LOCALIZATION OF LEARNING PROCESS 201 



and there is no reason to assume that the latter field 

 is related with these habits differently from the rest of 

 the cortex. 



We are now confronted with the very remarkable 

 fact that if a rat with cortex intact is taught a visual 

 discrimination habit and then the occipital cortex is 

 removed the habit is totally lost. It can be reacquired, 

 but this takes about as many trials as were necessary 

 in the first instance. Removal of no other part of the 

 cortex has this effect. This demonstrates a localized 

 cortical participation in this habit of peculiar charac- 

 ter, and if the nature of this participation can be dis- 

 covered we may have the key to the secrets of cortical 

 function on the rat's biological level. 



The anatomical connections between the thalamus 

 and the occipital cortex which are probably involved 

 in this behavior are shown in greatly simplified form 

 in Figure 50, diagram III. In addition to the thalamic 

 arcs shown in diagram II there is a cortical loop con- 

 nected with the thalamic arcs, as the electrician would 

 say, in parallel. Both circuits are open. It is clear 

 that in the absence of the cortical loop the thalamic 

 connections are adequate. It is equally clear that if in 

 the initial training the cortical loop is employed, its 

 further participation in the reaction is essential. What 

 function, then, does the cortical loop perform? We 

 have no satisfactory answer to this question. 



One may approach the problem indirectly from 

 several directions. Let us look at it as a problem in 



