Red cells transfused into a case of haemolysing blackwater fever are destroyed just as readily as the patient's own cells, making it appear that it is not the red cells that are at fault in this condition, but that there is some circulating haemolysin. By an improved spectrophotometric technique it has now been shown that, like those of haemolytic jaundice, the red cells from blackwater fever have an increased fragility to lyso-lecithin, although their fragility to saline is normal, unlike that of the cells of haemolytic jaundice. These two facts taken together make it seem that in blackwater fever the cells may possess some abnormal feature, but that this is probably secondary to more fundamental changes that take place in the cells' environment. By using van Boros' formula it is possible to demonstrate that spherocytosis occurs in blackwater fever and is accompanied by decreases in diameter-thickness ratio, volumes and areas; that these changes in blackwater fever are intermediate between those taking place in haemolytic jaundice and normal controls, and are not related in blackwater fever to changes in hypotonic saline fragility. It is probable that the initial stage in the destruction of the cells is a change in the permeability of the cell membrane, which allows haemoglobin to escape. The cells have later been seen to swell, become “transparent,” and then to disappear. That they have not been broken up can be shown by resuspending them in saline when the “ghosts” reappear. This phenomenon takes place in both lyso-lecithin and snake venom haemolysis. If the process is allowed to continue the cells are finally disrupted, and will not reappear on saline resuspension. In haemolytic jaundice it seems that there may be some defect in the red cell as well as some “splenic factor,” shown by the fact that splenectomy fails to alter the abnormal fragility of the cells to saline, but does stop the periodic haemolysis. A combination of both abnormal cells and splenic factors would seem to be necessary for the destruction of the red cells in this condition. The same cannot be determined for blackwater fever. The enlargement of the spleen in so many of these haemolytic conditions is regarded by many as of considerable significance. The production of lyso-lecithin as a result of the separation of the cells and plasma in this organ has led to the suggestion that this powerfully haemolytic substance may play a part in the haemolyses of certain of these conditions. It is possible that lyso-lecithin fragility is in some way related to the action of this substance on the lipoids or lipo-protein complex in the red cell membrane. It has been shown that the amount of lipoid present in the cell envelope is closely related to the dimensions of the cell, and that the ability of the cell to increase its volume without being disrupted is related to the lipo-protein ratio in its membrane. The kinetics of lyso-lecithin haemolysis have not been worked out, but there is some ground for regarding the process as one of enzyme action where the enzyme lecithinase acting on a substrate of red cells or plasma lecithin results in the production of lyso-lecithin or some allied phosphatide. Lecithinase has been shown to be present in many conditions where haemolytic reactions take place, such as in favism, snake poisoning, Cl. welchii filtrates as well as in the spleen and peripheral blood. The degradation products of the action of lecithinase on lecithin are not always the same, nor are the various lecithinases alike. Although the end product of the action of snake venom lecithinase on lecithin is lyso-lecithin, yet this substance when injected into baboons in relatively huge amounts produces no haemolysis, and only a slight shift of the Price-Jones curve to the right. This is in sharp contrast to the haemolysis that follows the injection of snake venom freed of its neurotoxic factor. It is suggested that this difference may be due to the fact that in the former case the lecithin in the cells and/or plasma is not disturbed, whilst in the case of snake venom injection of the lecithinase present in the venom will split the lecithin of the cells and/or plasma and so disturb the intracellular lipo-protein complex. The part played by the reticulo-endothelial system in the haemolytic process is briefly discussed—the resulting decision being that no precise data are available to enable any conclusions to be reached either upon the part that this system plays in actually destroying red cells, or elaborating haemolysins that later lyse the cells. Proteolytic enzymes may be produced in the macrophages of the reticulo-endothelial system and these have been shown to be capable of destroying red and white cells.