swell its volume, though only one of any consequence empties into it from the north; and it is worthy of remark that whilst there is a regular decline of the plain from Chacabuco to the Chonos Archipelago, the principal tributaries of all the large rivers are from the southward. In its general extent, the Itata is regarded by those who know it as a wider and deeper river than the Maule, with less current than the latter, though interrupted by frequent ledges of rocks. Mr. Miers says that its banks are rocky and precipitous, preventing the use of its waters for irrigation—a natural obstacle of some consequence, as in this latitude there are times when agriculture would derive much benefit from such contribution. A bar across its mouth excludes vessels of all classes, and there is no shelter near, where they could safely embark the abundant products of the fertile country on its banks. Early writers speak of all these rivers as navigable; and it may be, that the gradual rising of the land and recession of the ocean, caused by continued earthquakes-facts now indisputably established-have reduced the depth of water to what it is at the present epoch; but unless we suppose a diminution in the volume from the Andes, an explanation seems more properly to be sought in the sand and shingle brought down by the rivers and deposited where the currents and swell of the ocean are in equilibrium.

In a statistical account of the province of Maule, made to the supreme government in 1845, the commissioners say: "The river Maule, which is the largest of all that water this province, has its origin among the mountains Descabezado and Campanario, situated at the centre of the cordilleras of the Andes. Thence it flows in a direction from east to west, until it discharges itself in the ocean, traversing consequently the widest part of the republic here, which is about forty leagues." They make no allusion to a lake there, nor is one of any extent mentioned in a detailed account of every visible object from the table-land at the base of the Descabezado given in the "Annales de la Universidad de Chile," by Professor Domeyko. He says: "At 3, P. M., we reached the upper plateau (meseta) of the Descabezado, which, at a spot where the bases of the greater and lesser Descabezados rest, is covered with perpetual snows, and from whence a small lake is seen at a little distance, as smooth and quiet as though it was in the most retired and sheltered valley of the world." This can scarcely be the ten square leagues of water here located by M. Gay, and it is to be regretted that his map of this vicinity is less satisfactory than those of even more southern districts.

The Descabezado, one of the summits in the fourth range counted from the plain, is still some miles to the southward of the dividing line of waters, from which, to the junction of the Loncomilla at the eastern base of the Western cordilleras, the Maule has few tributaries, and flows in a serpentine line with a resultant direction west by south. At the same time, here, as well as at many other points of the Andes, the hills separating the waters from those that fall to the Atlantic are invariably less elevated than the line which would connect the great cones or peaks. Deriving its supply in summer wholly from melting snows, the stream is deeper and more rapid during the earliest warm days, when the sun's heat is first powerful in the lower and sheltered ravines. After January the volume gradually diminishes; and even where it crosses the great plain, although the bed sometimes occupied is more than a mile wide, the rapidity of its current alone renders the Maule a river of note. There it is composed of several streams spread over the extent mentioned, with islands of shingle and sand occupying at least three fourths of the space between them. Its main volume does not exceed forty yards in width, with an average depth of two feet; and probably three fourths of all the water at this season passes through this channel. According to a MS. map in the archives at Santiago, its frequent distribution into several streams continues almost to the base of the Andes.

The Loncomilla, after gathering nearly all the water from the Andes and plain north of the Nuble and the Western cordilleras, falls into the Maule six leagues south of Talca. Its three The first of these principal sources are the Perquilaoquen, the Longaví, and the Achihueno. has its origin within the Andes, near Cerro Florido, receiving many mountain brooks in its descent to the plain, and after a most circuitous course of twenty-six leagues unites with the

Longaví, the two being subsequently called Loncomilla. The Longaví and Achihueno also rise within the Andes, one to the north and the other to the south of the Cerro Nevado. Though neither of so great volume nor so long as the Perquilaoquen, they are much more rapid. Only the lower half of the Loncomilla was visited by me. Generally, that has steep rocky or alluvial banks, separated from fifty to one hundred yards according to the locality, and which, in the same manner, vary in height from fifteen to forty feet. Its length is only twenty miles, through ten of which it is navigable even in the dry season by launches carrying forty tons. For these ten miles it runs through and parallel with the mountains composing the Western cordilleras, its water clear, and current scarcely exceeding two miles per hour. As might be inferred from its sluggishness, there is very little shingle or sand, and its bed partakes much of the rocky character of its banks.

After draining the plain to the north and northeast, through nearly seventy miles, the Claro empties into the Maule twelve miles below the Loncomilla. This, also, is a stream of clear water from forty to fifty yards wide in the wet season and early summer, when it is navigable as high up as Talca; but which between January and June, at a league from its mouth, is only a brook, never exceeding twenty yards wide nor two feet deep.

