6/4 'Journal of Comparative Neurology and Psychology. 



Was the animal following his own trail ? We first met this pos- 

 sible difficulty by providing for each set of colored glasses six food- 

 glasses all of the same color. After five trials with one glass we re- 

 moved it from the room and put a second one in its place. After 

 five or six trials more, a third food-glass was used and so on. Had 

 the animal been selecting the food-glass by any odor attached to it 

 (except that of its pigment) this exchange of food-glasses should have 

 confused him. It did not do so, and in the later experiments we 

 changed the food-glass only every tenth trial. Again, if the animal 

 were selecting the food-color by any odor except that of the pigment, 

 he should learn to select a food-glass from a group, all of which were 

 covered with the same color. On July 15th ISTo. 2 was given a set 

 of glasses covered with OYS 1, and he pulled at all of them. At the 

 end of half an hour he had so scratched the yellow papers that the 

 scratched places would soon have served as distinguishing marks. 

 The result was the same with ISTo. 3. He was tested with the six 

 yellow glasses on the same day. We repeated such tests with the 

 food-color of seve rah later groups. We also filled the glass holder 

 with tumblers, all of which were covered with Gray 5, and gave the 

 animals opportunity for olfactory discrimination. In the first thirty 

 trials Raccoon iSTo. 2 made seventy-eight pulls at the glasses, eleven 

 of which Avere at the food-glass, but in six of these eleven cases the 

 food-glass was the first one to which the animal came. He passed 

 by one of the glasses only nine times in the seventy-eight attempts. 

 We restate these facts in the same order in the first line of Table 18. 



These records show that the animals soon ceased to try to dis- 

 tinguish one glass from another and pulled blindly at almost all of 

 them. By going but part way down the row and then returning to 

 the point of beginning, it was possible for the animals to pull re- 

 peatedly at several of the glasses without coming to the one which 

 contained food, and thus to make such records as those of the last 

 series in each table. The failure of the animals in this experiment 

 gives further evidence that they were not using smell in the color 

 tests. 



On continuing these experiments with ITo. 2, though he was very 

 hungry, the large number of selections which yielded no food seemed, 

 to a great extent, to inhibit the impulse to pull at the glasses. 



