DISTRIBUTION OF EFFORT IX LEARNING IN WHITE RAT 23 



Before leaving the discussion of the maze, it will be well to 

 compare the results and the curves of the latch box with those 

 of the maze. In both cases learning was more economically 

 attained (with respect to number of trials) when trials were 

 distributed over several days, than when accumulated in one 

 day. On the other hand, fewer days were required to learn 

 a problem when several trials were given. The curves of the 

 maze show some interesting differences from those of the latch 

 box; the individual curves of the maze are longer and more 

 irregular than those of the latter. The increased length of the 

 curves of the maze is, beyond doubt, due in part to the fact 

 that an additional criterion, " error free," has been added to 

 that of time in determining the norm of the maze, and it also 

 is due to the fact that the maze is a more difficult problem to 

 learn, though it is said to be one where more natural movements 

 are demanded. It is certain that the rats would have required 

 fewer trials and the curves, as a consequence, would have been 

 shorter, if time alone had been considered. The length of the 

 curves of the maze would then probably approximate those of 

 the latch box. The irregularities in the curves, however, would 

 still remain. 



II. THE EFFECT OF DIFFERENT DISTRIBUTIONS OF PRACTISE 

 WHEN THREE PROBLEMS ARE LEARNED ABREAST 



The previous section shows that when a single problem is 

 learned, one trial daily or one trial even at longer intervals, is 

 more advantageous than a larger number of trials given every 

 day. The question next arises as to whether this same relation 

 obtains under more complicated conditions of learning. In the 

 education of children several subjects have to be learned simul- 

 taneously or abreast. It is easy to arrange similar conditions 

 in the animal work by forcing them to learn three problems 

 concurrently. We chose for this work, first the latch box, second 

 the maze, both of which have already been described, and third 

 a new problem, viz., the inclined plane box. 



The inclined plane box, Fig. Ill, in general construction and 

 dimensions resembles the latch box, with the exception that 

 the latch was inside of the box and was released in a different 

 manner. A wire chain connected with the latch passed over two 

 pulleys attached to opposite corners of the upper frame work 



