78 MECHANISM OF INFECTION 



of the evolution of anthrax in the white rat has given us the 

 most interesting side lights on the nature of the reciprocal 

 relations between bacteria and organism. We know by the 

 studies of Savtchenko that anthrax bacilli are very rapidly 

 destroyed in rat sera. However, by starting the first 

 cultures in a mixture of a very small quantity of serum with 

 a large quantity of ordinary bouillon, and by continuing 

 passages through mixtures containing relatively larger and 

 larger quantities of serum, we finally obtain a fairly abund- 

 ant culture in pure-rat serum; that is, we obtain a race of 

 bacteria which resists the destructive action of this serum. 

 If we again transplant this race into ordinary bouillon and 

 if we make several transplants in it, we find that even in 

 spite of the change in media, that the properties of serum 

 resistance are obtained. 



If one of these twenty-four-hour cultures is filtered through 

 a porcelain bougie so as to completely remove the bacteria 

 and if a small amount of the filtrate is added to rat serum, 

 and this mixture is added to broth which is then inoculated 

 with a non-serum-resistant strain, a growth will be obtained 

 quite as abundant as in ordinary broth. Anthrax bacilli 

 can thus acquire the ability little by little to fix and digest 

 rat serum by increasing the quantity of a "fixing substance." 

 The ability to produce this substance is retained even when 

 there is no longer any rat serum to stimulate it. The excess 

 given off to the exterior and when added to rat serum, can 

 neutralize in vitro the bactericidal properties of the serum, 

 and again render it all the more assimilable for a non-serum- 

 resistant race. Finally a filtered broth culture originating 

 from a non-serum-resistant culture mixed in equal propor- 

 tions with rat serum will hinder the bactericidal action of 

 the latter, but in a much less degree. 



Anthrax bacilli can produce then a specific antibody 

 under different conditions, but by the same process as para- 

 typhoid as studied above. Moreover, contrary to suppo- 

 sition, a culture of anthrax avirulent for rats does not 

 become virulent when rendered serum-resistant; so that we 

 must conclude that it is not by acquiring an affinity for 

 any animal substance that a bacterium can become patho- 

 genic for the animal, and that it is not by virtue of its 



