the branch rivers of the Amazons which flow from the south through the interior of Brazil, with the first cataracts. We started on the 19th; our direction on that day being generally south-west. On the 20th our course was southerly and south-easterly. This morning (August 21st) we arrived at the Indian settlement, the first house of which lies about thirty-one miles above the sitio of Joaõ Aracú. The river at this place is from sixty to seventy yards wide, and runs in a zigzag course between steep clayey banks twenty to fifty feet in height. The houses of the Mundurucús to the number of about thirty are scattered along the banks for a distance of six or seven miles. The owners appear to have chosen all the most picturesque sites—tracts of level ground at the foot of wooded heights, or little havens with bits of white sandy beach—as if they had an appreciation of natural beauty. Most of the dwellings are conical huts, with walls of framework filled in with mud and thatched with palm leaves, the broad eaves reaching halfway to the ground. Some are quadrangular, and do not differ in structure from those of the semi-civilised settlers in other parts; others are open sheds or ranchos. They seem generally to contain not more than one or two families each.
At the first house we learnt that all the fighting men had this morning returned from a two days' pursuit of a wandering horde of savages of the Parárauáte tribe, who had strayed this way from the interior lands and robbed the plantations. A little further on we came to the house of the Tushaúa or chief, situated on the top of a high bank, which we had to ascend by
