ANT STRUCTURES 471 



directly related to the age of the colony. At Serrmha, on the Sao Fran- 

 cisco Eailway, I was told that mounds about 2 meters high and having a 

 base of about 5 meters were probably as much as a hundred years old. 

 This was an expression of views based simply upon a general impression 

 and not upon records. 



M. Gounelle^^ says in regard to the age of ant-hills : 



"The age of these ant-hills is rather difficult to determine; however, the 

 curate of the city mentioned above (Condeuba, Bahia) has assured me that 

 three nests like those in the photograph, located at the entrance to the ceme- 

 tery, were not more than one metre high when he went to that country, about 

 sixty years ago. Taking a hundred years as an average for the building of 

 these gigantic ant-hills, it does not seem rash to suppose that in these regions 

 the soil must have been worked over by the ants to a certain depth several 

 times in the course of geologic periods." 



UNDERGROUND WORK 



So far as I can learn, there has never been any careful examination or 

 study of the character, extent, and uses of the underground excavations 

 made by ants in the tropics. What is known about them has been 

 learned accidentally, and our knowledge of the passages is, therefore, 

 fragmentary. I have frequently dug into the mounds, but always with- 

 out the time necessary for satisfactory results. The most I have been 

 able to make out in these hasty explorations is that the superficial mounds 

 are penetrated in every direction with passageways. The large mounds 

 were in no case opened down to the original surface of the ground; but 

 when small mounds were -opened they were found to connect through 

 small tunnels with the underground excavations. 



A pit started by the removal of a large ant-hill east of Timbo, in the 

 interior of Bahia, and continued to a depth of about 4 meters, showed 

 the arrangement of the underground tunnels better than I have seen it 

 elsewhere. The section did not pass through the main shaft or tunnel 

 that connected the ant-hill with the subterranean excavations, but a little 

 to one side of it. The upper layer of the earth, to a depth of half a 

 meter, was undisturbed; then there was one tunnel with a flat floor, 

 about 20 to 25 centimeters across, and having a low arched roof; below 

 this, at a distance of about 25 centimeters, were two tunnels at the same 

 level and of about the same size and shape; below these, at a further 

 depth of about 25 centimeters, were three similar openings. This ar- 

 rangement continued to a depth of nearly two meters, the tunnels being 

 more numerous alwavs at the lower levels. The tunnels at the lowest 



*> E. Gounelle : Transport de terres effectue par des fourmis au Bresil. Bull. Soc. 

 Entomologique de France. 7me ser., no. 6, p. 332. Paris, 1896. 



I. 



