the southeast trade hanging far from the southward, and strong westerly currents, have made the Brazil coast, about Cape Roque, or farther to the westward, which greatly prolonged their voyage.

In the summer months, particularly when the sun is in the northern hemisphere, outward bound ships should not run too far to the westward; for in this season, it has sometimes happened, that the northeast winds have continued longer with ships in long. 19° to 23° W., than with others which had separated from them, and lost the trade in 26° and 27° west longitude.

On whatever side of the Cape Verd Islands ships may pass, the most eligible situation at losing the northeast trade, is probably from long. 18° to 23° W.

When the sun is near the northern tropic, the trade often fails ships near, or in sight of these islands; it is certainly best to pass to the westward of them at such times, at 8 or 10 leagues distance at least, to preserve the steady wind, and prevent delay, as light eddy winds prevail near, and amongst them in this season, When to the southward of the Cape Verd Islands, steer to the southeastward, if the wind permit, and endeavour to get into longitude 18° to 23° W., at losing the northeast trade. If then the southerly winds commence, take advantage of the shifts to stand on the tack which gains most southing, and endeavour to cross the equator from 18° to 23° W., if the winds permit; but do not be induced to make a long tack either eastward or westward, with a dead southerly wind, in hopes of meeting a better, unless the wind should veer so far as to gain much southing.

The southeast trade, generally at its northern limit inclines far to the southward, particularly in July, August, and September; and the same has been known in other months. When

a ship meets this trade, she should not be kept too close to the wind, or she will make little progress, but ought to be kept clean full, to enable her to make good way through the water to the southwestward, by which means she will soon get to the southward of the limits of the westerly* current prevailing about the equator, and to 4° or 5° north latitude: it also extends to 3° or 4° south latitude about Fernando Noronha; and from longitude about 27° W. to Cape Roque, it runs very strong, particularly from September to March.

In proceeding to the southward, the wind will draw more to

* In winter, the currents sometimes from the Cape Verd Islands set easterly, and sometimes westerly to 4° or 5o N. latitude, at other times they are variable; but to the southward of 3° or 4° N. latitude, and westward of 20o or 22° West longitude, the equatorial current perpetually runs to the westward.

the southeast, and finally to east and east-northeast at the southern limit of the trade.

WARLEY'S SHOAL, is described by Captain Collins, of that ship, to be a small coral bank (which she passed over, at 7 A. M. May 7th, 1813) about 100 feet long and 50 feet broad, which was too distinctly seen to admit of any mistake; for its edges were clearly delineated, and upon it several ridges of rock appeared, with sand between them. The ship passed too quickly over it to admit of time to sound, as it was accidentally seen by Captain Collins, when looking over the quarter. He thinks there may be full 7 fathoms water over the shoalest part; and a quarter-master, who also saw it, thinks the least water on this shoal may probably be 10 or 12 fathoms.

The fleet at this time consisted of 8 ships, including H. M. S. Salsette, their convoy; and by mean of all the observations and chronometers of those 8 ships, this doubtful rocky bank is situated in lat. 5° 4′ 23′′ N., lon. 21° 25′ 40′′ W. It is matter of regret, that it was not carefully examined, so as to have established its real existence.

It might have been a shoal of Devil-fish, which the Warley passed over, as they are gregarious, and very large near the Equator; and as they swim at great depths, their variegated backs appear exactly like coral rocks.

ST. PAUL'S ISLAND, called also Panedo, and St. Peter's, in lat. 0° 55′ N. lon. 29° 15′ W. by mean of many ships' chronometers and lunar observations, is now correctly determined, as this small island has been seen by ships both outward and homeward bound, although it is considerably to the westward of the common route of the latter, and no ship bound to the southward, should cross the equator so far west as this island.

