CANNIBALISM IN AMOEBA 681 



soon as the organism was able to adjust the physical state of 

 its protoplasm to that of the fluid around it. 



It may also be suggested that, if the vacuole in which the 

 sphere had been enclosed were a food vacuole, the sphere, 

 when set free, would be suffering from the effects of the attempt 

 of the amoeba to digest it and would therefore naturally be in 

 a pathological condition, and that the contractile vacuoles would 

 be among the first of the organellae to betray this condition. 



The contractile vacuoles arose deep in the protoplasm of the 

 sphere and could be seen to move to the surface, when they 

 were ready to burst. Often in doing so they glided between 

 the granules in the protoplasm, and were then compressed into 

 a dumb-bell shape when they passed between the granules, much 

 as an air bubble is distorted when it is pressed under a cover-glass. 



The pulsation rate was not very regular. It varied from as 

 much as one contraction every quarter of a minute to one every 

 six and a quarter minutes, the average being about every 

 minute or rather less. Doflein's experiments (9) showed that 

 high concentration in the medium, such as would be hkely to 

 occur in a hanging drop, induces slow pulsation and a decrease 

 in size of the contractile vacuole. Some such influence probably 

 in part explains the irregularity observed here ; but no definite 

 evidence can be offered either in support of or against this view. 

 The slowest pulsation seemed to occur in the spheres with several 

 contractile vacuoles. 



When several contractile vacuoles were present they often 

 burst simultaneously, leaving the sphere free from them ; but 

 when only two were present they seemed to alternate, one 

 bursting while the other one grew, so that the sphere always 

 contained one. Further, when several were present, two half- 

 grown ones often fused to form one larger one, which then 

 moved to the surface and burst. 



The contractile vacuoles did not appear constantly in any 

 one position in the sphere, but, after bursting, might reappear 

 anywhere. That this is not a false impression produced by 

 rolling over of the sphere is shown by the fact that it was 

 observed in perfectly motionless spheres, and also by the fact 

 that when th© bursting of one set of four was delayed the second 



