o26 ANNUAL, REPORT SMITHSONIAN INSTITUTION, 1911. 



in order to observe the workers repair them. In every case observed 

 the repairs were made by building up a wall or covering of excre- 

 ment or something of the kind. At least it is voided from the poste- 

 rior part of the body in a plastic condition, and is smoothed down 

 on the sides so that the later layers always override the earlier ones 

 on both sides of the wall. An examination of their construction, 

 however, shows that they are made partly of clay or the earth about 

 the nest and partly of woody fiber. These two substances are 

 variously mixed in structure, sometimes one being more abundant, 

 sometimes the other. An examination of the materials of the out- 

 •side part of the Targe and old nests, however, shows that this part 

 of the nest at least contains fragments of quartz, sand grains, and 

 such like rock fragments that could not possibly have passed through 

 the bodies of the insects. The structure of some of the nest walls 

 suggests that these walls are constructed partly of earth and rock 

 fragments brought up from beneath the ground and built into the 

 nests by cementing them together with excrement or some other 

 adhesive- substance. 



The outer parts of the nests, when they stand on the ground, are, so 

 far as my observations go, always made of earth cemented in a thick, 

 hard wall. In the inner portions of the nest the partitions are thinner, 

 and though they are made largely of an easy-spreading clay, they are 

 often made partly, or at least overspread, with a dark, friable sub- 

 stance that has the appearance of being masticated wood, leaves, or 

 other organic matter. 



The openings through the mass of the nests are pretty uniform in 

 size, being from 3 to 10 millimeters in diameter and averaging close to 

 5 or 6 millimeters. The openings within the nests sometimes have the 

 appearance of being arranged in rude tiers; sometimes they are appar- 

 ently haphazard labyrinths. 



The external forms of the nests vary considerably, but unfortu- 

 nately I do not know whether this variation is due to difference in the 

 species of termites, to difference in the nature of the ground, or to 

 other causes. 



As a rule, the mounds are rudely domed, rounded or conical, and the 

 method of adding to the outside gives them a bumpy, lumpy appear- 

 ance, so that, as Burmeister suggests, they resemble gigantic Irish 

 potatoes. In some localities they are mostly tall and slender. Most 

 of the tall, slender forms observed have been in wet ground or on 

 ground that is sometimes overflowed. For this reason it is inferred 

 that these forms are due to the presence of water rather than to a dif- 

 ferent species of termites. In size they also vary greatly. I have 

 seen them as much as 6 meters high and 8 meters in circumference, but 

 these very large ones are exceptional. 



