298 PROTEIN POISONS 



twice with salt solution, and then incubated with the 

 serum of a normal guinea-pig. The ferment in the serum 

 split up the precipitate with the liberation of the poison, 

 and when the serum containing the poison was injected 

 into a fresh guinea-pig the animal promptly died, with all 

 the symptoms of anaphylactic shock. Washed stroma 

 treated in a similar way yielded the same poison. With 

 whole red corpuscles the results are complicated by the 

 poisonous action of the liberated hemoglobin, which is an 

 active poison without further cleavage. Friedberger desig- 

 nates the precipitates, stroma, and corpuscles subjected to 

 the action of the normal serum as antigens. We have a 

 marked antipathy to the use of this term in discussing 

 the phenomena of anaphylaxis, and would designate the 

 substances submitted to the action of the normal serum as 

 substrates, and regard the serum as containing the ferment. 

 It might be said that the terms we use are of but little 

 importance, and the meaning is the one important thing. 

 This is true, but we use words to express ideas, and we 

 hold that the term "antigen" in this connection confuses 

 and tends to lead to gross misconception. As has been 

 shown by Friedberger and others, anaphylatoxin may be 

 obtained by incubating various proteins with normal 

 serum, and w r hy should we call the substances thus split 

 up by the ferment in the serum an antigen? Would it not 

 be just as proper to denominate starch which is converted 

 into sugar by amylase an "antigen?" But this is a digres- 

 sion, and we will return to the work of Friedberger and 

 Vallardi. They found that when specific precipitates and 

 stroma were used the amount of the poison obtained was 

 not in proportion to the amount of substrate used. With 

 a larger amount of substrate, that of the serum (ferment) 

 being constant, they obtained no poison. With the amount 

 of substrate very small, they obtained either no poison or 

 at least not enough to demonstrate its presence by its 

 effect on the animal. They obtained positive results, as 

 shown by anaphylactic death only when the amount of 

 substrate but slightly exceeded that necessary to kill a 



