THE PROBLEM OF LEARNING 1 79 



his arms, — all of which are movements to which 

 we have agreed to attach a certain meaning. 

 There is some difficulty in always understanding 

 our human definitions but this is a minor difficulty 

 in comparison with the study of animals where all 

 of our conclusions must be drawn from a study of 

 their movements to which we and not they give a 

 meaning. 



When we find that animals move toward a 

 given substance as bees do toward cane sugar, and 

 are indifferent to saccharine, we conclude that bees 

 like cane sugar. In the experiments which are to 

 follow illustrating the study of learning in animals, 

 note that movement is the key used in unlocking 

 this problem. 



In times past much was written about the part 

 played by reflex action as an elementary form of 

 nervous activity and many animals were studied in 

 the expectation of finding that somewhere the re- 

 lations were so simple that the physiological 

 aspects of reflex action could be analyzed. Parker's 

 numerous studies on the elementary nervous sys- 

 tem of animals throw no new light on the real 

 character of a reflex act. As soon as such simple 

 animals as earthworms are studied the nervous 

 mechanism is found to be far too complex to per- 

 mit of simple reflexes between a receptor and ef- 

 fector such as Parker describes in the Coelenterate 

 Metridium. Between these two structures, receptor 



