At half-past one we passed within three miles of the point which had been the furthest extreme at noon; it is low and sandy, and a ledge of rocks extends from it to the north-east. I named it Point Malcolm, in honour of Captain Pultney Malcolm of the navy. The depth diminished from 20 to 10 fathoms, in passing near a sunken rock two miles to the south-east of the point, and upon which the sea breaks only at times. The coast from thence trended rapidly to the northward; and in following its direction at from three to five miles distance, we left eight islands of the Eastern Group on the starbord and two on the larbord hand. These, with the exception of the southernmost, which has a hill at each end and some vegetation, are little better than low sterile rocks.

At seven in the evening, the water being smotth, we anchored in 8 fathoms, sandy bottom, three or four miles from the shore; where our calculated situation and the bearings of the land were as follows:

From Cape Pasley to the northern extreme the coast is sandy and low, presenting, with trifling exceptions, a continued beach. On the north side of Point Malcolm it stretches north, and then eastward, forming a bight five miles within the land; after which the general trending is north-north-east, with very little sinuosity. Four or five miles behind the shore, and running parallel with it, is a bank of moderately high and level land, over which the tops of some barren-looking mountains were occasionally seen. The most remarkable of these is Mount Ragged, lying N. 8° W. nine or ten leagues from Cape Pasley.

[SOUTH COAST. BETWEEN THE ARCHIPELAGOS.]

We had now altogether lost sight of the Archipelago of the Recherche. The chart which I have constructed of this extensive mass of dangers is much more full, and in many parts should be more accurate than that of D'Entrecasteaux; but I dare by no means assert that the very great number of islands, rocks, and reefs therein contained are the whole that exist; nor that every individual one is correctly placed, although the greatest care was taken to obtain correctness. All the islands seem to be more or less frequented by seals; but I think not in numbers sufficient to make a speculation from Europe advisable on their account; certainly not for the China market, the seals being mostly of the hair kind, and the fur of such others as were seen was red and coarse. There is, besides, a risk of being caught in the archipelago with strong south or western winds, in which case destruction would be almost inevitable, for I know of no place where a ship might take refuge in a gale. The shelter in Thistle's Cove is, indeed, complete, when a vessel is once placed; but the cove is too small to be entered except under favourable circumstances, and the shelter in the western corner could not be attained with winds blowing strong out of it. The archipelago should not, therefore, be entered without the assurance of carrying fine weather to the proposed anchorage.

During the night of the 17th there was no current or set of tide past the ship. Every thing was kept prepared for getting under way at a moment's notice; but the wind blew gently off the land, and the people of the watch occupied themselves successfully in catching dog-fish. At daybreak [MONDAY 18 JANUARY 1802] we made all sail to the north-eastward, along the same low and, if possible, more sandy coast. The wind was light, and at nine it fell calm. This was succeeded by a sea breeze at east-south-east, and we trimmed close to it, keeping on our former course until four in the afternoon; when the land being one mile and a half distant, we tacked in 12 fathoms, and stretched to the southward.

The shore curved round here, and took a more eastern direction; and the bank of level land, which continued to run along behind it, approached very near to the water side. Three leagues further on it formed cliffs upon the coast; and a projecting part of them, which I called Point Culver, bore N. 77° E. four leagues: this was the furthest land in sight.

This afternoon we passed a number of pale red medusas, such as I had usually seen on the East Coast at the entrances of rivers, and which, on being touched, produced a sensation like the stinging of a nettle. There was also a red scum on the water, and some of it was taken up to be examined by Mr. Brown in a microscope. It consisted of minute particles not more than half a line in length, and each appeared to be composed of several cohering fibres which were jointed; the joints being of an uniform thickness, and nearly as broad as long. These fibres were generally of unequal length, and the extremities of the compound particle thence appeared somewhat torn. The particles exhibited no motion when in salt water; and the sole effect produced by immersing them in spirit of wine was the separation of each into its component fibres.

Until daybreak next morning the wind was unfavourable; but it then veered round to the south, and enabled us to pass Point Culver. Our situation at noon, and the bearings taken were these:

Our course along the shore was so favoured by the wind that at seven in the evening we had passed another projecting part of the cliffs, named POINT DOVER, distant from Point Culver fifty miles; and the extreme in sight ahead was twenty miles further, and still cliffy. The nearest part was two or three leagues distant; and the wind being still at south, we hauled up to it, and at nine o'clock stood back to the westward.

