From the earliest days of the modern study of malaria the phenomenon of relapses in this disease has been invested with a mystery which has been responsible for various hypothetical solutions made in an attempt to explain the known facts. The first of these facts is that periods of patent infection accompanied by the demonstrable presence of parasites in the peripheral blood are often succeeded by more or less prolonged periods when no such parasitaemia can be demonstrated. This condition is described as latent malaria, and it may supervene in the natural course of the disease or may be produced artificially by the use of antimalarial therapy which destroys the erythrocytic infection. Another fact is that such relapses are often precipitated by any happening which causes a temporary lowering of the host's resistance, and the problem here is the source of the parasites producing the relapse. A third fact is that the various species of malaria parasites vary in their tendency to relapse and in the period of years during which relapses may recur. A final fact relates to artificially produced infections. Those which are produced as the result of inoculation of blood show no tendency to relapse after antimalarial treatment, while those induced by sporozoite inoculations, either artificially or by mosquito bite, tend to produce relapses after the cessation of treatment. The extent to which these facts are explained by the finding that is the subject of this communication is dealt with in the discussion. Up to the present time three commonly adduced explanations of relapses have held the field, but, in the absence of accurate knowledge of the life-cycle of malaria parasites, none could claim greater weight than a mere guess. These theories are: (a) The continued existence of a low-grade erythrocytic infection kept in abeyance by the host's immune mechanism but flaring up on any impairment of the latter; (b) the theory of the parthenogenetic development of the female gametocyte (Grassi, 1900; Schaudinn, 1902) ; (c) the existence of a cryptic stage in the internal organs capable of producing an erythrocytic invasion on any lowering of the host's resistance. The third of these theories gained added weight with the discovery of the exo-erythrocytic cycle in bird malaria, where these forms play an important, if not the essential, part in the causation of relapses. The possibility that the hypothetical exo-erythrocytic cycle in mammalian malaria played a similar part in the production of relapses has been put forward at various times by many workers, and the names given in brackets are only a representative selection from workers who have made such a suggestion, after the first establishment of a definite exo-erythrocytic cycle in bird malaria by James and Tate in 1937. (James and Tate, 1937; Shortt, Menon, and Iyer, 1940; Fairley, 1945; Huff, 1947; Cooper, Ruhe, and Coatuey (personal communication to Huff), 1947; Shortt and Garnham, 1948.) The discovery recorded in this paper of the continued existence in monkey malaria of the exo-erythrocytic cycle after establishment of the blood infection would appear to constitute the strongest proof that this form of the parasite is the aetiological agent concerned in the production of relapses. When we (Shortt and Garnham, 1948) published a detailed description of the pre-erythrocytic cycle in P. cynomolgi and P. vivax we suggested that if this cycle was found to persist after establishment of the blood infection it might play an essential part in the maintenance of the infection over long periods and in the production of relapses, so that the present work is the logical sequence to the experiment establishing the pre-erythrocytic cycle in mammalian malaria. To investigate the theory of the persistence of the exo-erythrocytic cycle and its relation to relapses, examination of infected monkeys a considerable time after the establishment of sporozoite-induced infections seemed to offer the most direct approach to the problem. …