moments in the street, even at a distance from their nests, we were sure to be overrun and severely punished, for the moment an ant touched the flesh, he secured himself with his jaws, doubled in his tail, and stung with all his might. When we were seated on chairs in the evenings in front of the house to enjoy a chat with our neighbours, we had stools to support our feet, the legs of which as well as those of the chairs, were well anointed with the balsam. The cords of hammocks are obliged to be smeared in the same way to prevent the ants from paying sleepers a visit.
The inhabitants declare that the fire-ant was unknown on the Tapajos, before the disorders of 1835–6, and believe that the hosts sprang up from the blood of the slaughtered Cabanas. They have, doubtless, increased since that time, but the cause lies in the depopulation of the villages and the rank growth of weeds in the previously cleared, well-kept spaces. I have already described the line of sediment formed on the sandy shores lower down the river by the dead bodies of the winged individuals of this species. The exodus from their nests of the males and females takes place at the end of the rainy season (June), when the swarms are blown into the river by squalls of wind, and subsequently cast ashore by the waves. I was told that this wholesale destruction of ant-life takes place annually, and that the same compact heap of dead bodies which I saw only in part, extends along the banks of the river for twelve or fifteen miles.
The forest behind Aveyros yielded me little except insects, but in these it was very rich. It is not too
