The geology, and even the actual area, of the watershed districts drained by these two great affluents are but imperfectly known. But we may almost venture to assert, from the evidence of these contrasted phenomena, that the Uruguay must be fed from a larger proportionate area of impermeable strata; while the main part of the water of Paraná is shown, by its steady flow, to have been subjected to the retarding and distributing action of subterranean storage and filtration. This, at least, is our European, and we may say, our English experience. A sudden and violent rainfall, over a clay basin, such as that of the Brent, raises, in a few hours, an insignificant brook into a river, swelling in its course into lakes of considerable extent. But a similar fall on the chalk downs and oolite ranges that feed the head waters of the Thames, produces comparatively little effect in the immediate volume of the river. The supply is at once carried, by Nature herself, to the credit of the general average of the year. It is thus that, in the investigation of physical phenomena, we may illustrate the unknown by reference to the known; and obtain from the phenomena of the streams of Middlesex, of Bucks, and of Wilts, more than a slight intimation of the geological characteristics of vast regions of almost unmapped continents.

The Paraná is the largest river on the surface of our planet of which we have any exact knowledge. For that knowledge we are indebted, in no ordinary degree, to the patience and skill of M. Révy. At a distance of 200 miles from the island of Martin Garcia, which is situate on the La Plata estuary, at nineteen miles below the entrance into the water of the Paraná Guazu, stands the town of Rosario, a place of some commercial importance, and the largest town on the banks of the river. 'It is built,' says M. Révy, ' on the margin of the bluff, about eighty feet above the river, and the town reaches from the main land down to the Paraná. Some desultory attempts at engineering stare at us in the shape of a number of strange structures, supposed to be piers; which a number ' of vessels anchored along the margin of the river seem rather 'to avoid than to court. Vessels drawing fifteen feet of water may come up as far as Rosario at all times, during the lowest level of the river. The rise of the Paraná from ordinarily low water to ordinary flood level is here about twelve feet, ' and the flood level is always maintained for three months.' Thirteen miles below Rosario occurs a straight reach of the Paraná, about five miles in length. The width is 4,787 feet from margin to margin at low water. The depth increases, by a gentle and regular slope, from that of a few inches, on the

left shore, to seventy-two feet, at a distance of about 1,100 feet from the right bank. Thence it rapidly shallows to about twelve feet, and then rises gradually to the foot of a vertical cliff forming the right-hand shore of the river.

At this point, three hundred and twenty miles above the junction of the Plata with the Atlantic, a series of measurements have been taken by M. Révy; to whom the students of hydraulics will hereafter owe no small debt of gratitude. The largest accurate measurement of any river section yet attempted, has there been carried out; under arrangements that reflect much credit on the engineers who have planned and accomplished the work. We wish to give our readers some idea of the grandeur of the river, before calling attention to the contribution made, in the present volume, to geographical, or rather hydrographical, science. We shall therefore return, a little later, to the description of the method adopted for gauging the volume of the stream; only remarking that the cross-section, in this locality, of the ordinary flood of the river, gives an area of 243,000 square feet. This large superficies, however, is far from representing the entire waterway of the Paraná; as the left bank is, on occasions when the flood level is attained, submerged for many miles; and although the current over the hidden bank and islands is small, yet a considerable volume of water passes over the great additional sectional area. But even the ascertained section is hard upon the double of that occupied by the October flood of the Uruguay, before cited. In January, when the observations were made, the low-water sectional area of the Paraná was found to be 184,858 superficial feet; or, in round numbers, three-fourths of the area in flood time. The average depth of the section taken in January was forty-seven feet six inches; the extreme depth, seventy-two feet. At times of exceptional rises of the water, such as occurred in 1858 and 1868, this great depth has been increased by twenty-four feet. But that increase, although amounting to thirty-nine per cent. of the smaller dimension, represents not only a factor, but also a function of the second order, of the increased volume of the flow. It is from the observations to which we now refer, that M. Révy has been enabled to indicate a new aspect of the great hydraulic law of the proportion of current to depth. We shall presently find that this ratio is, at all events approximately, represented in this instance, by the ratio of the squares of the depths. That is to say, in round numbers, that the velocity of the Paraná in flood, is to its low-water velocity as 1 to 7.

If we endeavour to estimate the volume of water thus

VOL. CXXXIX. NO. CCLXXXIV.

G G

poured into the Atlantic, by any definite measures of capacity, we shall run the risk of attempting to grasp ideas, and finding that we have only succeeded in accumulating figures. The lowest unit that seems to be applicable to such enormous volumes is that of a million of tons. But who can present to his mind any accurate idea of a million of tons of water? We shall find a more illustrative, though still a very rough, mode of comparison by taking, as in the case of the Uruguay, the discharge of the Thames for a unit of comparison. We labour under the disadvantage that the most important discovery made by M. Révy, in the pursuit of his very elaborate survey, that of the law of the relation of depths and currents, was arrived at late in the course of his observations. The observations are sufficient to show the inaccuracy of the ordinary mode of calculation, and thus to prevent us from attempting to gauge the flow of a river by taking any percentage of the surface velocity as the mean velocity of the stream. Unfortunately, however, they are not accompanied by those measurements of the bottom-current which this very experience shows to be so needful. It is therefore only approximately that we can estimate, from the details before us, the low-water flow of the Paraná, at 600 times that of the Uruguay; a movement of water which is from 2.5 to 3 times that of the Thames, according to the highest estimate of its flow. In other words, the total body of water which the Thames pours into the Channel, in the course of a twelvemonth, is only equal to that which passes the Rosario section of the Paraná, in less than twenty-four hours. Again we seem to arrive at the limits of the receptive power of the imagination, in dealing with such enormous volumes.

