quantity of rock in
its western part. Many of the boulders showed a foliation in their strata
with a dip of 45 deg. east. The accumulation of boulders formed a formidable
barrier before we reached an island most beautiful to gaze upon, so
luxuriant was the vegetation on it.

This particular island was 200 m. long; next to it was another 150 m.
long; then, joined to this by a link of high rocks to the south-east, was
a third, also of considerable beauty. So charming were these islands that
I called the group the Three Graces Islands.

The river turned due west from that point in a channel of continuous
rapids and violent eddies for some 3,000 m. We went down, the canoe being
knocked about in a most alarming way on one or two occasions, and
shipping so much water as to reach almost up to our knees inside it.

It was fortunate that all my photographic plates, note-books and
instruments were in water-tight boxes, or they certainly would have been
damaged beyond saving. This was not the case with my clothes, shoes, and
bedding, which had now been wet for many days with no possibility of
drying them, as we were travelling all day long and every day, and during
the night the heavy dew prevented them from getting dry. Why we did not
get rheumatism I do not know, as not only did we wear wet things all day
long, but we slept in blankets soaked with moisture.

The moment I dreaded most was that in which we emerged from the rapid
into the whirlpool which always followed, and in which the canoe swerved
with such terrific force that it was all we could do to hold on and not
be flung clean out of her--owing, of course, to the centrifugal force as
she revolved quickly.

Making a survey of the river was getting to be a complicated and serious
job, what with the numberless islands we encountered, the continuous
rapids, and the constant changes of direction. I was busy writing, as
fast as I could--only interrupted momentarily by involuntary
shower-baths--prismatic compass and watch in hand all the time, the
latter in order to measure the distances as accurately as possible.

We had now come to another group of islands in a line in the centre of
the river. They had been at one time evidently all one, which had
subsequently been eroded into five separate islands and an extensive bank
of gravel and sand. Taken in succession from south to north, there was
first an oblong island, thickly wooded, 120 m. long--Nina Island--having
on its w