DOWN AN UNKNOWN RIVER 289 



way down the rapids Lyra and Kermit, who were oversee- 

 ing the work as well as doing their share of the pushing 

 and hauling, got them into a canal of smooth water, which 

 saved much severe labor. As our food supply lowered we 

 were constantly more desirous of economizing the strength 

 of the men. One day more would complete a month since 

 we had embarked on the Duvida as we had started in 

 February, the lunar and calendar months coincided. We 

 had used up over half our provisions. We had come only 

 a trifle over 160 kilometres, thanks to the character and 

 number of the rapids. We believed we had three or four 

 times the distance yet to go before coming to a part of 

 the river where we might hope to meet assistance, either 

 from rubber-gatherers, or from Pyrineus, if he were really 

 coming up the river which we were going down. If the 

 rapids continued to be as they had been it could not be 

 much more than three weeks before we were in straits for 

 food, aside from the ever-present danger of accident in 

 the rapids; and if our progress were no faster than it had 

 been and we were straining to do our best we would in 

 such event still have several hundreds of kilometres of un- 

 known river before us. We could not even hazard a guess 

 at what was in front. The river was now a really big 

 river, and it seemed impossible that it could flow either 

 into the Gy-Parana or the Tapajos. It was possible that 

 it went into the Canuma, a big affluent of the Madeira low 

 down, and next to the Tapajos. It was more probable that 

 it was the headwaters of the Aripuanan, a river which, as 

 I have said, was not even named on the excellent English 

 map of Brazil I carried. Nothing but the mouth had been 

 known to any geographer; but the lower course had long 



