106 K. S. LASHLEY 



vented from reaching the food. At a later stage he hesitates 

 before choosing an alley, advances toward the negative stimulus 

 hesitatingly and turns back before he reaches it. Later he may 

 stand at the entrance to the two alleys and sway back and forth 

 between them, finally choosing the positive stimulus. Such be- 

 havior is easily recognizable and is the invariable precursor of 

 accurate discrimination. In the following records it is referred 

 to as " comparison-mo vements." 



After the postoperative retention tests the animals were killed 

 and their brains examined in serial sections. The records of the 

 separate tests follow. 



Group 1. Frontal pole of cerebrum destroyed. 



Number 5. Small male, 109 days old. Trained on brightness 

 discrimination. 



Number of trials required for learning 100 



Preliminary retention tests, per cent error 4 



The frontal pole of the cerebrum was injured by transverse incision. 

 On the following day the animal was very weak, scarcely able to reach 

 the food, but his reactions were unhesitating, with comparison move- 

 ments. Only one error was made in ten trials. On the second day 

 he was stuporous and was not tested. On the third day his reactions 

 were quick and discrimination was unmistakable. 



Postoperative retention tests, per cent error 10 



Lesion: Plate III, figure 5. Both frontal poles were completely 

 destroyed by a section passing diagonally forward from the fronto- 

 parietal suture along the forceps of the callosum to the base of the 

 olfactory bulbs. 



Number 6. Small male (castrated), 108 days old. Trained in 

 brightness discrimination. 



Number of trials required for learning 50 



Preliminary retention tests, per cent error 5 



Operation was as in the case of number 5. On the day following 

 operation his reactions were quick. One error was made in ten trials 

 with comparison-movements and negative reactions to darknes? 

 Later trials verified the first day's record. 



Postoperative retention tests, per cent error 10 



