DISCRIMINATION EXPERIMENT WITH WHITE RATS 167 



observed in rats A, C and D. While C and D started toward 

 the still light at times, I have no records of paths of repeated 

 starts and halts with change of direction as in the case of A. 



Testing results. — To determine whether some clue given by 

 the experimenter might not be the ground upon which the rats 

 discriminated, Dr. Karl S. Lashley kindly tested the rats in 

 my absence. There was an average of 80% of correct choices. 

 In order to test further the significance of the results the rats 

 which had been trained to go to the still light were placed as 

 usual but with both lights still. They went equally to either 

 box. A strange reaction occurred, for upon coming to the door- 

 way of the box whose lamp had been swinging, when the vibrisse 

 lightly touched the edges, rat A stopped, squealed, turned back 

 and went over to the box which had had the still light. Ap- 

 parently some sensation from contact dominated his reaction. 

 The same halt at the doorway occurred when the rat C was 

 presented with both lights swinging and he went to the box 

 which had had the still light. One day when the rats trained 

 to go to a swinging light were presented with both lights swinging 

 in the middle of the day's series, the rat C for the four tests 

 given, went to the box where accustomed' to feed, but when 

 the lamp which had been stationary was set swinging, while 

 the light previously made to swing was still, this rat went to 

 the box with the swinging light. D, when both lights were 

 swinging, went to either box. The per cent of choices of the 

 food box changed for this rat from 94 under standard conditions 

 to 40 when both lamps were in motion. In other words, they 

 went freely to the box which they had consistently avoided when 

 the light stimuli were reversed. This fact together with the 

 fact of the halt upon touching with the vibrisse the wrong box 

 seemed conclusive evidence that the lights and not some other 

 factor had been the effective one in making the choice when at 

 a distance. After this halt at the doorway I noted that the 

 small roughnesses of the edges of the doorway made by the saw 

 were, of course, not exactly alike. I suspected that by chance 

 or from attraction of the swinging lamp the rat C went to that 

 box often and was then able to track himself but the evidence 

 seems conclusive that the moving stimulus came to be the 

 stimulus depended upon in reaching the box and food. 



