It seems to be clear that iso-agglutinins and iso-hæmolysins, that is to say, serum reactions among the individuals of a species, are to be found distributed widely through the animal kingdom. The phenomenon is, however, weak in operation compared with that found among human beings, and it is very much more difficult to demonstrate. The facts have not been investigated for very many species of animals. Some of the earliest attempts to investigate the distribution of iso-agglutinins among animals were made by Hektoen in 1907. He tested the blood of rabbits, guinea-pigs, dogs, horses, and cattle; his results were negative in every case, but probably his technique was imperfect or an insufficient number of animals was tested. Grouping has been found among goats by Ehrlich. Ottenberg and others believe that they have demonstrated the existence of three groups among steers, and of four groups among rabbits. Von Dungern has shown that there are four groups among dogs. Agglutination reactions were found by Ingebrigtsen and by Ottenberg among cats, but they were not constant, and it was not found possible to distinguish any grouping. The same was found to be true of rats. I have not been able to discover any record of research upon iso-agglutinins in birds or reptiles. The phenomenon of blood groups has a possible bearing on the success or failure of experimental transplantations of tissue, whether healthy or diseased, from one animal to another of the same species. From this point of view an investigation of the blood reactions among mice was carried out by B. R. G. Russell in the laboratories of the Imperial Cancer Research Fund, but he was unable to find any sort of grouping. Ingebrigtsen has made an attempt to correlate the results of the transplantation of arteries in cats with their serum reactions, but he was unable to do so. His results were equally bad whether iso-agglutinins were present or not. Nevertheless, it is highly probable that the success of tissue transplantation in man will be found to be largely dependent upon compatibility of blood groups in donor and recipient. The problem is one that cannot easily be investigated by experiment on animals, among which natural incompatibility is evidently much less well marked than it is in man. A method of overcoming this unsuitability is suggested by the experiments of Ottenberg and Thalimer. These observers, as already mentioned, found that in cats iso-agglutinins were present, though inconstant; on the other hand, iso-hæmolysins were seldom if ever found in normal cats, though they often appeared in the recipients of transfusions. Grafting experiments might therefore be preceded by transfusions designed to stimulate artificially incompatibility of the tissue fluids.