THE MADEIRA RIVER 
saw in the Madeira high, gently sloping banks, covered 
with verdant grass and neat trees and palms along the top 
of them; whereas along the Amazon the trees stood almost 
in the water on the recently formed islands and banks. 
The left bank of the Madeira was of grey and reddish 
clay (grey below, red above), cut vertically, sometimes 
actually in steps. Blocks of a rectangular shape, in get¬ 
ting dried up, split and fell over, leaving the banks verti¬ 
cal. The right bank, on the contrary, was gently sloping, 
descending with a beautiful carpet of green grass into the 
stream. The islands were charming, with lovely lawns all 
round. Blackish and deep red rock, vertical and fluted, 
and with innumerable perforations, could be seen here 
and there, covered over with a padding of earth from ten 
to twenty feet deep. 
The journey up the Madeira River had no great in¬ 
terest. By seven o’clock in the evening we arrived at the 
mouth of the Canuma River — or rather at a channel con¬ 
necting the Madeira River with the river Canuma, which 
river actually has its proper mouth about half way be¬ 
tween Itaquatiara and Santarem, at a place called Parin- 
tins. By way of the connecting channel the two rivers 
were only a short distance apart, but that connecting chan¬ 
nel was not always navigable. The steam launch, which 
drew little water, would have difficulty in going through, 
even at that time, when the water was fairly high. 
We therefore thought we would stay for the night 
at the mouth of the channel, and start on our journey 
by that difficult passage in broad daylight the next day. 
There was a house on the right-hand side of the mouth of 
the channel. While we made preparations to make our¬ 
selves comfortable for the night on the launch, the pilot 
went up to the house in order to get an expert at that 
place to take us through the dangerous channel. 
I was just in the middle of my dinner when the pilot 
sent down a message for me to go up to the house at 
855 
