Pomona College, Claremont, California 49 



movements more than the tips of the arms. After it has been at 

 rest for a time however the ej^e-spots are usually so placed as to 

 be protected from the light. The starfish in each case (Jennings) 

 moves towards that part of the body that is least illuminated. 



There are a number of ways in which starfish right them- 

 selves according to Jennings : 



1. The simplest method. Two adjacent arms twist their tips 

 with ventral faces inwards. 



2. Two arms, the ventral faces not inwards but facing in the 

 same direction. 



3. Three adjacent rays attack and usually turn by twisting 

 the outward rays. 



4. Four rays take hold, two to right, two to left. Fifth ray 

 helped up, and swings over. 



5. All rays attack release later of certain rays. 



6. One ray twists and rights the whole. 



7. Righting without attaching tube feet of any of the rays. 

 Raises disc strands on tips of arms then topples over. 



If a starfish begins a reaction in a certain way it usually con- 

 tinues in the same way even in spite of opposing conditions. 

 When the starfish gets started it continues in the same way. The 

 variability of form in starfish that are righting themselves is 

 very great. No species rights itself in one way alone. When cer- 

 tain tube-feet are prevented from acting in righting movements 

 the others carry on the movements. In righting if one method 

 does not help another is used. 



Habit Formation 



Preyer, 1886, Jennings, 1907, have brought further information 

 as the results of experiments to test habit formations in starfish. 

 By perventing certain rays to act in the righting reactions in star- 

 fish Jennings showed that he could establish temporary habits 

 and the slower formation of more lasting habits. The many 

 factors which determine the righting reactions have not a constant 

 tendency to make starfish turn on one given pair of rays. On the 

 contrary, they must sometimes act in one way, sometimes in an- 

 other. Therefore no very fixed habits are formed under normal 

 conditions. 



In the righting reactions the impulse tends towards the ac- 

 complishment of the general turning of the starfi.sh as a whole and 

 given parts sacrifice their own direction or even prevent it in the 

 general result. 



We cannot assume single specific external stimuli as the deter- 

 mining factors for each separate movement, yet in some way the 

 relation of the organism to its environment has set in operation a 

 uniform action of which separate movements are parts. 



