1. The sera of rabbits injected repeatedly with the nucleoproteins, globulins, and albumins of the liver and kidney of the dog give no evidence in vitro or in vivo experiments of organ affinity. The precipitin test offers no proof of the specificity of these sera for the proteins employed as antigens.
2. The anaphylaxis reaction applied to the same proteins indicates a slight relative organ affinity but no specificity as far as the respective protein fractions are concerned. The relative organ affinity resides, rather, in the globulin and albumin fractions than in the nucleoprotein fraction. Dog serum used both as a sensitizing and an intoxicating agent gives rise to very active cross reactions with organ proteins, thus failing to support the theory of organ or of protein specificity.
3. These results do not support the view put forward that nucleoproteins play an important part in the course of production of cytotoxic immune sera.