deg. 57' and 1 deg. 30' east; and the mean 1 deg.
44'. The variation had therefore _decreased_ considerably since leaving
Kangaroo Island, contrary to the natural order; which proves that the
quick increase on passing Yorke's Peninsula, was owing to some peculiar
attraction, either in that or the neighbouring lands. Whilst beating
through the Back-stairs Passage, I had observed an amplitude when the
ship's head was south-south-west, which gave the extraordinary variation
of 2 deg. 41' east, or reduced to the meridian, 1 deg. 27' east; although we were
then not so much as four miles from the anchorage where it had been found
4 deg. 13' east. Another amplitude was observed at eight leagues to the east
of Cape Willoughby, when the head was north-east-half-east, and gave 2 deg.
5' east variation, or reduced, 4 deg. 36'. This last is correspondent with
what was observed near Kangaroo Head and in the Gulph of St. Vincent; but
the variation of 1 deg. 27' in the passage is totally irregular, and must I
think be ascribed to an attraction either in Cape Jervis to the
north-east, or in the east end of Kangaroo Island to the south-east, or
to both. When the great variation Of 4 deg. 36' was obtained, both these
lands were to the west; and when afterwards the 1 deg. 57' and 1 deg. 30' were
observed, the nearest land was again to the eastward of the ship; and
nearest in the last case.

The winds continued to be light and unfavourable; but by taking advantage
of the changes in direction, and keeping further from the land, out of
the tide or current, we had gained eight leagues by noon of the 11th
[SUNDAY 11 APRIL 1802]. About twenty miles of coast beyond what had been
set as the furthest extreme on the preceding day, was then in sight
(Atlas Plate V.); and our situation and bearings were as follow:

Latitude by corrected log, 36 deg. 11'
Longitude by time keepers. 139 291/2
Northern extreme, from the mast head, N. 10 E.
Nearest part, distant 7 or 8 miles, N. 59 E.
A broad patch of white sand, N. 78 E.
Southern extreme, from the mast head, S. 66 E.

At one o'clock we bore away along the coast with a light breeze from the
north-eastward; and having run five leagues, tacked to seaward soon after
dark. Next morning [MONDAY 12 APRIL 1802] we again followed the coast at
the distance of from five to three miles; and at noon a somewhat
projecting part, which appears to be the _Cape Bernouilli_ o