A GENERAL RESUME OF IMMUNITY. 519 



substances in the mixture. A comparison with a process of dyeing 

 is very apt; the molecules of the toxin would "stain" more or less 

 deeply by the antitoxin molecules, and the complexes that result 

 in the various instances are less toxic in proportion as they contain 

 more antitoxin and less toxin. I brought out this point of view 

 clearly in 1903 and discussed all the details of it so that no further 

 insistence on the subject is necessary. It may be added that this 

 interpretation has met with the support of various writers, among 

 whom may be mentioned Grassberger and Schattenfroh, Biltz and 

 Pauli, who have accepted it in its entirety. 



The essential point of difference between these different concep- 

 tions evidently lies in attributing a different composition to the mix- 

 ture of toxin and antitoxin and particularly to such mixtures as 

 contain too small a dose of the antidote to bring about complete 

 neutralization. In order to simplify it we may choose a toxic fluid 

 which contains a single poison and add to it a relatively small 

 amount of suitable antitoxin. In such a case are we to suppose 

 that the mixture includes both well neutralized toxin and an excess 

 of a free and intact toxin, or are we rather to believe, as I thought, 

 that the antitoxin is shared by the totality of the toxin present, 

 but saturates it only partially in the same way that a small 

 amount of dye does a large quantity of the substance to be dyed, by 

 spreading over the whole of it and staining it faintly? The complex 

 thus obtained, which is weak in antitoxin, would still have certain 

 toxic properties and would act like toxon. 



But how, it may be objected, can we determine with certainty 

 whether the toxic activity of such a mixture is due to a certain 

 amount of intact toxin or the presence of a partially saturated com- 

 plex? It is here that we must distinguish very carefully in the 

 effects of a poison between quality and quantity. I may recall 

 the observations which I offered in 1903 that showed that a hemoly- 

 sin may manifest toxicity in two different ways; when employed 

 in a small dose without being in any way affected by an antitoxin, 

 it hemolyzes a small amount of corpuscles rapidly, but owing to 

 the fact that it is not present in sufficient quantities it cannot 

 destroy many of them. On the other hand, a large dose of the same 

 hemolysin to which a little of its antitoxin has been added and 

 which has been transformed thereby, according to my idea, into an 



