154 MODIFICATIONS OF BEHAVIOR 
bent over to the right. Now when stimulated it contracted 
as before, then bent regularly, in extending, over to the 
right. It seemed to have acquired a new habit—bending 
to the right instead of to the left.”” Observation showed 
that the bend of the body persisted in the contracted as well 
as the extended state, and that all parts extended propor- 
tionately and thus led to a repetition of the previous action. 
The habit of action in this case had a persistent structural 
basis in the bend of the body. If now the anemone was 
caused to contract very strongly in all parts so that it was 
no longer bent to one side in the contracted state it would 
lose its habit of bending when it subsequently expanded. 
The réle of the nervous system in this case is a very doubtful 
one and it is not improbable that the so-called habit is 
merely a result of purely mechanical factors. The same 
may also be true of the habit of irregular bendings acquired 
as a result of living in crevices between the rocks, which 
might be compared to the difficulty experienced by a person 
whose body and limbs have been confined for a long time 
in any one position in making new movements. 
Jennings has performed numerous experiments on the 
formation of righting habits in the starfish Asterias forret. 
Starfish when placed on their backs have several methods 
of turning over. Different individuals have their personal 
peculiarities in this as in other respects, and in most speci- 
mens there is a tendency to use one particular ray or set of 
rays upon which to turn. These differences as Moore 
has pointed out may be due to inequalities in the size of 
the arms, injuries to certain arms, or any initial twist an 
arm may have had due to its previous position. Before 
any attempt to train a starfish was made the animal was 
put through a set of tests to determine the rays most com- 
monly employed in the righting reaction. After this the 
