1. Studies in vitro and transfusion experiments indicate that the "cause" of nocturnal hemoglobinuria is an abnormality of the erythrocytes. In vitro, patient’s corpuscles undergo hemolysis in fresh human sera, but only within a pH range of 6 to 8, a range more restricted on the alkaline side than the limits within which isohemolysis will take place.
2. The nonspecific nature of the acid-serum test is emphasized. In addition to nocturnal hemoglobinuria, positive tests for hemolysis may be obtained in this way with certain "warm" and "cold" hemolysins, and in the presence of marked spherocytosis. The control observations necessary for the diagnosis of nocturnal hemoglobinuria are described.
3. The erythrocytes in nocturnal hemoglobinuria are remarkably sensitive to hemolysis by anti-A (or anti-B) but are not hemolyzed by anti-Rh.
4. It is suggested as a hypothesis that the same abnormality, presumably at the corpuscular surface, which is the cause of the increased sensitivity to hemolysis by anti-A results in the erythrocytes being fatally sensitive in vivo to a hemolytic factor distinct from complement and normally present in serum.