Between the junction of the Claro and Loncomilla, the Maule may be said to enter the Western cordilleras; the first-named stream having washed the base of these mountains during the last forty miles of its course. From thence to the ocean, a distance of near eighty miles by the windings of the stream, the mountains on both sides are broken ranges of hills, never less. than 200 nor more than 1,000 feet high. In some places they slope to the water; in others, terminate in short and narrow terraces or plateaus twenty feet above its level; and, except where occasional strata of rocks crop out, the whole is covered with forest-trees and plants in great luxuriance and variety. Until quite near to its mouth, the river apparently increases very little in width, its mean breadth from the Claro to within three leagues of the ocean not exceeding one hundred yards. Below the Claro, islands and deposites of shingle become much less frequent. The river bed, however, is by no means a uniformly inclined plane, but in places. is obstructed by strata of rocks or accumulations of rolled stones, over which the rapidity of the current is much increased. If we consider the elevation of Talca (derived from six barometrical observations on four different days) to be the same as that of the mouth of the Loncomilla, (and it cannot differ very greatly,) the average fall of the Maule will be nearly eight feet per mile.* As more detailed information is given in Chapter XV, it need only be added here, that the river is navigable, at all seasons, by vessels of 300 tons to Constitucion, a secure port within its mouth, and by boats drawing eighteen inches water as high as half way up the Loncomilla.

This is the most northern of the rivers of Chile which is useful, except for irrigation or drainage; and the others will be briefly mentioned. Having their sources at greater elevations, and crossing portions of the plain more above the sea-level as we advance, whilst the distance from their sources to the ocean is not increased proportionately, they partake more and more of the nature of torrents loaded with detritus, whose places of deposite are constantly changing, and whose momentum no boat could resist. West of the ranges of mountains bounding the plain, there are ferries across all of them south of Santiago; and very near the sea, boat navigation would be practicable if there were suitable anchorages in the vicinity to render such mode of transport useful. But there is only one indifferent shelter for vessels between Constitucion and Valparaiso, and the products of the adjoining provinces must find their way to market through one or the other of these ports. Of all, the water from the Andes is divided into two or more streams spread over beds of shingle and silt, sometimes more than a mile

* On a map, in his notices of "Araucania y sus habitantes," published by Professor Domeyko in 1846, the height of Talca is 374 feet; and in the “Annales des Mines" for 1848, he says only 311 feet. The number of observations from which the results are derived is not mentioned; but he used the same barometer as myself, and the extreme differences of my observations were only 0.201 inch, or less than 200 feet.

wide. This, from its nature, cannot be permanent. Every freshet changes the channels, and thus a great obstacle is interposed to the erection of any but extensive and costly bridges to preserve communication between the north and south-improvements which the scarcity of suitable materials and the revenue of a young country scarcely authorize it to undertake. Only one such bridge has as yet been erected-that across the Maypu-at six leagues from the capital, and where a natural configuration of the shores favored the undertaking. Consequently, there are often days when travel to the south is wholly interrupted by streams ordinarily having only the volume of brooks, and even the highway between Santiago and Valparaiso becomes impassable.

The Mataquito, formed by the union of the Lontue and Teno, empties into the ocean in latitude 34° 48′. Its longest tributary—the Lontue-takes its rise about the centre of the Andes, and near Cerro del Medio, a snow-covered mountain, midway between a line joining the Planchon, Cerro Azul, and Cerro Nevado of Chillan, and the point occupied by the Descabezado. Cerro del Medio is on the dividing line of waters here, two leagues to the eastward of that which would pass over the crests mentioned. Passing through valley Grande, the stream descends almost parallel to the sources supplying Lake Mondaca, with whose surplus water it unites to the westward, and the two form the principal volume of the Lontue. Two leagues westward of Curicó, and near the western limit of the plain, the Teno, a much smaller stream from the northeast, unites with it, the two being called, from thence to the ocean, the Mataquito. Where the last penetrates the Central cordilleras their height exceeds 1,000 feet, with hills to the north and south of more than treble that altitude. In comparison with the rivers mentioned, the course of the Mataquito is quite straight, and conforms to the resultant of the two inclinations of the plain. From the source of the Lontue, by the windings to the sea, the distance exceeds forty leagues. Some years ago it was proposed to open a canal between this river and the Maule, so as to throw its whole volume into the latter stream, in the hope that its channel might be deepened and the bar across its mouth be broken down. But wiser counsels prevailed, and whenever journeying to Talca, one still crosses its principal arm over a lasso bridge.

