21° 59′, lon. 121° 38′ N. (centre). It is high, 3 or 4 miles in extent, and seen from S.S.W. or N.N.E., appears like a saddle, visible 16 or 17 leagues from the masthead.

The North Bashee Islands form the South limit of this channel. The North island is in lat. 21° 9′ N., lon. 122° 0′ E. There is another smaller one to the South of it like it, not to be seen very far off, and to the North of the larger high island, in lat. 21° 3′ 30′′ N. This generally makes in the form of a peak, and may be discerned 13 leagues off. Ships passing through the Bashee Channel should keep well over toward these islands, to avoid the Vela Rete and Gadd's Reef. Typhoons happen in both monsoons, and the weather is generally very unsettled hereabout.

It is of coral,

Farther to the West is the dangerous Prata Shoal and Island. level with the water in many parts, but slightly covered in others. The island in the centre is a low coral patch, of considerable extent, covered with coarse grass and shrubs. It is in lat. 20° 42′ 55′′, lon. 116° 45′ E.; N.E. part of the shoal, lat. 20° 47′ N., lon. 116° 42′ 15′′ E.; West part, lat. 20° 43′ N., lon. 116° 41′ 45′′ E. The currents are very strong and irregular, and the greatest caution is necessary to avoid them. H.M. screw-ship Reynard was totally wrecked here in 1851, in consequence of their unknown influence.

There does not appear to be any dangers in the open ocean to the East of Canton beside this.

The ENTRANCE to the CANTON RIVER is marked by the Grand Ladrone Island, in lat. 21° 57′ 10′′ N., lon. 113° 44'. It is a steep and bold island, the N.W. part forming a round mount or dome, higher than the other part, and visible 9 leagues from a ship's deck. Although most of the islands here are high, this is unlike them, but being the outermost and different in appearance, it is made the standard position for vessels going to or leaving Canton.

The Little Ladrone Island is just to the West of it; the channel between is too narrow for use. The western side of the channel is the island and flats of Samchow. On the East side of the channel is Potoe or Passage Island, in lat. 22° 2′ 6′′, bearing N.W. by N. from the N.W. end of Little Ladrone Island, 4} miles off. It is a flat, sloping rock, but should not be approached too closely.

In approaching Canton River during the S.W. monsoon, ships endeavour to make the Grand Ladrone, bearing about N. or N. by E., and then run up into the river by the western channel; but late in the season, when the winds incline easterly, or in coming from the North or eastward, as would be the case from the northern portions of the Pacific, it is prudent to make the Great Lema Island, and so in by the Lema Channel.

The Lema Islands consist of three large and one small island, the largest 6 miles long by 1 miles broad, of moderate height and undulating appearance. The N.E. end is in lat. 22° 4' 45", lon. 114° 18' 30" E. The Lema Channel enters between this and the Poo-toy Islands on the North. It is about 2 leagues wide, and very safe. This channel should be always adopted in the N.E. monsoon.

From the North end of the Great Lema, the course is about West to pass to the northward of Lingting Island toward the Lantao Passage, above Macao. Lingting is of considerable height, terminating in a conical peak. It may be passed on

either side, but to the northward is preferable. Having passed 1 or 2 miles distant, steer West for the Lantao Passage, between the A-chow and Chichow Islands, and then round the S.W. end of Tyho or Lantao Island, to the northward for Lingting, or the westward for Macao, as occasion may require.

MACAO, as is well known, lies at the entrance of the Canton River, on the West side, in lat. 22° 10' 30", lon. 113° 32′ E. Capt. P. Heywood's excellent plan, published by Mr. Laurie, will afford quite sufficient guide to run through the Typa into the harbour. Here pilots may be had for Canton; but it may be stated that the river appears the safest to navigate in the world.

CANTON FACTORIES are on the North side of the river, in lat. 23° 7′ 10′′ N., lon. 113° 15' 0" E.

HONG KONG, the British settlement, lies to the North of the Great Lema Island.

Its greatest length is 9 miles, its breadth 5 miles. It is very uneven, rising into innumerable ridges and peaks; the highest, Victoria Peak, is 1,825 feet high. Victoria is the capital of the colony, and is on the North side. To the East of it there is a hillock, near the house of the Morison Education Society, which was carefully ascertained by Sir Edward Belcher to be in lat. 22° 16′ 30′′ N., lon. 114° 8' 30". The approach to it from the S.E. of the Fathon Channel is open, and inside the anchorage is perfectly safe.

With this we will close this chapter, with a feeling of the imperfection of the remarks contained in it. Our acquaintance with the most ready means of making a quick passage is certainly very far below present requirements. The Pacific navigation is now assuming a new character, and its traffic will, ere long, assume a regularity approaching to that of the western oceans; but until a more detailed experience is acquired of this, as of other features of the great South Seas, it is needless to enter into crude speculations, founded as they must be on imperfect data.

