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



easily, and second, for ventilating the subterranean portions of the nest. 

 In the nests of Aita texana I have been unable to detect two kinds of craters, 

 one used as entrances, the other for ejecting the exhausted portions of the 

 fungus gardens, as Forel has observed in the Colombian Atta cephalotes 

 and as I have observed at Cuernavaca, Mexico, in the nests of A. mexicana. 

 All the craters when fresh, consist of large, uniform pellets of earth or sand, 

 3-5 mm. in diameter, which are carefully compacted and carried to the sur- 

 face by the workers. The grains of sand or earth seem to be held together 

 merely by the moisture that permeates the soil at the depth from which they 

 are dug, rather than by any salivary secretion such as von Ihering supposes 



■v.. 





^^.: 



mmmm 



Fig. 9. One of the craters of the Atta texana nest represented in the preceding figure, about 

 J natural size. (Photograph by Mr. C. G. Hartraann.) 



the Brazilian A. sexdens to employ for this purpose. The pellets disinte- 

 grate in the first rain, so that the walls of the craters become lower and more 

 rounded and fuse with one another to form the low mound of older nests. 

 The ants usually work at only a few of the craters at a time, and as only one 

 or two of the openings are used when the ants are busily engaged carrying 

 in leaves, it seems probable that the greater number of craters is con- 

 structed for the aeration of the nest and not for entrance or exit. 



The depth and extent of the excavations vary, of course, with the size 

 of the colony, its age, and the character of the soil. This is evident from the 

 following; notes on three nests examined at different seasons of the vear. 



