CONCLUSIONS 275 



such a connection exists has been established with a 

 high degree of probability by clinical observers, but 

 the actual details of the relationship have not been 

 satisfactorily established. In our discussion of the 

 relation of bacterial infection to cytost intoxication, 

 it was pointed out that bacterial infection, although 

 localized in a particular area, may cause the liber- 

 ation of considerable quantities of cytost in much 

 the same manner as would any other localized tissue 

 damage. 



If such a condition persists for some time, the ani- 

 mal is continually subjected to a concentration of 

 cytost analogous to that which occurs when sublethal 

 quantities of this substance are injected at frequent 

 intervals during a protracted period of time. In con- 

 sequence, the cytost liberated by the activities of the 

 bacteria in foci of infection may be expected to bring 

 about a degeneration of the tissues of endodermal 

 origin in much the same manner as was found to be 

 the case in our experiments. The writer feels confi- 

 dent that an investigation of this mechanism would 

 yield results of considerable interest in the interpre- 

 tation of some disease processes. 



When a large number of animals is injected with 

 equal quantities of the same extract of autolyzed tis- 

 sue, it is found that they react to the injection in vary- 

 ing degrees. If the quantity is sufficiently great, the 

 larger proportion of the animals develops shock and 

 dies. Certain other members of the group develop a 

 severe shock, and recover, apparently none the worse 

 for their experience, while a lesser number reacts but 

 slightly to the injection. This, of course, is a common 

 finding in all types of biological experimentation, and 



