164 TJniversity of California Publications in Zoology. [Vol. 4 



There was at first the same difficulty in getting the starfish B 

 to use a and d that was found inducing the starfish A to use 

 b and c. In B, it was the ray c that was particularly obstinate. 

 In time this was overcome, and the rays c and d took hold 

 readily. But their tips usually turned both in the same direction, 

 making the turning very difficult. The experimenter was there- 

 fore compelled to spend much time in drilling the animal to place 

 these rays properly. For about a week the results of this were 

 discouraging, but at the end of about that period, the effects of 

 the training showed clearly, and the starfish when turned on its 

 back usually at once took hold w^ith c and d in a coordinated way, 

 and turned itself neatly. Now it was often unnecessary in the 

 course of the lessons to interfere with the animal at all; it at 

 once turned in the way desired. However, for the present I 

 kept close watch and interfered as soon as any other ray showed 

 an inclination to meddle in the matter. 



After seven days of training, I determined to test the results. 

 The starfish was allowed to go twenty-four hours without lessons 

 or experiments of any sort. It was then given ten trials, being 

 permitted to turn freely, without interference. 



In these ten trials it turned neatly "five times on c and d, 

 while the next highest number for any pair was two, for a and b. 

 This result appears remarkable in view of the fact that the com- 

 bination c -\- d was never used at the beginning ; and it becomes 

 especially striking when ^e recall M^hat a struggle was required 

 for several days to make c and d work together at all. 



Now ten lessons a day were given for another week. Then 

 after forty-eight hours ten free trials were given. The starfish 

 turned five times out of the ten on c -\- d. 



The starfish was now allowed to rest 5 days, so that it had 

 gone seven days without training. Then, on the morning of 

 August 16, it was given ten trials, turning four times on c -\- d. 

 The effects of the training were thus still very evident. 



On the afternoon of this same day 15 more trials were given, 

 and now the results of training did not appear so plainly. The 

 combination c -\- d was used only once, while d -\- e and a -\-b 

 M'ere each employed five times. In the entire 25 trials given on 

 this day, the different rays were employed as follows : 



