These matters being arranged, I pressed captain Cumming to depart, fearing that a change of wind might expose the Rolla to danger; but finding him desirous to take off more provisions and stores, I made sail for a bank or rather islet seven miles distant at the eastern extremity of Wreck Reef, for the purpose of collecting seabirds eggs, and if possible taking a turtle.  The Rolla joined on the following day [TUESDAY 11 OCTOBER 1803], and I went on board to take leave of Messrs. Fowler and Flinders and the other officers and gentlemen; at noon we parted company with three cheers, the Rolla steering north-eastward for China, whilst my course was directed for Torres’ Strait.

With the time keeper, Earnshaw’s No. 520, I had received from lieutenant Flinders an account of its error from mean Greenwich time at noon there Oct. 6, and its rate of going during the fourteen preceding days, which were as under.

No. 520, slow 0h 9’ 49.35” and losing 34.13” per day.

The latitude of Wreck-Reef Bank was ascertained from eight meridian observations from the sea, and four from an artificial horizon:  the mean of the latter, which are considered the best, is 22 deg. 11’ 23” S.

Longitude from sixty sets of lunar distances, of which the individual results are given in Table VIII. of the Appendix No.  I. to this volume, 155 deg. 18’ 50.5” E.

The longitude of the bank, as given by Earnshaw’s No. 520 on Aug. 28, eleven days after the shipwreck, was 155 deg. 4’ 14.6” with the Port Jackson rate, or 14’ 35.9” less than the lunar observations.  In laying down the Porpoise’s track on the chart, this error has been corrected by an equal proportion, according to the time of each observation for the longitude.

Mr. Flinders deduced the variation of the compass from observing the sun’s magnetic azimuth a. m. and p. m., when equal altitudes were taken, and comparing the mean azimuth at corresponding altitudes with the true meridian; this method is probably not the best, and the results from two compasses differed considerably; Walker’s compass, marked No. 1, giving 9 deg. 17’ east from ten observations, and that marked No. 2, 13 deg. 54’ from five observations.  The first is undoubtedly the best, though possibly not very correct.

There are here two regular tides daily, and it was high water on the day of full moon at 8h 50’ in the morning; the rise was six feet two inches, but the night tide will probably reach to eight, or perhaps nine feet at the height of the springs.

Some account was given of Wreck-Reef Bank before quitting it in the boat, but I had not then acquired a knowledge of the whole extent of the reef.  It is about twenty miles long, and from a quarter, to one mile and a half in breadth; and consists of many distinct patches of different magnitudes, the six principal of which are from four to eight or ten miles in circuit.  They are separated by channels of one mile to near