.lAM/'//17..lA7^' OH ALLERiiY 195 



an antitoxic iniiuunity. In the absence of complement, or when the 

 complement fixation is prevented by strong salt solution, ^^ the poison 

 (anaplu'latoxin) does not develop, so that the anaphylactic reaction 

 falls into the same class as the lytic reactions, in which the non-specific 

 serum complement is united to a cell by the specific amboceptor, and 

 then causes lysis of the cell; in anaphylaxis not an organized cell but 

 a comi^lex protein molecule is disintegrated by the complement, but 

 in either case a poisonous substance may be liberated. 



This agrees with Vaughan's hypothesis in ascribing the poisoning 

 to products of protein disintegration formed by enzyme action, but 

 differs in that specific intermediary substances or amboceptors are sup- 

 posed to be developed by sensitization, rather than specific enzymes. 

 Friedberger is of the opinion that many or all the different immunity 

 reactions depend upon a single antibod}', the different reactions merely 

 being different methods of demonstrating the presence of the antibody 

 in the serum. The precipitin reaction differs from the anaphylactic 

 reaction, he contends, only in that in the latter the specific pre- 

 cipitate is dissolved by complement, yielding the anaphylatoxin. 

 There are many objections'^ to accepting this idea in its entirety 

 which we shall discuss later, but the formation of a poison resembling 

 that of anaphylaxis, by a digestive action of complement fixed to the 

 antigen by the antibody, seems to be well established, both as regards 

 in vitro and in vivo reactions. 



It would seem probable that proteins may yield a similar poison in 

 whatever way their hydrolysis is brought about, provided the cleav- 

 age is not too deep-seated. For example, Rosenow^' has found that 

 pneumococci and other bacteria, permitted to autolyze for a proper 

 length of time, produce poisonous substances with all the toxicologic 

 characters of the anaphylatoxin. Too extensive autolj^sis again de- 

 stroys the poison, which is also produced by digestion of pneumo- 

 cocci w'ith serum from normal guinea-pigs, and more rapidly with serum 

 from sensitizeti animals, w'hich likewise causes a demonstrably more 

 rapid proteolysis. The pneumococcus anaphylatoxic poison is soluble 

 in ether and seems to be a base, containing amino-acids, but Fried- 

 berger chd not find anaphylatoxin made from serum proteins to be 

 soluble in ether or alcohol, nor was it precipitated with the globulins. 

 The so-called "Abderhalden method" of sero-diagnosis of pregnancy, 

 which depends on the presence of specific proteolytic properties in the 

 blood, is an especially studied instance of these principles, and is 

 discussed later. 



Presumably anaphylactic intoxication is hut an exaggeration of the 

 normal process of defense of the body against foreign proteins {including 



'^'^ Friedberger's explanation of the inhibiting effect of salt as interference with 

 complement action, has been questioned. (See Zinsser, Arch. Int. Med., 1915 

 (16). 238.) 



2- See Besredka et al, Zeit. Immunitat., 1912 (16), 249. 



" Jour. Infec. Dis., 1912 (11), 94 and 235. 



