h of the Arinos River.]

There was a nasty-looking rapid close to the camp. We had to let the
empty canoe down carefully by means of ropes, my men on that particular
occasion donning their lifebelts again, although they walked on dry land
when they were taking the canoe along. When I asked them why they put
them on, they said that perhaps the canoe might drag them into the
water and they had no wish to get drowned.

We left that camp late in the afternoon--at three o'clock--having wasted
the entire morning conveying the canoe to a spot of safety and then
carrying all the baggage along overland.

After having gone some 2 kil. farther we came to another rapid and a
_pedraria_ with nasty rocks right across the channel, the only passage I
could see possible for our canoe being in the centre of the stream. That
channel was only a few metres wide, and had in the centre of it a large
rock just under the surface, which flung the water up in the air. We just
managed to shoot that rapid safely, although with trembling hearts.

Farther down, rocks innumerable, rising only two or three feet above
water, spread half-way across the channel from the right side. Then
rapids and strong eddies were encountered. For 700 m. the river showed
foliated rock strewn all along on both banks, and great volcanic boulders
of a more rounded shape. The foliation showed a dip westward of 45 deg..

We were delighted when we discovered in that region many _solveira_ or
_sorveira_ trees, or milk trees, exuding when incised milk most delicious
to drink. Then there were plenty of _figueiras_ or _gameilleiras_ and
wild bananas. We wasted much time extracting milk from the _solveiras_
and eating wild fruit.

Monkeys were to be seen in that part. They seemed most astonished on
perceiving us, and came quite close, gazing at us in the most inquisitive
manner.

We felt that we had come to a real heaven on earth, except for the
river, which could have given points to the River Styx of infernal fame.

When we returned to the canoe we found obstructions of all kinds in the
stream. Small rapid succeeded small rapid. Rocky islets and scattered
rocks rendered navigation complicated.

Where the river turned sharply to the N.N.E. another dangerous rapid was
reached, with rocks scattered all over the channel, some just submerged.
We tried to shoot that rapid on the east side, but we got badly stuck on
a submerged rock, and once more the canoe filled with water.