328 BACTERIOLOGY. 



Ehrlich sense, and, upon injection of the mix- 

 ture into white mice no influence of the toxin 

 was to be noted. Had it been a real neutrali- 

 zation we should expect that a correspondingly 

 large quantity of this mixture would have had 

 no effect upon other animals. When the mix- 

 ture was inoculated into guinea-pigs, however, 

 signs of tetanus poisoning made their appear- 

 ance. In reality, therefore, toxin and antitoxin 

 could not have acted directly one upon another, 

 but both substances must have exerted their 

 effect in the animal organism side by side. A 

 partial nullification and prevention of the toxic 

 action does appear, to be sure, in such cases, but 

 I was able to bring about the same result by 

 mixing enzymes with the toxin. A mixture 

 of canine serum and rabbit serum acts less 

 powerfully upon the typhoid bacteria than does 

 either serum by itself (Buchner). One kind 

 of active proteid body, therefore, seems to act 

 to some extent upon another, but this action 

 is not always sufficient to destroy the toxic 

 power of a mixture, and on the other hand an 

 antagonistic action of this sort is not usually 

 specific, but is wholly general in character ; for, 

 according to Behring, it is possible to dimin- 

 ish the toxic power of the culture fluids with 

 use of iodine terchloride, and even, according 

 to Vaillard, with simple iodine. 



