THE NORMAL FATE OF ERYTHROCYTES : I. THE FINDINGS IN HEALTHY ANIMALS.

The phagocytosis of red corpuscles, while frequent in the normal dog, rat, and guinea pig, is slight in man, the rhesus monkey, and many rabbits. In cats it is always negligible in amount and frequently absent. Phagocytosis will not suffice as a general explanation of normal blood destruction. When the liver, spleen, and bone marrow of the cat, dog, rabbit, or monkey are slowly perfused with defibrinated blood or Locke"s solution, bodies are given off into the fluid which have the appearance of red corpuscles that have lost their hemoglobin but retained the rest of their cell substance. These bodies possess many of the properties supposedly distinctive of red corpuscles. They are the product of disordered parenchymal cells. By a special method, it has proved possible to search the body, organ by organ, and the circulating blood also, for disintegrating red corpuscles. Shadows of red cells are not present anywhere, nor are hemolyzing red cells found. A hemolytic process, in the ordinary sense of the term, can scarcely play an. important part in normal blood destruction. Instead, it is certain that some red corpuscles, at least, are destroyed in another way; namely, by fragmentation. Normal blood regularly contains small numbers of fragmentation forms-microcytes and poikilocytes-and accumulations of them are regularly present in the spleen, but are found only inconstantly in the other organs. The fragments are in evident process of further subdivision. They occur not only in species in which phagocytosis as a means of cell destruction is negligible (cats), but also in animals in which it is an important process (dogs, some rabbits). The method of study that we have employed is well suited to disclose how the blood is destroyed. The importance of cell fragmentation in this connection is indicated by our failure to find any other means of destruction, save only the phagocytosis already known. Further facts indicating the importance of fragmentation are presented in our second paper, where a general discussion will also be found.