316 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1911. 



The figures obtained are given in the table below: 



Table of areas and cubical contents of mounds of different sizes xoithin an area of 10,000 



square meters. 



{All measurements are in meters.] 



This estimate makes the area actually covered by the mounds close 

 to one-fifth of the total area under consideration. My notes show 

 that within areas of a few acres the ground covered by the mounds is 

 sometimes as high as one-half of the total area. The cubical con- 

 tents of the mounds, if evenly distributed over the entire 10,000 

 square meters, would have a thickness of 22.25 centimeters. 



Although the mounds within the area here considered were large, 

 they were not the biggest I have seen, nor do they average as large as 

 can be found. The largest ones measured were on the upper drainage 

 of Rio Utinga; several of these were found to be 5 meters high and 

 16 and 17 meters in diameter at the base, and each contained, there- 

 fore, about 340 cubic meters of earth. There were no other mounds 

 closer to these than 10 or 15 meters. 



The reader should be reminded, however, that this sort of thing 

 is not to be seen in all parts of the country, by any manner of means. 

 So far as my own observations go, ant mounds are unusually large 

 and unusually abundant in this particular part of Brazil. 



Age of the mounds. — The amount of work done by these ants in a 

 region where they seem to be favorably located is fairly well shown in 

 the preceding table. Trustworthy data for calculating the time 

 required to build a mound of a given size or to do any given amount 

 of work are lacking. Necessarily the time must vary with the size of 

 the colonies, other things being equal. The colonies, however, 

 appear to have their ups and downs, for while some of them increase 

 in numbers and continue to add to the mounds for long periods, 

 others appear to be less active, while still others disappear, whether 

 by migrating or through the death or captivity of the members 

 is not certainly known at present. It is interesting to note that the 

 Brazilians generally regard the size of the ant hill as directly related 



