796 Bulletin American Museum of Natural History. [Vol. XXIII, 



species discovered by Smith may have been nesting in the abandoned fungus- 

 stained galleries of ambrosia beetles, or the dark color of the walls may 

 have been due to other causes. I may say also that in no colonies of the 

 various species of Strum igenijs VN'hich I have found in the United States and 

 West Indies were there any traces of fungi. These ants live in rather small 

 communities under stones or in rotten wood and feed on insects. Many of 

 our species live as thief ants, after the manner of Solenopsis molesta Say, 

 in the nests of larger ants. Rhopalothrix seems to have similar habits, to 

 judge from some field notes accompanying a colony of an undescribed species 

 taken, with all its larvae and pupse, under a stone in Jamaica. 



Forel's view, hoAvever, contains an interesting suggestion, for the nature 

 of the substratum on which the fungi are grown may be supposed to throw 

 some light on the origin of the habit under discussion. In all the fungicolous 

 insects there is an unmistakable tendency to employ vegetable substances 

 that have passed through the alimentary tract of insects. This is the case in 

 all fungus-growing termites, and in the ambrosia beetles. Among the Attii, 

 as I have shov\'n, this tendency is apparent in nearly all the species that have 

 been closely observed. Though most pronounced in the lower genera and 

 subgenera (Cyphomyrmex, Apterostigma, Mycocepurus, Trachymyrmex), it 

 is not wholly lost even in the leaf-cutting Atta\ and the method employed 

 by the Atta queens in manuring their incipient fungus-gardens suggests that 

 the food plant may have been originally grown on fecal substances. It is 

 cpiite possible, however, that in the Attii this habit is secondary and that it 

 was preceded phylogenetically by culture on some other substance since 

 generally abandoned as less suited to the purpose. This leads us to a con- 

 sideration of another view on the origin of the fungus-growing habit. 



Von Ihering (1894) advances the following opinion: "We know quite 

 a number of ants, like the species of Pheidok, Pogonomyrmex and further- 

 more species of Aphanogasier and even of Losius, which carry in grain and 

 seeds to be stored as food. Such grain carried in while still unripe, would 

 necessarily mould and the ants feeding upon it would eat portions of the 

 fungus. In doing this they might easily come to prefer the fungi to the seeds. 

 If Atta lundi still garners grass seeds and in even greater than the natural 

 proportion to the grass blades, this can only be regarded as a custom which 

 has survived from a previous cultural stage." Thus von Ihering would 

 explain the origin of fungus cultivation and the supervention of the leaf- 

 cutting habit. 



This view, like Ford's, is, of coin-se, purely hypothetical. There are, 

 however, a few facts which indicate that the Attii may have developed from 

 grain-storing species allied to the Tetramorii (Meranoplus and Tetramorium) 

 as Emery has suggested. That certain harvesting species form nests and 



