Maurel's(56) investigations are of interest in this connection, as his work embraces a systematic study of the toxicity of a large number of substances in the rabbit, pigeon, and frog when given by mouth, subcutaneously, intravenously, or when injected into the muscles. According to this investigator the range of variation of the toxicity of a substance is widest when given by mouth. Potassium sulphocyanid, for example, is about 2.5 times as toxic for the frog as for the rabbit when given by mouth. Quinin hydrobromid is three times as toxic for the frog as for the pigeon, while for the rabbit it is twice as toxic as for the pigeon. When given by hypodermic injection the toxic dose per kilo weight is practically the same for all three species. The difference of resistance according to the mode of administration is even more marked for spartein sulphate. When given by mouth the toxicity for the rabbit is six times as great as for the frog, but when injected subcutaneously the toxic dose is about the same for the rabbit and for the frog. The relation of the mode of administration to toxicity is further shown in the following substances: For the rabbit the minimum fatal dose per kilo of quinin hydrobromid is 1.5 grams administered by mouth, 0.5 gram when injected subcutaneously, and 0.07 gram by the intravenous path, while strychnin sulphate is twice as toxic administered intravenously as subcutaneously, and six times as toxic as when administered by mouth. The mode of introduction, however, does not always affect the toxicity of a substance. This is made evident by the action of strychnin on frogs in which, according to Maurel(56), the toxic dose is the same whether given by mouth or injected into the subcutaneous tissues. This appears to hold true also for other animals as demonstrated by the experiments of Hatcher(35) on the cat, in which he observed that strychnin is as readily absorbed from a full stomach as from the subcutaneous tissues. These findings are extremely interesting, especially in view of Maurel's(57) work on the subject, according to which he finds that a substance is much less toxic when given by mouth than when administered by hypodermic injection or intravenously. That this generalization does not admit, however, of universal application is made evident by the work of various experimenters. Claude Bernard(10) observed that curara is as poisonous for the pigeon when given by mouth as when injected subcutaneously, while Zalesky(86) found that samandarin is more toxic for frogs when introduced into the stomach than by injection into the lymph sacs. Our experiments with caffein likewise show that Maurel's generalization does not always hold good, since it was found in experiments with gray rabbits that the minimum fatal dose is but moderately greater by mouth than by the subcutaneous path.