It has long been known that if the blood of one species of animal is injected into the circulation of another species, the corpuscles of the foreign blood are at once destroyed, their contained hæmoglobin being set free. This process of hæmolysis is under such circumstances rapid and complete, and hæmoglobin may appear in the urine in a short time. The precise nature of the reaction is obscure and need not be discussed here in detail. The present bearing of the phenomenon is the fact that a similar, or analogous, reaction may occur when the bloods of certain individuals are mixed with the bloods of certain others even of the same species. It was the observation of this fact that first led to the discovery of the so-called “blood groups” among human beings, and so to the partial elucidation of the cause of the previously unexplained fatalities following blood transfusion. In 1901 Landsteiner had detected the presence of hæmolysins and iso-hæmolysins in blood and classified three groups in human beings. In 1907 it was shown by Jansky that human beings may be divided into four groups, the blood of the members of each group having a certain definite relation to the blood of the other groups as determined by the manner of their interaction. The work was repeated and confirmed by Moss in 1910. The reaction takes place between the serum of one group and the corpuscles of the other groups, and is evidenced by the agglutination or hæmolysis of the corpuscles that are being acted upon. In the course of his researches Moss showed that hæmolysis, or the breaking up of the corpuscles, is always preceded by agglutination or the clumping together of the corpuscles. The process does not necessarily go as far as the destruction of the corpuscles, but may be arrested at the stage of agglutination. It may, on the other hand, be as rapid and complete as if the bloods belonged to different species, and the appearance of hæmoglobin in the urine may quickly give evidence of this. The groups have been arbitrarily numbered, and it is now usual to refer to them by the Roman numerals I, II, III, and IV. According to the accepted convention, the reactions of these four groups are as follows:[5] The corpuscles of Group I are agglutinated by the sera of II, III, IV. The corpuscles of Group II are agglutinated by the sera of III, IV. The corpuscles of Group III are agglutinated by the sera of II, IV. The corpuscles of Group IV are not agglutinated by any of the other groups. On the other hand: The serum of Group I agglutinates no other corpuscles. The serum of Group II agglutinates the corpuscles of Groups I, III. The serum of Group III agglutinates the corpuscles of Groups I, II. The serum of Group IV agglutinates the corpuscles of Groups I, II, III. This may be represented more graphically by the following table, a + indicating agglutination, a - indicating no reaction: