Turner, Homing of Ants. 423 



conveying pupae to the nest, and yet another set pihng the pupae in 

 the center of the stage. 



Time and again, when ants were busy conveying pupae from 

 the stage to the nest, I have dropped pupae in the path of ants 

 returning from the nest. I have tried this experiment with all 

 the species of ants considered in this paper; and, without excep- 

 tion, the pupa was picked up and carried into the nest. 



In the Hght of these facts, let us consider Herr Gredler's 

 account. In this case the ants had learned the way from the sugar, 

 by way of the string, wall, window-sill and wall to their nest. 

 Now had a small number of these ants begun to pile the sugar on 

 the edge of the bottle, from which most of it would surely have 

 fallen to the window-sill, the case would have been similar to the 

 one where one portion of one of my colonies conveyed pupae from 

 the stage to the nest, while another portion piled the pupae at 

 a certain place on the stage. But the falling of a portion of the 

 sugar to the window-sill, which was a portion of the pathway of 

 the ants returning from the nest, would introduce a modifying 

 variation. As soon as a returning ant encountered a grain of 

 sugar, it would pick it up and carry it to the nest. And if the 

 grains fell from the rim of the bottle fast enough, we should 

 soon have no ants taking the trip up the string. Thus would be 

 produced a behavior which simulates mutual cooperation. Since 

 Herr Gredler failed to observe the stages by which the contin- 

 uous line of ants was transformed into two distant yet cooperating 

 lines, this anecdote has no value as evidence proving that ants 

 rationally cooperate. 



To sum up the results of this section, it is quite probable that 

 division of labor of the type mentioned above is of rather common 

 occurrence among certain ants; but until new data are forthcom- 

 ing, we must consider all such cases as coincidences, rather than 

 as proofs of rational cooperation. 



VII. CONCLUSIONS. 



1. In their journeys, the movements of ants are not tropisms. 



2. Ants are not as slavishly guided by the scent of their foot- 

 prints as is usually believed, for all of the species examined by me 

 could be trained to pass over at least twelve inches of an unscented 

 path. This discovery furnishes an easy means of investigating 

 many problems in ant psychology. 



