4l6 'Journal of Comparative Neurology and Psychology. 



ants climbed the slides and mounted the incline. Similar stacks 

 of slides placed just to one side of the path at other points along 

 the trail were not thus mounted by the ants. The road by which 

 the ants had learned to reach the stage from which the gap now 

 separated them, had no vertical walls that needed to be climbed. 

 This w^all of slides was not placed across the path, but a little to 

 one side of it. Immediately, the ants mounted it and thereafter 

 used it as a means of passing from island to incline and from 

 incline to island. 



Then, too, it seems to me that ants use light as a means to an 

 end. In a former section, an attempt was made to prove that 

 light is not for ants a tropic stimulus, yet, by a repetition, under 

 control, of one of Lubbock's experiments, I have shown that the 

 direction of the rays of light does have a guiding influence. 



In each of those experiments, I used a cardboard stage, which 

 was connected with the Lubbock island by means of a single 

 incline. In some experiments, I placed the light (a i6 c.p. incan- 

 descent electric lamp) near the side to which the incline was at- 

 tached, in other cases the light was placed on the opposite side. 

 There was no other light in the room. After waiting untrl the 

 ant had thoroughly learned the way down the incline,^ I trans- 

 ferred the light to the opposite side. Invariably, as shown by its 

 movements, the ant would be very much disturbed by the change. 

 Now to my mind, this disturbance seems due to the fact that the 

 ants use the direction of the rays of light as reference data. In 

 other words, the light is responded to as a means to an end and not 

 as itself an end. If this contention be valid, then ants learn, not 

 by the method of trial and error, but perceptually, and they have 

 associative memory. As the result of careful observation, I am 

 convinced that ants use such things as irregularities of the surface, 

 edges of flat surfaces, the edges of shadows, etc., as reference data. 



On several occasions, after a marked worker had been carrying 

 pupae for a long time, it was imprisoned from one to several hours 

 and then returned to the stage. In some cases the ant failed com- 

 pletely to find the way to the nest, in a few others it went to the nest 

 immediately; but in most cases, it took some time to find the way 

 to the nest. Usually the time required to re-solve the problem was 



^ When the ant was not disturbed in its movements by the substitution of a new incline of the same 

 kind for the old, the ant was judged to be thoroughly acquainted with the way. This was the test I 

 always used in such cases. 



