chart. The following is Capt. Hudson's (of the United States' Exploring ship. Peacock) account of them :

There are two islands known under this name; the largest is called by the natives Taritari (Touching Island), and the smallest Makin (Pitt Island). The latitude of the southern port of Taritari is 3° 8' N., longitude 172° 48′ E. This island is of the figure of a triangle, with its apex to the South, and its sides are above 14 miles in length. The S.E. is a continuous grove of cocoa-nut and pandanus, with some undergrowth. On the two other sides is a reef, which is awash, excepting the N. W. point, in which there is a small inlet.

Makin is of much smaller dimensions, being but 6 miles long; it varies in width from half a mile to a mile. Its northern point lies in lat. 3° 20′ 43′′ N., lon. 172° 57′ E. The entrance into the lagoon has 4 fathoms of water, and is about one-third of a mile in width. This small island is the seat of government, and the natives now unite both names under the one of Makin. When the Peacock approached, it soon became evident that the island was thickly inhabited; for, on reaching the lee side, above twenty canoes came off, with from five to ten natives in each; but in one of them was a white man, Robert Wood or Grey, who had been left on the island by an English whaler, at his own request, seven years before. He had seen no white man since he landed. He had always been treated kindly; at first they almost deified him. They have no wars, and very few arms, and seldom quarrel, except about their women. They are very different in appearance and character to those of the other islands. Their features are regular, with glossy black hair, flowing in ringlets. They were also of a lighter colour than the rest of the natives with whom they are grouped; their figures are, for the most part, rotund, and they seem to have an abundance of food to become fat upon. In walking they appeared like a moving mass of jelly. The men are very handsomely tattooed; the women are well treated, and never offered for traffic. Wood represented the women as outnumbering the men, and said they were very handsome. There are five towns on the island, which, according to the same authority, contain about 5,000 inhabitants.

There were no quadrupeds found, except rats, which were in great plenty. There were but few birds.*

MARSHALL ARCHIPELAGO.

This extensive collection of islands lies between lat. 4° 45′ N. to 12° 0′ N., and is separated from the Gilbert Islands by a channel 50 leagues broad. The islands were discovered in 1788, by Capts. Marshall and Gilbert, of the English navy, in the ships Scarborough and Charlotte, in their voyage from Port Jackson to China, which voyage originated in the first colonization of Australia. As Capt. Marshall was the principal officer of the expedition, his name has been applied to the whole archipelago by Krusenstern and others; that of Capt. Gilbert has been given to that to the southward. There is considerable confusion in the original accounts, Capt. Gilbert having named some of the islands already

• United States' Exploring Expedition, vol. v. pp. 73–75.

named by Capt. Marshall; and Capt. Bishop, who ten years later discovered some of them, has also named them, which has caused great difficulties in the nomenclature, that are not decreased by the great discrepancies between the respective longitudes given by these officers.

Besides the discoveries of Capt. Marshall, there is another range of islands, occupying the same extent of latitude, at 3 degrees to the westward of them. These islands have been accidentally discovered by various navigators proceeding to India. Capt. Kotzebue, in his first voyage, in 1817, intended to have examined them, but he passed through the range in a dark and stormy night, and was afterwards driven so rapidly to the West, by a strong current, that he could not see them. In his second voyage he discovered three groups belonging to this chain. Capt. Duperrey also determined one group; since which they have been visited by Capt. Hagemeister in 1831; by Capt. Chramtschenko in 1829 and 1832; and by the United States' Exploring Expedition in 1840. From the accounts contained in these voyages we have drawn up the following imperfect description.

The archipelago is separated into two parallel ranges, which Kotzebue tells us are called by the natives RADACK, the easternmost, and RALICK, the western range of islands. They are almost without exception of the usual character of coral islands; a narrow belt of low land or reef, unfathomable on the outer shore, enclosing a lagoon more or less shallow. The productions or capabilities of the group do not appear to be of very great importance, or their amount has not been very well ascertained. The natives on some of the groups are very numerous, and, like other savages, are not to be trusted. Their canoes have the same admirable sailing and seaworthy qualities which distinguish those of the Western Pacific. The description of the archipelago commences from the southward.

MULGRAVE or MILLE ISLANDS.-The second of these names was given by their discoverer, Capt. Marshall, in 1788. They were afterwards examined by Capt. Duperrey and Capt. Chramtschenko. The following recent account, by M. Dutaillis, is the most explicit.

The Mulgrave Islands appear to form a chain of atolls, the extent of which is not yet determined. Together they form a sort of quadrilateral figure, on one of the sides of which (that on the East) is a double belt of islands.

These islands, almost all of which are connected with each other at low water, are sometimes separated at high tides by the sea, which more often covering these reefs does not leave sufficient depth for a boat to be able to clear it for an extent of nearly 40 miles; for after a careful search, only one pass for a large ship and another for boats could be found.

