IMMUNITY. 



22 5 



haptophores of antitoxin occupied by the toxin previously added. In this case 

 any additional toxin remains uncombined, and, if such a mixture is injected 

 into a guinea-pig, the animal is killed. 



Bordet's* explanation differs from Ehrlich's. Bordet does not admit the exist- 

 ence of toxons, and regards the paralysis attributed by Ehrlich to the action of this 

 hypothetical substance as due to weakened toxin. He explains the peculiar behav- 

 ior of a neutralized mixture of crude toxin with antitoxin, stated above, by assuming 

 that antitoxin is capable of taking up and neutralizing varying amounts of toxin. 



Pro I 'OK Old 



Toxen. 



C 60 'CO 2.00 



FIG. 54. " Spectrum " of very fresh crude toxin. 



He compares the effect of mixing toxin and antitoxin to that of mixing starch and 

 iodine: the more iodine added to the starch, the bluer the color. Let A repre- 

 sent, then, a certain amount of antitoxin; let A be capable of combining r, 2, 3, 

 4, 5, different amounts of toxin; call these amounts of toxin T I} T 2 , T 3 , T 4 , T s ; 

 and assume that a combination in which all the A's are combined with T's in 



Synloxoid 



y////////////// 



6O /GO 2&> 



FIG. 55. " Spectrum" of crude toxin as it is supposed practically 

 always to occur. 



the proportion of AT ly is neutral, that it has no poisonous properties; that a 

 combination represented by AT 2 also has no toxic properties, but that AT 3 , 

 would begin to show toxic properties, and that A T 4 is distinctly toxic, and that 

 AT S is very toxic. Starting with toxin, then, if. just enough antitoxin is added 

 to neutralize its poisonous properties, AT-i is first formed, which is not toxic; 

 now add more toxin, and none of this remains free, but, on the contrary, A T 2 is 

 formed, which is not toxic; on adding still more, when AT 4 or AT S is reached 



*Bordet. Toxines et Antitoxines. Annales de I'Institut Pasteur. 1903. 

 p. 161 et seq. 



15 



