land of New Guinea is very low, and can hardly be distinguished at the distance

of 4 or 5 leagues.

"For this reason, and for fear of being drifted on this low land, I decided on keeping under weigh all night. From 9h till 11h we were passing not more than half a league from the Arimoa Isles, and we saw that they answered completely to those that Bougainville saw, August 14, 1768. The one in the centre is only an islet, and the two others are not more than 3 or 4 miles in extent. The highest is that to the West.

Although it was nearly calm, the current still continued to bear us to the West. When day appeared, we saw that the coast continued to run to the W.N.W., always very low, and covered with trees. Its aspect also gave rise to the conjecture that the portion we saw was only islands lying before the coast. In the interior, at a great distance, some peaks showed themselves here and there. "Throughout the day (August 15) the breeze was so light and uncertain that we made but little progress, and frequent squalls prevented us from approaching nearer the land as we could have desired. Thus the configuration of the coast is far from being given in a particular manner. If I had had anchors to cast, I should not have hesitated to touch on some point of the coast, to gain some insight into the country and its inhabitants.

"The night was rainy, and at 6 A.м. (16th), the sky clearing, we saw the land, like last night, at 3 or 4 leagues to the S. W., low, woody, and looking more than ever like numerous islands before the coast. Neither mountains nor any eminence whatever could be seen in the interior. It is true that the horizon was not very clear, and the haze might have obscured the distant lands.

"At 8, the sea in our track assuming a greenish tint, I sent a small canoe to sound in the space. No bottom was found at the depth of 45 fathoms, and I

directed the corvette into the centre of this band of discoloured water. At 101 a fresh change of colour, much more distinct, appeared at half a cable's length before us. This time the water was quite yellow and muddy, and above it flew numerous flocks of birds, and fish were seen to leap and sport on the surface of the water. The sea seemed to break lightly along this suspected belt, and the look-out at the mast-head, and M. Guilbert himself, thought that they could see dry spots here and there.

"Although I was quite convinced that this appearance of the sea was only the effect of current, I sent M. Gressen to sound, but he found no bottom with 45 fathoms; at 11 we were across this discoloured water, and traversing several veins of current very rapid and very turbulent. A quantity of branches of trees, of fruits, and pieces of plants, fish, and mollusca, floated on the surface. At 11, in the spot where the tint was most muddy, the water we drew up was but very slightly salt. Throughout this part we had no bottom with 45 fathoms.

"This proves to me that this discoloration was but superficial, that the vessel, in her progress disturbing the water to a certain depth, left her wake of the natural blue colour. The muddy stratum seemed to have at most a depth of 2 or 3 feet. In general the lines of these currents were directed from S.E. to N.W., and the current itself followed the same direction.

"I consider, then, that it is a positive fact that these waters proceed from some considerable river, which discharges itself into the sea on this part of the coast. Bougainville observed the same thing in this part, and from it drew the same deduction. Precisely in this part of New Guinea the land forms a low point, Point D'Urville, very far advanced into the sea, and everything tends to the belief that it is formed by the outlet of a considerable stream."*

POINT D'URVILLE, according to the observations made in the Astrolabe by M. Jacquinot, is in lat. 1° 24′ S., lon. 137° 47′ E. The land, or, what is more likely, the islands forming this point, are certainly the same that Bougainville indicates to the N. W. of Mount Moulineaux. This was not seen from the Astro- . labe, but doubtless the haze obscured it.

GEELWINK BAY is the very extensive bay of which Point D'Urville may be said to be the easternmost point. We shall not describe it here, but shall terminate our notice of New Guinea with Port Doreï or Dory, at the N.W. point of the bay. Off its mouth are some very large islands, Jobie Island, Mysory or Schouten Island, and others. The last named is very imperfectly known.

From Point D'Urville the coast of New Guinea turns to the S.W., preserving the same aspect, that is to say, very low, and always lined with large trees, very close to each other, which give it the appearance of an immense wall. Here and there some tufts of trees, twice as high as their neighbours, resemble rounded bastions, intended to flank this gigantic wall. Some natives came off in two canoes, with double outriggers.

Point Geelwink terminates this portion of the coast, and here, perhaps, the immense bay of the same name may be said to commence. West of this is the extensive Island of Jobie, and between its West end and Point Geelwink is a smaller one, which nearly closes the passage, and was named by D'Urville, after one of his officers, Quoy Island.

QUOY ISLAND is 8 miles long, and moderately elevated, well wooded, and pleasing in appearance. A channel of only 3 miles in breadth separates it from Point Geelwink, formed by a moderately high hill. A channel of the same breadth separates it from the West end of Jobie, which descends in a very gentle slope to the sea.

