[Illustration: Sketch map of the south-central part of the Amazon drainage system.
The party of which the writer was a member descended the Rio Commemoração and the Gy-Paraná to the Madeira. From here Manaos on the Amazon was reached by the regular steamer route down the Madeira.
The drainage between the Upper Paraguay and the Madeira is based on the surveys of the Brazilian Telegraph Commission, so far as available. Note the recently explored course of the Rio Ananás (_The Geographical Review_, January, 1916, p. 50, and February, pp. 143–144), and the completed telegraph-line from Cuyabá to Santo Antonio (_Bulletin American Geographical Society_, September, 1915, p. 693). The latter is taken from a map in the Rio de Janeiro newspaper _A Noite_ for October 25, 1915. ]
Our party consisted of Captain Amilcar de Magalhães, a remarkably skilful and wholly tireless leader, and Lieutenant Joaquim de Melho, of the Brazilian Army, Doctor Euzebio Paulo de Oliveira, a geologist, and Señor Henrique Heinisch, a taxidermist, all of the Brazilian Telegraph Commission, besides myself; then there were some thirty-odd _camaradas_, or native assistants. We had a very large pack-train of mules and oxen, as that wing of the expedition, in charge of Captain Amilcar, which had hitherto travelled ahead of the main party, was to proceed with us from this point. Our plan was to continue overland to the headwaters of the Gy-Paraná and to descend that stream to the Madeira, taking observations as we went, for, in common with many of the rivers of the South American continent, the course of this stream has not been accurately mapped.[1] Zoologically speaking, we were in a most interesting and almost unknown country, and no opportunity could be lost to add to our already large and constantly growing collection of both mammals and birds.
[1] The Gy-Paraná had been descended by two parties which Colonel Rondon detached for this purpose from his main expedition of 1909. The first, under the zoologist Alipio de Mirando Ribeiro, went down the Pimentá Bueno and the Gy-Paraná to the Madeira; the second, under Lieutenant Antonio Pyrineus de Souza, descended the Jarú and the Gy-Paraná (see map). It was the reconnoissance survey made by the first party that established the fact that the Gy-Paraná, instead of flowing northwest throughout its course, as until then supposed, turns abruptly north in 11½° south latitude, and flows in this direction for nearly three degrees, until, at another abrupt bend in 9° south, it turns west and empties into the Madeira.