The Tellicherry passed within 5 miles of it, May 17th, 1802, bound for India. A view was taken, when it bore from N. 30° W. to N. 37° W. distant 5 or 6 miles; by this view, St. Paul's seems to be a heap of rugged rocks, having low gaps between some of them; the northernmost is a small pyramidal rock, not so high as the others. The description annexed to the view in the journal, says, "This island is all rocks, about the height of a ship's mast out of the water.*

*

Mons. de Landeneuf, in the ship Le Curieux, was sent to explore this island in 1768. His account and the Tellicherry's are similar. He found it consisted only of a heap of steep rocks, covered with birds' dung, without verdure, and had no place fit for anchoring, nor convenient for landing.

The variation off St. Paul's was 6° W. in 1802.

It is about 35 feet elevated above the sea, and consists of a group of seve ral rocks adjoining each other, with soundings of 30 to 80 fathoms near them, as found by a commander of the navy, who surveyed and landed on it in 1813.

FERNANDO NORONHA, has not unfrequently been visited or seen by ships bound to India, occasioned by the currents having horsed them to the westward, after the failure of the northeast trade. This island has on it a high rocky peak, called the Pyramid, which is very remarkable, and seems to lean or overhang to the eastward, when it bears S. S. W. The S. W. point is perforated, off which is a sunken rock at a considerable distance, dangerous to approach. From the S. E. part of the island a reef extends to seaward, and some sunken rocks at nearly a league's distance from the shore. There is also said to be a reef, on which the sea always breaks, about 3 miles from the east part of the island, with a channel of 10 to 15 fathoms within it, and that the pyramid is shut in with the highest hill when upon the rocks.

The currents generally run strong to the westward about Fernando Noronha, therefore ships intending to anchor here, should always pass round the north end of the island, which is formed by a chain of several small islets, very near each other, having forts on some of them that command the anchorage.*

This island extends nearly 10 miles about S. W. and N. E. and is about 2 miles broad; the shore is rocky and the surf frequently high; at such times there is no landing. It is not adviseable to touch at this island, except in cases of necessity; for it appears that water is a scarce article in the dry season, and when procurable, cannot always be got off from shore on account of the surf. There is little rain, and they have been sometimes two years without any, then the rivulets were dried up, and vegetation quite parched; at such times, it cannot be supposed a ship would obtain much benefit by stopping at this place.

The General Stuart anchored at Fernando Noronha, September 15th, 1803, (outward bound) in 18 fathoms water, the N. E. end of Wood Isle E. N. E., the S. W. end of Fernando Noronha S. W. by W., the Peak S. by W., Water Bay S. E. off shore about 2 miles. She remained here four days, and could procure only nine casks of water, the well being nearly dry.

November 20th, 1805, the Ann, outward bound, anchored in 17 fathoms shells and rocky bottom, extremity of Fernando Noronha from E. N. E. to S. W. by W., the Peak S. by W., the Church and Round Castle S. by E., the large Fort E. S. E., off shore 2 or 24 miles.

November 22d, 1805, the Tigris anchored in a less 9 fathoms, sand and rocky bottom, Cloven Rock N. E. & N., Fort Island N. E. by E. 4 E., Fort Remedios S. S. E. 4 E., Py. ramid S. W. S., western extreme S. W. by W. W., off shore about half a mile. These ships sailed in company 24th; the

Tigris received 3 bullocks, the Ann received some stock and 12 butts of water, but they found great difficulty in getting the water from the shore, the surf being very high.

There is good anchorage in 13 fathoms, fine white sand, off shore about 1 mile, with Fort St. Antonio E. by S. S., Fort Remedios S. by W., Fort Conception S. S. W. W., Pyramid S. 42° W. The road of Fernando Noronha is unsafe to lie in, with northerly or northwest winds, which are said to prevail from December to April; at other times, they are mostly southeast or easterly, and sometimes at northeast. The well which supplies ships with water is near the governor's house, but landing the casks and getting off the water is inconvenient. The wood is cut on a little island near the north point of the large one, but is not conveniently got into the boats on account of the rocky shore.

Fernando Noronha is peopled with exiles from the coast of Brazil, and is well defended by forts built on the places most eligible for its security. It is hilly uneven land, and seen at 10 leagues distance in clear weather.

