338 J. M. D. OLMSTED 



roundings in all directions. If some twenty or more individu- 

 als are placed in a finger-bowl, they at first cling together in a 

 dense wriggling mass. Presently they begin to separate. The 

 tentacles attach to the bottom, then to the sides of the dish, but 

 their posterior ends remain hooked into one another. The tenta- 

 cles then carry the oral ends up to the top of the water, the pos- 

 terior ends still holding to one an other, until the mass of animals is 

 spread out in a sheet, each individual like a spoke in a wheel. 

 This would seem to indicate a positive reaction to mechanical 

 stimuli at the posterior end of the animal, yet in trials in which 

 this portion was gently touched with a blunt needle, only nega- 

 tive reactions were given. 



Although holothurians exhibit the usual radial symmetry com- 

 mon to all echinoderms, they are decidedly more bilateral than 

 other classes of the phylum. The surfaces to which are attached 

 mesenteries that hold the intestine in place, are called dorsal and 

 ventral. The only author who shows that this bilateralsymmetry 

 in the apodous holothurians is connected with their behavior is 

 Semon ('87). He found that Synapta digitata and S. hispida 

 normally creep with one side, the dorsal, uppermost. In Syn- 

 aptula hydriformis, likewise, there is a tendency to keep one 

 side away from the substrate. This surface is decidedly darker 

 and contains more miliary granules than the opposite one. Clark 

 ('98) found no such difference in specimens of this species which 

 he obtained in Jamaica, but it is certainly very pronounced in 

 those of Bermuda. If a Synaptula which is moving horizontally 

 along the bottom of a dish is gently turned over so that the lighter 

 colored side is uppermost, the animal may draw in the tentacles 

 completely, contract the body, roll slightly to the right or left, 

 and then move on again with the dark side up. Usually, how- 

 ever, the animal, when turned over, continues locomotion using 

 the tentacles of the lighter side to pull itself along, rolling over 

 gradually until the dark side again appears uppermost. The 

 average time required for a Synaptula to right itself in this way 

 was 45 seconds. In ordinary undisturbed locomotion a few cases 

 were noted where an individual rotated on its long axis, but this 

 was by no means the usual procedure, either when crawling up 



