62 L. W. SACKETT 



ness difference is probably less than lo shades in the Nendel's 

 gray series which have been shown to be near the discrimina- 

 tion limen for these two animals. One would expect them to 

 get 50% correct choices by chance alone. In reality No. 3 

 got 62% and No. 10 only 54% correct. Their behavior was 

 even more convincing of the fact that they were not making 

 any discrimination between the boxes. When No. 3 was given 

 dark instead of Hght blue with the Hght green he was able to 

 discriminate at once. No. 10 required some time to break his 

 habit of always going to the right hand side. When the dark 

 green was put in with the dark blue there was another confu- 

 sion resulting rather seriously for No. 10 but from which he 

 recovered. Explanation should be made of why the experi- 

 menters did not conclude that No. 10 had failed with the darker 

 shades when for five series he made only a little more than 

 chance would give him. The explanation lies in his general 

 behavior which the table cannot show and in the fact that he 

 had 80.5% correct choices in the first series. Even with that 

 long struggle there was never a doubt about the final outcome. 

 In the case where the hght tints were used each of the animals 

 convinced the experimenters in five series that they could not, 

 or at least would not extricate themselves from the difficulty. 

 This is not a personal factor of the experimenter but is an 

 observation arising from long experience with the porcupines. 

 In final comment on color and brightness reactions it can 

 only be said that the results cannot warrant the conclusion 

 that the porcupine can discriminate and react to colors in an 

 intelligent manner. They can distinguish boxes which are blue 

 to the human eye from boxes which are green to the human 

 eye, both in the standard Bradley colors and in the darker 

 shades; but it is most probable that the dift'erence in bright- 

 ness in these two colors is also above their threshold of bright- 

 ness discrimination. No satisfactory means has been devised 

 for determining by physical methods just how much grayness 

 there is in each of these colors. Various photometric determi- 

 nations have been made in twilight, contrast rings, comparisons 

 by flicker photometry, etc., but no trustworthy results have 

 been obtained. Furthermore, as the grays which appear to 

 correspond with the colors used vary with each change in inten- 

 sity of light, and as the experiments with the animals were 



