DISCRIMINATION OF SIZE AND FORM IN THE RAT 327 



ness of the discrimination apparatus does not seem, then, to be 

 a very serious one. 



EVIDENCE FROM THE RAT'S BEHAVIOR IN THE 



DISCRIMINATION COMPARTMENT 



i 



In the descriptions of the experimental work the percentage 

 of error has been employed to express the rat's ability to dis- 

 criminate, since a great number of correct choices forms the 

 most certain evidence, but in all the experiments, where posi- 

 tive results were obtained at all, the rats showed a recognition 

 of the difference between the forms before the percentage of 

 error was reduced below 50. In the unsuccessful experiments, 

 after the first 100 trials, the animals became almost machine 

 like in the regularity of their movements. There was seldom a 

 pause in the discrimination compartment and the choice of a 

 passage usually followed some definite rhythm. In experiments 

 3, 5 and 7 the beginning of discrimination was marked by long 

 hesitation with a swaying back and forth between the passages. 

 In the early stages of learning a correct choice more frequently 

 followed where the animal hesitated than when it made a rapid 

 choice. Later in the experiment choice by negation seemed to 

 become the more important method of eliminating error. The 

 rats chose a passage quickly, and if it were wrong, turned back 

 when about eight inches from the stimulus. Choice by affirma- 

 tion never seemed to be very effective, even with the best trained 

 rats. They would frequently make a correct choice, then hesi- 

 tate, turn back, and compare the forms. 



When responding to visual stimuli they showed a careful 

 attention to the illuminated forms, frequently returning from 

 the food compartment to sniff at them after an incorrect choice. 

 Failures more often followed after evident inattention. 



At the introduction of any new element into the discrimina- 

 tion box, as when swinging doors were placed in the food boxes 

 or when the passages were repainted, the rats became very much 

 excited, refusing to leave the starting box or to approach the 

 new object. This type of behavior was very marked in experi- 

 ment 2 when the revolving sector was introduced, and shows 

 unmistakably that the rats perceived the moving shadow, al- 

 though they did not associate it with the motive. Changes in 

 the brightness relations of different parts of the box were fol- 

 lowed similarly by confused reactions. 



