Greater success is claimed by Hooker, who reported that in five cases of pyogenic infection the results were distinctly favourable. He used immunized blood, but has formed the impression that the transfusion even of normal blood is of value in septicæmia by correcting the anæmia and helping to restore the normal resistance. He recommends that if the patient has a good blood volume and a high bacterial content in the blood, he should be bled by venesection before transfusion. A striking case of staphylococcal septicæmia has been recorded by Little, who believed that the patient’s recovery was directly due to the treatment. Four transfusions were given, the blood for three of these being taken from donors who had each received, four days previously, an injection of vaccine made from the patient’s own infection. Ottenberg and Libmann have treated ten cases of pyogenic infections with transfusions. All the patients were extremely ill and six died. It is stated that the four who recovered “probably owe their lives to the transfusion,” but obviously it is difficult to control the results. The same observers have used transfusion in the treatment of infective endocarditis, but unsuccessfully. Some experimental work on this subject has been carried out by Kahn. A bacterial infection was introduced into the peritoneal cavities of several dogs. Continuous transfusion between an infected dog and a healthy dog was then performed, the blood passing to and fro between the animals, sometimes for over an hour. It was found that all the transfused animals fared better than those that were not. The experiment suggests that resistance to infection is heightened if two bodies can combat the infection present in one; but continuous transfusion is scarcely practicable in man.