BEHAVIOR OF ANTS AND MYRMECOPHILES 409 



which was followed 30 or 40 meters. Then they spread out, 

 some climbing weeds in search of seeds. When food was found 

 the ants returned by the same trail, which was then being 

 traversed by two files of ants moving in opposite directions. 

 The seeds gathered were of numerous varieties. Some of the 

 foragers returned carrying lumps of earth, small stones or other 

 worthless objects, a fact tending to show T variability in the 

 grain-collecting instinct. At the nest entrance several large 

 workers acted as guards and examined with their antennae the 

 returning workers before allowing them to enter the nest. The 

 husks from seeds were brought out of the nest very quickly 

 and taken to some distance. 



Among the foods collected by the ants were many tubers 

 from the root stalks of Cyperns bulbosus, which are eaten also 

 by the Abyssinians. Their carrying these about may possibly 

 be of importance in the dissemination of the plant species. 



Goeldi (25), in an account of the structure and character of 

 the ant nest, introduces the term Gonepitropy for the assump- 

 tion of the reproductive functions by the queen. The workers 

 have taken over the functions of food-gathering and nursing 

 for the entire colony, and this Goeldi calls Ergepitropy. The 

 workers Goeldi does not consider to be degenerate females, but 

 rather a distinct caste, fitted for the work that is done by them. 



Krausse (26) records a few observations made on Aphaeno- 

 gaster sardoa, a characteristic Mediterranean ant. The holes in 

 which this ant lives are not dug by it, but have been made 

 previously by other animals or natural means. In the fall the 

 ants cluster into a ball, holding on to each other with mandibles 

 and tarsi, with the eggs, larvae and pupae in the center, and 

 remain in a semi-dormant condition. This habit is analogous 

 to the hibernation of northern ants. In summer the clustering 

 habit to some extent persists, but the ants are much more 

 active. Krausse placed a number of workers with pupae in a 

 glass nest and after a half hour found them in a compact cluster, 

 the pupae in the middle. When dispersed, they again built a 

 cluster. No myrmecophiles were found with sardoa, and Krausse 

 considers their absence related to the non-harvesting and non- 

 cell-building habits of the ant. The larva-robbing Dermap- 

 teron, Enborellia moesta, which steals from the nests of other 

 ants in the same vicinity, was not found with Aphaenogaster, 