At a future and early day it is sincerely hoped that this, and many other defects which these volumes must necessarily contain, may be amply remedied.

APPENDIX.

I. PROPOSED COMMUNICATIONS BETWEEN THE ATLANTIC AND PACIFIC OCEANS.

THIS question, which has attracted particular attention of late, is not a new one. It originated very soon after the Spaniards had ascertained that a narrow isthmus only separated the two oceans; and innumerable plans have been proposed, but the desired object, that of a navigable canal, seems, as yet, to be visionary.

The great impetus given to commerce, and the consequent addition to the importance of this subject, arising from the Californian movement, have led to at least a portion of the proposed communications being effected, in the railway now constructing to Panamá from near Chagres; but the great desideratum is still that of a canal, by the means of which ships may pass from either ocean without encountering the dangerous and circuitous route around Cape Horn.

The whole subject has been fully treated by Capt. R. FitzRoy before the Royal Geographical Society; and his paper appears in vol. xx., 1851, page 161, of their Journal. We therefore extract the following summary of what has been written or done in connexion with it from that paper.

The principal object in view is a navigable channel between the two oceans, through which the largest ships may pass, without breaking bulk or being lightened; the least object contemplated is a common waggon road, and between these limits other methods of transit will be noticed, which may be classed under four heads-namely, ship canals, boat canals, railroads, and waggon roads.

Men, horses, and mules have hitherto carried all that has been transported from sea to sea, where barges (bongos) or canoes have not floated. Much traffic through continued woodland, the rapid growth of vegetation, a soft, tenacious soil, and an extraordinary amount of rain, have hitherto destroyed and almost effaced such attempts at road-making as were effected in former days by the exertions of the early Spaniards, who employed slaves, and also availed themselves of Indian labour unscrupulously. In these regions roads must be made with large logs of wood (corduroy), or paved with stone, or else macadamized with an unusually thick layer of “metal," until an embankment (or “battery ") fit for a railway can be constructed.

Excessive wetness, rather than humidity of climate, may be considered as the principal impediment to constructing roads, bridges, and the solid works of canals. This wetness causes a quick growth of vegetation, and a very rapid decay of all ligneous substances.

Miasmatic exhalations, or malaria, are engendered in low situations, especially near the confluence of fresh and salt water. Intermittent fevers, ague, and, at times, the pestiferous yellow fever, are common in such situations, but proper treatment cures the majority of cases, and those persons who have been thus "seasoned" do not again suffer from the same malady.

Another serious impediment to the permanence of solid works is, the liability of the greater

part of these countries to the destructive effects of volcanic convulsions; but Panamá and the narrow isthmus of Darien are free from their effects, which are felt on either side of them. The southern part of Central America is not so subject to violent tempests as the northern, but they are sometimes experienced.

A few diminished tribes of Indians still inhabit the mountainous ranges of some districts. The Indians of Darien visit the adjacent islands and reefs. They are the descendants of the Symerotes," who were never conquered by the Spaniards. This tribe still opposes, as of old, the exploration of certain tracts. There is yet another hindrance to the undertaking of any great works on the isthmus, and that is the great instability of the local government. Irrespective of all these considerations, there is one chief requisite, one main point, to be insisted on, in connexion with any route-the indispensable adjunct of a good port. It would not be practicable to sail into any canal direct from the open sea, nor could the entrance of a canal be kept open without some protection from the heavy surf and sand or shingle.

Four principal lines for a canal have been hitherto recommended, to which may now be added three more, which ought not to be undervalued, and some others of less importance. Taking the four principal lines in order, from North to South, the first is that between the Gulf of Mexico and Tehuantepec, and which may be called the Mexican line. The second is through the great Lake of Nicaragua, and is well known as the Nicaragua line. The third route crosses the isthmus, near Panamá, and is of course termed the Panamá line. The fourth, from the Gulf of Darien, by the Atrato River, to Cupica (Tupica ?) Bay, is known as the Atrato and Cupica line.

Of the three next in importance, one is from Chiriqui Lake to Dulce Gulf; another, from the Gulf of San Blas, or Mandinga, to Chepo, in the Bay of Panamá; and a third, from Caledonia Bay, in the Gulf of Darien, to the Gulf of San Miguel.

The Mexican line has been alluded to on pages 259-60, part i. It leads over the Isthmus of Huasacoalcos, or Coatzacoalcos, from the mouth of a river so called, in lat. 18°, to the Gulf of Tehuantepec, near lat. 16°. It is about 120 miles across, in a straight line from sea to sea, and is nowhere very much elevated: about 700 feet above the sea is, however, the lowest summit level. The rivers are unimportant for navigation, and the adjacent coasts are frequently troubled by furious storms, and there is no port. Nevertheless, so much has been thought of this locality, that two surveys have been made of it; one, in 1825, by D. J. de Orbegoso and Don T. Ortiz, and another, in 1842-43, by Don J. Garay and Senhor G. Moro; and at the present time the legislature of New Orleans have considered the practicability of carrying out these plans.

