*No . of bacteria in 50 neutrophile leucocytes. 

 **Animal providing serum. 



While this table reveals a considerable variation in the 

 "cytophagic" indices of the normal and infected animals, the 

 nature of this variation is by no means constant. Sometimes 

 the index for the treated animal is lower for the non-specific 

 bacteria than in the control animal but quite as often it is 

 higher. Altogether the variations observed are no greater than 

 the v;ork of Fleming (34) would lead us to expect for two lots 

 of normal leucocytes. V/hile experiments of this kind are exceed- 

 ingly difficult to control and dangerous to interpret, it is 

 contended that the results set forth in the above table are 

 sufficiently definite to indicate that there is no invariable 

 decline in the function of the phagocytes during the last stages 



of fatal infection. 



(54) 



