Chap. I. ANT COMMUNITIES. 29 



servant told me that they would carry off the whole 

 contents of the two baskets (about two bushels) in the 

 course of the night, if they were not driven off ; so we 

 tried to exterminate them by killing them with our 

 wooden clogs. It was impossible, however, to prevent 

 fresh hosts coming in as fast as we killed their com- 

 panions. They returned the next night ; and I was then 

 obliged to lay trains of gunpowder along their line, and 

 blow them up. This, repeated many times, at last 

 seemed to intimidate them, for we were free from their 

 visits during the remainder of my residence at the 

 place. What they did with the hard dry grains of 

 mandioca I was never able to ascertain, and cannot even 

 conjecture. The meal contains no gluten, and there- 

 fore would be useless as cement. It contains only a 

 small relative portion of starch, and, when mixed with 

 water, it separates and falls away like so much earthy 

 matter. It may serve as food for the subterranean 

 workers. But the young or larvae of ants are usually 

 fed by juices secreted by the worker nurses. 



Ants, it is scarcely necessary to observe, consist, in 

 each species, of three sets of individuals, or, as some 

 express it, of three sexes — namely, males, females, and 

 workers ; the last-mentioned being undeveloped females. 

 The perfect sexes are winged on their first attaining the 

 adult state ; they alone propagate their kind, flying 

 away, previous to the act of reproduction, from the 

 nest in which they have been reared. This winged 

 state of the perfect males and females, and the habit of 

 flying abroad before pairing, are very important points 

 in the economy of ants ; for they are thus enabled to 



