REWARD AND PUNISHMENT IN HABIT FORMATION 249 



TABLE 5 Continued 



hunger forty-one hours. Detailed results are given in tables 5 

 to 12 inclusive. These tables are constructed on the same plan 

 as tables in the preliminary experiments. 



Retention. On the twenty-first day after the subject had per- 

 fected the habit a retention series of ten trials was given. This 

 series was given under the same conditions and in the same 

 manner as the training series except that the subject was neither 

 given food for right choice nor electric shock for wrong choice. 



Retraining. Three days after the retention test retraining was 

 begun and each individual was retrained in the same way that 

 it had been trained. The habit was considered perfected, as it 

 was in the training series, when the rat made a perfect series of 

 ten successive trials. 



CURVES OF LEARNING 



Figures 2 and 3 show the characteristic differences in the curves 

 of learning and curves of re-learning with electric shock and 

 hunger. These curves represent the average number of errors 

 made in each training series as given in next to the last column 

 of tables 5 to 12 inclusive. The curve marked one hundred and 

 fifty units is based on the average number of errors for the sub- 



