278 LOCAL ILSSIE REACTIVITY 



sufticiciiL anioiiiu ol aiilibody icniained in the blood sticain at 

 tlie time oi the subsequent injection ol the antigen. 



It sliould be noted here that there were tested various ])atches 

 ol anti-human serum derived from the same horse during different 

 stages of inmiunization. Since every bleeding did not give positive 

 residts, it seemed of interest to determine the precipitation titers. 

 Comparison of the respective residts estaljHslied no direct rela- 

 tionship between the precipitin contents and the ability of the 

 serum to induce reacting factors upon coml)inati<)n with the pre- 

 cipitinogen in vivo (Shwartzman, 1932^) . 



A NEW METHOD FOR DEMONSTRATION OF ANTIGEN -|- ANTIBODY 



INTERACTION 



Inasmuch as the reactions in sites prepared with bacterial fil- 

 trates depend on a specific antigen-f-antibody interaction in vivo, 

 the observations offer a new method for demonstration of this 

 interaction. It is easily carried out as follows: 



A rabbit sensitized to scjme animal protein receives an intra- 

 dermal injection of 0.25 c.c. of undiluted bacterial filtrate of as- 

 certained skin-preparatory potency. Twenty-four hours after the 

 skin preparation the rabbit is injected intravenously Avith the 

 same animal protein. From four to five hours later, severe hemor- 

 rhagic necrosis at the })repared skin site is fully developed. 



Since the necessary state of sensitization can be obtained by a 

 single intravenous injection of a protein and a one \veek incuba- 

 tion ]3eriod is suflic ient, the method offers the achantages oi speed 

 and simj)licity. The readings are reliable and clear-cut because 

 the incidence of positive rabbits is high (i.e., for some proteins 

 about (S5 j)er cent of animals tested after a single sensitizing in- 

 jection) , and the severe hemorrhagic necrosis makes the reaction 

 unmistakable. The test is highly sensiti\'e and strictly specific 

 unless rejjeated sensitizing injections are made of large doses of 

 the antigen. Anaphylactic shock in rabbits is difficult to elicit, 

 and the test injections of the proteins do not have a lethal effect 

 on these animals. 



It is also possible to elicit severe reactions in the prepared 

 skin sites of non-sensitized rabbits upon separate intra\'enous in- 

 jections of antigen and antibody, i.e., passive transfer. 



The test described is clearly differentiated from the Arthus 

 phenomenon by the folloAving important features: 1. In mc^st in- 

 stances, a single sensitizing injection is sufficient. 2. The reaction 



