564 'Journal of Comparative Neurology and Psychology. 



and a considerable space behind them, were screened in by wire net- 

 ting. Sliding wire partitions served to confine the animal in any 

 desired portion of the enclosed space during the adjustment of the 

 apparatus. The passageway between the boxes as they stood in 

 position was five inches wide. The animal entered from behind 

 and passed between the boxes to the open space in front, and there 

 had the choice of turning into the opening of either box. 



It will be seen that these tests were made upon few animals. All 

 of the animals, however, were in good condition and the experimenter 

 feels that the results, as far as they go, are representative. 



The first set of tests was made upon the male squirrel. The 

 conditions of this test were as follows: Box A, the standard (which 

 contained the food in all the tests), was kept at a temperature of 

 40° ± 2° C. Box B, the variable, during the formation of the dis- 

 crimination habit was kept at the temperature of 15° ±2° C. A 

 record of this test is shown in the first three columns of Table V. 



After discrimination had become definite, the temperature of the 

 variable box was raised to 25° ± 2° C. Columns 4 and 5 of the same 

 table show the results of this change. 



The second set of tests was made upon a female. In this case, 

 as in the one above, the standard box A was kept at a temperature of 

 40° ± 2° C, but the variable B was kept at a temperature of 

 25° ± 2° C. After the association had become definite, the tempera- 

 ture of the variable was raised to 30° ± 2° C. The latter part of 

 this test (discrimination between 40° and 30°) was not completed, 

 but the work was carried far enough to leave no doubt that the dis- 

 crimination of this difi'erence was possible. Table V, columns 6, Y, 

 8 and 9, show the results of the whole test. 



A similar set of tests was made upon three white rats. Table VI 

 shows the records of these animals and the temperatures used. 



(a) Discussion of results of tests upon squirrels. — The behavior 

 of the squirrels in this test varies little from recorded descriptions 

 of like experiments in other sensory fields. The "controls" used 

 were such as to eliminate sensory factors other than that of tem- 

 perature. It may be well to mention in some detail the precautions 

 taken. 



