1907.] Wheeler, Fungus-growing Ants of North America. 683 



alone are sufficient for its food at that period of its life. But it is not so 

 when the larvae have increased so much in size, that the pout can be seen 

 without a glass, for then the whole piece after having been manipulated by 

 the nurse's mandibles into a ball, in the same manner as the leaves are 

 served, when they are first brought into the nest, is placed in its throat and 

 if that is not sufficient the pout continues when the next one and even the 

 next passing proceeds with the feeding, till the pout is withdrawn, showing 

 that it is satisfied. No further notice is then taken of it by the feeders, until 

 it agains asks for a meal by pouting later on in the day." 



In 1893 a nephew of the celebrated Fritz ]Miiller, Alfred ]Moeller, who 

 was given a grant of 5,000 marks by the Berlin Academy of Sciences for 

 the purpose of studying the habits of the Attii at Blumenau in the province 

 of Rio Grande do Sul, published the most important of existing works on 

 these insects and their relations to the fungi which they cultivate. He 

 studied several species of Atta belonging to the subgenus Acromyrmex (dis- 

 cigera, coronata, octos'pinosa, moslleri) and of the genera Apterostigma (pilo- 

 suvi, viosUeri, wasmanni, and an undetermined species) and Cyphomyrmex 

 {auritus, strigatus). A. octospinosa and discigera, which nest in the woods, 

 form truncated cones of dead leaves and twigs, beneath which they excavate 

 a single chamber containing a large fungus garden sometimes 1^ meters long. 

 A. viwUeri has similar habits, but coronata resembles the species of the sub- 

 genus Atta s. str. in forming several chambers, each with its own fungus 

 garden. In all of these species the garden is built up on the floor of the 

 chamber in the form of a loose sponge-work of triturated leaf -fragments 

 permeated with fungus h^^ihae which he describes as follows: "Over all 

 portions of the surface of the garden are seen round, white corpuscles about 

 \ mm. in diameter on an average, although some of them are fully ^ mm. 

 and sometimes adjacent corpuscles fuse to form masses 1 mm. across and 

 of irregular form. After a little experience one learns to detect these cor- 

 puscles Avith the naked eye as pale, white points which are everywhere abun- 

 dant in all the nests. Under the lens they sometimes have a glistening 

 appearance like drops of Avater. They are absent from the youngest, most 

 recently established portions of the garden, but elsewhere uniformly distri- 

 buted, so that it is impossible to remove with the fingers a particle too small 

 to contain some of the white bodies. I call these the 'kohlrabi clusters' 

 of the ants' nests. They constitute the principal, if not the only food of 

 the species of A tta." These clusters are made up of the " heads of Kohlrabi," 

 which are small terminal dilatations of the hyphse of a spherical or oval form. 

 Moeller confirmed Belt's observations on the solicitude of the ants for their 

 gardens, and showed that these insects in artificial nests will completely 

 rebuild these structures within 12 hours after they have been disintegrated 



