coming to one's
face; and without the assistance of both hands to keep them off, they
will creep into one's nostrils, and mouth too, if the lips are not shut
very close."

Sir Edward Pellew's Group, as will be seen by a reference to the plan,
affords numerous anchorages against both the south-east and north-west
monsoon; but unless it should be within the two small isles near the
south-west side of Vanderlin's Island, where the depth was not well
ascertained, there is not a single harbour, the different bays and coves
being too shallow to admit a ship. Wood for fuel is easy to be procured;
and water may be had in December, and probably as late as April or May,
but I think not afterwards. The most accessible watering place we could
find, was at the back of the mangroves near our principal anchorage,
within the east point of North Island, where, with some trouble, our
casks were filled; and at a beach there, left dry at low water, the seine
was hauled with some success. At Vanderlin's Island there are many
beaches fit for the seine; and indeed it seemed superior to the other
islands as well for this, as for every other purpose, when a ship can lie
there; it is also the most frequented by the Indians, and may probably
have fixed inhabitants.

The _latitude_ of Observation Island, from two meridian altitudes to the
north and south, is 15 deg. 36' 46" S.

_Longitude_ from six sets of distances of the sun east of the moon, given
in Table IV. of Appendix No. 1, 137 deg. 6' 42"; but by the time keeper No.
543 corrected, it is preferably 137 deg. 3' 15" E.

The rates of the time keepers were found from afternoon's altitudes in an
artificial horizon, between the 16th and 26th; and the means, with their
errors from mean Greenwich time, at noon there on the last day of
observation, were as under:

Earnshaw's No. 543, slow 2h 29' 11.17" and losing 14.93" per day
Earnshaw's No. 520, slow 4h 11' 37.59" and losing 28.25" per day

This rate of No. 543 is only 0.19" more than that found at Sweers'
Island, and so far as the six sets of lunars may be relied on, the
longitude by this time keeper was not far from the truth; the letting
down on the passage therefore did not seem to have produced any
change; but in No. 520, the rate is more than 8" greater, and the
longitude was getting 11/2' per day too much to the east, as well before as
after it was let down. The coast from Sweers' to Observation Island is
consequently laid down by