Cuap. VII. TOBACCO. 165 
ettes. It is the inner bark of a tree, which separates into thin papery 
layers. Many trees yield it, amongst them the Courataria Guianensis 
and the Sapucaya nut-tree (Lecythis ollaria), both belonging to the same 
natural order. The bark is cut in long strips, of a breadth suitable 
for folding the tobacco; the inner portion is then separated, boiled, 
hammered with a wooden mallet, and exposed to the air for a few 
hours. Some kinds have a reddish colour and an astringent taste, but 
the sort prepared by our host was of a beautiful satiny-white hue, and 
perfectly tasteless. He obtained sixty, eighty, and sometimes a hundred 
layers from the same strip of bark. The best tobacco in Brazil is grown 
in the neighbourhood of Borba, on the Madeira, where the soil is a 
rich black loam ; but tobacco of very good quality was grown by Joao 
Trinidade and his neighbours along this coast, on similar soil. It is 
made up into slender rolls, an inch and a half in diameter, and six feet 
in length, tapering at each end. When the leaves are gathered and 
partially dried, layers of them, after the midribs are plucked out, are 
placed on a mat and rolled up into the required shape. This is done 
by the women and children, who also manage the planting, weeding, 
and gathering of the tobacco. The process of tightening the rolls is a 
long and heavy task, and can be done only by men. The cords used 
for this purpose are of very great strength. They are made of the inner 
bark of a peculiar light-wooded and siender tree, called Uaissima, which 
yields, when beaten out, a great quantity of most beautiful silky fibre, 
many feet in length. I think this might be turned to some use by 
English manufacturers, if they could obtain it in large quantity. The 
tree is abundant on light soils on the southern side of the Lower 
Amazons, and grows very rapidly. When the rolls are sufficiently well 
pressed, they are bound round with narrow thongs of remarkable tough- 
ness cut from the bark of the climbing Jacitara palm-tree (Desmoncus 
macracanthus), and are then ready for sale or use. 
A narrow channel runs close by this house, which communicates at 
a distance of six hours’ journey (about eighteen miles) with the Urubu, 
a large and almost unknown river, flowing through the interior of 
Guiana. Our host told me the Urubt presented an expanse of clear 
dark water, in some places a league in width, and was surrounded by 
an undulating country, partly forest and partly campo. Its banks are 
fringed with white sandy beaches, and peopled only by a few families 
of Miira savages. The family now in his employ, and who were living 
gipsy fashion, the only way they can be induced to live, under a wretched 
shed on his grounds, were brought from this river six months previously. 
The channel was navigable by montarias only in the rainy season ; it 
was now a half-dry watercourse, the mouth lying about eight feet above 
the present level of the Amazons. The principal mouth of the Urubt 
lies between this place and Serpa. The river communicates with the 
lake of Saraca, but I could make out nothing clearly as to its precise 
geographical relations with that large sheet of water, which is ten or 
twelve leagues in length and one to two in breadth, and has an old- 
established Brazilian settlement, called Silves, on its banks. 
It was very pleasant to roam in our host’s cacaoal. The ground was 
clear of underwood, the trees were about thirty feet in height, and 