The Rapel is formed from the waters of the Tinguiririca and Cachapual, both streams of considerable volume, whose sources are at very great elevations. Those of the former stream have their origin in the melting snows about the extinct volcano of San Fernando, and flow nearly west to the vicinity of the town of that name, where, after penetrating a range of hills belonging to the Central cordilleras, the stream is deflected to the northward and unites with the Cachapual. This last takes its rise on the eastern slope of the Cruz de Piedra (stone cross) group, a part of the culminating line of the Andes, here more than 17,000 feet high. Descending rapidly through a transversal ravine whose direction is somewhat north of west, and which terminates at the foot of the mountains, it thence turns northwest as far as Rancagua; next southwest by west, to a point formed by a prolongation of the central chain; and, having divided the latter, passes to the ocean in a nearly northwest line. Its origin is about latitude 34° 30′, longitude 69° 44′; its mouth, latitude 33° 53′, longitude 71° 51′; between which points its sinuosities measure 155 miles, with an average fall during the last hundred rather exceeding twenty-six feet per mile. At its entrance on the plain of Rancagua it divides into two principal arms, enclosing an island above the village of Peumo, which is above thirty miles in length. This is entirely composed of alluvium, and is noted for its fertility, as indeed is every part of both provinces irrigated by these streams.

The Maypu originates at an elevation of more than 11,000 feet above the Pacific in a portezuelo or depression between the extinct volcano of the same name and the Cruz de Piedra range. Popular opinion located its source in Lake Diamante, which occupies the bottom of a vast basin just to the eastward of the portezuelo, through whose volcanic strata the waters were supposed to filter; but by a series of levels Señor Pissis ascertained that the surface of the lake is actually seventy-five feet below the Maypu springs. For the first sixty miles the stream rushes

through an exceedingly narrow and deep defile between two ranges of mountains, whose general direction is N. 28° W., and to which its course across the plain and through the central chain is nearly perpendicular. The thirty miles next the sea are in a line inclined a few degrees to the north of west. Its mouth is in latitude 33° 39', longitude 71° 41'. Estimating its length at 150 miles, its mean fall is one foot in about seventy-three; of which it descends 1 for 35 during thirty miles, 1 for 28 in the next 25, 1 for 174 through the following fifty, and 1 for 390 from thence to the sea. A bar extends across its mouth, which is nearly two miles from the land and parallel with the coast. Three miles north is San Antonio covę, a small place affording indifferent shelter to a few coasters and fishermen; the former finding occasional freights of produce from the estates on the lower part of the river.

Its principal tributaries are the Colorado and Mapocho. The former, whose name is derived from the reddish color of its always muddy waters, is alimented by snows from some of the loftiest mountains in America, four of the summits that overshadow its basin ranging from 17,000 to 22,000 feet in height. Scarcely forty miles in length, its whole course is within the Andes, along whose rocky defiles it rushes, urged by a momentum attained by falling one foot in every twenty-four of its trajet. The Mapocho also falls into it on the north shore. This is a clear stream, which proves the lower origin of most of its small affluents. It drains the ravines to the northward of the Colorado, enters the plain just to the eastward of Santiago, and crosses it in a nearly east and west line, as far as the Central cordilleras, along which it flows southwest until its junction with the Maypu. As its supply in very dry seasons is often small, and additional fields brought under cultivation required an additional supply, a canal cut from the Maypu along the very base of the Andes constantly pours a large stream into it just above the city of Santiago. The whole plain to the westward can be irrigated at will from this canal. Notwithstanding this addition to its volume, such is the nature of its bed, the extraordinary dryness of the air, and consequent consumption by the porous soil of the vicinity, that the small remaining rivulet not unfrequently disappears in the shingle a league west from the capital. It again appears, however, where the harder sub-strata, a mile or two nearer the Central cordilleras, force it above the surface. Its length is about seventy miles, and elevation of the headwaters above the junction with the Maypu, 12,000 feet.

