AN AMICRONUCLEATE OXYTRICHA 497 



supplied with fresh culture medium or not. From this time on, 

 however, since no fresh medium was added, both the supply of 

 food became scanty and the excretion products of the animals 

 accumulated, a balanced relation of the animals and their envi- 

 ronment was reached, and little or no further multiplication by 

 fission took place. Just at the time this condition was reached, 

 pairs were observed sticking together and cannibals began to 

 appear. In all cases pairs and cannibals were isolated as soon 

 as observed. In the second subculture cannibals appeared on 

 the same day. 



It will be seen that the appearance of cannibals in both the 

 first and second subcultures is due essentially to the same cause, 

 since in the latter case the medium was changed daily and the 

 removal of waste products plus the added food supply gave a 

 more rapid increase of the animals. This rapid increase, which 

 in turn tended to bring about a state of balance rapidly by the 

 using up of the food supply and the increase of excretion prod- 

 ucts, is most marked in mass cultures 2B and 2C. In mass 

 culture 2A an accident on the seventh day led to the loss of over 

 half the animals. The sudden drop in division rate in mass 

 culture 2D I am at a loss to explain. Mass cultures 2B and 2C 

 were most instructive. In these cultures there was an initial 

 rapid multiplication. Just after this period of multiplication 

 had reached its climax, but before a state of equilibrium had 

 been reached, for numerous divisions by fission were still taking 

 place, cannibalism occurred and the animals showed also a 

 strong tendency to become united in pairs. Addition of fresh 

 culture medium prevented a permanent balanced state such as 

 that in the first set of cultures with the result that cannibal 

 formation proceeded steadily until, as in the case of 2B, 20 

 per cent of the animals on the slide were cannibals. 



The divison rate for subcultures 3 (see curve on p. 501) and 

 4 show that in the former set for the first five-day period repro- 

 duction took place more rapidly. This seemed somewhat sur- 

 prising in view of the results obtained by Woodruff ('11) in 

 studying the effect of excretion products on the division rate of 

 Paramecium. Evidently in the present study excretion products 



