GEOLOGIC WORK OF ANTS BRANNER. 329 



filled with smaller chambers made of materials like that used to make 

 the chambers of the mounds above ground. An examination of the 

 thin chamber walls found in some of the underground excavations 

 shows that they have been constructed of soft, plastic materials, so 

 piled up that each later addition overlaps the preceding one on both 

 sides of the wall. The materials are partly of reddish clay like that 

 of the ground in which the nest is made and partly of a dark brown 

 substance that I take to be organic matter — probably masticated 

 plants. 



I have never seen the excavations made by the white ants more than 

 a meter and a half below the surface, but I have heard of them being 

 found considerably deeper. My friend, Dr. Joaquim Lustosa, of 

 Lafayette, State of Minas Geraes, writes me on this subject: "As for 

 the depth to which they penetrate the ground, it is my impression 

 that it is but little more than 3 meters." 



RELATION OF NESTS TO THE SOIL. 



The white ants do not seem to be so dependent on the character of 

 the soil as do the true ants. This is probably due to the fact that 

 when the true ants excavate their tunnels in the earth they depend 

 on the character of the ground and the form of the excavations to 

 support the structures. The white ants, on the other hand, depend 

 partly on the nature of the soil, but partly on their method of cement- 

 ing the materials of which their nests are made. 



The preference, however, of the termites for certain soils and certain 

 localities is very evident in some districts. On the upper Paraguay 

 places have been seen where the nests are quite thick over certain 

 areas, while there were none, or but few, on an adjoining area. Wher- 

 ever these marked contrasts have been observed, however, they have 

 apparently been due to a difference in the amount of moisture in the 

 ground. I have thought that the white ants are sometimes found in 

 rather wet ground, because they are there comparatively free from 

 the attacks of their enemies, the true ants. 



Opinions of Brazilians in regard to the distribution of the termites' 

 nests vary considerably; some think they arc more abundant in the 

 open campo regions than in the forests; others think they prefer 

 fields; still others think they are favored by a dry climate. All of 

 these views appear to have more or less support. I have much doubt, 

 however, about the theory of their preferences for campos. It is true 

 that they do appear very abundant in the campo regions, but I am of 

 the opinion that the apparent abundance is deceptive and due to the 

 fact that all the nests are visible at once over a wide area (pi. 1, fig. 2), 

 while in a forest-covered area no nests, or but few nests, can be seen on 

 account of their being concealed by the dense vegetation. This im- 



