ON THE AMAZON. : 207 
on the 10th of September, and would cease at the end of the year, by 
which time all the young turtles would have taken to the water. I 
found him familiar with the Trombétas and its principal tributaries. 
He had ascended the main river some months before as far as the 
cachoeiras, at the head of a party who were in quest of runaway 
negroes. They left their canoes on reaching the first cataract, where 
he estimated the fall of the water to be about fifteen feet, and prose- 
cuted their journey by land for some days, during which time they passed 
seven cachoeiras, some being distinct falls, and others merely rapids 
among rocks. At the rate we were travelling, he estimated the distance 
to the first cachoeira of the Trombétas at eight days, and to that of 
the Aripecurü at four days. I should have mentioned previously, that 
in consequence of information received from Senhor Angelo Bentes, 
I had decided on changing my original intention of proceeding: to the 
cachoeiras of the main river, and on visiting those of the Rio Aripecurá 
instead. From him I learnt that the falls of the latter river were more 
abrupt, and much more numerous, and that the adjacent serras were 
far more lofty. All this was confirmed by the commander of the 
Estacamento. Another reason for this change in my destination was, 
that the falls of the Aripecurü could be reached in half the time, a 
consideration of the first importance, as we had unmistakeable intima- 
tions that the rainy season was commencing. Last night was dry 
enough to allow us to sling our hammocks under the trees, but so 
gloomy that I was unable to get any observation for determining the 
position of the junction of the Cumina and Trombétas, which I much 
desired to do. 
This morning we found ourselves in a wide bay, formed by the 
junction of the Trombétas, Cumina, and Aripecuri, and appearing 
quite shut in on the north side by a long line of coast stretching east 
and west. The Trombétas enters this bay from west by north ; but, 
so far as I have been able to ascertain, its general course is from the 
north-west. The Cumina enters from east-north-east, and just within 
its mouth enters the Aripecurü from north north-east, by two mouths, 
with a long narrow island between them. In less than an hour we are 
fairly within the Aripecurü. The land on our right swells into serras 
densely covered with mato. Here and there are deep bays, with 
valleys running up from them, each, probably, affording a passage to 
an igarapé, but so overhung with trees as not to be discernible from the 
