86 E. A. ANDREWS. 



In this connection it should be pointed out that in several 

 instances it was observed that free-swimmers intruded into 

 dwellings already occupied and nevertheless when the time for 

 transformation arrived, these intruders, one in each tube, secreted 

 a new house within the host's house; thus showing that the 

 presence of a surrounding house need not necessarily inhibit the 

 making of a house within a house. In such instances also there 

 was no intelligent result, since the new house blocked the passage 

 and neither host nor guest was ever able to feed nor to escape 

 from the predicament (except in one instance when a tear in the 

 side of the old tube enabled both animals to protrude their 



feeding arms). ( 



SUMMARY. 



In the protozoan Folliculina an individual imprisoned con- 

 tinued to pass from differentiated to dedifferentiated phases, 

 though both were exposed to the same restricted conditions. 



Imprisoned in the night of August 28 the feeding adult de- 

 differentiated into a swimming form which was active within 

 the prison all the forenoon of August 29. 



It then passed into the sedentary phase and omitting to con- 

 struct the usual dwelling, being inside the old one, differentiated 

 again into the perfect form by ten o'clock. 



During that night it again dedifferentiated into the swimming 

 form, still within the old dwelling to the bottom of which it had 

 twice fastened itself. 



At 8 140 in the morning of August 30 this swimming form had 

 partly emerged through a rent in the dwelling. 



This protruding anterior part of the animal swam violently, 

 held tethered to the dying posterior half left within the dwelling. 



At ten forty the free-swimming part settled down on the 

 outside of the dwelling and carried out the normal series of acts 

 in fabricating a new dwelling attached to the old one and almost 

 completed at 12:40. 



But at one P.M. the animal was destroyed and rhythmic 

 alternation thus ended. 



It is inferred that these alternations of form with differentiation 

 and dedifferentiation are not directly dependent upon such 

 external conditions as exposure to long stretches of water, feeding, 



