XVIII CANELOS TO BANGS 151 
to be met with, nor even a palm-leaf wherewith to 
thatch a rancho. Our Indians therefore made the 
roofs by tying long slender sticks across each other, 
so as to form small squares, and then overlaying 
them with such large leaves of terrestrial and 
epiphytal Aroideae (chiefly species of Anthurium) 
as they could meet with. Roofs so constructed 
are not proof against heavy rains, and the leaves 
soon begin to shrivel and rot. Our huts being put 
up, we cooked our humble supper and lay down to 
sleep. At 9 P.M. heavy rain came on and continued 
without intermission till daybreak (5 a.m.) of the 
27th. When we looked out in the morning we 
saw that the river had risen still higher, and there 
was no hope of getting across this day. Our pro- 
visions began to run low. The Indians had drank 
their last chicha, and they had all along kept 
robbing me of such eatable things as I could not 
keep under lock and key, so that my stock of salt 
fowls was reduced to three, and I had only besides 
a few dried plantains in a tin secured by a padlock ; 
with their usual carelessness for the morrow, 
they had already eaten up the large monkeys killed 
at Casha-urcu, and all their provision consisted of 
a few baked plantains. 
The day continued gloomy, but no more rain 
fell. I sallied forth along the river-bank to see if I 
could meet with anything eatable. Rude granite 
blocks, often with quartz veins, and here and there 
small masses of pure quartz, were so heaped up as 
not to be passed without difficulty and danger. 
Among them grew scattered plants of a small 
Cardamine, of which I gathered all I could find to 
eat as salad. I then struck into the forest and 
