[* At the Cape of Good Hope, in 1810, His Excellency the Earl of Caledon favoured me with the following extract from the log book of the sloop Columbus—­Long, master; returning to the Cape from the coast of Brazil.

“September 22, 1809, at five p.m., saw the island of Saxonburg, bearing E. S. E., first about 41 leagues distant:  clear weather.  Steered for the said island, and found it to be in the latitude of 30 deg. 18’ south, longitude 28 deg. 20’ west, or thereabout.

“The island of Saxonburg is about four leagues in length, N. W. and S. E, and about 21/2 miles in breadth.  The N. W. end is a high bluff of about 70 feet, perpendicular form, and runs along to the south-east about 8 miles.  You will see trees at about a mile and a half distance, and a sandy beach.”

The situation of Saxemberg in the common tables and charts, was 30 deg. 45’ south and 19 deg. 40’ west, almost 9 deg. of longitude too little; and therefore it is not surprising that ships have missed it.  At the time so many birds were seen, on the 28th, the Investigator was not more than eighty miles from the position of the island, as above given from Mr. Long.]

SUNDAY 4 OCTOBER 1801

The fresh western winds continued, with short intervals of calm, as far as the latitude 33 deg. 23’, and longitude 13 deg. 0’ west; when they died away, and a breeze sprung up from the eastward.  With this wind we could do little more than look up for the isles of Tristan d’Acunha, whose bearing was then S. 16 deg.  E., and distance seventy-seven leagues.  From the description given by sir Erasmus Gower* of the anchorage, and the convenience with which water may be obtained, and his account of the animals which resort there, I should not have considered it to be lost time, had the wind made it advisable to put in at Tristan d’Acunha, for a few days; but it veered round to the north-west, on the [TUESDAY] 6th. and we resumed our former course to the Cape of Good Hope.

[* Lord Macartney’s Embassy to China, by sir G. Staunton, Vol.  I. p. 198-201.]

WEDNESDAY 14 OCTOBER 1801

In the morning of the 14th, the variation by mean of amplitude and azimuth, was 25 deg. 10’ west; the ship’s head being E. by S., and our latitude 35 deg. 4’ south, and longitude 12 deg. 50’ east.  It is worthy of being mentioned, that in the year 1797, and near the same place, I observed the variation to be 19 deg. 40’ west, on board His Majesty’s ship Reliance; and as the compass was upon the binnacle in both cases, the sole cause to which I can attribute this great difference is, that the ship’s head was west, instead of E. by S. The true variation could not be far from the mean of the two observations, since it was 26 deg. at the Cape of Good Hope.  In the English Channel, the compass on the binnacle had shown nearly 4 deg. too much west variation, when the ship’s head was at west; but here, it gives at least 2 deg. too much, with the head in an opposite direction!  This difference in the two hemispheres merits particular notice; it is part of a series of apparent anomalies in the compass which have hitherto remained unaccounted for; but which seem reducible to one general cause, as I have attempted to show in the Appendix No.  II. to the second volume.