The elevation of these cliffs appeared to be about five hundred feet, and nothing of the back country was seen above them. In the upper part they are brown, in the lower part nearly white, and the two strata, as also the small layers of which each is composed, are nearly horizontal. They were judged to be calcareous, as was the white, grey, and brown sand which the lead brought up when the bottom was not of coral.

A surveyor finds almost no object here whose bearing can be set a second time. Each small projection presents the appearance of a steep cape as it opens out in sailing along; but before the ship arrives abreast of it, it is lost in the general uniformity of the coast, and the latitude, longitude, and distance of the nearest cliffs are all the documents that remain for the construction of a chart. Point Culver and Point Dover are exceptions to the general uniformity; but it requires a ship to be near the land before even these are distinguishable. The latter point was somewhat whiter than the cliffs on each side, which probably arose from the front having lately fallen off into the water.

TUESDAY 19 JANUARY 1802

In the night of the 19th the wind shifted round to the eastward, and continued there for three days; and during this time we beat to windward without making much progress. Several observations were taken here for the variation of the compass: with the ship's head east-by-north, azimuths gave 7° 15' west, and at south, 4° 26'; five leagues further eastward they gave 6° 13' with the head north-east, and eight leagues further, an amplitude 4° 18' at south-by-east. These being corrected would be 4° 13', 4° 26', 4° 2', and 3° 42' west; so that the variation had now reassumed a tolerably regular course of diminution. The mean of the whole is 4° 6' west variation in the longitude of 125° 51' east.

FRIDAY 22 JANUARY 1802

At the end of three days beating our latitude in the evening of the 22nd was 32° 22', and longitude 126° 23', the depth in that situation was 7 fathoms at two miles from the land, and the furthest extremes visible through the haze bore west-half-north and east, the latter being distant four or five miles. The bank which before formed the cliffs had retired to a little distance from the coast, and left a front screed of low, sandy shore. Several smokes arose from behind the bank, and were the first seen after quitting the archipelago.

The barometer had kept up nearly to 30 inches during the east and south-east winds, but it now fell to 29,65; and we stretched off for the night in the expectation of a change of wind, and probably of blowing weather. At ten the sails were taken aback by a breeze from the westward; but at daylight [SATURDAY 23 JANUARY 1802] it had veered to south-by-west, and the mercury was rising. We then bore away for the land; and having reached in with the low, sandy point which had borne east in the evening, steered along the coast at three or four miles distance in from 7 to 11 fathoms water. The latitude at noon from very indifferent observations was 32° 22½', and longitude 127° 2'; the coast, four miles distant to the northward, was low and sandy, but rose quickly to the level bank, upon which there were some shrubs and small trees. Nothing of the interior country could be seen above the bank; but this might possibly have been owing to the haze, which was so thick that no extremes of the land could be defined. The wind was fresh at south-south-west, and by seven in the evening our longitude was augmented 55'; the land was then distant six or seven miles, trending east-north-eastward; and we hauled to the wind, which had increased in strength though the barometer was fast rising.

Having stood to the south-east till midnight, we then tacked to the westward; and at five next morning [SUNDAY 24 JANUARY 1802] bore away north for the land, the wind being then at south-by-east, and the barometer announcing by its elevation a return of foul winds. At six we steered eastward, along the same kind of shore as seen on the preceding day; but the wind coming more unfavourable, and depth diminishing to 5 fathoms soon after eight o'clock, made it necessary to stretch off to sea. The coast in latitude 32° 1' and longitude 128° 12' was three miles distant to the north. A league further on it took a more northern direction, but without much changing its aspect; it continued to be the same sandy beach, with a bank behind it of level land topped with small trees and shrubs as before described.

MONDAY 25 JANUARY 1802

The rest of the day and the whole of the 25th were taken up in beating fruitlessly against an eastern wind. Azimuths observed when the ship's head was east-by-north gave variation 6° 4'; and ten miles to the south a little eastward they gave 3° 8' west, at south-by-east; corrected 3° 2' and 2° 32', and the mean 2° 47' for the true variation, showing a decrease since the last of 1° 19' for 2° 11' of longitude.

At ten in the evening our situation was less advanced than on the morning of the 24th, when we tacked off shore; but the mercury was again descending, and during the night the wind veered to north-east, to north, and at eight in the morning [TUESDAY 26 JANUARY 1802] to west-by-north, when we steered in for the land. At ten the shore was eight or nine miles distant, and our course was north-east, nearly as it trended. The latitude at noon, from observations to the north and south, was 31° 51' 34", and longitude by timekeepers 128° 41'; the beach was distant three or four miles in the north-north-west, and the bank behind it lay two or three miles inland and was somewhat higher, but had less wood upon it than further westward. The wind was fresh at south-west, and the mercury was rising; but the haziness of the weather was such that no extremes of the land could be set.

