DISTRIBUTION OF EFFORT IN LEARNING IN WHITE RAT 25 



inside of the box, then through the mesh of the wire netting 

 covering the box to an incHned plane behind the box. A rat 

 stepping on the plane would release the latch, and the door 

 would open inward. The plane was 17 cm. long and 7.5 cm. 

 wide and was hinged to the top of a table on which the box 

 rested. It was about 7 cm. behind the box, and made an angle 

 of 15 degrees with the table top when the door was closed. Like 

 the latch box, the inclined plane box was covered with a wire hood. 



We arranged three groups of animals to work upon these 

 problems. One group was given one trial a day on each of the 

 three problems in the order: latch box, maze, inclined plane box. 

 The second group was given three trials per day on each prob- 

 lem, and the third group was given five trials per day on each 

 problem. The rats used in this experiment were of a hardy 

 strain, descendants of those employed in the first experiment on 

 the latch box. Each group contained eighteen animals. In the 

 beginning it was feared that the rats that were running five trials 

 per day on each problem (i.e., fifteen trials in all) could not 

 stand the strain of so many tests; but this did not prove to be 

 the case. The rats worked well and there was no indication of 

 fatigue or apathy at any time (see p. 48 also). Only slight rest 

 was allowed between the problems. Time was given after each 

 trial to allow the animal to nibble a little food. Then the next 

 trial was given, etc. No time was lost in transferring the rats 

 from one problem to the next. 



With the three problems no methods were adopted that had 

 not been employed with the maze and latch box when these 

 were under experiment alone. The same norms were used, an 

 average of one second in two successive days for the former, 

 and six seconds, without error, for the latter. For the new 

 problem, the inclined plane box, a norm of two seconds was 

 adopted. The rats were required to work at these three prob- 

 lems until all three norms were reached. When this had been 

 attained, they were forced to exercise daily in the 3S cm. run- 

 way until ready for retention test. 



Two sets of data ought to appear in making these tests: 

 (1) data on the effect of the different distributions and (2) an 

 answer to the question as to whether, from the standpoint of 

 the number of trials involved it is more efficient to learn one 

 problem at a time or to learn several abreast. 



