CHANGES WROUGHT IN PATHOLOGY. 127 



facts won 'through a study of immunity. But the facts 

 of colloidal chemistry, together with advances in the 

 investigation of immunity, show clearly, it seems to me, 

 that the EHRLICH assumptions, in spite of the many 

 variables introduced into them, do not at all suffice to 

 explain the phenomena actually observed. 



It is, for example, an easy matter to foretell even now 

 that through changes in the manner and the rapidity with 

 which toxin and antitoxin are mixed, and through a 

 proper choice of animals, the varieties of toxins that 

 have been assumed to exist by EHRLICH might easily be 

 increased indefinitely. 



It can be readily seen, too, how difficulties arise in the 

 expansion of any theory that is based upon crystalloidal 

 chemistry. With EHRLICH it is the application of syn- 

 thetic organic chemistry, with his opponents, the appli- 

 cation of a special case of the dissociation of salts that 

 finally constitutes too narrow a frame to receive the 

 entire picture of facts. Insufficient also was the attempt 

 of DANYSZ and BORDET to explain the behavior of toxins. 

 The latter especially tried in his brilliant way to sup- 

 port his theory through an analogy with the process of 

 dyeing, which we now know to be a colloidal reaction. 

 In this theory the correct assumption that toxin and 

 antitoxin are able to unite in different proportions was 

 made; it could, however, be of value only as a hypo- 

 thetical objection, as it included only some of the possi- 

 bilities and was unable to explain, while lacking the broad 

 base of other facts in colloidal chemistry, the variety of 

 observations made on toxins. If we disregard my first 

 brief suggestion, pointing out the relation between im- 

 munity reactions and colloidal changes in state, 