Our course from noon was nearly east at the distance of five or six miles from the shore; and we ran at the rate of between seven and eight knots, under double-reefed top-sails and foresail. Abreast of our situation at half-past two the level bank again closed in upon the shore, and formed cliffs very similar to those along which we had before run thirty leagues. Their elevation appeared to be from four to six hundred feet, the upper part was brown, and the lower two-thirds white; but as we advanced, the upper brown stratum was observed to augment in proportional quantity. We could not distinguish, as before, the smaller layers in the two strata; and from the number of excavations in the white part, apparently from pieces having fallen down (see Mr. Westall's sketch, Atlas, Plate XVII. View 6.), I was led to think the lower portion of these cliffs to be grit stone rather than calcareous rock. The bank was not covered with shrubs, as before it came to the water side, but was nearly destitute of vegetation, and almost as level as the horizon of the sea.

At dusk we hauled up south-east-by-south to the wind, at one in the morning [WEDNESDAY 27 JANUARY 1802] tacked to the westward, and at four bore away north for the land. Having reached within six miles of the cliffs, we steered eastward again, with a fair breeze; and at noon were in latitude 31° 40' 52' and longitude 130° 59'; the cliffs were then distant seven miles to the northward, and at N. 9° E. was their termination.

The length of these cliffs, from their second commencement, is thirty-three leagues; and that of the level bank, from near Cape Pasley where it was first seen from the sea, is no less than one hundred and forty-five leagues. The height of this extraordinary bank is nearly the same throughout, being no where less, by estimation, than four hundred, nor any where more than six hundred feet. In the first twenty leagues the ragged tops of some inland mountains were visible over it; but during the remainder of its long course the bank was the limit of our view.

This equality of elevation for so great an extent, and the evidently calcareous nature of the bank, at least in the upper two hundred feet, would bespeak it to have been the exterior line of a vast coral reef, which is always more elevated than the interior parts, and commonly level with high-water mark. From the gradual subsiding of the sea, or perhaps by a sudden convulsion of nature, this bank may have attained its present height above the surface; and however extraordinary such a change may appear, yet, when it is recollected that branches of coral still exist upon Bald Head, at the elevation of four hundred or more feet, this supposition assumes a great degree of probability; and it would further seem that the subsiding of the waters has not been at a period very remote, since these frail branches have yet neither been all beaten down nor mouldered away by the wind and weather.

If this supposition be well founded, it may, with the fact of no hill or other object having been perceived above the bank in the greater part of its course, assist in forming some conjecture of what may be within it; which cannot, as I judge in such case, be other than flat, sandy plains, or water. The bank may even be a narrow barrier between an interior and the exterior sea, and much do I regret the not having formed an idea of this probability at the time; for notwithstanding the great difficulty and risk, I should certainly have attempted a landing upon some part of the coast to ascertain a fact of so much importance.

At the termination of the bank and of the second range of cliffs the coast became sandy, and trended north-eastward about three leagues; after which it turned south-east-by-east, and formed the head of the Great Australian Bight, whose latitude I make to be 31° 29' south, and longitude 131° 10' east. In the chart of admiral D'Entrecasteaux the head of the Great Bight is placed in 31° 36' and 131° 27'; but I think there is an error at least in the latitude, for the admiral says, "At daybreak I steered to get in with the land; and the wind having returned to south-east, we hauled our starbord tacks on board, being then four or five leagues from the coast. At eleven o'clock the land was seen ahead and we veered ship in 32 fathoms, fine sand."* The latitude observed at noon, as appears by the route table, was 31° 38' 58"; and if we suppose the ship, lying up south-south-west, to have made 2' of southing in the hour, as marked in the chart, she must have been in 31° 37' at eleven o'clock; which is within one mile of the latitude assigned to the head of the bight, where the shore curves to the south-east-by-east. This does not accord with the land being only then seen ahead, since the weather appears to have admitted the sight of it at the distance of four or five leagues. If we suppose the admiral, when he veered, to have been eight, instead of one mile from the head of the Great Bight, and the account strongly favours the supposition, it will then agree with my latitude. I had only 27 fathoms in crossing the head, and although it is possible there may be 30 closer in, yet in such a place as this the probability is, that the ship having the greatest depth of water was the furthest from the land.

