414 WILLIAM MORTON WHEELER 



that these insects in their forays may suddenly change the 

 direction of their route at wiU and do not slavishly adhere to 

 an odor-trail. Observation "appears to indicate that the sense 

 of smell may be superseded any moment by that of sight or 

 by the individual's will to fluctuate in its adherence to scent. 

 It appears in sanguinea that the sight of the many or some 

 social responsiveness of the many (a responsiveness considered 

 usually to be communicated through the sense of touch in the 

 antennae) may supersede the trail scent equally well." Burrill 

 finds, moreover, that the files of ants do not always take the 

 shortest path between their own nest and that of the fusca 

 which they are plundering, but sometimes make detours to 

 avoid uneven or arduous paths. 



Cornetz **"■ ^' made an elaborate study of more than one 

 hundred trails to and from the nest in the following North 

 African ants: Myrmecocystus hicolor and its subspecies mega- 

 locola, Messor harhanis and its variety sanctus, Aphoonogaster 

 testaceopilosa, Monomorium pharaonis and Tapinoma erraticum. 

 These trails were carefully plotted on paper, and, together 

 with minute observations of the behavior of each ant by which 

 they were laid down, constitute the basis of fact from which 

 the author draws his conclusions. The species studied are dis- 

 tinguished as "superior" and "inferior." To the former class 

 belong the Myrmecocysti, large, active ants, with well-developed 

 eyes and ocelli. They forage singly. Messor, Monomorium, 

 and Tapinoma, which belong to the inferior group, are usually 

 seen foraging in files. The range of distinct vision in Tap- 

 inoma can scarcely exceed 2 mm., in Messor harhanis 4-5 mm., 

 whereas in Myrmecocystus it is 5-6 cm.- The formation of straight 

 files in the inferior forms, from the nest to a definite food-supply 

 and home again, is a derivative phenomenon which is preceded 

 by the development of the trail of the individual worker, which 

 leaves the nest, discovers the food-suppl}^ and returns without 

 relying on any preestablished trail. The author therefore 

 correctly insists that the homing behavior must be studied, 

 not on files of ants which are following a developed trail, but 

 on the single exploring worker When such a worker is studied, 

 she is seen to depart from the nest opening in a definite direction 

 and, after going a short distance, to halt and move about, often 

 in a complicated path, exploring for food. If none is found 



