35^ "Journal of Comparative Neurology and Psychology. 



the cage, but never once to the chute. She then settled down to 

 her usual behavior and after ten minutes the stick was replaced. Slie 

 pulled the string once and tlie sticlv vv^as again removed. She then resumed 

 her usual conduct for almost ten minutes more. Then, when under the 

 chute, she looked up and her eyes accidentally fell on the chute. She rushed 

 up the front of the cage and leaped to the chute, swung herself down and 

 worked the device. This slie repeated several times, although in a much 

 less skillful way than No. 2 was able to do it. 



2. Chute Experiment B. 

 A. Description of Device. 



This apparatus was a moditication of the one used in Chute Experi- 

 ment A. 



Fr»m the top of the experiment cage (fig. 3), 30 cm. from the board side 

 and 40 cm. from the board end, a hollow wooden chute, a, 5 cm. square, 

 projected into the cage a distance of 70 cm. Inside this chute, 40 cm. 

 from the lower end was a trap door, b, hinged to drop downward, but held 

 up by a rubber elastic. To the bottom of this trap door was attached a 

 string, c, which hung down to within 10 cm. of the lower edge of the chute. 

 A coiled wire spring, d, was tied to the lower end of the string to serve for a 

 hand-hold for the animals. In the top of the cage was a tube leading into 

 the cTiute from a small feeder (fig. 19, /,) adjusted on top of the cage. 

 By pulling a string, attached to the feeder, the experimenter could drop food 

 (sunflower seed and chopped peanuts) upon the trap door.* Two horizontal 

 rungs nailed to the sides of the chute helped the animal to support himself on 

 the chute. 



The problem set for the animal was to leap from the wire front or end 

 of the cage and, while holding to the chute, to swing his head and shoulders 

 down, thrust one hand up the inside of the chute, grasp the coiled spring, and 

 pull the trap door down. The food on the top of the door would then fall to 

 the floor, unless checked by striking the body of the animal. In either case 

 the animal could get it. 



B. Bclutvior of No. 2. 

 No. 2 had learned to work the mechanism in Chute Experiment A. But 

 when he was flrst put into the new cage, four months after liis last experi- 

 ence in the old one, he apparently had no memory of the chute. Only after 

 several minutes did he go to it. He jumped to it from the front wire. He 

 stopped to examine the opening in the end before he thrust his hand into 

 it. In the old cage he usually thrust his hand into the opening without 



*The essential part of the feeder was a copper plate, 3 em. wide and 6 mm. 

 thick, arranged to slide back and forth in grooves beneath a food hopper. In 

 the plate was a circular opening into which the food dropped from the hopper. 

 When the string was pulled this opening, full of food, was drawn over a tube 

 leading into the chute, into which the food dropped. 



