DANGER OF CANYONS 289 



whatever to following even a fairly dangerous river 

 which has been thoroughly explored and has become 

 in some sort a highway, so that experienced pilots can 

 be secured as guides, while the portages have been 

 pioneered and trails chopped out, and .every dangerous 

 feature of the rapids is known beforehand. In this case 

 no one could foretell that the river would cleave its way 

 through steep mountain chains, cutting narrow clefts 

 in which the cliff walls rose almost sheer on either hand. 

 When a rushing river thus "canyons," as we used to 

 say out West, and the mountains are very steep, it 

 becomes almost impossible to bring the canoes down 

 the river itself, and utterly impossible to portage them 

 along the cliff sides, while even to bring the loads over 

 the mountain is a task of extraordinary labour and 

 difficulty. Moreover, no one can tell how many times 

 the task wiU have to be repeated, or when it will end, 

 or whether the food will hold out ; every hour of work 

 in the rapids is fraught mth the possibility of the 

 gravest disaster, and yet it is imperatively necessary to 

 attempt it ; and all this is done in an uninhabited 

 wilderness, or else a wilderness tenanted only by un- 

 friendly savages, where failure to get through means 

 death by disease and starvation. Wholesale disasters 

 to South American exploring parties have been frequent. 

 The first recent effort to descend one of the unknown 

 rivers to the Amazon from the Brazilian highlands 

 resulted in such a disaster. It was undertaken in 1889 

 by a party about as large as ours under a Brazilian 

 engineer officer. Colonel Telles Peres. In descending 

 some rapids they lost everything — canoes, food, 

 medicine, implements — everything. Fever smote them, 

 and then starvation. All of them died except one 

 officer and two men, who were rescued months later. 



19 



