The effective dose in these experiments showed considerable variation. One dog (No. 38) died after a dose of 0.12 gram caffein per kilo, while some subjects survived doses of 0.2 and 0.23 gram per kilo. In the 12 experiments given in Table 12, page 62, it will be noticed that from 0.12 to 0.152 gram per kilo proved fatal to three dogs, while three others survived the same amounts in proportion to the body weight. The results were the same with larger doses. It may be observed in this connection that in the case of the five dogs in which vomiting was noticed some time during the 24 hours following the administration of caffein, four survived, No. 38 being the exception. The greater toxicity of caffein in this case is in all probability due to some morbid process, the presence of which was indicated by the high temperature of this subject. That vomiting may avert a fatal issue after larger doses of caffein is made further probable by experiment on dog No. 48, for which, in the absence of vomiting, a dose of 0.2 gram of caffein per kilo proved fatal. On this supposition the discrepancy in the results obtained in this series may be readily explained. The smallest doses which proved fatal in these experiments were 0.145 and 0.152 gram per kilo. No. 38, which died from a dose of 0.12 gram per kilo, may be considered as an exception, as this subject was not normal. Experiments with caffein on dogs were made at various other times in this laboratory but failed to show that smaller doses of caffein, even when vomiting did not occur after its administration, were fatal, although toxic effects were observed. The conclusion is therefore justified that the minimum fatal dose of caffein for the normal dog is about 0.15 gram per kilo when given by mouth. SUBCUTANEOUS INJECTION. SERIES B. To determine the toxicity of caffein more accurately, especially for comparison with animals of other species, the subcutaneous method of administration was also used. The injections were made with a syringe of 20 cc capacity, the contents of which were introduced into contiguous areas. The results of experiments on six dogs indicate that approximately 150 to 160 mg per kilo is the minimum fatal dose, since such doses proved fatal to two out of the three animals receiving this amount, while three others which received doses of from 143 to 160 mg per kilo survived. EXPERIMENTS ON PUPPIES. SERIES C. In these experiments the resistance of young growing puppies to caffein was studied. Caffein was given by mouth to all the subjects except one, to which it was administered subcutaneously. The protocols, only a few of which are given, and the tabulated data of the experiments (p. 62) show that the age of the animal has a decided influence on the toxicity of caffein. _Dog 11. Weight, 1,260 grams._ August 2: At 10 a. m. given 12.5 cc of 2 per cent of caffein through stomach tube; 2 p. m., had convulsions, diarrhea, salivation, and stiffness of limbs.