1907] Jennings. — Behavior of the Starfish. 171 



largely determined internally, so that what each part of the star- 

 fish does may be said in brief to depend mainly on the direction 

 of the impnlse. 



Effect of External Conditions.— Yet the external conditions 

 of course do affect the course of the action to a certain degree. 

 A certain ray, for example, feels about in the water till it comes 

 in contact with the bottom ; thereupon it takes hold and now be- 

 gins to pull ; if it did not come in contact with an external object 

 it of course would not thus change its action. But under the uni- 

 fied impulse there is this peculiarity in the reaction to external 

 conditions : each part reacts to the given external stimulus in such 

 a way as to assist in the righting. One ray, therefore, reacts to 

 contact by holding and pulling, another by releasing, another by 

 walking, etc., as the progress of the reaction requires. It is 

 therefore evident that the external stimulus is not the easential 

 determining factor in these actions. How then are they deter- 

 mined, and wherein consists the unity of the action? 



Action of Parts in the Interest of the Whole. — A tempting 

 suggestion comes to the mind, one which if correct might at least 

 lead to a simplified formulation of the unity of action, with a 

 consequent simplification of the problem which it presents. This 

 suggestion is, that every part of the starfish performs that move- 

 ment which most directly restores it {i.e., this part) to its normal 

 position, with ventral side down. Every movement performed by 

 a given ray, for example, would tend directly to restore the nor- 

 mal position of that ray, and a given ray would perform no move- 

 ment that tends to oppose this regaining of its normal position. 

 If this statement were correct we should at least be able to ac- 

 count for the movements of each part mainly by the present con- 

 dition of this part, as determined by its own direct relation to the 

 environment. We should be able to avoid such statements as that 

 one part acts in the interest, not of itself, but of other parts, or of 

 the organism as a whole ; or that there is a sort of worked-out 

 plan in the behavior, requiring the subordination of certain parts 

 to others. The reaction would still be in many respects difficult 

 to understand, but much less so than in case this suggestion turns 

 out a mistaken one. 



The view suggested above could perhaps be maintained for 



