Summary and Conclusions The present work has involved the use of a total of 250 birds, of which 154 have been included in the experimental series. One third of the latter number were given quinine, one third plasmochin, and one third served as controls. Five strains of avian malaria have been used: one strain of Plasmodium elongatum, one of Plasmodium cathemerium, and three of Plasmodium praecox. One of the latter came from Germany, one from New York, and the third from Virginia. This last was originally described by Hartman as Plasmodium inconstans. The conditions of the experiment have been made as uniform as possible, so that such differences as might seem to appear could be justly laid to differences in the treatment or in the strain of malaria. Equivalent inocula of parasites have been used for all the birds in any given series, and in most of the cases the dosage has been varied with the weight of the bird. The duration of treatment was fourteen days in all cases, except for five series of six birds each infected with Plasmodium praecox (type strain). Ten of these birds were given quinine and ten others plasmochin. In each instance two birds died during treatment—probably as the result of it—but treatment of the others was continued for eight weeks in the case of those receiving quinine and for seven with those receiving plasmochin. The latter were treated for a shorter time because subinoculations made at the end of the first month proved negative, except in 1 case, and it was thought that sterilization had been accomplished. The single case which gave a positive result was treated for another week. The results justify the following conclusions: 1. The different strains of Plasmodium praecox react to treatment in essentially the same way for either plasmochin or quinine, but the three species show marked differences. Apparently they stand in the following order, the most susceptible being at the head of the list: Quinine Plasmochin 1. Plasmodium praecoz 1. Plasmodium elongatum 2. Plasmodium cathemerium 2. Plasmodium praecox 3. Plasmodium elongatum 3. Plasmodium cathemerium Plasmodium cathemerium and Plasmodium praecox, the two species which resemble each other most morphologically, also differ least in their susceptibility to treatment. 2. There is no evidence that the gametocytes of any of the three species are more resistant to the action of plasmochin and quinine than are the asexual stages. What differences there may be seem to be of the reverse sort. Thus it appears that the avian malarias differ in this respect from the malarial fevers of man. This finding is opposed to that of Kopenaris (1911) who used quinine in experiments with bird malaria. 3. Treatment during the incubation period is considerably more effective than that started during the acute stage. Treatment with plasmochin for two weeks will often entirely prevent infection with Plasmodium praecox when started the day after inoculation, but if postponed until parasites appear in the peripheral blood seven or eight weeks of treatment is seldom sufficient to cause complete recovery. 4. A system of treatment comparable to the Bass system recommended for human malaria was followed with Plasmodium praecox infections of birds, but in only 1 case was sterilization attained. This occurred in one of the plasmochin-treated birds. Quinine, although administered to eight birds for an equal number of weeks, failed to prevent relapse or cause sterilization in a single case. 5. Sterilization can apparently be regularly produced by plasmochin when used against Plasmodium elongatum, and it does not matter whether treatment is commenced the day after inoculation or postponed until parasites are demonstrable in the blood. Fourteen cases in all have been cured in this way. Quinine seems to be relatively ineffective against this species but it did prevent infection in one bird out of six. In no case was sterilization obtained in infections of Plasmodium cathemerium. 6. Except in cases infected with Plasmodium cathemerium freedom from relapse during the first month after treatment proved an almost sure indication of sterilization. 7. Birds which have been sterilized by the use of quinine or plasmochin were as susceptible to reinfection as if they had never been infected to begin with, or had recovered naturally. 8. Plasmochin is uniformly superior to quinine in the treatment of bird malaria, but the degree of superiority differs with the species of parasite responsible for the infection. Roehl (1926) estimated that it was sixty times as effective as quinine, but it is also more than six times as toxic to the host. Even when this allowance is made for the difference in toxicity it does not seem to be nearly as much better than quinine as Roehl believed, except in the case of Plasmodium elongatum. Roehl did not work with this species however. 9. Increasing the dose to the limit tolerated by the bird does not seem to add greatly to the efficacy of either of these drugs. 10. Relapses occurring after treatment are usually considerably milder than untreated acute attacks. 11. Chronic infections of Plasmodium praecox differ very much in the individual parasite levels established, and previous treatment has no effect on them. 12. Plasmochin is able to reduce the parasite level to a point lower than quinine can when used against Plasmodium praecox, and in both cases the level is forced below that established in untreated chronic infections. That this difference exists is shown by the fact that the blood of quinine-treated cases remains infective when inoculated into clean birds, whereas in most cases that of birds being treated with plasmochin is not.