526 BTUDIES IN IMMUNITY. 



We know how to recognize a given bacterium by means of a 

 specific scrum. Hut is the character which this organism { >ossesses 

 of reacting with the appropriate scrum constantly present? May 

 it not disappear when the organism has been subjected to certain 

 exigencies? What, for example, is the effect ot change in culture 

 medium? As early as 1896 Metchnikoff and I noted that the 

 cholera vibrio as a result of certain vicissitudes, such as remaining 

 with leucocytes, comes to resist the agglutinating effect of cholera 

 serum markedly, and since that time many analogous instances 

 have been collected. The remarkable work of Grassberger and 

 Schattenfroh on symptomatic anthrax gives most interesting infor- 

 mation on this point. These authors have found that a serum that 

 is capable of agglutinating the organism when it is cultivated in a 

 given medium, has scarcely any agglutinating effect on it when 

 grown in a different medium. 



\Xc could understand this fact by supposing that the organism 

 has, on account of certain vital reactions, succeeded in resisting the 

 effects of the agglutinin, although it has kept its property of com- 

 bining with it; in other words, we would be dealing with the simple 

 phenomenon of adaptation to a harmful effect by the formation of 

 a refractory condition. But as we shall see, the modification is 

 still more profound. As a matter of fact, owing to certain con- 

 ditions of life, the organism may completely lose the antigen, 

 which combines with the antibody and which is its principal char- 

 acteristic. In other words, the distinctive sign in serum diagnosis 

 may be lost. As far as the action of the serum is concerned, the 

 organism thenceforth acts as if it belonged to a different species, 

 which proves distinctly that specific sera (or at least the agglutinins) 

 do not act primarily on those substances of the bacteria which are 

 essential to their life and characteristic of their particular species 

 but on certain of the accessory substances of the organism which 

 may occur, but the existence of which is in no way part of the fixed 

 hereditary character, which gives the living organism its particular 

 appearance and autonomy. 



The whooping-cough bacillus is very suitable for a study of this 

 sort. We find that this organism is very variable, not only as 

 regards the appearance of its growth, but, more important still, as 

 regards its antigen, in accordance with the conditions under which 



