Cole and Long, Visual Discrimination in Raccoons. 68 1 



as a food-glass do not appear in the table. The records for other 

 grays are those obtained after the animals had been brought to avoid 

 Gray 5, by having been trained on VBT 2 as a food-glass in the 

 group to which Gray 5 belonged. 



TABLE 25, 



If by the use of the flicker method we did not secure colors which 

 were equally bright, for the animals, with the gray which was used 

 as a standard, they should not have made wrong choices by selecting 

 the gray. Or, at most, they should have made very few such errors. 



The first half of Table 25 shows almost as many mistakes on the 

 gray as on any single wrong color, though some hundreds of trials 

 which should have discouraged the tendency to make these mistakes 

 on gray, preceded those which are tabulated (see pp. 667 and 668). 

 In the latter half of the table the animals seem to have profited by the 

 fact that they had never found food in the gray. This might be 

 because of a difi^erence in brightness, or because, after long practice 

 with food in colored glasses, the animals began to pay more attention 

 to the colored papers and less to the grays. 



Comparison of the tables, then, reduces the question to this. If 

 the papers were not equally bright for the animals they should have 

 made fewer mistakes than they did. This is shown by their records 

 on the consecutive grays. On the other hand, if the papers were 

 equally bright for the animals and yet they were color-blind, the 

 excellent records they made on the food-glasses finds no explanation. 



(6) So far, then, as evidence can be gained by the use of reflected 

 light, we think it probable that the raccoon can be made to discrimi- 

 nate objects by their color alone. We do not think that in their native 

 state they are often called upon to make pure color discriminations. 

 Samojloff and Pheophilaktowa concluded, only, that the dog can 



