520 BRAIN MECHANISMS AND LEARNING 



continuous bars) separated from the negative trials (vertical interrupted 

 bars). In the graphs of temporal tracings, the first bar of each pair indicates 

 the number of trials that showed no decreased amplitude, and the second 

 bar, the number of trials that showed a decreased amplitude. For the visual 

 area EEG, the first bar is the number of trials with no photic driving, the 

 second bar, the number of trials with photic driving. The percentage of 

 correct responses during the daily session was indicated at the bottom of 

 the graph. 



These graphs show that consistent changes occurred only in negative 

 trials, but not in positive trials. This probably reflected the fact that the 

 animals were initially trained to reach for the stimulus every time. The 

 number of negative trials showing a depressed amplitude in the temporal 

 EEG increased as the animal started to learn, but returned to the hiitial 

 level when criterion was reached. Such transient changes occurred in 

 eight out of eleven cases of discrimination learning. A similar change in the 

 number of trials showing photic driving was illustrated in one oi the 

 graphs for the visual area. Such visual area EEG changes appeared, how- 

 ever, in only three out of eleven cases. Thus, it seems that the only signi- 

 ficant electroencephalographic alterations in this study were contmed to 

 the temporal cortex which is known to be related to visual learning. The 

 fict that the EEG change is transitory argues against it being a direct 

 neural expression of the learning process. Rather, it probably represents 

 some alerting or attention factor. When the animal began to be aware of 

 the problem it had to concentrate not to respond to the negative stimulus. 

 Once that was learned little effort was needed to respond appropriately. 



LEARNING AND RETENTION DURING ELECTROGRAPHIC AFTER- 

 DISCHARGES 



The same six monkeys in the same testing situation described in the last 

 section were used later to study the effect of local EEG after-discharges of 

 the temporal cortex on learning and retention. Presumably, seizure after- 

 discharges disrupt neuronal circuits. The present experiment was designed 

 to show whether or not some such neuronal circuits are necessary for 

 learning and memory processes. 



The after-discharges were induced either unilaterally or bilaterally by 

 bipolar electrical stimulation of the temporal cortex. A Grass stimulator 

 was used to deliver monophasic square wave stimuli (3-5 ms. 50 c.p.s., 

 4-25 volts). Each trial was presented as follows: The EEG was turned on, 

 the light was changed to flashes, the EEG was turned oft, the electrical 



