BEHAVIOR IN LOWER METAZOA 239 



a notable example of behavior that is not stereotyped, but is flexible and 

 variable. The usual course of this reaction is as follows: After the 

 starfish has been placed on its back, it extends its tube feet and moves 

 them about in all directions. At the same time the tips of the arms be- 

 come twisted, so that some of the tube feet are directed downward. In 

 this way, after a time, some of the feet become attached to the bottom. 

 These begin to pull on the arm to which they belong, turning it farther 

 over and bringing other tube feet into contact with the bottom; these 

 now assist in the process. If two or three adjacent rays become thus 

 attached, the other rays cease their searching, twisting movements, and 

 allow themselves to be turned over by the activities of the tube feet of 

 the attached rays. If two or more opposite rays become attached to 

 the bottom in such a way that they oppose each other, then one releases 

 its hold, and allows the turning to be accomplished by the opposing rays. 

 It is evident that the reaction is an example of the performance of varied 

 movements under stimulation, with selection from the conditions re- 

 sulting from these movements. Certain features in this reaction are of 

 special interest. At first all the tube feet and rays try to find an attach- 

 ment. When certain ones have succeeded, this is in some way recog- 

 nized by those parts whose action would oppose the movement, for these 

 cease their attempts, or even release the hold already attained. In some 

 way the physiological state corresponding to "success" in -certain rays 

 is transmitted to the other rays, and they change their behavior accord- 

 ingly. 



Variability and flexibility are the essence of such behavior. This 

 is well illustrated by study of repetitions of the righting reaction in the 

 starfish. It is by no means always the same arm or combination of 

 arms that initiates and finally brings about the turning. The essential 

 point is to get started in some way, then to continue on the basis of the 

 start made. Preyer (1886) studied this behavior in the starfish with 

 great care. He says : " Neither in one [species] nor the other is the method 

 of turning always the same. I have likewise seen Asterias glacialis, 

 which was several times in succession turned on its back without change 

 in the outer conditions, right itself sometimes in one manner, sometimes 

 in another. The spirals of the twisted arms do not work each time in 

 corresponding directions, but at first the neighboring arms often oppose 

 each other. But soon the correction takes place, in that the attached 

 feet stop those that are disturbing the turning, and the wrongly twisted 

 radii straighten out again. . . . The variability of form in starfish 

 that are righting themselves is great, and no species rights itself in only 

 one way. . . . But here, too, it is true that no Astropecten rights itself 

 twice in succession in exactly the same way. An adaptation to the sur- 