The scenery of this mighty river, which in its winter flow is thus of a 360-Thames power, is of appropriate grandeur. On leaving the island of Martin Garcia, which commands the entrance of the Uruguay and the Paraná Guazu into the Plata, the horizon to the west is one vast sea, and nothing is seen of the low islands forming the delta of the Paraná. After an hour's sail, a fine line becomes visible, which is the outline of the luxuriant vegetation that fringes the delta, where its gradual slope first rises above the tide. The nineteen miles of water-way, from this coast-line to the granitic formation of the island of Martin Garcia, are gradually silting up by the deposit borne down by the rivers. The Admiralty charts show a channel of about a mile in width, and from thirty to forty feet deep; but no indication of this channel is apparent on the surface, unless it be that furnished by the varying rapidity of the current.

The character of the mouth of the Guazu is not less extraordinary than the general magnitude of the river. Its width is comparatively small, being only about half a mile from bank to bank; a fact which is explained by the repeated division of the river in passing through its delta. It breaks up into numerous branches, each of which forms a great river by itself. The land rises only about a couple of feet above the river. It is firm, and covered with trees, with a thick undergrowth of rushes. The Guazu has swept for its water so deep a channel, that the steamer which plies on its surface can touch the boughs on either shore, while a depth of several fathoms is under her keel.

A profound stillness reigns in these wild regions. The river glides along without a ripple, and its surface seems that of a lake rather than of a stream. The banks are covered with dense forests of a glossy-leaved tree, known by the name of 'Seibo.' It is to be regretted that so few botanical names have yet been attached to the numerous trees of this part of the world. Specimens of seventy different kinds of wood were sent to the Paris Exhibition of 1867, by the Argentine Confederation. Among the most important of these is the Algarrobo Negro (Prosopis dulcis), useful for furniture and carriagebuilding, of which the edible fruit furnishes' chicha.' The seeds of the cones of the Araucaria imbricata are edible (as are those of the well-known umbrella pine which forms so marked a feature in the landscapes of Southern Italy). The Quiniquina furnishes the odoriferous balsam of Tolu. Seibo is described by M. Révy as resembling an oak short of leaves, having numerous stout branches of very crooked growth; so that it would be difficult to find a straight portion of ten feet long, in either trunk or branches of a tree sixty feet high. The leaves resemble those of the laurel. In spring, the brilliant crimson flowers are as numerous and as large as the leaves; and the mixture of dark green and bright red gives an aspect of great beauty to the forest.

The

Even the aërial tribes seem to shun the vast solitudes of these silent waters. A short distance inland, where marshes and lagoons stretch for mile after mile, swans, geese, ducks, cranes, snipes, and the beautiful native turkey, are abundant, and, as yet, without fear of man. But over the river itself, the flight of some carnivorous bird, or of a little smoke-like wreath of water-fowl, is alone, and even that rarely, to be seen.

At fifty-four miles from the mouth of the Guazu, the loopline of the Paranacito, which separates from the main stream about eighty geographical miles farther up, rejoins the Paraná.

At seventy-two and a half miles, occurs the bifurcation of the river into the two main branches of the Guazu and the Palmas rivers. The Paraná proper is here reached; the whole volume of the stream, with the exception of that which flows down the Ibicuy branch, or Paranacito, above-mentioned, being conveyed in one channel. The scenery undergoes but little change; the trees are, however, less numerous, and a long coarse grass covers the islands.

At ninety-eight and a half miles, the actual continent, or mainland, on the right bank of the Paraná, is first sighted. A bluff of seventy feet rises above the river, and the town of San Pedro is built on the margin. The width of the Paraná is here about 4,000 feet; the river is deep, and runs with a strong current. The channel is a mile distant from the town, and a lagoon, of about three miles long, forms the only natural harbour on the shores of either the Plata or the Paraná.

This

great self-formed place of shelter contains upwards of 300 acres in area, with a minimum depth of eighteen feet at low water. A bank of clay, of only 100 yards wide, separates this noble natural dock from the Paraná, which here runs, for a reach of nine miles in length, at right angles to its general course. A stream called the Baradero continues the right line of the river from San Pedro to a junction with the Palma branch at Rancho, a distance of some twenty-four miles; but the main channel bends abruptly to the north-east. As far as the hydrography of the region is concerned, it would be difficult to point to any other spot in the world that should seem better fitted, by Nature herself, to become the seat of a great commercial emporium.

At 108 miles, the channel contracts between steep bluff banks to about one-half its former width. This locality, known as the Straits of Obligado, is swept by a powerful current, the depth of the channel being 150 feet. It is a place where future observations of the velocity of the under-currents may be expected to afford information of the utmost value to hydraulic science. The high table-land of the province of Buenos Ayres now forms the right bank of the Paraná for more than 100 miles, although large islands sometimes divert the channel for miles from the shore. The left bank is an immense swamp, of from fifteen to thirty miles in width, commencing at the junction of the Guazu and the Ibicuy (at fifty-four miles), and terminating at the division of the latter channel from the main stream, where the Ibicuy bears the name of the Pavon channel. The mainland, rising as a bluff, now forms the left bank of the channel, for the first time, at 253 miles from the mouth of the Paraná. At this place, which is marked by the site of a village