1907] Jennings. — Behavior of tJie Starfish. 161 



First, however, it was decided to test whether the effects of 

 the training: thus far given would last. The starfish was left for 

 two days undisturbed, then it was given ten trials in succession 

 at three minute intervals. The result showed that the effect of 

 the training had been lost. The ray e was used 8 times; d, 7 

 times ; a, 3 times ; c, once and b only once. 



Now, on July 24, a long series of lessons was begun, ten every 

 day at three minute intervals. In these experiments the starfish 

 was so far as possible driven to turn on b and c, by preventing 

 a, d and e from attaching themselves, or by pulling the latter 

 away if they became attached. 



In the first lessons there was again much difficulty in getting 

 b to play the active role ; in several experiments it absolutely 

 refused, so that c was forced to turn the animal unaided. But 

 after four lessons, b began to take hold more readily. 



Now there appeared again the difficulty caused by the unco- 

 ordinated twisting of the tips of b and c, these usually facing in 

 opposite directions. It was frequently necessary to release one 

 or both of these rays after they had become attached, and to 

 turn the starfish into such a position that the proper side of the 

 ray first touched the bottom. By holding the animal in this posi- 

 tion for a time, the tube feet were induced to come out and 

 attach themselves on the proper side. Thus by a tedious process 

 the two rays were induced to take hold in a coordinated way, 

 and to turn the starfish neatly. 



At the end of the first ten lessons of this new series, b and c 

 had acquired a fairly ready inclination to attach themselves, 

 though not in a coordinated way. But if permitted, the other 

 rays attached themselves later, so that the righting reaction be- 

 came slow, complicated and comparatively ineffective. The star- 

 fish was in a transitional stage, when neither the old nor the new 

 way was neatly used, but the two interfered somewhat. 



On July 25, b and c still took hold readily, though usually in 

 an uncoordinated way. The rays just mentioned usually at- 

 tached themselves first; then other rays took hold if permitted, 

 confusing the reaction. The ray e was still bound to play its 

 original part, interfering with the turning inaugurated by b 

 and c. 



