488 IMMUNOLOGY 



important papers cited by Wells are those of Ilaiizlik and De Eds 

 (1926), who observed a marked change in endothelial permea- 

 bility in anaphylactic shock, and Paterson and Levinson, who 

 noted not only an augmented flow but an increased protein con- 

 tent of thoracic lymph during severe shock. More recently 

 Kuslmarjew (1930) calls attention to changes in the blood cal- 

 cium and potassium during shock. We have suggested that since 

 electrolj'tes enter into antigen-antibody reactions in general, the 

 disturbance of electrolyte balance in or on the tissue cells may be 

 an important factor. 



The Nature of Anaphylaxis. — The evidence presented in the 

 preceding pages, which seems to prove conclusively that the type 

 of hypersensitiveness, discovered independently by Richet and 

 Portier and by Theobald Smith, respectively, and named by the 

 former anaphylaxis, is mediated by an antigen-antibody mecha- 

 nism, may be summarized as follows : 



1. Active sensitization can be produced by introducing into 

 the tissues a true antigen only, 



2. The incubation period which must elapse before specific 

 hypersensitiveness develops corresponds to the required incuba- 

 tion period for the development by the body of specific antibodies 

 and their fixation by the tissues. 



3. Anaphylactic shock, as previously defined, is elicited by nor- 

 mally nontoxic doses of the specific antigen only, or, in the case 

 of bacterial anaphylaxis, it may also be elicited by the specific 

 bacterial polysaccharide used likewise in amounts nontoxic for the 

 normal animal. 



4. Blood from a sensitive animal is capable of passively sensi- 

 tizing a normal animal only when it is obtained from the former 

 at a time at which experience has indicated antibodies should be 

 present. Where blood from sensitized rabbits is employed, it has 

 been shown that its sensitizing power parallels its precipitin 

 (antibody) content. 



5. Weil showed that the washed precipitate obtained in the 

 precipitin reaction can passively sensitize as well as actively sen- 

 sitize a normal guinea pig. This suggests the identity of sensitizer 

 and precipitin, since the precipitate is known to contain the latter 

 bound to antigen. 



