Following the divisions before used, the results presented in the preceding pages may be briefly stated. I. The particular method of sensitization and the place where the test injection is made have an important bearing on the results obtained by various workers. Comparing the results obtained by the various methods, we may conclude that the incubation period of the hypersensitive reaction is not sharply limited, but that there is a progressive increase in sensitiveness from the sixth day, and presumably before that, extending over a period of several weeks. It seems very probable that the degree of hypersensitiveness attained where the sensitizing dose consists of a mixture of diphtheria toxin and serum is greater than when a single dose of the same small quantity of serum is given alone. II. Our early experiments, the first in this field, are in thorough agreement with those first reported by Otto, and shortly after him by Rosenau and Anderson. III. This hypersensitive reaction is transmissible from mother to offspring. The transmission is probably not equally effective in all cases, and individual young guinea-pigs probably vary greatly in the rate with which they lose their ability to react. As a result not all of the young of a hypersensitive mother react to a subcutaneous dose of five cubic centimeters of serum given when they are four or five weeks old. The reaction in the young animals differs quite markedly from that in those actively sensitized. These differences are such as to indicate that in the mother there is a considerable localization of the reaction in tissues and organs whose destruction does not cause sudden death. This local reaction is a protective factor and is not transmitted to the same degree as the factors involved in the fatal acute reaction. IV. The hypersensitive reaction to horse serum depends on the development of a special anti-body during the incubation period, which anti-body may be passively transferred to a fresh animal. If the dose of hypersensitive serum be sufficient, and the intoxicating injection be given directly into the circulation, this passive hypersensitiveness may be enough so that the animal will die when tested. There is also in the serum of hypersensitive guinea-pigs an uneliminated horse serum element or "rest," which is distinct from this antibody, and probably without influence on the course of the acute reaction. V. The anti-body on which the hypersensitive reaction depends may be entirely neutralized by horse serum without causing symptoms. The gradual introduction of increasing doses over a total period of twenty-four hours suffices for this. The animal is then, properly speaking, neither immune nor refractory, but is essentially in the condition of a normal animal which has recently had a large dose of horse serum. This rapid neutralization is made possible by the great binding power which the subcutaneous and other relatively unimportant tissues have for the toxic element of the serum. The so-called "Phenomenon of Arthus" is probably the same reaction for the rabbit that we have here dealt with in the guinea-pig. The fact that the manifestation is more prominently a local one depends on racial differences. I have encountered cases in the guinea-pig in which the conditions in the rabbit are closely simulated.