568 JourJial of Comparative Neurology and Psychology. 



to the further training of the same animal in problems which pre- 

 sent a large number of similar conditions. Different methods of 

 opening the same problem box constitute such unfavorable conditions. 

 A latch problem box and the maze, for example, do not present 

 enough identical conditions to interfere with the learning process of 

 the one when the other has been learned first. 



4. The grey squirrel and the white rat can form associations upon 

 the basis of a temj^erature sense. Actual tests show that the squirrel 

 can discriminate between two boxes when they differ in temperature 

 by 25°. In the case of one animal tested, a difference of 10° in 

 the temperature of the box was found discriminable. Further tests 

 with a more accurate apparatus would in all probability show that 

 the discriminable difference in temperature may be much less. The 

 experiments upon the white rats show that a difference of 16° is 

 easily discriminated. 



5. Incentives which may be used easily in further investigations 

 with the squirrel are: himger, disagreeable odors and tastes (bad 

 nuts, onions, etc. ) , g-nawing impulses, and love of exercise and greater 

 freedom. This last incentive is especially strong in the squirrel. 



The ability of the squirrel to detect an edible nut from a faulty 

 one probably contains the sense, or senses, in which the keenest dis- 

 criminative power of the squirrel is reached. 



