1907] Jennings. — Be]iavior of tlie Starfish. 133 



is perhaps the most significant thing seen in the entire series of 

 phenomena. As we have just seen, it is sometimes present at the 

 beginning. In other cases it appears later during the reaction, but 

 it is practically always evident before the righting is accomplished. 

 Only in 'the first stages of the behavior is there opposition of 

 action ; then suddenly this unified impulse appears and all opposi- 

 tion and incoordination is gone. This unified impulse is clearly 

 evident to the eye, and the precise moment of its appearance is 

 apparent, through the peculiar action of the tube feet. So long as 

 the diflferent rays are not acting in unity, the tube feet are directed 

 in all sorts of ways ; the unattached ones of a given ray are likely 

 to be extended in the direction in which that ray is trying to 

 pull the starfish, while those of the central disk are extended di- 

 rectly upward. But suddenly all this changes ; all the free tube 

 feet incline in a certain direction; a visible unified impulse to 

 turn toward a certain side appears at a stroke. This unified 

 direction of the feet is shown in fig. 13, while in fig. 12 it has not 

 yet appeared. Often at the moment this unified impulse appears, 

 it is the only indication as to the direction toward which the 

 starfish will turn, but it can be depended on to give a sure pre- 

 diction as to the next move. As soon as it appears, the attached 

 tube feet of those rays on the side toward which the starfish is 

 to turn, which I may call the goal (the right in fig. 13), begin to 

 pull with redoubled vigor, and the other tube feet of these rays 

 extend out and take hold as rapidly as possible. The attached 

 tube feet of the opposite rays (those of the off side, as we may 

 call them) at the same time let go their hold and extend toward 

 the goal (the side to which they are to swing), as if preparatory 

 to meeting the bottom. In the intermediate rays, lying trans- 

 versely, the tube feet participate in the coordinated movement 

 in one of two ways. Sometimes they release their hold, extend 

 toward the goal (or side toward which the turning is to occur), 

 and allow the rays bearing them to be carried over in the general 

 somersault. In other cases the tube feet of these transverse rays 

 release on the off side and bend toward the goal. Those on the 

 goal side retain their attachment, and new ones become attached 

 on this side. All pull toward the goal, aiding the turning, and 

 often causing these transverse rays to walk toward the goal side. 



