282 ANTS. 



but arc usually >mallcr and more often situated under large stones. 

 Uoth species are omnivorous with an evident preference for fruits and 

 seeds. At Monahans, I found the craters of cockcrelli covered with 

 the disjointed pods of the mesquite (1'rosopis juli flora) which had 

 been carried into the nest, deprived of the sweet pulp enclosing the 

 hard seeds and then rejected. On the kitchen middens I also recog- 

 nized the legs and elytra of three species of Eleodcs and of several 

 other beetles. At Fort Davis, the workers of albisctosns were seen 

 carrying in the dried seeds of umbelliferous plants, grasses and cotton- 

 wood (Popuhis fremonti), and occasionally stopping to collect pieces 

 of insects (shards of Podof>hylla and Coccinella) and bits of cow-dung, 

 or even bird-droppings. None of these substances, however, is stored 

 in the nests, but merely carried in and then rejected. These ants are 

 not, therefore, highly developed harvesters like those of the allied sub- 

 genus Messor, but resemble more closely the northern species of Apluc- 

 nof/aster. In Connecticut I have often seen A. picea collecting and 

 temporarily storing in its nests small flowers, green seeds or the pulp- 

 covered akenes of raspberries, and Emery long ago made a similar 

 observation on an Italian variety of A. testaceopilosa. Of this ant he 

 says : " It is not predaceous but collects soft vegetable substances, such 

 as petals of flowers and green seeds, which it carries into the nest and 

 then rejects, after having extracted from them any utilizable sub- 

 stances," and he adds in a foot-note: " In a courtyard of the University 

 of Palermo I saw this ant daily collecting the petals of roses that were 

 somewhat dried but still of the natural tint, and later rejecting them, 

 soiled and crumpled, and of a yellow color, as if they had been tritu- 

 rated. The typical A. testaceopilosa, which I have observed in Sar- 

 dinia, has intermediate habits and lives partly on prey, partly on 

 vegetable substances." During the summer of 1907 I was able to 

 confirm Emery's observations on several colonies of A. testaceopilosa 

 in the Alameda at Gibraltar. These observations are of considerable 

 interest in connection with the habits of the fungus-growing ants to be 

 considered in the next chapter. 



The most characteristic American harvesters are the large or 

 medium-sized, black or red ants of the genus Pogonomyrmex, which is 

 closely related to Mynnica of the temperate and boreal portions of the 

 Northern hemisphere. In most of the species of Pogonomyrmex the 

 head of the workers and females bears on its under side a conspicuous 

 beard of long, curved hairs (ammochaetc'e ). a character to which the 

 generic name ("bearded ant") refers and one which occurs also in 

 many species of Messor, in Holconivnnc.r and in other desert ants 

 (see p. 16). The genus ranges from British America to Patagonia, 



