22 HARRY H. WYLIE 



tain control series were used. After the animal had apparently- 

 learned the problem, it was given one or two series without 

 the light stimulus. The result was a drop in efficiency of 

 from 25% to 70% with an average drop of 40%, leaving an 

 average of 60% correct. This is not much above what 

 chance would enable the animal to do. Upon the light 

 being restored again an immediate return to the required 

 efficiency appeared again. This was taken as conclusive 

 evidence that the light stimulus was the dominant stimulus 

 in making the correct response when the percentage of 

 efficiency had reached 95. 



(2) Sound— Five groups of animals. III, IV, V, VI, VII, 

 were trained with the sound stimulus preparatory to trans- 

 ferring them to the light. In all 32 animals were trained in 

 these groups. The response required was the same simple 

 positive response required in the case of the experiments 

 with light. Groups III and IV used the sounder mentioned 

 in the description of the apparatus. Group V was trained 

 with the telephone receiver as sounder in place of the 

 regular electric sounder. Groups VI and VII were trained 

 with the telephone receiver as sounder and with pain as 

 punishment for a wrong choice. Groups III and V were 

 older animals than those used in the other groups and had 

 had training on the maze. They were all practically full 

 grown animals. 



The results can be summarized very briefly. All the mem- 

 bers of group III were given 85 series or 1700 trials without 

 learning the response. All the members of group IV received 

 60 series or 1200 trials without learning the response. All the 

 members of group V received 40 series or 800 trials without 

 learning the response. Three members of group VI received 

 40 series or 800 trials, one 35 series or 700 trials, and two 30 

 series or 600 trials. None learned the response. Four 

 members of group VII received 60 series or 1200 trials, 

 two 55 series or 1100 trials, and none learned the response. 



In order to bring to light the significant features of these 

 results we need to notice the method and results more in 

 detail. 



In the main, the method with the sound stimulus was the 

 same as with the light. The stimulus was given in a more 

 or less regular way, that is, the sounder was 'clicked' with 



