402 WILLIAM M. MANN 



in searching for food it deviates but always resumes the original 

 direction. To an ant (Messor) at several meters distant from 

 the nest, Cornetz gave a seed. It took this in its mandibles, 

 turned about and returned directly to the nest. This experi- 

 ment Was repeated a great many times, always with the same 

 result. The line of return was more direct than the outgoing 

 line, which was rather sinuous, with frequent deviations. The 

 axes of these curves were followed in returning, and these led, 

 with slight inaccuracies, toward the nest. In the immediate 

 vicinity of the nest the ant stopped and searched for the en- 

 trance, showing here a realization of having gone a certain dis- 

 tance. This sense is quite variable. Sometim.es the ant stops 

 altogether too soon. When the ant first experimented upon 

 was away from the nest Cornetz changed the aspect of the 

 returning line by sweeping, digging ditches and bringing piles 

 of earth or rubbish. The ant returned without hesitation, along 

 a nearly direct line toward the nest. This experiment, which 

 was repeated many times, shows conclusively that sight, smell, 

 touch, chemotropism or topochemism play absolutely no role 

 in the return of the foraging ant to the nest, except in the imme- 

 diate vicinity of the entrance, when these senses may be brought 

 into play. As far as these experiments went they seemed to 

 show the existence of some mysterious "homing" sense, but 

 further experiments were made that utterly disproved any such 

 supposition. 



On a plantain leaf, left in the vicinity of a nest until the 

 ants had become accustomed to its presence, Cornetz placed a 

 number of seeds. An ant (a large Messor) climbed upon this 

 and seized one of the seeds. Cornetz carried the leaf several 

 meters aw T ay and placed it again on the ground. The ant 

 was then absolutely incapable of returning directly to the nest, 

 but began to wander about, increasing the extent of its search- 

 ing, but not further on one side of the point than on the other, 

 until it approached the immediate vicinity of the nest. This 

 experiment, repeated 150 times, always gave the same result. 

 When carried to a distance of two or three meters from the 

 nest, it took from a half to three-quarters of an hour before the 

 ant could find the nest entrance, which, had the ant walked 

 such a distance from the nest, would have taken only a minute 

 or two. From this it is evident that the ant possesses no sense 



