332 THROUGH THE BRAZILIAN WILDERNESS 



by actual exploration and investigation discovered that the 

 river whose upper portion had been called the Duvida on 

 the maps of the Telegraphic Commission and the unknown 

 major part of which we had just traversed, and the river 

 known to a few rubber-men, but to no one else, as the 

 Castanho, and the lower part of the river known to the 

 rubber-men as the Aripuanan (which did not appear on 

 the maps save as its mouth was sometimes indicated, with 

 no hint of its size) were all parts of one and the same river; 

 and that by order of the Brazilian Government this river, 

 the largest affluent of the Madeira, with its source near the 

 1 3th degree and its mouth a little south of the 5th degree, 

 hitherto utterly unknown to cartographers and in large 

 part utterly unknown to any save the local tribes of In- 

 dians, had been named the Rio Roosevelt. 



We left Rondon, Lyra, and Pyrineus to take observa- 

 tions, and the rest of us embarked for the last time on the 

 canoes, and, borne swiftly on the rapid current, we passed 

 over one set of not very important rapids and ran down 

 to Senhor Caripe's little hamlet of Sao Joao, which we 

 reached about one o'clock on April 27, just before a heavy 

 afternoon rain set in. We had run nearly eight hundred 

 kilometres during the sixty days we had spent in the canoes. 

 Here we found and boarded Pyrineus's river steamer, which 

 seemed in our eyes extremely comfortable. In the senhor's 

 pleasant house we were greeted by the senhora, and they 

 were both more than thoughtful and generous in their 

 hospitality. Ahead of us lay merely thirty-six hours by 

 steamer to Manaos. Such a trip as that we had taken 

 tries men as if by fire. Cherrie had more than stood every 

 test; and in him Kermit and I had come to recognize a 



