CANADA PORCUPINE 59 



his selection by choosing the black while No. lo learned to select 

 by avoiding the white — the one reacting positively and the 

 other negatively. The reason for this is evident in the records. 

 When the food box remained constant and the comparison box 

 varied, No. 3 was not confused in the least while No. 10 had to 

 relearn the task after every change. In the last part where the 

 white remained constant No. 10 had very little difficulty with 

 the changes for the first time in his experiences. No. 3 also had 

 little trouble in the last half and that is all that is lacking to 

 make it a thoroughly good h}^othesis, though his experience 

 may account for the improved results in the last part of the 

 experiment. Yerkes (49) reports a similar experience with a 

 mouse which had to relearn grays when they were changed 

 though the relative brightness was not greatly altered. The 

 results in each case illustrate the very strong influence of every 

 condition of the environment as well as the concreteness of the 

 reactions of the animals and the inability of the animal to gen- 

 eralize e^'en in the most obvious cases. 



Discrimination of Color following Experiments with Brightness. 

 — After completing the brightness tests reported above the 

 experiments with blue and green stimuli were repeated. The 

 same animals were used as subjects. Both had failed in the 

 earlier series but after the experience with the brightness in 

 which they had formed a habit of going to the darker box and 

 with food in the blue box, one chose blue 1 7 times and the other 

 18 times on the first series. While No. 3 was much more certain 

 of his choice, there was never any doubt about No. lo's success. 

 Following this, control experiments were made to determine if 

 possible whether the animal was choosing the blue box because 

 it was blue or because it was the darker of the two. 



A pair of boxes covered with blue "shade 2 " and green "tint 

 2 " of the Bradley colors, also a pair with blue "tint 2" and 

 green "shade 2" were used. The first pair was a very dark 

 blue and a very light green while the second pair was the re- 

 verse, very dark green and \'ery light blue. These pairs were 

 presented to the animal in alternate individual experiments, 

 feeding in the dark blue with one pair and in the dark green 

 with the other pair. Colors in each experiment were alternated 

 and at the same time the darker box continued to be the food 

 box. No. 10 made 84% correct choices in four series and No. 



