BACTERIAL PROTEINS 31 



do so. This was observed when we were ad- 

 ministering the substance by intraperitoneal in- 

 jection. Then we found that the more finely 

 the substance was ground the smaller was the 

 fatal dose. "When the substance was only 

 coarsely ground in a porcelain mortar and sus- 

 pended in water it did not kill guinea-pigs on in- 

 traperitoneal injection in doses less than 1 to 

 40,000 parts body weight. When the same pow- 

 der was more finely ground in an agate mortar, 

 it killed 15 out of 16 animals at 1 to 75,000; 9 

 out of 28 at 1 to 100,000; 5 out of 8 at 1 to 

 200,000 ; 4 out of 34 at 1 to 2,000,000 body 

 weight. We observed that when heavy suspen- 

 sions were used, lumps of the substance re- 

 mained undissolved in the peritoneal cavity af- 

 ter death or recovery. In these observations 

 we found the solution of our puzzle. The poi- 

 sonous action of the cellular substance is in pro- 

 portion to the extent to which and the rapidity 

 with which, it is split up by the secretions of the 

 body cells and this cleavage is determined by 

 the relative surface exposure of the substance 

 to the action of the cleavage agents. I dare 

 say that the difference in susceptibility as 

 shown among the individual animals is due to 

 the abundance and effectiveness of the secre- 

 tions elaborated by the body cells. 