[* Voyage de D'Entrecasteaux, par M. de Rossel, Tome I. page 220. The 32 fathoms are, I believe, of five French feet each, making very nearly 30 fathoms English measure.]

After steering east-north-east, east, and east-south-east, and having seen the beach all round the head of the Great Bight, we hauled up parallel to the new direction of the coast, at the distance of six miles; and at five o'clock were abreast of the furthest part seen by the French admiral when he quitted the examination. The coast is a sandy beach in front; but the land rises gradually from thence, and at three or four miles back is of moderate elevation, but still sandy and barren. According to the chart of Nuyts, an extensive reef lay a little beyond this part. (Atlas, Plate IV.) It was not seen by D'Entrecasteaux, but we were anxiously looking out for it when, at six o'clock, breakers were seen from the mast head bearing S. 43° E. some distance open from the land. We kept on our course for them, with the wind at south-south-west, until eight o'clock, and then tacked to the westward in 27 fathoms; and the ship's way being stopped by a head swell, we did not veer towards the land until three in the morning, at which time it fell calm.

THURSDAY 28 JANUARY 1802

On a light breeze springing up from the northward we steered in for the coast; and at noon were in the following situation:

The breakers lie five or six miles from the land, and did not appear to have any connection with it, nor with two other sets of small reefs which came in sight to the east and east-south-east, soon afterward. At two o'clock our situation was betwixt these last reefs. The southernmost patches are two or three miles in length, and there are large rocks upon them, standing above water; the northern patches extend eight miles along the coast, from which they are distant three miles, and on the eastern parts there are also some rocks above water, but there were none upon the western reef first seen. It may be doubted whether the western reef were known to Nuyts, but there can be no doubt concerning these last; and I call the whole NUYTS' REEFS.

The aspect of the shore to the northward was nearly the same as that seen the preceding afternoon, but behind the second reefs it began to assume a more rocky appearance. A high cliffy cape is formed a little further eastward; it has a pyramidal rock near it, and the coast there takes a direction somewhat on the north side of east. This remarkable projection, being within a few leagues of the furthest part of the main coast discovered by the Dutch, I have called CAPE NUYTS: its latitude is 32° 2' south, and longitude 132° 18' east.

After clearing Nuyts' Reefs we steered east-north-east, past the cape, to look for anchorage in two bights, but there were rocks in both, and they were open to the southward. Beyond them was a low, cliffy point, lying E. 3° N. seven or eight miles from Cape Nuyts; and seeing a bay behind it which promised shelter from south-west and south winds, we hauled round the point at half-past five. The water shoaled gradually from 11 to 3 fathoms, on which I hove the sails aback and sent the master ahead to sound; and as he did not make the signal for deeper water, as we were already in tolerable shelter, the anchor was dropped in 3¼ fathoms, sandy bottom. We had then the following bearings:

Between the first and the last of these bearings we were exposed to the sea, but sheltered at all other points of the compass.

Being arrived at the extremity of that part of the south coast of Terra Australis which had been previously explored, it may be useful, before entering on the unknown part, to compare my examination of it with what was contained in former charts. It will thence appear that the employment of fifteen days in running along the coast, more than would probably have been required had I kept at a distance, was not without some advantage to geography and navigation.

(Atlas, Plate II.)

At Cape Leeuwin, the largest Ile St. Alouarn of D'Entrecasteaux was seen to be joined to the main, and to form the south-western extremity of Leeuwin's Land, and of Terra Australis. The coast from thence to King George's Sound was more accurately investigated than the French admiral had an opportunity of doing and his omission of soundings supplied. Captain Vancouver's chart is superior to that of the French from Cape Chatham to the Sound; but that officer's distance from some parts prevented him from seeing them correctly. In the Sound, no particular advantage will be derived from the new survey, the plan given by Vancouver being sufficiently correct for nautical purposes, with the exception of the bar to Oyster Harbour, over which he had marked seventeen feet, but where thirteen now appeared to be the greatest depth. From King George's Sound to Point Hood the coast had been very indistinctly, and sometimes not at all seen by Vancouver; but I found it, speaking generally, to be laid down by D'Entrecasteaux with accuracy, though the bights in the land are marked somewhat too deep, from his distance not allowing the low beaches to be always distinguished. These trifling inaccuracies were remedied, the passages between Bald and Doubtful Islands and the main land opposite to them ascertained to be safe, and the omission of soundings along the coast remedied.