Having been formed successively from coral, marine productions, and the debris of vegetation, they have in the course of ages acquired their great breadth. Here and there are grouped some trees, among which are the native houses. These smaller masses of verdure, intersected by sandy beaches towards the interior, and by coral blocks towards the outside, give to the chain the appearance of small islands, on which cocoa-nut trees, bread-fruit trees, &c., form clumps of beautiful verdure, having a most agreeable and cheerful appearance.

The parts covered with trees, being the largest, generally form the points around which the current runs very strong.

In general, the whole chain is very steep-to on the outside, and only increases on the side of the interior lagoon, where it is less disturbed by the sea, and where the coral banks are in course of formation, some of them already uncovered at low water, and projecting frequently a cable's length off the land. When they reach the level of the water, they become, like the islands already formed, covered in their turn by sand and some vegetables, and will in the lapse of time be of importance.

The space enclosed by these islands is a real sea, navigable for every description of vessels. The bottom, generally at the depth of 22 to 27 fathoms, rises now and then nearer the surface, and shows in white patches, which thus indi. cating the dangers, also point out the points where the anchor may be dropped. These banks bestrewed with rocks are nevertheless dangerous to anchors and moorings, because they will break the one and chafe the other, of both of which M. Dutaillis says he had sad experience.

To take the best position, attention ought to be paid to the rise of the tide; its level has great influence on the quality of the bottom, and it ought not to be chosen if possible but at low water. The passage for large ships is between Barr Island (pronounced as if three r's) and the Island of Tukoeoa, the first to the East and the other to the West of the entrance.

To reach the anchorage with winds from the eastern quarter, the only point for attention is to take up such a position as will allow you to range as near as possible to the pitch of the eastern reef, and rounding the bank attached to Barr Island as near as you please; it is shown by the whitish water; by this means you will avoid a small patch which is to the S.E., and over which the flood tide runs with a velocity of 1 knots. This patch is, until half tide, indicated by very strong ripplings, and becomes nearly awash at low water.

In general it is better to enter or leave with the tide, unless indeed the breeze is fair and of sufficient force to overcome the strength of the current. The less distance you are withinside the anchorage, the less difficulty there will be in getting out with westerly winds.

There is another passage, but for boats only, between Anil Island and Bouguenieu. Bouguenieu is the first islet to the West of Tokoeoa. The distance between these two passes may be about 1 miles. Position of the anchorage, lat. 6° 14' 37" N., lon. 171° 56′ 6′′ E.; variation, 9° 45' N.E.

Capt. Wilkes says they are in lat. 5° 59′ 15′′, lon. 172° 2′ 33′′, but his account will not coincide with that given above.

ARROWSMITH or MEDURO ISLANDS.-This is a discovery of Capts. Marshall and Gilbert, though their account and chart of them are not very definite. Capt. Chramtschenko examined them in detail. Their length, W.N.W. and E.S.E., is 18 miles, their breadth being 11 miles. They are of the usual coral formation, with a lagoon, and are inhabited. According to the U.S. Exploring

* Annales Hydrographiques, vol. i. pp. 154-156.

Expedition, their S.E. point is in lat. 7° 5' N., lon. 171° 23' 54" E., identical with previous observation.

DANIEL ISLAND; PEDDER ISLAND.-To the West of the foregoing the same ships discovered two groups, to which these names were applied, calling the broad and open channel separating them from the Arrowsmith Islands Fordyce Passage. Capt. Chramtschenko did not see them. Capt. Hudson, U.S. Exploring Expedition, passed along the West side of Pedder Island, and through Fordyce Passage. He says that Pedder and Daniel Islands are of coral formation, and are inhabited. They are called Arno on Kotzebue's chart.

AUR, IBBETSON, or TRAVERSEY ISLANDS.-Ou Arrowsmith's chart a group, Ibbetson Islands, is marked, which beyond doubt are the same named by Kotzebue after the Marquis de Traversey, and by the natives Aur (or Aour).

On the sur

This group is 13 miles long N. W. and S.E., and 6 miles broad. rounding reef are thirty-two islands. The population is numerous when compared with the other islands, but scanty for its size; it is between 300 and 400. Stobual Island has a very pleasant aspect. The anchoring place was in lat. 8° 18′ 42", lon. 171° 12′ E. (by chronometer, 171° 8' 14"); variation, 11° 58' 30" E. procured great quantities of cocoa-nuts here.

KAVEN or CALVERT ISLANDS, BASS-REEF-TIED OF ARAKTCHEEFF ISLANDS. This group was discovered on board the ships Scarborough and Charlotte, June 29, 1788, and were named Calvert Islands, probably by Capt. Gilbert. The Nautilus saw some islands, July 5, 1799, which were then named the Bass-Reef-tied Islands, which are identical with the Calvert Islands. Capt. Kotzebue says their native name is Kaven (or Kawen), and he applies the name of Araktcheeff (or Araksheef) to them.

