BEHAVIORISTIC STUDY OF THE RAT 
35 
leading into a revolving drum, the other into a 
food box in which food was left all the time. 
The animal could move freely from one cage to 
another as it chose. 
What then under these conditions happens to the 
periods of activity and inactivity set up through¬ 
out the 650 days of the animaFs life on the out¬ 
side? The results of this experiment are shown 
schematically in figure 19. In this figure only the 
activity in the drums is recorded. The first line 
gives the alternating periods of light and darkness 
in twelve hour periods 8:00 p.m. to 8:00 a.m. and 
8:00 a.m. to 8:00 p.m. for the 650 days of the 
animaFs life preceding the experiment. Also it 
shows the period of constant darkness during the 
experiment. The second line gives the probable 
relation of activity of the animal to the changes of 
light and darkness for the period of life before the 
experiment. It also gives schematically for twelve 
days the time relations of the revolutions of the 
drum made by the animal when placed in the 
environment of constant darkness. At the end of 
the twelve days the experiment had to be discon¬ 
tinued for unavoidable reasons. It may be seen 
that during these twelve days the regular daily 
alternation of activity and inactivity is fairly 
definitely maintained despite the constant dark¬ 
ness. How much longer this regular alternation 
of periods would have continued is difficult to 
conjecture. 
The fact of the persistence of rhythms of activity 
after the withdrawal of the rhythmic stimulus has 
been observed in a number of different animals in 
the past. Bohn and Keeble have interested them¬ 
selves very much with this phenomenon especially 
with relation to the life and habits of a small 
animal Convoluta roscoffensis which has its habitat 
