Now, these harmless serums become poisonous when they have been taken from animals which have been first treated with the blood or the organs of other species of animals. For instance, the blood serum of a sheep which has been treated with the blood of a rabbit becomes poisonous because it has acquired the power of dissolving the red blood corpuscles of the rabbit. It is a poison in the case of the rabbit, but is harmless to most other animals. The injection of the rabbit’s blood into the sheep has conferred on the sheep a new property which comes into operation only with regard to the red blood corpuscles of the rabbit. We have here to do with something analogous to what has been observed in the cases of serums used to arrest infectious disease. When the bacilli of diphtheria, or their products, have been injected into horses, there is produced an anti-diphtheric serum, capable of curing diphtheria, but powerless against tetanus or plague. After M. J. M. Bordet of the Pasteur Institute had made his discovery of serums that had acquired the power of dissolving the red blood corpuscles of other animals, the attempt was made to prepare similar serums directed against all the other elements of the body, such as white blood corpuscles, renal and nervous cells. In the course of these investigations it was proved to be necessary to employ a certain dose of the serum in order to obtain the poisonous result. If smaller quantities of the poisonous dose were used, the reverse effect was produced. Thus a serum, strong doses of which dissolved the red blood corpuscles and so made them less numerous in the blood, increased the number of these when given in very small doses. M. Cantacuzène was the first to establish this fact in the case of the rabbit, whilst M. Besredka and I myself did it in the case of man.[117] Since then M. Bélonovsky of Cronstadt has confirmed the result on anæmic patients, treating them with small quantities of serum. He has been able to produce in them an increase in the number of the red blood corpuscles, and in the quantity of the red colouring matter (hæmoglobin) in the blood. Later on M. André[118] devoted much attention to this matter at Lyons. He prepared a serum by injecting human blood into animals and made use of it in the case of several persons who suffered from anæmia from different causes. In the case of patients, the anæmic condition of which had hitherto remained stationary, Dr. André found a sudden increase in the number of red corpuscles after injecting small doses of the serum. M. Besredka, in the case of laboratory animals, increased the number of white corpuscles by injecting them with a small quantity of a serum, strong doses of which destroyed these cells.