256 



Vniversity of California Publications in Zoology [Vol. 20 



But it was decided to conduct one more experiment, for corroboration, 

 under slightly different conditions and with especial care in fixation 

 of the material. It had been noted that though often both types 

 occurred in the same pig (which was the case in most of the samples 

 used in the first measurements), still one type might be greatly in 

 excess, or there might be only one type present. In getting material 



Fig. B. Graphic representation of the difference in ratio of length to breadth 

 between two carefully selected lots, of 100 individuals each, of Balantidium from 

 separate hosts. The continuous line represents those from one pig (nearly all are 

 Balantidium coli) ; the broken line, those from the other pig (nearly pure infection 

 with Balantidium s-ais). 



Fig. C. This graph shows the variation in the ratio of length to breadth among 

 100 Balantidium secured from a case of balantidiasis in man. 



for this second set of measurements, it seemed best to take it from pigs 

 which had, as nearly as could be determined, pure infections of the 

 respective types. Fortunately these requirements were fulfilled in 

 the next lot of pigs examined. Both samples of material, the one con- 

 taining the ovoid and the one containing the elongate type, were 

 treated in exactly the same wav. They were killed at the same time 



