CEREBRAL FUNCTION IN LEARNING 105 



days, a total of 20 successive errorless trials. Entrance into the 

 darkened alley was counted as an error. 



Since the operation entails a period of from one to five days 

 without practice it was necessary to determine the loss of the / / 

 habit which this rest alone would produce. Therefore, when each u 

 annual had learned the problem he was kept for seven days 

 without practice and was then tested for retention of the habit. 

 These tests are recorded as " preliminary retention tests" in the 

 records of experiments. In case a loss of habit appeared after 

 this rest period the animal was retrained until he made 20 suc- 

 cessive errorless trials. When these were completed he was 

 subjected to operation. 



For the experiments the animals were divided into four groups. 

 The number of animals in each group and the character of the 



TABLE 7 

 Distribution of animals in experiments on discrimination of brightness 



operation are shown in table 7. As soon as they had recovered 

 from the shock of operation, in from 24 to 96 hours, they were 

 tested in the discrimination box for evidence of retention. At 

 least 30 trials were given in the postoperative retention tests. 

 Two sorts of evidence are important in interpreting the results. 

 First, the postoperative retention tests may show no more error 

 than the preliminary retention tests. They then give conclusive 

 evidence of retention. Second, when the postoperative reten- 

 tion tests show numerous errors the animal may yet give evidence 

 of retention by selective reactions to the stimuli. 



At one stage in the formation of a visual discrimination habit 

 the rat begins to give evidence of attention to the lights. The 

 first trials are evidently random dashes and in case of a wrong 

 choice the rat does not turn back until he has actually been pre- 