The TRAITOR'S ISLANDS lie to the North of Jobie, and off the East point of Mysory, forming the North side of the Strait of Jobie. They are small and low.

JOBIE ISLAND, which lies in the opening of Geelwink Bay, according to M. D'Urville's chart, is about 90 miles in length, lying nearly East and West. The Astrolabe only sailed along its northern side. The land on this coast is high, steep, and covered with woods, without any openings. High mountains. form the central ridge. This aspect is preserved in the most uniform manner, and throughout its whole extent it did not appear to offer a single cove or a single creek fit to receive a large ship. But few smokes were seen, and no traces of cultivation, signs of a scanty population. The mountains of Jobie decrease in altitude on nearing its western extreme, and its West cape is separated from Bultig by a channel of 6 miles in breadth.

* Voyage de L'Astrolabe, tome iii. pp. 565–568.

BULTIG OF HUMP ISLAND is hilly, of an irregular form, and 10 or 12 miles in length, but not more than 4 miles broad. Three rounded islets, called the Three Sisters, lie near its East point, and before its West point are two similar islets, named by D'Urville the Brothers.*

LONG ISLAND was only explored on its North side. The island is tolerably large, and the land in general is but little elevated. Numerous clumps of cocoanut trees grow near the shore; under one of them a group of natives was seen assembled, but they did not attempt to come off.t

"It is fortunate," says M. D'Urville, " that the squalls which are felt off the North coast of New Guinea, though very violent, are of very short duration;" otherwise his navigation would have been perilous, and his experience proved, that during that season (August, 1827), at least, the bad weather was of short duration. I

PORT DOREÏ or DORY.-In describing this, the best-known portion of New Guinea, we shall follow the account given in the Voyage of the Astrolabe, by M. D'Urville, who, having visited it at two separate intervals, is entitled to the greatest consideration.

Notwithstanding that Saavedra, Gaëtan, Schouten, Tasman, and Dampier, have by turns explored some portion or other of the North coast of New Guinea, the relation of their voyages have left us but a vague notice of the inhabitants of this great island. The English commander Forrest, who visited the harbour of Doreï in February, 1775, was the first who gave to Europe more exact notices of the customs of the Papuas, and the productions of the country. After him, Duperrey visited the same point in the month of August, 1824, and passed fifteen days at the anchorage. The officers of La Coquille made an exact plan of the port and its environs, and the naturalists of the expedition collected a mass of materials for science in the various natural kingdoms, and their observations may be seen in the account of the voyage of M. Duperrey.

The Harbour of Dorei is situated immediately to the South of Cape Mamori, which forms the westernmost and outer point of the entrance of the great Bay of Geelwink.

The harbour is entered by a channel 3 miles in length, formed on one side by the Peninsula of Mamori, and on the other by the Islands of Mana-Souari, and Masmapi, and two banks, which are awash. The harbour itself is not more than half a mile deep, and 200 yards in breadth, with a regular depth of 10 fathoms, sand and shells. Notwithstanding the confined extent of this basin, ships of any class may calculate on a safe anchorage, and sheltered from the winds and swell from the offing. But as it is surrounded by deep forests, and at the bottom of the harbour there are many mud-banks, often dry, a long stay here would doubtless be unhealthy for Europeans, especially in the rainy season.

All the environs of the harbour, properly so called, are occupied by forests in a state of nature, standing on a coralline soil, which rises with a very gentle slope. But the beds of the torrents are bestrewed with numerous boulders of a granitic

Voyage de L'Astrolabe, tome iii. p. 573.

+ Ibid. tome iii. p. 574.

Ibid. tome iii. p. 572.

ISLANDS BETWEEN THE EQUATOR AND LAT. 10° N. 1045

nature, probably brought down from a higher level. The whole of the mountains of Arfak appear to be of primitive formation, according to the geologists.

The cultivated lands do not commence but at the villages, and extend all along the bank of the North channel. The land is of so rich nature that it is sufficient to stir it and pull up the weeds to obtain most abundant crops. But the Papuas are as indolent as they are wanting in intelligence in the art of cultivation, and the vegetables for food are often choked by parasitic plants. The plantations of the arum alone seemed to be a little better attended to.