In Doubtful Island Bay the French chart does not give the north-western part sufficiently deep; but the coast from thence to the Archipelago of the Recherche, as also the reefs and rocks, were well distinguished, better perhaps than by me; but the usual want of soundings, with the exception of some distant ones by Vancouver, still continued. D'Entrecasteaux's chart appeared to be excellent in the western part of the archipelago, and good in the positions of the islands on the outskirts; so that I have, in some cases, borrowed from it. With respect to the inner islands and the main coast, it was necessarily defective, from the French ships having sailed round the archipelago, and not through the middle of it as I did in the Investigator. Here, my survey, though far from complete in the details, will afford much new information and useful also, since it has brought to light a well-sheltered cove affording wood and water, and two other tolerable anchorages at which some refreshments may be procured, and at one, quantities of salt in the summer season.

(Atlas, Plate III.)

From the archipelago eastward the examination of the coast was prosecuted by D'Entrecasteaux with much care, and with some trifling exceptions very closely; but as far as the 127th degree of longitude from Greenwich no soundings were given. These have been supplied, and a more minute description given of the coast. At the 129th degree the French ships seem to have been closer in with the land than was the Investigator; and it would appear by the track that they were also closer at the 30th, and at the head of the Great Bight, but these last are not corroborated by the soundings. From thence to the bay in which we anchored on the 28th, the Dutch chart of 1627 was the sole authority; and making allowances for the state of navigation at that time, it is as correct in form as could reasonably have been expected.

The latitudes and longitudes of the points and islands along the coast have been either verified or corrected, for there are commonly some differences between any longitudes and those of Vancouver and D'Entrecasteaux. The observations by which certain places, taken as fixed points, are settled in longitude, are mentioned at those places, as also are the corrections applied to the time-keepers for laying down the intermediate parts; and both are more particularly specified in the Appendix to this volume.

Monsieur Beautemps Beaupré, geographical engineer on board La Recherche, was the constructor of the French charts; and they must be allowed to do him great credit. Perhaps no chart of a coast so little known as this was will bear a comparison with its original better than those of M. Beaupré. That the Plates II and III in the accompanying Atlas, are offered as being more full and somewhat more correct, does neither arise from a wish to depreciate those of my predecessor in the investigation, nor from an assumption of superior merit; there is, indeed, very little due to any superiority they may be found to possess; but there would be room for reproach if, after having followed with an outline of his chart in my hand, improvements should not have been made in all or some of those parts where circumstances had not before admitted a close examination.

Longitude of the point, deduced from twenty-two sets of distances (see Table III of the Appendix to this volume) is 132° 30'; but that given by time keepers with accelerated rates and supplemental correction, as explained at the end of Chap. VI, and in the Appendix, is preferred, and is 132° 27' east.

The variation observed upon the binnacle, with the ship's head east-south-east, was 3° 11' west by the surveying compass; and in the offing, with the head north-north-east, it was 1° 41' west. These, corrected, will be 0° 19' and 0° 30'; and therefore the variation allowed upon the bearings on shore was 0° 25' west.

The wind was at south-east-by-south at one in the afternoon, when the anchor was weighed to beat out of the bay. At half past five we were three miles from a cliffy head which had been taken for an island at the anchorage, and set at N 77° E. The shore forms a small bight on the east side of this head, and then stretches south-south-eastward in a sandy beach, with a ridge of barren land behind. At sunset we passed to windward of Point Fowler, and stood off to sea for the night.

[SOUTH COAST. NUYTS' ARCHIPELAGO.]

SATURDAY 30 JANUARY 1802

Cape Nuyts bore north, two or three leagues, soon after daylight, and the wind was then at east; but as the day advanced it veered to the south-east, and permitted us to make a stretch toward the furthest land. At five in the evening we tacked near some low, whitish cliffs, which had been seen from the mast head when in Fowler's Bay; they were two or three miles off, and the furthest land visible from the deck bore S. 63° E. at no great distance. The coast here is broken into sandy beaches and small, cliffy points, and the same ridge of barren land runs behind it, but the elevation is not great.

SUNDAY 31 JANUARY 1802

At three in the afternoon of the 31st we reached in again with the coast, about four leagues beyond our situation on the preceding day. The depth at two miles off shore was 7 fathoms on a coral bottom; the northern extreme bore N. 58° W., and a low point on the other side, named Point Bell, S. 45° E., seven miles. To seaward, a flat rock bore W. 3° S., one mile and a half; it is the largest of four which were called Sinclair's Rocks, and lie scattered at the distance of two or three miles off the coast. We stood off at this time; but so little could be gained upon the south-east winds that when we came in next morning [MONDAY 1 FEBRUARY 1802] it was almost exactly in the same spot, and Point Bell was not passed until late in the afternoon; the weather, also, was adverse to the examination, being so hazy that the highest land could not be seen beyond three or four leagues.

