by which the groups are determined.