The arguments in favour of this are the level nature of the intervening country; but its clevation would require a great number of locks. A railroad may soon become locally valuable; but for the present a good waggon road, from river to river, appears to be all that the insuperable disadvantages of the position will admit of. Other arguments are more favourable to this route: its climate is less unhealthy; supplies of stone and mortar are procurable; and native labour is to be had. Besides this, the proposed line is entirely within one territory. But this line can never interest the world at large. The Gulf of Mexico is not in the line of general intercourse between the two great oceans. This last circumstance will prevent it becoming more than of local value for the traffic between New Orleans and California.

The Nicaragua line has long attracted much attention; and public opinion in the United States, as well as in Europe, is at present rather in favour of this route. There is no geographical objection to the line; but most of the proposed terminations of this route are unfit for large vessels, except during fine weather, and they are exposed to gales of wind. Six branch lines from the great Lake of Nicaragua have been advocated, which will all be noticed.

Nicaragua Lake is so large, being nearly 90 miles long, and about 30 miles in width, that

artificial harbours, at the mouths of any canals opening into it, will be necessary. In the comparatively still water of the lake they might be constructed, however impracticable in the open sea, where no shelter exists for such operations. The depth of the lake seems to be irregular, from 2 to 40 fathoms and upwards.*

In the River San Juan there are rapids, and in the River Panaloya (or Tipitapa), which leads from Leon (or Managua) Lake, there is a sudden fall of 13 feet, caused by a dyke of recent lava. In both rivers there are places almost dry across during the summer; and in both the winter rains cause great floods, overflowing their banks, and inundating the low lands. It is evident, therefore, that these rivers could only serve as the principal feeders of canals, and that they themselves cannot be made navigable for large vessels.

It is said that the surface of Nicaragua is about 125 feet above the mean level of the Pacific Ocean, and that the surface of Managua Lake is about 28 feet above that of Nicaragua. From the Atlantic, at the mouth of the River San Juan, it is more than 80 miles to the lake, in nearly the mean direction of the river; and from Fonseca or Conchagua Gulf to the north-west part of Nicaragua the very shortest line that could be taken for a canal would measure more than 90 miles. It is more likely, however, that each of those distances would exceed 100 miles, including all bends; and, as the ground is very irregular, some idea may be formed, off hand, of the vast number of locks that would be required, of the delay, labour, and expense of towing ships through, and of the time which would be occupied. Between Realejo, which is by no means so good a port as had been usually supposed,† and Managua Lake, about 45 miles of canal would be required; and the highest level necessary would be about 212 feet above the Pacific; but at that elevation it is doubtful whether a sufficient supply of water could be then secured without an immense cutting, 11 miles in length, and not less than 70 feet deep (wide enough also for a canal), to convey a stream from Managua. From one ocean to the other not much less than 100 miles of canal would be indispensable, if the shortest lines were adopted; but probably nearer double that extent would be required if the longer route, to Conchagua, were followed.

The third proposed branch canal, from Lake Managua to Tamarinda, on the Pacific, has not been examined. The fourth proposition is from Nicaragua Lake to San Juan del Sur, on the Pacific; but many difficulties appear in the way—a range of high land intervening, through which it would be necessary to cut a tunnel, 1 to 2 miles long, 120 feet above the lake, and of dimensions hitherto unattempted. How a sufficient supply of water could be obtained is a matter of mere conjecture. Besides these grave objections, the district abounds in volcanoes, and is subject to earthquakes.

The next branch line from Nicaragua is from the mouth of the Sapoa River, at the S.W. part of the great lake, to Salinas Bay (see page 230, part i.). It is said to be about 15 miles in length, half of which is by the river, and now navigable by boats. No exact survey has yet been made, but estimated levels and distances have been made by Dr. Oersted, in 1847. The lowest summit level is supposed to be only 130 feet above the lake, and about double that elevation above the sea. There is sufficient water, it is stated, to supply a canal at that elevation. Salinas Bay is surveyed, and known to be a good harbour.

The sixth proposition is to connect Nicoya Gulf (see page 227, part i.) with the southern part of Nicaragua, or directly with the River San Juan, by the help of the River San Carlos. In this case an available port exists in the Pacific, but a large extent of land intervenes between the oceans, which has not been surveyed, and may be found unavailable.

The respective peculiarities of these six branches, as they may be called, of the Nicaragua line, cannot yet be described minutely, because insufficiently explored, excepting that of the branch of San Juan del Sur. Enough, however, is known to discourage any attempt to construct either canal or railway, unless the Sapoa track should prove as eligible as

* See page 233, part i., where the lake and the proposed canals are described. It is described in pages 235-240, part i.

See page 230, part i.