The inhabitants of Doreï are distributed in four villages on the borders of the sea; two are on the North side of the harbour, and the two others on the Islands Mana-Souari and Masmapi. Each village contains from eight to fifteen houses, built on piles; but each house contains a range of distinct cells or apartments, and holds several families. Some of these houses contain a double range of these cells, separated by a corridor, which extends from one end to the other. These edifices, entirely constructed of wood, rudely shaped, are open to the weather in every part, and frequently shake under the feet of the visitor. The total population of Doreï cannot exceed 1,500 souls.

All the inhabitants of Doreï recognise the sovereignty of the sultan of Tidore; and, notwithstanding the distance, a vessel is despatched every year to carry to the sultan the tribute and homage of his Doreïan subjects. These tributes consist of slaves of both sexes, tortoise-shell, birds of paradise, wax, &c.

"The result of the observations of M. Jacquinot has placed the observatory at Doreï in lat. 0° 51′ 43′′ S., lon. 133° 59′ 52" E. Our longitude differs 6' from that of M. Duperrey, and this difference partly arises from that officer having adopted a greater longitude for the westernmost of the Mispalu Islands than we have done, and also that he made a greater difference between the longitude of Mispalu and that of Doreï than we did. More lengthened and rigorous observations would reconcile these slight differences."*

CHAPTER XXXI.

ISLANDS BETWEEN THE EQUATOR AND LAT. 10° N.

FOLLOWING the plan adopted in the preceding pages, we commence from the coast of America, and include in this chapter the group of islands, the Gilbert Archipelago, which lie on the Equator, and the Marshall Archipelago to the North of it. The Caroline Archipelago, lying within the same belt of latitude, will form the ensuing chapter.

Voyage de L'Astrolabe, tome iii. p. 600.

MALPELO ISLAND, in lat. 4° 0' N., lon. 81° 32′ W., is a barren, high, perpendicular rock, which may be seen in clear weather at the distance of 20 leagues, the summit being 1,200 feet above the sea level. A small quantity of green moss, and a few dwarf bushes, which grow in its cracks and gullies, afford the only verdure that it possesses. It is surrounded with islets, and the whole may extend about 9 or 10 miles from North to South. The centre of this island bears a resemblance, in several points of view, to the crown of a head, and its being barren accounts naturally enough for the name (bald head) which the Spaniards have bestowed on it. It is surrounded, as it were, by a strong current, having much the appearance of breakers, which setting into the gulf, and being accompanied by light winds, with thick and hazy weather, Colnett did not think it deserving of any further attention. The current was found to set N.E. by E. 23 miles an hour.* Another statement† is made that they run violently to the southward and westward near it, a difference possibly owing to the different seasons they have been observed in. Colnett's was in July, 1793. The rock itself has 40 fathoms alongside of it, and 110 fathoms at a quarter of a mile distant.

RIVADENEYRA SHOAL.-" Being on board the steamer Peru, abreast of Puná, October 22, 1842, and hearing that there was a terrible yellow fever raging at Guayaquil, the steamer put back, and I was placed on board a small schooner going to Realejo. On the 28th, in the middle of the day, the sea calm, we had caught a large turtle, when I observed at a few fathoms off a slight swell on the sea; we took the boat and went to it, when we sounded, and to our astonishment found only 16 feet (French?) of water; in the centre of this circular spot was only 10 feet depth; we then found 14, 16, 27, 56 feet, and then no bottom. By our very imperfect instruments we made it to be in lat. 4° 15′ N., lon. (87° 30′) 85° 10′ W. of Greenwich; but this we considered nearly correct, as we hastened on to Realejo."

COCOS ISLAND.-The discovery of this island is involved in obscurity. It is mentioned as being well known by early navigators, Lionel Wafer, Dampier, &c. Its more exact position and character appear to have been first ascertained by the Spanish exploring ships, the Descubierta and Atrevida, in 1791. It was then visited, and, it is stated, surveyed, by Capt. Colnett in 1793. In 1795 it was visited by Vancouver, who also examined it. There are some singular discrepancies in the accounts given by these different visitors, more particularly in those of the two last named. Vancouver states it to be 4 miles in length N.E. and S.W., while Colnett states it to be 12 miles; and the respective plans given also coincide with the descriptions. These discrepancies were decided in Vancouver's favour by Sir Edward Belcher in 1838. He places the observatory at the head of Chatham Bay, at the N.E. part of the island, in lat. 5° 32′ 57′′ N., lon. 86° 58′ 22′′ W.; var. 8° 24′ E.

Nautical Magazine, September, 1837, p. 611.

+ Colnett, p. 66.

+ Substance of an extract of a letter from M. Rivadeneyra to the Societé de Geographie de Paris, Bulletin, &c., tome ix. 1848, pp. 125–127, 323.