February 9: Found dead. _Autopsy_: Liver, spleen, and kidneys congested; large intestines hemorrhagic; omentum congested and showed the presence of small caseous nodules; liver showed adhesion to diaphragm; viscera presented the appearance of intraabdominal infection. Of the five rabbits of this series three died as a result of the administration of caffein. The other two which survived showed mild symptoms only, such as increased reflexes, but no evidence of severe poisoning such as was observed after the injection of the same doses of caffein in series A when a 2 per cent solution of caffein was injected. Convulsions were noticed in one case only (No. 568); paresis in two cases (Nos. 570 and 571). The nervous symptoms even in this group, therefore, were much milder than in series A. The percentage of death, however, was greater than in series A, in which the concentration of caffein was four times as great. It is quite probable that the strain on the heart due to the sudden increase in volume of the blood and its dilution might be an important factor in increasing the toxicity of caffein. It is conceivable that doses just sufficiently large to depress the normal heart may cause paralysis of an already overstrained organ. SERIES E. In the two experiments of this series the rate of injection as a possible factor influencing the toxicity of caffein was tested. A 2 per cent caffein solution was injected at the rate of 1 cc in two and one-half to three minutes. _Rabbit 572. Gray male. Weight, 1,770 grams. Diet, oats._ February 8: Injection began at 3 p. m., discontinued at 3.37 p. m., and resumed at 3.38 p. m.; rabbit was restless; injection finished at 3.52 p. m. Total quantity received, 17.4 cc 2 per cent caffein intravenously in 52 minutes; struggled intermittently during the injection; anterior legs paralyzed. February 9: Found dead. _Rabbit 573. Gray male. Weight, 1,810 grams. Diet, oats._ February 8: Started to inject at 1.35 and discontinued at 2.27 p. m.; received 18 cc 2 per cent caffein intravenously in 52 minutes; reflexes markedly increased soon after; 2.45, passed bloody urine; 4.30 p. m. reflexes increased; no other symptoms. February 9: 9 a. m., found dead. It will be observed that some retardation of the onset of symptoms was caused by slower injection, but the final result was the same as when the injections were made more rapidly. It is quite probable, therefore, that a much slower rate of injection may lessen considerably the toxicity of caffein.