88 
NOTES OF A BOTANIST 
compact mat or yapa, impervious to any rain. 
When required for thatch, the midrib of the frond 
is spHt nearly through along its whole length, and 
the two halves turned over to one side, when the 
leaflets of the one half fall over the interstices of 
the other half. The fronds are then spread out to 
bleach and dry, when they become perfectly white 
or light straw-colour ; nor do the leaflets curl up in 
the least ; so that a house newly thatched with 
Pindoba has a very neat and pretty appearance. 
We had now to leave the Trombetas and turn 
into the Aripecuru, which we found to have two 
islands in its mouth. We took the channel between 
the islands ; towards its farther extremity it was 
nearly choked up with the Luziola (an aquatic grass), 
through which we had some difficulty in pushing 
with poles. Other islands succeeded, and we went 
on threading narrow channels, walled in by lofty 
trees which were festooned with climbers from 
base to summit, until after midday, when we 
emerged into water clear of islands, where the 
river was more than 500 yards wide. At this point 
there was a little sandstone rock exposed on the 
banks, resembling that of Obidos. Sandbanks 
began to peep out, and some way farther on the 
river was so obstructed by them that we had 
difficulty in finding a passage. Towards evening 
we came on a very long beach, about 200 yards 
wide, standing high and dry out of the water, of 
which there was only a narrow strip along its 
western side. This is known as the Playa grande 
de tartaruga or Great Turtle-bank ; and we drew 
up alongside it as night closed in. Our men 
