_Rabbit 226. Gray male. Diet, carrots._ October 28: Weight, 1,045 grams; 10 cc 2 per cent caffein injected subcutaneously at 1.50 p. m.; 4.30 p. m., tremors observed, but no other symptoms. October 29: Rabbit in good condition. November 1: Weight, 950 grams; 10.55 a. m., 11 cc 2 per cent caffein injected subcutaneously (231 mg per kilo). November 4: Weight, 930 grams; 2.50 p. m., 12 cc 2 per cent caffein injected subcutaneously (258 mg caffein per kilo). November 6: Weight, 945 grams; 11.45 a. m., 15 cc 2 per cent caffein (313 mg per kilo) injected subcutaneously. November 17: Rabbit still alive; weight, 890 grams. The results of these experiments indicate that when sufficient time is allowed between two successive injections, susceptibility to caffein is not increased. The rabbit, on the contrary, seems to acquire a tolerance for the drug, for the fourth dose was 15 per cent larger than the minimum fatal dose of caffein. This is in all probability due to the better elimination of caffein and its products of decomposition and to recovery from the deleterious effects of each dose, made possible by long intervals between injections. The results of these experiments may be briefly summed up by stating that subminimum doses of caffein given to the rabbit daily or at intervals (not too long) do not produce any symptoms such as were observed in acute caffein intoxication, namely, increased reflexes and convulsions, or increased rate of respiration, thus showing that it is not cumulative. But evidence of undoubted summation of effect was adduced to show that if the administration of subminimum doses of caffein be continued daily for a period of 11 to 18 days the result is fatal. Tolerance, however, may be acquired, although to a limited extent only, provided sufficiently long intervals between injections are allowed to give time for repair of the injury done by the drug and to develop a mechanism for its better decomposition and elimination. Furthermore, the evidence just given indicates that the elimination of subminimum doses of caffein and its products of decomposition is probably accomplished within 24 hours or thereabouts. That the elimination of larger doses is not accomplished in this interval is made probable by the following experiment: _Gray rabbit 455. Female. Diet, oats._ October 12: Weight, 1,185 grams; 3.30 p. m., 11.5 cc 2 of per cent caffein injected into the lumbar muscles; 3 p. m., reflexes increased. October 13: 10 a. m., rabbit weighed 1,070 grams; no symptoms of caffein poisoning, reflexes normal; 10.30 a. m., 10 cc 2 per cent caffein injected into the lumbar muscles; 11.30 a. m., rabbit jumped off the table, had convulsions, and died. EXPERIMENTS ON DOGS.