This group is 30 miles long N.W. and S.E., and 11 miles broad. The whole cluster consists of sixty-four islands, and was well surveyed by Kotzebue. The group and its people are precisely similar to those of the Otdia, to the northward. Kaven or Araktcheeff, to the N.W., is the largest island of them. It is 2 miles long and three-quarters of a mile broad. It is in lat. 8° 54′ 21′′ N., lon. 170° 49' E. The S.E. island is in lat. 8° 29′ 30′′, lon. 171° 11' E. High water, full and change, 1" 52'; greatest rise, 4 feet. Kotzebue found very good water in pits on some of the islands, but provisions were not very abundant.+

EGERUP or BISHOP JUNCTION ISLANDS.--This is four of two groups, which together were called the Chatham Islands, from the voyage of the Scarborough and Charlotte, probably by Capt. Gilbert. The other group adjoining, Otdia, appears to have been considered as a portion of it, and Capt. Marshall took them for the Barbadoes Islands, marked upon Anson's chart. They were surveyed by Kotzebue, who states the native name to be Egerup, and also gives the name of Tschitschagoff to them.

Egerup lies to the southward of Otdia, and is considerably smaller. Its length is 24 miles, and its breadth is 4 miles. The whole circle consists of one reef, and contains but very few islands. The South point of it is an island,

Voyage of Governor Phillip; Narrative of the United States' Exploring Expedition, vol. v. p. 107. + Kotzebue's Voyage to the South Seas, vol. i. p. 357.

probably named Egerup, the only one seen with cocoa-nut trees and people; but these, it was said, were limited to one man and two old women. A passage near this perhaps may be navigable, but is dangerous from its numerous turnings; there appears but little inducement for visiting the group.

OTDIA or ROMANZOFF ISLANDS lie to the North of the preceding. Both these names are applied by Kotzebue, but the remarks relating to their early discovery also belong to those of the Egerup group, as above stated.

This group was minutely examined by Kotzebue, in January, 1817, and he has given a large and detailed chart of it. It is of an irregular oval form, 28 miles long in a W.S.W. and E.N.E. direction, by about 10 miles in breadth. It consists of the usual encircling reef, on which are distributed sixty-five islands, of various magnitudes. Otdia is the easternmost and largest, about 2 miles long. The anchoring place inside the island, which was called Christmas Harbour (or Port Noël), was ascertained to be in lat. 9° 28′ 9′′ N., lon. 170° 16' 5" W.; variation, 11° 38′ 30′′ E. As the low land here has no influence on the atmosphere, the barometer falls and rises as uniformly as it generally does between the tropics. The mean of the tidal observations at Otdia gave for the time of high water, full and change, 2 30'. The greatest range was 7 feet. The population on it is scanty, and indeed all the southern islands are uninhabited. The people were very friendly to Kotzebue, who remained here a considerable time. To the South of it is an island called by Kotzebue Egmedio, near to the S.E. angle of the reef; to the South of it is a small high island.

The reef is quite continuous to the N.W. of Otdia, and on it stands a connected chain of small islands, reaching as far as Ormed Island, at the North part of the reef, 8 miles from Otdia. The anchorage in this is in lat. 9° 33′ 16′′, lon. 170° 10′ 58′′ E.; variation, 12° 14′ E.

From Ormed, the North side of the reef, also quite continuous, runs to W.S.W. 8 miles to Bird Island, and 9 miles still farther to Goat Island, from whence to the West end of the reef is 5 miles.

There are several channels through the reef, all on the lee side. The first is 13 miles S.E. of the West point, narrow and impracticable; the next is Rurick Strait, 5 miles farther round the reef, by which Kotzebue entered. From hence the reef continues, without islands, nearly 20 miles, to Schischmareff Strait, in every way preferable to the Rurick Strait, as a ship can beat in or out with the usual trade-wind. To the East of this again is Lagediak Strait, 4 miles from the S.E. point of the group.

LEGIEP or COUNT HEIDEN ISLANDS.-This group, first seen by Capt. Kotzebue, November 5, 1817, perfectly resembles the others, though considerably smaller. Its greatest extent was 19 miles. There is a passage wide enough for a ship on the western side. The centre of the group is in lat. 9° 51′ 30′′ N., lon. 169° 13′ 30′′ E.; variation, 10° 56′ E. The natives of the islands are tall, handsome, robust men, advantageously distinguished from the other Radackers. They live chiefly on fish. In his second voyage Kotzebue

• Kotzebue's Voyage to the South Seas, vol. ii. pp. 12-87.

↑ Kotzebue's First Voyage, vol. ii. p. 222.