At half-past six in the evening, when we tacked to stand off for the night, Point Bell bore N. 68° W. four miles. It lies in 32° 16½' south and 133° 5' east; and there is a broad, flat rock, surrounded with breakers, one mile to the westward. The main coast beyond the point forms some bights, and is divided betwixt sand and rock, as before described: its general trending is nearly east. A small island, somewhat elevated, lies six miles to the south-east of Point Bell, and has a ledge of rocks and islets extending from it a league to the north-eastward, and a separate islet one or two miles to the east: these obtained the name of Purdie's Isles. After we had tacked in 9 fathoms, a wave was perceived to break upon a sunken rock within less than half a mile of the ship; and I think it would be dangerous to pass between Point Bell and Purdie's Isles.

TUESDAY 2 FEBRUARY 1802

At noon of the 2nd February no land was in sight. The weather was still hazy, and the wind at south-east; but in the afternoon it favoured us two points, and we got sight of a higher and larger island than any before seen on this part of the coast. At half-past four, being then near a smaller isle and several rocks, we tacked towards the large island which was six or seven miles to the southward; and soon after eight in the evening got to an anchor in a little sandy bay on its north side. The depth was 6 fathoms in passing the north-west point of the bay, but 10 within side, on a fine sandy bottom, where the anchor was dropped. At daylight [WEDNESDAY 3 FEBRUARY 1802] we found ourselves hall a mile from the shore, and the extremes bearing from N. 32° W., round by the west and south, to S. 77° E.; and at the distance of two miles we were sheltered by four small islands, extending from N. 41° to 88° E. The master was sent to sound in the bay; but the bottom was everywhere good, and nothing found to injure the cables. The scientific gentlemen landed upon their respective pursuits; and I followed them to take angles for my survey, and see what could be procured for the ship's company.

The island is nearly three miles long, north-west and south-east, and is moderately high and cliffy at the ends; the middle part is a sandy isthmus, not more than half a mile broad, but the breadth of the higher ends is from one-and-half to two miles. This island is the central one of a group; for besides the four small isles to the north-east, there are two close to the west end, and two others, something larger, lying off to the southward. I call these the ISLES OF ST. FRANCIS; in the persuasion that the central one is that named St. Francis by Nuyts. Independently of the eight isles and a rock, surrounding this Isle St. Francis, I set from the north-east point three other islands. The first, named Lacy's Isle, bore N. 28° E., seven miles; and two miles from it to the north-west there is an islet and a separate rock above water surrounded with breakers, the same near which we had tacked at half-past four on the preceding evening. The second was called Evans' Isle, and bore N. 49° E. eleven miles, and the third to which the name of Franklin was given, bore N. 81° E. sixteen miles. All these are much inferior in magnitude to the central island of St. Francis.

For several days before anchoring here we had observed large flocks of sooty petrels; and I found the surface of the island, where it was sandy and produced small shrubs, to be full of their burrows. Penguins, similar to those of Furneaux's Islands, had their burrows nearer to the water-side. A small species of kangaroo, was also found, and at some preceding season the island had been frequented by geese; but at this time, the vegetation being almost burnt up, they seemed to have quitted it from want of food. The heat was, indeed, such as to make walking a great fatigue; and this was augmented by frequently sinking into the bird holes and falling upon the sand. The thermometer stood at 98° in the shade, whilst it was at 78° on board the ship.

Where the surface is not of sand it consists of calcareous rock, mostly in loose pieces; but the stone which forms the basis of the island is heavy and of a close grain, and was judged to be porphyry. In the crevices of a low calcareous cliff, at the south-east side of the bay, I found some thin cakes of good salt, incrusted upon a stone containing laminae of quartz.

A party was sent on shore at dusk to collect petrels, and in less than two hours returned with sufficient to give four birds to every man in the ship. Early in the morning [THURSDAY 4 FEBRUARY 1802] the boats were again sent upon the same errand, and to haul the seine; but the birds were gone off to sea for the day, and no fish were caught. A small kangaroo was brought off, as also a yellow snake, which was the second killed on this island. The great heat deterred the naturalists from going on shore this morning, for the very little variety in the vegetable productions presented no inducement to a repetition of their fatigue. I landed to see what further could be discovered of the neighbouring islands; and we then prepared to get under way so soon as the breeze set in from the south-eastward, which it usually did about noon, after a few hours of calm or of light airs.

