472 



J. C. BRANNER GEOLOGIC WORK OF ANTS 



level did not form a complete row, but the work seemed to have been 

 commenced at the outside. 



This same arrangement of the tunnels has been seen frequently in 

 railway cuts and ditches, but nowhere else have I seen so man}^ levels or 

 such a clearly defined plan in the placing of the excavations. 



In some other cases noted the number of tunnels connecting the above- 

 ground mounds with the underground galleries seemed to vary with the 

 size of the mounds — that is, the more ground the mound covered, the 

 more passageways there seemed to be to connect with the galleries 

 beneath. 



Figure 4. — Nest of leaf-cutting Ant 

 After Belt. "The Naturalist in Nicaragua," p. 80 



Bates gives the following regarding the sauhas. 



.31 



"The entrances are small and numerous ; in the large hillocks it would re- 

 quire a great amount of excavation to get at the main galleries ; but I suc- 

 ceeded in removing portions of the dome in smaller hillocks, and then I found 

 that the minor entrances converged, at the depth of about two feet, to one 

 broad, elaborately worked gallery, or mine, which w^as four or five inches in 

 diameter." 



The section through the burrows given by Belt is reproduced in figure 

 4. This section, however, is diagrammatic, and does not claim to show 

 the great extent of the galleries. Belt tells, however, of galleries 1.5 

 meters in depth (page 76). The best evidence I have been able to gather 

 in regard to the depth to which the ants penetrate has been obtained in 

 cuts along railways and canals, and in deep ditches often dug to serve as 

 fences. On Eio do Peixe, near Serro, in the State of Minas Geraes, I 

 found the galleries as deep as 2.5 meters at several places along a canal 

 under construction. Most of them, however, were only about 1.5 meters 

 below the surface at the deepest points exposed. At Bomfim, on the 



"i H. W. Bates : The Naturalist on the Amazons. 



