/ 



190^] Jennings. — Behavior of the Starpsh. 125 



ward" the rays a and b likewise signifies that these are the rays 

 that attach themselves and pull the others over. 



First, it may be said that my observations fully confirm Prey- 

 er's statement as to the variability of the righting reaction. It 

 could probably be said, in a word, that the starfish may, and does, 

 in different cases, right itself in any conceivable way, — and in- 

 deed, in many ways that would not readily be conceived before 

 they were observed. But a mere statement of the fact of varia- 

 bility carries little insight. The variations may be grouped about 

 a number of typical methods, and a statement of these with their 

 modifications will give an insight into the nature of the reaction 

 and the kind and extent of the variations. 



The main types of the righting reaction, with the variations 

 of each, are the following: 



1. The simplest and neatest method of turning is the follow- 

 ing: Two adjacent rays twist their tips in such a way that the 

 ventral surfaces of the two face each other (fig. 12) ; then the 

 tube feet of these rays attach themselves and throw the starfish 

 over (fig. 13, 14, 15) in a neat somersault. Any two adjacent 

 rays may act in this way, so that in a given starfish five different 

 variations are possible. 



2. The tips of the two adjacent rays may so twist that the 

 ventral surfaces do not face each other, but both face in the same 

 direction. The tube feet then take hold and throw the starfish 

 over, — twisting it about an axis which passes lengthwise through 

 one of the attached rays. This method of turning is extremely 

 difficult and awkward, but is seen at times. Usually when two 

 rays become attached in the way described, a third ray takes 

 hold and aids the turning, the method then forming a transition 

 to that given next. Any two of the rays may play the part we 

 have described, so that there are five possible variations. 



3. Three adjacent rays twist, attach themselves, and remain 

 attached, all pulling throughout the reaction. Usually the animal 

 turns primarily by the aid of the two outer rays, while the mid- 

 dle one is relatively passive, and is compelled to double back un- 

 der as the animal turns. Often this middle ray walks backward 

 beneath one of the other rays, or the other walks actively over 

 its surface or there is a combination of these two movements, till 