The small bay in the Isle St. Francis, which I call Petrel Bay, affords excellent shelter for two or three ships; but no fresh water, not even to rinse our mouths, could be found at this time; and a few scattered bushes were the nearest approach to wood upon the island. Petrels, penguins, and a few hair seals may be procured, and probably some geese in the wet season.

I had hitherto observed upon this coast that the south-east and east winds produced the same effect upon the barometer as at the Cape of Good Hope, in keeping the mercury high, commonly at or above 30 inches and the more fresh was the wind, the higher it stood; but within the last few days the barometer was much lower with the same winds, and at this time was at 29.74. The dense haze which prevailed might possibly have caused the change, but I suspected another reason for it. Winds coming off the land, I had remarked, had a tendency to depress the mercury, and sea winds to make it rise, though no change took place in the weather; and it therefore seemed probable, as the trending of the coast beyond these islands was unknown, that the south-east and east winds came off the land, and not from the sea, as before; in which case the unknown coast would be found trending to the southward, a conjecture which, it will be seen, was verified. That there was no entrance to a strait, nor any large inlet near these islands, was almost demonstrated by the insignificance of the tides; for neither in Fowler's Bay nor at this Isle St. Francis could any set be perceived; nor was there any rise by the shore worthy of notice.

At half-past one we left Petrel Bay; and having passed between the small isles to the north-east, steered for Evans' Island, and toward the Isles of St. Peter, which were expected to lie beyond it. At five o'clock, we passed between Evans' Island and some rocks above water, with breakers round them, lying three miles to the eastward. An island, equally high with that of St. Francis, was then seen to the north, and low land extended from it to N. 45° E., which had some appearance of being part of the main. We steered for these lands; and seeing an opening between them at sunset, I attempted it in the hope of getting anchorage for the night; but the water shoaled suddenly, from 4 fathoms to sixteen feet upon rocks, and obliged me to veer on the instant. We then stood back to the southward till eight o'clock, and nothing being perceived in the way of the ship's drift, hove to for the night.

FRIDAY 5 FEBRUARY 1802

The wind was north-east in the morning; and at half-past four o'clock we filled the sails and steered eastward until eight, when the central island of St. Francis bore N. 71° W., and Franklin's Isles, for there are two, besides rocks, were distant four leagues, the small opening between them bearing N. 28° W. To the south-eastward of these islands, at the distance of eleven miles, is a low projection of the main land, to which the name of Point Brown was given, in compliment to the naturalist; and four leagues further, in the same line, was a cliffy head, called Cape Bauer after the painter of natural history. Between these projections there was a wide space where no land was visible, and for which we accordingly steered on the wind veering more to the northward. The atmosphere was still hazy, more especially about the horizon, and no observations worthy of confidence could be taken for either latitude or longitude. At noon,

No land was yet visible ahead; and there being much refuse from the shore, as well as seaweed floating about, some hopes of finding a river were entertained. At half-past two, however, low, sandy land was seen from the mast head, nearly all round, the depth had diminished from 19 to 7 fathoms, and the water was much discoloured in streaks at less than a mile from the ship. Smokes Were rising in three different places; but as the wind was unfavourable, and there was no prospect of any opening sufficiently large to admit the Investigator, I gave up the further examination of this place, and called it STREAKY BAY.

There remained nearly forty miles of space between Point Bell and Point Brown, in which the main coast had not been seen. This it was necessary to explore; but the wind being then at north-north-east, I steered to the southward, to gain some further knowledge of the coast in that direction before dark.

West of Cape Bauer, and distant four miles, there is a low island, extensively surrounded with rocks and breakers, which I called Olive's Island. We passed between it and the cape, and observed the cliffs of the latter to be stratified, and apparently calcareous. Another cliffy and somewhat higher projection opened from it at S. 1° W., distant seven miles, the intermediate low land forming a bight four or five miles deep, which is mostly skirted by a sandy beach. This projection I named Point Westall, in compliment to the landscape painter; and at six in the evening, when it bore north-east-by-east two or three miles, we veered round to the northward. Beyond Point Westall the coast takes a more eastern direction, the first land which opened out from it being at S. 43° E.: this was a third cliffy projection, terminating another sandy bight in the coast. No hill nor anything behind the shore could be perceived, but it does not certainly follow that there are no hills in the back country, for the haze was too thick to admit of the sight extending beyond four or five leagues.

