28 JOHN LINCK ULRICH 



of movements had to be laid down before response could be 

 immediate. When this had taken place, learning was rapid. 



The data obtained from these experiments are quite similar 

 to those obtained when the problems were given singly. One 

 trial daily was the most economical proceedure when either one 

 problem or several were offered. The number of trials and the 

 number of days required by each rat to learn the problems, are 

 given on the distribution curves, Plates IV, V, and VI. There 

 is one feature noticeable on these curves that was not so evident 

 on previous distribution curves; individual differences are more 

 strikingly shown, and from this we infer that learning is 

 more difficult when several problems are taken concurrently 

 than when one is learned. The number of trials and days re- 

 quired for the first rat to complete learning with the one, three, 

 and five trial groups for the latch box was respectively as 

 follow^s: trials 11, 24, and 40; days 11, 8, and 8; for the maze 

 the number of trials w^as 27, v^6, and 50, and the number of days 

 27, 12, and 10; for the inclined plane box the number of trials 

 w'as 22, 33, and 50, and the number of days 22, 11, and 10. 



The curves for the integration of movements, Curves IV, V, 

 and VI, have been constructed from averages presented in 

 Tables IV, V, and VI. These curves of the latch box, maze, 

 and inclined plane box vary as do others of their kind in respect 

 to their complexity and irregularity, and Curve VI of the in- 

 clined plane box shows these variations most clearly. It has 

 proved to be the most difficult of the problems. The rats, in 

 order to learn this problem, must form, as has been said, two 

 integrations for precise movements; first, one for pushing down 

 the plane, and, secondly, one for responding to the noise of 

 the opening of the door in order to obtain food. The maze, 

 seems to be less dif^cult than the inclined plane box, and the 

 latch box the least dif^cult when making like comparisons with 

 their curves. 



