1802.
February.
It was here confirmed by satisfactory observations on shore, that our former latitudes and longitudes taken on board the ship were erroneous; and the consequent necessity of reconstructing my chart of these islands, induced me to remain at anchor the rest of the day. A boat was sent to fish with hook and line, and had some success; and at dusk a sufficient number of sooty petrels were taken from the burrows to give nine to every man; making, with those before caught, more than twelve hundred birds. These were inferior to the teal shot at the western Isle of St. Peter, and by most persons would be thought not eatable, on account of their fishy taste; but they made a very acceptable supply to men who had been many months confined to an allowance of salt meat.
The latitude of our anchorage in Petrel Bay proved to be 32° 33⅓′ south, and corrected longitude, by time keepers, 133° 15½′ east. The variation of the compass on the binnacle, with the ship's head south-eastwardly, but the exact point not noted, was 2° 23′ west. Other azimuths, taken five leagues to the north-westward, with the head south-half-west, gave 0° 19′ east; and six leagues to the eastward, the head being north-half-west, we had 0° 16′ east. All these observations, being corrected, and supposing the ship's head in the first case to have been south-east-half-east, as is probable, would agree in showing that the true and magnetic meridians exactly coincided at the Isles of St. Francis in 1802.
Being about to quit this archipelago, it may be expected that I should make some general remarks upon it. The basis stone of the islands where we landed, and that of the others, as also of the projecting parts of the main, appeared to be similar, was either porphyry or granite; but this was generally covered with a stratum, more or less thick, of calcareous rock. The arid sterility of the two largest islands has been already mentioned; and yet they appeared superior to any of the smaller isles, where there was no probability that the small kanguroos could exist in the dry season. The surface of the continent seemed to be almost equally destitute of