TRANSFER OF RESPONSE IN THE WHITE RAT 19 



curves have been given instead. The curves for number 22 

 of Group One, (Figure III) and for number 25 of Group 

 Two (Figure IV) are given. These curves show not only 

 the learning of the response to the light stimulus but they 

 also indicate the record for the control series in each case. 



These curves indicate neither a rapid initial rise nor a long 

 period of final perfecting of response. They rather show a 

 gradual, though very irregular, rise from the beginning to 

 the end of the learning process. 



The unusually large number of trials required by Numbers 

 16 and 17, for learning the light finds its explanation, 

 apparently, largely in terms of distracting olfactory stimuli. 

 The males in these two groups were kept in a cage by them- 

 selves and were given their daily stint of work immediately 

 after the females of the two groups. The two individuals 

 mentioned, being unusually vigorous animals, were con- 

 stantly attracted apparently by the odors left by the females 

 and would often wander about the box sniffing here and 

 there to the neglect of the light stimulus. So even after the 

 response had been apparently learned, it took a greater 

 number of series to meet the required standard. 



The important thing about learning the light was, of 

 course, to be as sure as possible that the animals were re- 

 sponding to the light and not to other stimuli. The order of 

 presentation of right and left pathways followed in the regu- 

 lar and control series has been given already on page 29. It 

 will easily be seen that position habits might account for a 

 rather large percentage of correct choices in such a series. 

 Of the position habits most likely to be learned the most 

 advantageous would be a double alternation between the 

 right and left pathways. Such a position habit, if acquired, 

 would account for a record in which 80% of the responses 

 were correct. The next most advantageous position habit 

 would be a simple alternation between right and left. That, 

 if acquired, would account for a record in which 70% of the 

 responses were correct. Chance, of course, in any series 

 where two possible choices must be presented an equal 

 number of times, will account, in the long run, for 50% of 

 the responses being correct. A position habit of always 

 going to one side would also give a record in which 50% of 



