sed a sharp whistling
and occasionally quite melodious notes.

I again took observations for latitude and longitude at this place, but I
was beginning to find the work too heavy--not the observing in itself,
but the computing of all the observations, at which I was not
particularly quick. (Lat. 9 deg. 24' S.; Long. 58 deg. 40' W.) Also, the great
care which I had to take of the chronometer under most difficult
circumstances was a trial to me, considering the numberless things I had
to look after. The only little comfort I had on that journey had been my
camp bed, on which I could, if not sleep soundly, at least rest my weary
bones for a few hours at night. That had now gone, and I was beginning to
feel the strain of the hard work, constant mental exertion, and the total
lack of rest.

We had passed a great number of islands in the morning: one 2,000 m.
long--Melusine Island; another 300 m.--Janus Island; a third 3,000
m.--Midas Island--by the side of which was another enormous island, some
6,000 m. in length--Miranda Island. Then little islets 200 and 250 m.
long, and another big island, 2,000 m. from end to end--A. Maso Island.

Most beautiful sandy beaches were now constantly seen, mostly, like the
one on which we had landed, composed of singing sand. (Some of those
beaches were 200 and 300 m. long.) The beach on which we had landed for
lunch was at the southern end of a great island, 5,700 m. long, which I
named Queen Mary Island.

We left again that afternoon, travelling fairly speedily, chiefly in
W.N.W. and S.S.W. directions, varying from 290 deg. b.m. to 230 deg. b.m. When we
came to the end of Queen Mary Island, after passing some really
remarkable beaches on which we found a great many turtles' eggs, we came
to a large basin, 1,800 m. across, with numberless rocks scattered on the
north and south sides of it. The river there flowed due west; in fact,
those rocks formed a kind of corona all around the great circle. A
crescent-shaped island, 2,800 m. long--Giselle Island--was next passed.
The channel through which we went was full of dangerous rocks, and had a
width of 280 m.

Soon after another basin 1,600 m. broad was reached, with a formidable
barrier of islets and rocks spreading from south to north. The river
there flowed in a perfectly straight course for 10 kil. to 310 deg. b.m. A
most extraordinary-looking islet with a circular terrace of rock on the
east side of it, which was passed in mid-stream, w