238 BRAINS OF RATS AND MEN 



credit him with some measure of this sort of insight. 

 In some of his behavior he seems to show various 

 short-cuts from tedious trial-and-error to more direct 

 understanding of key frxtors in problem solution. 

 But the evidence for such an opinion is, so far as I 

 know, still in the "anecdotal" stage. If the dog learns 

 only by trial-and-error, some of the processes are 

 certainly much foreshortened. 



It is, in fact, very difficult to devise experiments 

 with desirable quantitative measurements adequate 

 to yield crucial evidence. The learning curve has been 

 appealed to here, learning by insight giving a steeper 

 curve than learning by overt trial. But so many other 

 factors may enter that this method is still of uncertain 

 value. 



Yerkes has published well-controlled observations 

 on monkeys and the orang-utan (191 6) and chim- 

 panzee (1925) that seem to give clear evidence of this 

 sort of learning. In Kohler's fascinating account 

 (1925) of his observations upon chimpanzees much 

 additional evidence is supplied. 



With the problem of insight and intentional con- 

 trol of behavior in mind, a review of the countless ex- 

 periments that have been made on rats' behavior dis- 

 closes very few which are well adapted to reveal this 

 sort of behavior, even if rats are competent to perform 

 it.' Most of the laboratory experiments so far re- 



* A necessary physiological background for such studies is accurate 

 knowledge of the natural spontaneous behavior, that is, the internally 

 excited activity (see p. 31 4). 



