A comparison of the fatal doses of caffein in the experiments on high and low protein diet does not show much difference in the resistance to caffein, since 175 mg per kilo proved fatal to Nos. 30 and 32, while No. 29 died after receiving 150 mg per kilo, and No. 24 received 125 to 185 mg per kilo. Moreover, No. 28, which was changed from low to high protein diet, succumbed when given 125 mg per kilo. Observations made during the experimental period indicate, however, greater toxicity of caffein in the subjects on low protein diet. Dog 30 showed the effects of the drug when the dose was increased to 125 mg of caffein per kilo, while in No. 32, 150 mg per kilo were received without any manifestation of symptoms. Dog 31, which was likewise on a high protein diet, is evidently an exception, and its low resistance to caffein may be accounted for by the condition found at autopsy. In other dogs on low protein diet symptoms of intoxication appeared early in the experiment. In Nos. 29 and 24 it was observed as soon as the amount of caffein was increased to 75 mg per kilo. In dog 28 the first dose of caffein 50 mg per kilo was toxic. The symptoms of gastro-intestinal disturbance were especially marked after caffein on low protein diet. This may seem to contradict the results of experiments on dogs 11 and 20, in which larger doses of caffein failed to induce symptoms of intoxication. But it should be observed that the diet, which consisted almost exclusively of carbohydrates, was given only during the administration of caffein, while in the experiments of series B the subjects received a low protein diet for several weeks before the administration of caffein was begun, and it was continued through the entire caffein period. It will be remarked that the absence of cumulative action in the experiments of the preceding series was also observed in dogs on high as well as on low protein diet. The appearance of symptoms after smaller doses of caffein in the latter experiments might suggest cumulative action, but since these symptoms disappeared on continued administration of the substance cumulation is clearly not indicated. The gastrointestinal lesions observed on post-mortem examination were, it will be recalled, also found in rabbits similarly treated. The explanation suggested probably applies also in the case of dogs. SERIES C. As already pointed out in the experiments on acute toxicity of caffein, young growing dogs are probably more resistant to caffein than adults. That this may also hold true in chronic caffein intoxication seemed indicated by the following experiments. _Dog 33. Black female puppy. Weight, 4 kilos. Had been continuously on a meat diet._ December 22: 2.30 p. m., received 0.69 gram of caffein (0.172 gram per kilo); 3.15 p. m., no symptoms except that feces were fetid. December 23: 11.30 a. m., received 0.79 gram of caffein (0.197 gram per kilo); 1.30 p. m., no symptoms.