CEREBRAL FUNCTION IN LEARNING 99 



I the mass action of the cerebrum, as held by Loeb ('00) or of the 

 formation of new functional reflex connections through any part 7 

 of the cortex which happens to be intact. If the first assumption 

 is correct, destruction of large areas of the cerebrum should 

 retard learning more than lesser destructions. 



The problem has been raised in a somewhat different way by 

 the apparent reduction in learning ability of human defectives 

 who show a deficiency in the cells of the pyramidal layer (Bolton, 

 '05). The present evidence bears only indirectly upon this con- 

 dition but it does seem that if deficiency in learning results from 

 a reduction of the number of functional cells in one cortical layer 

 it should also result from a reduction in the total number of cells, 

 provided all are destroyed in like proportions. 



To determine the percentage of the cortex destroyed by the 

 various operations the total area of the cortex shown on the dia- 

 gram used in plates I and II was measured. The area of the 

 plotted lesions was then determined and, after correction for 

 overlapping of the dorsal and lateral figures, was expressed as a 

 percentage of the total area. This method gives only a rough 

 measure of the extent of the lesion. As an absolute measure of 

 the area destroyed it would be quite misleading, but since the 

 same method was employed for all the data, it seems to give 

 comparable figures for the different brains. The per cent of the 

 total area of the cortex destroyed in each animal is given in table 

 6, ranked from the least to the greatest, together with the num- 

 ber of trials required for learning by each animal. 



The average amount of destruction in all animals is 28.4 per 

 cent. The average number of trials required for learning by all 

 the animals with less than this amount of destruction is 88.1. 

 The average number of trials required by those with more than 

 the average destruction is 69.9. This gives a difference of 18.2 

 trials, or 20.6 per cent, in favor of the animals with the greater 

 injury. But this analysis includes the animals with paresis 

 (table 4). If we exclude them the averages become: for the 

 group with less than average destruction, 82.8 trials, for the 

 group with greater than average destruction, 93.2 trials, a differ- 

 ence of 11.1 per cent in favor of the group with lesser injury. 



