ography of these isles, in the narrative of
my former voyage, I shall now only add, that we found the latitude of
Oaiti-piha Bay, in Otaheite, to be 17 deg. 43' 26" south, and the longitude 0 deg.
21' 25" 1/2 east from Point Venus; or 149 deg. 13' 24" west from Greenwich. The
difference both of latitude and longitude, between Point Venus and Oaiti-
piha, is greater than I supposed it to be, when I made the circuit of the
island in 1769, by two miles, and 4-3/4 miles respectively. It is therefore
highly probable, that the whole island is of a greater extent than I, at
that time, estimated it to be. The astronomers set up their observatory,
and made their observations on Point Venus, the latitude of which they
found to be 17 deg. 29' 13" south. This differs but two seconds from that which
Mr Green and I found; and its longitude, viz. 149 deg. 34' 49" 1/2 west, for
any thing that is yet known to the contrary, is as exact.

Mr Kendal's watch was found to be gaining on mean time 8" 863 per day,
which is only 0" 142 less than at Queen Charlotte's Sound, consequently its
error in longitude was trifling.

A VOYAGE TOWARDS THE SOUTH POLE, AND ROUND THE WORLD.

BOOK II.

FROM OUR DEPARTURE FROM THE SOCIETY ISLES, TO OUR RETURN TO AND LEAVING
THEM THE SECOND TIME.

CHAPTER I.

_Passage from Ulietea to the Friendly Islands, with an Account of the
Discovery of Hervey's Island, and the Incidents that happened at
Middleburg._

1773 September

After leaving Ulietea, as before mentioned, I steered to the west,
inclining to the south, to get clear of the tracts of former navigators,
and to get into the latitude of the islands of Middleburgh and Amsterdam;
for I intended to run as far west as these islands, and to touch there if I
found it convenient, before I hauled up for New Zealand. I generally lay-to
every night, lest we might pass any land in the dark. Part of the 21st and
22d the wind blew from N.W., attended with thunder, lightning, and rain,
having a large swell from S.S.E. and S., which kept up for several days,--
an indication that no land was near us in that direction.

On the 23d, at ten o'clock in the morning, land was seen from the top-mast
head, and at noon from the deck, extending from S. by W. to S.W. by S. We
hauled up for it with the wind at S.E., and found it to consist of two or
three small islets, connected together by breakers like most of the low
isles in the sea, lying in a triangular form, and about si