Cole and Long, Visual Discrimination in Raccoons. 683 



thirty trials and was perfect in the second thirty. He thus selected 

 the food-glass three times more in each of the two re-learning series 

 than in the learning series. One hundred and eight days later, 

 and meantime without practice, he selected this color 15, 25, and 29 

 times in each series. The utmost that can be said, therefore, is that 

 in the case of the two animals, relearning was a little more rapidly 

 accomplished than learning. So, also, must the fact of much more 

 rapid learning in the later experiments be given some weight after 

 due allowance is made for the darker colors being easier for the ani- 

 mals to discriminate. Group 2, for example, which w^as not dark, 

 was evidently learned mucli more quickly than it would have been 

 without previous training. The animals' behavior is thus greatly 

 modified by past experience, but the effect of having learned to 

 discriminate objects by means of a specific brightness or color differ- 

 ence, at any rate if that difference be slight, does not last more than 

 a few days. This might be expected since in its native state the 

 raccoon is probably not called upon either to detect or remember such 

 slight differences as were used in these experiments. 



Finally, it may be remarked that the animal which gave the better 

 "memory" record above for discrimination of colored papers was the 

 one which gave evidence of superior motor memory for fastenings. 

 This animal also required the greatest number of trials in both cases 

 for the original formation of the associations. So much evidence for 

 a mechanical law of association in animal psychology. 