The Pyramid is in lat. 3° 55' S., long. 32° 35′ 30′′ W., by mean of many ships' observations and chronometers. The tide rises about 6 feet, and flows to 4 hours on full and change of the moon. There is very little variation at present, 1816.

ROCCAS, is a very dangerous low isle or reef, a little above water. Ships which pass between Fernando Noronha, and the Brazil coast, should be cautious in the night, if not certain of their relative position from Fernando Noronha; for the strong westerly currents are liable to carry them more to leeward than may be apprehended. As this shoal has proved fatal to the East India Company's ship Britannia, and King George transport, which ships were wrecked on it, at 4 A. M., November 2d, 1805, it may be proper to describe and show the true situation of this dangerous reef.

The Earl Elgin saw it in July, 1761, having first seen Fernando Noronha on the 13th, and on the 19th she had soundings on the bank off Cape Roque; at noon 23d, the Roccas bore E. N. to E. & S. distant 4 miles, latitude observed 3° 50′ S. This ship's longitude, by account, placed the Roccas 2o 12' E. from Fernando Noronha, whereas it is about 50 miles west of the island; she had therefore experienced a westerly set of 3° 2' in 10 days. In the Earl Elgin, they call it a low island, or more properly a shoal, that cannot be seen at 3 leagues distance; a sand bank, surrounded by rocks, with high breakers mostly all round, and a projecting point of breakers at the north and south ends of the shoal.

By the Portuguese, the Roccas is said to bear west, a little northerly, distance 15 leagues from Fernando Noronha.

Captain I. Birch, who commanded the Britannia, says, "the Roccas are certainly not laid down right in any of the charts; they are only distant from Fernando Noronha 45 miles; their latitude the same as that island; the rocks most dangerous, are to the northward and northeastward; the whole extent may be about 5 miles; the current set 24 miles per hour to the westward; rise and fall of tide 6 feet."

In the fleet, several ships narrowly escaped the fate of the Britannia and King George, having separated several days before. The Leda frigate, with one division, led past the shoal, and just cleared it, when the Britannia and King George were wrecked. Several ships of the other division, under Sir Home Popham, saw the shoal on the following morning.

The Northampton's journal describes it as a dangerous shoal, very little above water,, with breakers all round, except on the southwest, or lee side, there appeared a white sandy beach, where a boat might land. The Glory's journal describes it as 2 low sand banks, when it bore S. S. E. 2 or 3 miles; and when on the west side of it, at 2 miles distance, she had ground 28 fathoms, coral rock.

By mean of the observations and chronometers of 10 different ships, the Roccas shoal is in lat. 3° 52′ S. lon. 33° 31′ W.*

MARTIN VAS ROCKS, are high and barren, the central one is largest, and may be seen from a large ship's poop at 11 leagues distance; this is a little more easterly than the other two, although they are nearly on the same meridian, as they are all in one, bearing south. The northernmost and central rocks are near each other, but between the latter and the southernmost, there is a channel, through which the Chesterfield passed in March, 1800, and observed the lat. 20° 28′ S. when in midchannel. When through, she hove to, in 12 fathoms, with the largest rock bearing E. N. E. about 1 mile distant, the bottom then visible, and caught plenty of rock-cod and other fish: the boat in sounding, found the depth decrease gradually over a rocky bottom, to 14 fathom close to the largest rock.

The north rock is small, and the most westerly of them; they are all steep and inaccessible, and the distance between the two extremes is about 3 miles.

The breadth of the channel between these rocks and the island Trinidad is about 8 leagues.

In some charts, a shoal is placed about 25 leagues southwestward, and another about 45 leagues W. N. W. from Fernando Noronha; probably these shoals do not exist, as several ships have passed over the places where they are laid down, without seeing any appearance of danger. The Sir Edward Hughes, in June, 1802, passed directly over the position assigned to the southcrnmost shoal; keeping a good look out, but saw no indication of danger.