DISTRIBUTION OF EFFORT IX LEARNING IN WHITE RAT 17 



free from error. Such standards, time and error-free, for learn- 

 ing have some interesting points worthy of notice, and their 

 relationship to one another is rather peculiar. In the first place, 

 since no definite agreement can be made as to what is an error, 

 difficulties immediately arise when we come to decide when 

 learning is complete. Gross errors, such as the animal turning 

 in the direction of a cul de sac or retracing its steps, can readily 

 be agreed upon as errors. Mere hesitation, however, cannot be 

 agreed upon as an error, and in this work it was not considered 

 such. And all the more interesting does the question of error 

 become when its relationship to time is considered. Frequently 

 runs in the maze were made in six seconds with several 

 errors and this usually persisted for several days before runs 

 in the maze were made error-free. It was not because six seconds 

 was too high a norm, for some rats could barely make a run in 

 six seconds and make no errors; and others, the more active, 

 could make the run in the same time and still be charged with 

 several errors. And this, of course, implies that speed was 

 first gained, then errors were eliminated. Integration for speed 

 seems to be first laid down, th-en that for the removal of errors. 

 Probably distracting stimuli of the sort which produce unneces- 

 sary responses interfere with the simultaneous integration for 

 speed and for runs that would be free from error. 



It is the dominance of hunger, acting intermittently, that 

 evokes responses in going from position to position while both 

 of these integrations are being laid down. Some deny that this 

 stimulus acts in the early trials. Hicks and Carr (12) maintain 

 in their work on the Hampton Court type of maze that the 

 animals are at this time " governed by caution and curiosity." 

 This scarcely proved true with this work on the circular maze. 

 In every case, scarcely had the rats made a turn into the second 

 or third path, and detected the odor of food, when there was 

 immediately a noticeable eagerness to obtain it. Their activities 

 were hardly governed by curiosity or fear. It must be admitted, 

 however, that these activities were decidedly sporadic. The 

 response of hunger is intermittent in character, and probably 

 it is more intermittent during the early trials, because of the 

 presence of numerous interfering stimuli. It is this interruption 

 of the dominance of hunger response that made it appear that 

 the rats in the early trials were governed by curiosity alone. 



