; MODE OF ACTION OF BACTERICIDAL SERA. 131 



" It has frequently happened to me that highly immunized guinea- 

 pigs died after an injection of moderate amounts of virus. On 

 section there were then found in the peritoneum living vibrios, some- 

 times even in considerable numbers. Notwithstanding this the 

 heart blood of the cadaver when introduced into new guinea-pigs 

 manifested the strongest power to dissolve vibrios." 



It is therefore conceivable that an individual can lose its natural 

 resistance by producing too large an amount of interbody in pro- 

 portion to the amount of its complement. Such an excess of inter- 

 body then would act injuriously rather than helpfully. 



This phenomenon is also of some theoretical significance. While 

 it can readily be explained by means of the views of Ehrlich and 

 Morgenroth, it appears, to us at least, to be absolutely irreconcilable 

 with the theory of Bordet. This author, as is well known, regards 

 Ehrlich's interbody as a substance capable of sensitizing the bac- 

 teria whereby they are made vulnerable to the action of the solvent 

 " alexin " (Ehrlich's complement). If this were the case it would 

 be absolutely incomprehensible how an excess of sensitizing sub- 

 stance could diminish the total effect; at the most such an excess 

 could only increase the sensitizing action, not decrease it. Since, 

 however, we have actually observed this decrease very frequently, 

 we must regard this as a weighty objection to Bordet's theory. 



Equally incomprehensible from Bordet's standpoint is the fol- 

 lowing observation. As has been shown above it is possible to over- 

 come the action of an equivalent mixture of interbody and com- 

 plement (the mixture itself being fatal) by adding a large excess 

 of interbody to it. When, however, through the addition of more 

 complement the equivalence of the mixture is again restored, the 

 action returns. This action therefore depends not only on the abso- 

 lute amounts of complement and interbody, but also and essentially 

 on the proportion in which these two substances are present. This 

 is true at least in the sense that not very much more interbody 

 should be present than is required. 



