284 PROTEIN POISONS 



It will be seen that the serum and organ extracts of 

 sensitized guinea-pigs contain an agent which when mixed 

 with homologous anaphylactogens in vitro in proper pro- 

 portions and incubated for the proper time produces 

 a poison which when injected intracardiacly into fresh 

 animals causes typical anaphylactic shock and sudden 

 death. This poison-producing agent is a ferment and is 

 inactivated by a temperature of 56 and reactivated on the 

 addition of serum or organ extracts from normal animals. 

 Like the toxins and many other ferments this one consists 

 of amboceptor and complement. The latter is destroyed 

 by a temperature of 56, but being a constituent of normal 

 serum and organ extracts, its loss is made good on the 

 addition of these substances. The ferments formed in 

 anaphylaxis are strictly specific, their specificity being 

 determined by the anaphylactogen and residing in the 

 amboceptor. The ferment elaborated in anaphylaxis, 

 like the toxins, consists of amboceptor and complement. 

 The anaphylactogen is not a toxin and the substance 

 produced in the body under its influence is not an antitoxin. 

 The anaphylactogen does not even contain a toxin group; 

 it contains. a poison, and it is this that is set free on reinjec- 

 tion. As we have stated, there is no more justification in 

 calling the anaphylactic ferment an antibody than there 

 would be in designating the proteolytic ferments of the 

 alimentary canal antibodies. 



The Poison. While anaphylactogens and anaphylactic 

 ferments are specific the poison is not specific. It is one 

 and the same thing whatever the anaphylactogen, and in 

 our opinion it is the poisonous group in the protein mole- 

 cule. Our studies on the protein poison, done before the 

 phenomena of anaphylaxis were known, demonstrated the 

 presence of a poisonous group in widely diversified proteins, 

 and it probably exists in all true proteins. We found it in 

 bacteria, both saprophytic and pathogenic, and, as has 

 been stated, we then were convinced that the pathogenicity 

 of bacteria bears no relation to the poison content of the 

 molecule of its cellular protein. The pathogenicity of a 



