In the course of our experiments on haemolytic sera we have met with the following phenomenon which appears of some interest. It consists in the agglutination of the corpuscles of an animal by its own complement through the medium of the corresponding immune-body derived from another animal, and was observed first in the case of ox's corpuscles, the immune-body used being obtained from the rabbit. The fundamental fact is that if a certain amount of immune-body and ox's complement be added to ox's corpuscles, scarcely any lysis of the corpuscles occurs but they become agglutinated into large masses which cannot be dissociated by shaking. The immune-serum from the rabbit contains some agglutinin, but the degree of agglutination produced by this is quite trifling compared with that seen when complement also is added. There is thus no doubt that the agglutination phenomenon depends on the cooperation of two substances in a manner comparable to what obtains in lysis. The following are the chief facts regarding the conditions of occurrence of the agglutination and the nature of the agglutinating substance in the ox's serum.