that the Coníbos practise the art of knitting cotton cloth, which they fashion into long cloaks. The cloth, of which I saw many specimens, forms a regular, durable, and not inelegant web of tolerably close texture. The Coníbos, like the Indians of Peru, do not grow the poisonous kind of mandioca, but simply the sweet kind, or Macasheira (Manihot Aypi). I estimate the length of the Jutahí at about 400 miles, and that of the Juruá at 600 miles.
We remained at anchor four days within the mouth of the Sapó, a small tributary of the Jutahí flowing from the south-east; Senhor Raiol having to send an igarité to the Cupatána, a large tributary some few miles further up the river, to fetch a cargo of salt fish. During this time we made several excursions in the montaria to various places in the neighbourhood. Our longest trip was to some Indian houses, a distance of fifteen or eighteen miles up the Sapó, a journey made with one Indian paddler, and occupying a whole day. The stream is not more than forty or fifty yards broad; its waters are darker in colour than those of the Jutahí, and flow, as in all these small rivers, partly under shade between two lofty walls of forest. We passed, in ascending, seven habitations, most of them hidden in the luxuriant foliage of the banks; their sites being known only by small openings in the compact wall of forest, and the presence of a canoe or two tied up in little shady ports. The inhabitants are chiefly Indians of the Marauá tribe, whose original territory comprised all the small by-streams lying between the Jutahí and the Juruá, near the mouths of both these great tribu-
