BEHAVIOR OF ANTS 421 



ber, and the aphids thus had a better chance of increasing." 

 Marsh found little evidence for the belief that Formica harbors 

 the aphids or their eggs in its nest during the winter or distri- 

 butes them over the vines. 



Neger '"' studied the harvesting habits of Messor harbanis 

 on the island of Arbe, in Dalmatia. This ant was found to 

 collect and store in its nest the seeds of a great number of differ- 

 ent plants, but often it seemed to mistake empty seed-pods, 

 husks, or other small objects for seeds and carried them into the 

 nest. These, however, were rejected by the more discriminating 

 individuals within the nest and carried out to the refuse heap. 

 The germination of the seeds is not prevented by the ants, as 

 so many observers have maintained since the days of Pliny. 

 When this process fails to occur, it may be due to the fact that 

 many seeds have to remain quiescent for a long period before 

 germination, even when they are kept moist, while others even 

 require exposure to the light. That germination was not pre- 

 vented was seen in all cases where seeds were carried out by the 

 ants and left to dry, usually on the refuse heap, for such seeds, 

 according to Neger, had all germinated. After drying they w^ere 

 carried back into the nest. Neger believes that germination is 

 really favored by the ants in order to bring about the dehiscence 

 and thus to facilitate the removal of the often ver}^ tough or 

 hard seed-coats. The drying, moreover, kills the seeds. That 

 this is not a malting process is shown by the fact that the seeds 

 are not permitted to germinate long enough to convert the 

 starch into maltose. Neger made the interesting discovery that 

 M. barbarus kneads the husked seeds into little pellets of dough, 

 which he calls "ant bread-crumbs." These are also dried in 

 the sun (sterilized?) when they become moist and are again 

 carried into the nest. Apparently these crumbs contain a fungus 

 (Aspergillus niger) the amylolytic and proteolytic action of 

 which renders them more easily assimilable by the ants and 

 their larvae. M. barbarus, like the American harvesters of the 

 genus Pogonomyrmex, mows down the plants growing on its 

 nest, but it carries their leaves down into the galleries, for what 

 purpose Neger was unable to determine. 



O'Brien '"* confirms the observations of former authors who 

 have studied the habits of the green tree ant (CEcophylla virides- 

 cens) of Queensland, in so far as these relate to the use of the 



