170 University of California Publications in Zoology. [Vol. 4 



The Problem. — After the impulse to turn in a certain way has 

 appeared, some of the rays bend down, twist, attach themselves 

 and pull ; others release, straighten, lift up and swing over ; some 

 walk in one direction, some in another,— all co-operating to pro- 

 duce the righting. Our present question is: What determines 

 each of these actions 1 



This is an entirely different question from that which asks of 

 the factors determining the action of the different parts before 

 the impulse is formed, and a much more difficult one. Before the 

 impulse is formed we can usually give some definite external 

 stinuilus or internal relation as the determining factor in the be- 

 havior of a given ray or other part; after the impulse appears 

 this can not readily be done. The external stimulus that pro- 

 duces a certain reaction before the impulse is set has no such 

 effect after the impulse appears. Thus, before the impulse, light 

 coming from one side causes the rays of that side to lift, of the 

 other side to bend down and take hold ; after the impulse has ap- 

 peared, the direction of the light has no effect. Before the im- 

 pulse, contact of the dorsal side of a ray with the bottom causes 

 that ray to twist, attach itself, and pull. After the impulse ap- 

 pears, this twisting and attaching occurs only in those rays that 

 have in some way been predetermined for this role, and they en- 

 deavor to perform the action whether they come in contact or not. 

 The other rays do not respond to this contact stimulus at all, or 

 they respond by lifting themselves from the contact. Before the 

 impulse, detaching certain rays causes neighboring rays to attach 

 themselves; after the impulse, the detached rays merely try in- 

 sistently to re-attach themselves, while the other rays persist in 

 trying to swing over freely. We could take almost every factor 

 mentioned in our list given on pages 135-142, and show that after 

 the impulse is formed its effect is either lost, or becomes entirely 

 different from what it was before the impulse. As we have seen. 

 before the impulse a certain configuration of external conditions 

 may determine a certain method of righting; after the impulse 

 this configuration may be completely changed, or its directive 

 action quite reversed, yet there is no effect on the method of right- 

 ing. The action becomes, as soon as the impulse appears, to a 

 considerable degree independent of external conditions; it is 



