16 HARRY H. WYLIE 



When the animal had entered either of these return alleys 

 the stimulus was given, that is, the light or sound or electric 

 shock, or a combination of any two of them as the case 

 might require, and the animal was required to go back and 

 come down the other return alley to the food box. In 

 fifteen out of every series of twenty trials the proper stimulus 

 was given and the animal was required to turn back and enter 

 the food box by the other alley. In the other five trials 

 the animal was allowed to continue on its chosen way and 

 return to the food box, no stimulus of any kind having been 

 given. These five where no stimulus was given were scattered 

 irregularly through the series of 20 to prevent any possibility 

 of the animal's learning a definite fixed order in the series. 

 The purpose was to train the animal to stop at a given stim- 

 ulus, to turn around, and to go back to the food box through 

 the other return alley but, when given no stimulus, to pro- 

 ceed on to the food box through the alley first chosen. When 

 the animal was first learning, the stimulus was given more or 

 less irregularly as to loudness and continuity from the time 

 the animal had fully entered the return alley until it had turned 

 back, even if the animal persisted in its course until it found 

 the door at the end of that alley closed. The fact is that 

 the usual course was for the animal at first to ignore com- 

 pletely the stimulus other than a short delay for examina- 

 tion and to proceed clear down to the closed door at the 

 end of the alley. After the first strangeness had worn itself 

 away, which happened after a few trials, there was little 

 indication, if any, to show that the stimulus was sensed at all 

 until actual learning of the required reponse began. However, 

 as accurate records as possible were kept of hesitations as well 

 as of the actual times of turning back, when the hesitations 

 were apparently due to the effect of the stimulus. Some- 

 times also in the learning process the speed of the reaction 

 was apparently much lessened by the disturbing effect of 

 the stimulus. This was particularly true when the light 

 stimulus was used, so a record of these was also kept. Both 

 doors, RD and R'D', were kept closed until a choice had been 

 made, when the proper one was opened as noiselessly as pos- 

 sible. An attempt was made to prevent the experimenter's 

 movements from being so evident, as to become cues for the 



