54 HARRY H. WYLIE 



the animals to learn the response to the sound is not to 

 train them directly with the sound right from the first, 

 but is to train them first with the light, then give a few 

 simultaneous light and sound series, and then give the 

 sound alone. The latter method required only 24 series 

 altogether for the light, for the simultaneous light and 

 sound, and for the sound, while the former method required 

 35.03 series all told for the sound only. 



With Group VIII after learning the response to the light 

 and the sound of the regular electric sounder, tests were 

 also made with the sound produced by striking a three 

 inch bell with a rubber hammer. This bell was held by 

 the experimenter in front of the box and below the vision 

 of the animal. After tests with the bell, tests were also 

 made with wooden pipes of 2048, 1024, 512 and 256 vibra- 

 tions. These pipes were sometimes blown by the experi- 

 menter who handled the animals during the experiments 

 and sometimes by an assistant who stood just behind the 

 regular experimenter. In sounding the bell or pipes no 

 attempt was made to make a regular continuous sound. 

 The only attempt was to keep the sound of moderate loud- 

 ness whether given irregularly or continuously. After a 

 few trials with these devices as sounders, it was the usual 

 thing for the animal to turn back at the first sound of the 

 bell or the first " toot " of the pipe, that is, the transfer 

 was practically perfect. 



It was in these trials with the sound after having learned 

 the light, that there was some evidence of " position in the 

 alley " functioning as a stimulus to cause the return of the 

 animal, that is, the animal would turn back without any 

 sound stimulus having been given. In cases where the 

 animal did turn back, when no stimulus was given by the 

 experimenter or before such stimulus had been given, and 

 entered the other return alley the stimulus was given in 

 the other alley and the animal required to return to the 

 alley first chosen to get back to the food box. Such cases, 

 however, were rare. A few typical records run as follows: 

 Number 77 made one false return out of 540 trials; number 

 78 made four false returns out of 580 trials; number 79 

 made three false returns out of 560 trials; number 80 made 

 three false returns out of 580 trials; number 102 made 



