130 S. J. HOLMES 



upon an extended, tube foot is to cause it to contract in an 

 irregular manner. The tube feet do not seem to show any 

 particular response to shadows. 



The reactions of the spines to light are independent of the 

 central nervous system. Pieces of shell from the inner side of 

 which all traces of the radial nerves were scraped away, were 

 exposed to a concentrated beam which had traversed several 

 inches of water. The spines would still bend toward the illu- 

 minated spot and could be made to bend over in any desired 

 direction by throwing the beam on different points. The response 

 of the spines to light depends upon the superficial nervous 

 plexus as Von Uexkull found to be the case for their reactions 

 to mechanical stimuli. How far these individual reactions of 

 the parts help us to explain the behavior of the animal as a 

 whole is a subject reserved for later discussion. 



EFFECT OF REMOVING THE SPINES 



In several individuals the spines were entirely removed from 

 the lower hemisphere of the test, leaving only spines which 

 were incapable of touching the bottom. The specimens were 

 then placed in a dish of sea water and exposed to the light. 

 Soon they began slowly crawling away from the light and all 

 assembled at the negative end of the dish. When the ends of 

 the dish were reversed the specimens crawled away from the 

 light as before and continued to do so in several other trials. 

 The rate of movement was considerably slower than that of 

 normal individuals. 



Removal of the spines and tube feet from one lateral half 

 of the body produced circus movements in most cases which 

 brought the animal slowly away from the light in an irregular 

 course. 



EFFECT OF REMOVING THE TUBE FEET 



Several times individuals were observed walking on the tips 

 of their spines with the test raised about a centimeter from 

 the bottom. Careful observation with a large hand lens, both 

 through the glass bottom of the dish and in other cases through 

 the sides when the specimen was placed near one edge, showed 

 that none of the tube feet reached the glass. To make it per- 

 fectly certain that none of the feet escaped observation amid 