The headwaters of the Quillota, or Aconcagua, are to be found in the basin, nearly enclosed between the Chacabuco ridge, which starts from Tupungato in a northwest direction, and the Cumbre ridge, which unites two of the Andean giants-Tupungato and Aconcagua. The most commonly travelled road from Mendoza passes near the banks of the principal stream for more than fifty miles, and travellers speak of it as a wild brook that is generally fordable early in the morning, and until the sun is high enough to dissolve the snows rapidly. It has but two tributaries of any note, the Colorado and Putaendo, both of which flow from the more immediate vicinity of the peak of Aconcagua. Although all the water produced by the melting snow on the western side of the main chain, and its more immediate lateral ramifications between the two great summits named, must pass into the valley of Aconcagua through one of these channels, yet, in the vicinity of the capital of the province, the stream is very little greater than the Mapocho, say fifty yards wide and two and a half feet deep in the centre. Its course is most serpentine and irregular; but, originating at an elevation of 10,500 feet in latitude 33° 05', longitude 69° 51', after winding 140 miles, it empties into the ocean in latitude 32° 55', longitude 71° 20′. From the fact that they are more extensively subdivided, the valleys of Aconcagua and Quillota, watered by it, are the best cultivated and most productive in Chile. Valparaiso looks almost wholly to the valley of Quillota for its supplies of vegetables and fruits in their seasons.

The Chuapa, the Limari, the Coquimbo, the Huasco, and the Copiapó, small streams of melted snow-water that tumble through craggy defiles of the higher Andes with gradually swelling volumes, until they reach an atmosphere so parched that evaporation almost equals the supplies, are all mere brooks when they have descended to levels where man can render

them useful. Over the tract they cross, between latitude 27° and 31°, the scanty products of the soil grown under their influence are of value almost proportionate to the seemingly inexhaustible mineral wealth buried within the rocks of that district, as though nature would preserve in equilibrium man's reward for the sweat of his brow. Higher and higher within the cordilleras, wherever a plateau can be found with a ribband of water over it, there the husbandman creeps, and we may find him cultivating figs, peaches, and melons, nearly 4,000 feet above the ocean, and wheat some 3,500 feet higher. At 10,800 feet in these parallels, the water is found frozen every morning at sunrise; and though it never rains at that altitude, snow-storms commence before the close of autumn. There, only a few herbaceous plants are found.

The region crossed by these streams varies greatly from the country further south in its topographical features. It is far more broken and rugged, the quebradas are deeper and more numerous, isolated and groups of hills more frequent; even the candelabra-like Cereus has ceased; and though there do exist plains, often of many square leagues in extent, the surfaces of the streams are usually so far below their levels as to keep them hopelessly barren. One such plain, more than a hundred miles long and ten miles wide, exists between Vallenar and Copiapó, at a distance of seven or eight leagues from the sea. Across the southern extremity the Huasco river flows; but the level of its water is nearly fifty yards below the plain, and the rivulet can only be used in irrigating a plateau some 250 yards wide on each side of it. No doubt exists that the plateau was once the bed of a noble river. Now, water can only be obtained at two places on the whole plain; and so small is the quantity, that a traveller with a dozen animals will not leave a drop for those who may come after him on the same day. As it rarely rains more than once in two or three years, vegetation lies dormant, and then pasturage is not to be found; the animals that pass over it being from fifty to sixty hours without other food than nibblings from the posts to which they are tied at night. If the volume of the Copiapó, the Huasco, or the Coquimbo, could be augmented to equal even that of the Mapocho, the mining proprietors would cheerfully pay a million dollars.

For the geographical position, courses, and lengths of the several streams north of latitude 32°, reference is given to the map; though it must not be inferred either that they are laid down with great accuracy, or that their volumes are proportionate to the space on the map occupied by the lines indicating them. Not a drop of water reaches the ocean through the old bed of the Copiapó, and no one would suppose that the brooklet he steps across at Huasco is the river of that name on the map. None but professional mine-hunters have ever explored them all, and their chorography is not always very reliable. Prof. Domeyko made journeys to the sources of the Copiapó and Coquimbo, and from his MS. map the relative courses and distances have been laid down, after locating the mouths in the latitudes given in the Appendix to the second volume of the "Voyages of the Adventure and Beagle."

BAYS AND HARBORS.

Were the eastern coast of the Pacific subject to such storms as constantly sweep over the corresponding shores of the Atlantic, if not actually diminished to a large extent, its present limited foreign trade would certainly be restricted to a smaller number of ports. Its proverbial tranquillity, however, authorizes vessels to anchor in roadsteads that are wholly open to the almost unvarying swell rolling across its broad expanse. Between the equator and Chilóe, anchorages entirely protected from the sea or gales are very rarely to be found; nor were the most secure of the few which nature has afforded always chosen by the settlers of the country as ports to the capitals and other cities they founded. They looked upon the sea only as a highway over which they must necessarily transport luxuries desired from home, or convey treasures they intended to amass in America; nor could they consider a harbor of any other utility than as a place where ships sometimes came for these purposes, or to land them succors. Whether five or five hundred miles from the sea, cities could only be founded where most of their treasures were to be the most speedily garnered, or the exigencies of conquest and control