ough
many bushes and some trees grew upon it, is little more than a bed of
sand. There were holes in the beach, made by turtle; and besides other
proofs of the islet being sometimes visited by the Indians, I found four
human skulls lying at the back of the shore.

From the woody islet I crossed over to the main land near the ship, and
took another set of bearings for the survey. Upon the shore were pieces
of bamboo, and other traces of the same foreign people of whom mention
has frequently been made; and three small huts were found, so entirely
covered with grass that no opening was left; but they were empty, and
nothing was buried underneath. On the borders of a small fresh lake the
botanists reaped a harvest of new plants, without molestation; indeed no
natives were seen any where; but several skeletons were found, standing
upright in the hollow stumps of trees; and the skulls and bones being
smeared or painted, partly red and partly white, made a very strange
appearance. Some kangaroos were perceived at a distance; and judging by
their foot-marks on the sand, they were rather numerous. The country near
the sea side is stony and barren; further back, it rises gently to a
small elevation, and seemed to be moderately well covered with grass and
wood.

WEDNESDAY 5 JANUARY 1803

In the morning of the 5th we got under way, and steered eastward for
Groote Eylandt, which I now intended to circumnavigate. In passing the
south side of Bickerton's Island, we observed in it a deep bight or bay
which would afford shelter in the north-west monsoon., if there be depth
sufficient for a ship; and the hills at the back being high and woody,
there was a probability of its receiving a stream of fresh water. The
country round the entrance of the bight, had the appearance of being
sandy and sterile.

Between the nearest parts of Groote and Bickerton's Islands is a space of
eight miles, which seemed to offer a perfectly safe passage, with
soundings, if I may judge from what we had in crossing the south side,
between 13 and 17 fathoms; nor can the rather high and woody isle, which
lies almost exactly in the middle of the opening, be considered as
presenting any obstacle. This isle, from its local position, would seem
to be the central one of three laid down in the Dutch chart between
Groote Eylandt and the main; but the latitude corresponds with the
southernmost. I call it _Connexion Island_; because my survey round
Groote Eylandt was