344 H. M. JOHNSON 



to ascertain what choice the animal made without watching the 

 animal in the act of choosing. I consider this feature important, 

 since it enables the experimenter to avoid giving certain secondary 

 cues to the animal. I adopted the device at the suggestion of 

 Professor Watson, who had been using a similar one for some time. 

 The entrance-doors to the food-boxes Rfb and Lfb are con- 

 trolled by rods indicated in the diagram, which pass to the experi- 

 menter's station. These doors are kept closed against the animal 

 until the latter shall have entered the proper entrance-alley, 

 which fact is indicated by the signal lamp. After the animal has 

 passed into the food-box, the entrance-doors are closed behind 

 him, and he is readmitted to the home-box through another 

 door opening directly into it. This door is also controlled by 

 rods which are not indicated in the sketch. It is unnecessary 

 for the experimenter to touch the animal during the daily series. 

 Except for the light from the signal lamps passing through 

 pin-holes made in the boxes in which the latter are placed, all 

 the surroundings of the test-fields were dark when the stimuli 

 were presented. (The walls and floors of the experiment-box 

 are painted dead black.) This does not indicate conditions of 

 dark adaptation, however. A 1 c.p. lamp with a frosted bulb 

 is mounted in the top of each food-box. It is protected by a 

 metal shield, so as to exclude direct light, as far as possible, from 

 the animal's eye, and is so placed that the animal is turned away 

 from it while eating. This lamp was not switched on until the 

 animal had made his choice. Two such lamps were similarly 

 mounted in the home-compartment H. They were switched on 

 and the lamps in the food-boxes switched off, after the animal 

 had finished eating and the settings had been made for the next 

 trial. The lamps in H were switched off before the exit-door 

 was opened. All the animals quickly learned to pass from the 

 dark compartment into the illuminated one, as soon as the doors 

 were opened, and without other attention from the experimenter. 

 The criterion of choice was the animal's entering a given alley. 

 Should the animal enter the wrong alley he was not disturbed 

 but was admitted to the proper food-box when he voluntarily 

 entered the proper alley. The choice in such case, however, 

 was recorded as incorrect. In this work the distance between 

 the test-object and the animal's eye is a necessary factor. The 

 experimenter prepared a number of stops extending across the 