The wind having veered to east-north-east, we kept to the northward all night, under easy sail; and at daylight [SATURDAY 6 FERBRUARY 1802], the lands around us were in the following bearings:

All sail was made to fetch between Franklin's Isles and Point Brown, in order to follow the course of the main land as close as possible; but finding, after several tacks, the impossibility of weathering the isles, we bore away; and at noon hauled up north-north-east round them. The wind was light at east, and the weather fine over head; but there was so dense a haze below that, the true horizon could not be distinguished from several false ones, and we had six or seven different latitudes from as many observers: those taken by me to the north and south differed 19 minutes. This dense haze, from its great refractive power, altered the appearance of objects in a surprising manner: a sandy beach seemed to be a chalky cliff, and the lowest islands to have steep shores. The thermometer stood, at this time, at 82° and the barometer at 29.60 inches.

On the north side of Point Brown the shore formed a large open bay, into which we hauled up as much as the wind would permit, passing near to a reef of rocks and breakers, two miles to the north-north-east of Franklin's Isles. At half-past two the water had shoaled to 5 fathoms; and not being able to distinguish any inlet, we then bore away westward along the land. The number of smokes rising from the shores of this wide, open place induced me to give it the name of SMOKY BAY.

At four o'clock we passed the small opening which had been unsuccessfully attempted in the evening of the 4th, and hauled up northward under the lee of the island forming its western side. The mainland then came in sight ahead; but between it and the islands was a space five or six miles wide, which had the appearance of being the entrance to a river. No land was visible to the north-east; and besides quantities of grass and branches of trees or bushes floating in the water, there was a number of long, gauze-winged insects topping about the surface, such as frequent fresh-water lakes and swamps. In order to form a judgment of how much fresh was mixed with the salt water, or whether any, I had some taken up for the purpose of ascertaining its specific gravity; but before the experiment could be made, the depth diminished to 3 fathoms, and low land was distinguished nearly all round. We then veered ship; and at seven o'clock came to an anchor in 6 fathoms, off a small beach on the north side of the western and smallest island, being sheltered at all points except between S. 58° and N. 80° W.

The specific gravity of the water taken up proved to be 1.034, or .008 greater than the water of the Southern Indian Ocean, westward of the Island Amsterdam, although the temperature in which it was weighed was higher by 14°. This circumstance, with the shallowness of the inlet and the land having been seen to close round so nearly, made me give up the intention of attempting to proceed any higher up, since no river of importance was to be expected.

Great flocks of sooty petrels were observed coming in from sea to the island, and at the first dawn next morning [SUNDAY 7 FEBRUARY 1802] a boat was sent to collect a quantity of them, and to kill seals; but the birds were already moving off, and no more than four seals, of the hair kind, were procured. The botanists preferred going on shore to the more eastern land, which, though low, was much more extensive than the island nearer to the ship; and in fact it was not yet ascertained whether it were not a part of the main. I went to the higher island with a theodolite to take bearings; and as the survey had shown that no dependence was to be placed in any observations taken on board the ship during the last five days, I took with me the necessary instruments for determining the latitude and longitude.

Granite was found to compose the rocks of the shore, and seemed to be the basis of the island; but it was covered with a crust of calcareous stone, in some places fifty feet thick. The soil at the top was little better than sand, but was overspread with shrubs, mostly of one kind, a whitish velvety plant--(artriplex reniformis of Brown),* nearly similar to what is called at Port Jackson, Botany Bay greens. Amongst these the petrels had everywhere undermined; and from the excessive heat of the sun, the reflection from the sand, and frequently stepping up to the mid-leg in the burrows, my strength was scarcely equal to reaching the highest hill near the middle of the island. I had no thermometer, but judged the temperature could scarcely be less than 120°; and there was not a breath of air stirring. My fatigue was, however, rewarded by an extensive set of bearings, and I overlooked the lower and larger island to the eastward, and saw the water behind it communicating with Smoky Bay. That low land and the island upon which I stood, being the north-easternmost of this archipelago, must, I conceive, be the ISLES OF ST. PETER in Nuyts' chart, notwithstanding their relatively small distance from those of St. Francis. The bay to the northward, between these islands and the mainland, I named DENIAL BAY, as well in allusion to St. Peter as to the deceptive hope we had formed of penetrating by it some distance into the interior country. The bearings most essential to the survey, taken from this station were these,