BOOK guedoc, the period of whose rising is each day fifty minute XIII. later than the preceding day.* The round fountain, on the road from Pontarlier to Touillon, in Franche-Comté, rises with a boiling appearance. The Bullerborn, in the bishopric of Paderborn, in Westphalia, rises with great noise. Near Brest, there is a well 75 feet from the sea, which sinks with the flow and rises with the ebbing of the sea. England furnishes many examples of these springs, one particularly near Torbay, in Devonshire, and one at Buxton, in Derbyshire. According to Gruner, there is one at Engstler, in the canton of Berne, which has a double intermission daily and annually. But of all these kinds of springs, of which many more examples might be adduced, none exhibits a perfectly regular course. These springs are accounted for by supposing, that in the lands where they are situated, there are reservoirs and conducting pipes in the form of syphons. It is perhaps unnecessary to explain, that the liquid begins to flow through the syphon as soon as the surface of the liquid in which one end of the tube is placed on a level with the curvature of the two branches; and the flowing continues as long as the fluid keeps above the orifice of the branch or end inserted in it. The moment the orifice ceases to be immersed in the liquid, the flowing ceases, and it recommences as soon as the reservoir is filled to the level of the bending. With respect to the reservoirs which supply these fountains, drought, rain, and the melting of the snow, may so affect them, as to render their periodical return more or less regular. The connection subsisting between the greater or less humidity of the atmosphere, and the reservoirs of intermitting fountains, justifies to a certain degree the conjectures which are sometimes formed from the movements of these springs as to the nature of the approaching season, conjecFountains tures which have given to some of them the names of founof dearth tain of dearth and plenty.§

and plenty.

*Astruc, Histoire Naturelle de Languedoc et de la Provence

+ Journal de Trévoux, 1728, Octobre.

Scheuchzer, Iter Alpin. 26, ii. 404.

Kant, Géographie-Physique, ii. part 2. p. 224.

Subter

raneous waters.

It is natural to imagine, that many channels of water not BOOK finding any other suitable outlet, flow into subterraneous XIII. cavities, are absorbed by the earth, or discharge themselves We may thus explain the oribelow ground into the sea. gin of those springs of fresh water that are to be seen spouting up even in the midst of the waves of the ocean. The water thrown up by volcanos, the sudden and terrible inundations of mines, the number of rivers which disappear, the mountains suddenly engulphed in the bosom of new lakes, all these facts combine in proving, that there are considerable subterraneous cavities, often filled with water. The necessity of supplying the scarcity of springs by digging wells, has procured us the knowledge of a fact still more interesting to physical geography. It appears, that there are lakes, or rather sheets of water, which extend under ground to considerable distances. Delamétherie relates,* that in the province of Artois, near Aire, in digging wells, they always come to a clayey bed, which, being pierced, the water gushes out in large bubbles, and, rising up, forms springs which continually flow. In the country of Modena, we find every where, at the depth of 63 fect, a bed of clay five feet thick, and beneath it water which spouts out with much force. In the interior of the country of Algiers, in the province of Wad-Reag, the inhabitants, after digging 200 fathoms deep, invariably meet with a stratum of slate, under which the water flows in such abundance, that they name it the subterraneous sea. We may easily conceive that one bed of clay may have sunk down horizontally by drying, whilst another bed of clay may have been forced upwards. The fissure horizontally formed in this manner, may have served as a channel for the lakes or rivers which constitute these extensive subterraneous waters.

The glaciers which crown the tops of the highest moun- Glaciers. tains, have a close connection and a common origin with

XIII.

BOOK springs. Snows accumulated for whole centuries, sink down and are compressed, and consolidated as much by evaporation, as by alternate thaws and frosts. Thus are formed immense caps which cover whole mountains or fields of frozen snow, which extend between the summits: The high valleys are filled at the same time with the snow which falls there, and with the icy waters which flow from the snowy summits. In fact, these flowings alone, joined to the avalanches, occasion those masses of pure ice, the branches of which extend even to the lower valleys. The latter masses of ice seem in some places to have continued increasing for a long series of years. They have, in Switzerland, filled up even whole valleys, buried villages, and shut up the pass between Le Valais and the Canton of Berne-but the diminutions on one side generally compensate for the increase of the other. A few warm seasons are sufficient to re-establish the equilibrium.*

The scenes which these bodies of ice exhibit are as various as their extent. At one time a great mass of water, congealed at the period of a tempest, presents waves resembling those of a lake-at other times these inequalities disappear, and leave nothing to be beheld by the astonished traveller, but one immense mirror of polished ice. Here superb portals of crystal appear fallen into rivers, and brilliant spires broken to pieces-in other places, avalanches of snow glide over a field of ice and then stop, and, reflecting the rays of the sun, display the form of pyramids and obelisks unseen before. These glaciers are of essential service in furnishing to the continents slowly, and in an almost regular manner, waters which, without this congelation, would be precipitated with impetuosity from the height of the mountains, so as to overflow and devastate the countries which they ought to fertilize. For this beneficial effect, we are indebted to the intense cold which converts the waters into snow and ice, and holds them suspended on the sides of the mountains, to supply

* Saussure, Voyages, Gruner, etc.

abundant and unfailing streams oozing from the bottom of BOOK these enormous masses, or from the bosom of their crystal

lized grottos.

XIII.

rents, and

The effusion of springs, and the flowings of melted ice, Streams, form little currents, more or less gentle, which are termed rivers, torrivulets. The water of great rains falls with more rapidi- rivulets. ty, and furrows the sides of the mountains by impetuous irregular torrents. The union of these currents forms streams, which, following the declivity of the ground, unite most frequently in a great canal which takes the name of river, and which conveys to the ocean the collected tribute of the earth.

The declivities (considered collectively, and as a whole,) from whence flow the streams and rivulets which discharge themselves into one particular river, are called the basin of that river, or its hydrographical region. It frequently hap- Hydrographical bapens, that the basins of two rivers almost touch. In Ame- sins. rica, the Cassiquiari, and some other rivers, actually unite the basin of the Oronoco, with that of the Amazon.* In Europe, the sources of the Duina, of the Niemen, and of the Borysthenes, nearly meet together in a marshy plain. Geology has been much employed in investigating the subject of basins; in general, the mineral beds and petrifactions of the same basin present a certain analogy; but, (according to the just observation of M. Desmarets,) it is also essential to distinguish with accuracy the hydrographical masses or groupes of mountains which furnish water to Masses, or hydrograrivers that receive no supply from any other quarter. The phical knowledge of these masses is indispensable to assist us in plateaux. explaining the nature of rivers. Calcareous soils produce waters of a very different nature from those which flow from the glaciers over sand or clay. The elevation of the springs determines the amount of the declivity, and this latter circumstance modifies the course of streams and rivers, rendering them rapid or gentle, regular or meandering.

Condamine, Voyage de la Rivere des Amazons, p. 119. Hartsink, Humboldt, etc.

BOOK

Beds of rivers.

The beds of rivers are the lowest parts of great chasins, XIII. formed by the same revolution which produced the mountains: The atmospheric waters have evidently brought down a portion of light soil which was adhering to the sides of the mountains; they may have formed by their sediments horizontal plains which occupy the bottom of certain valleys ;* but never could a river by its own force alone have opened for itself a passage through solid rocks, similar to those which border the upper Rhine; it must at the first have found the outline of its course deeply marked out. Running waters unceasingly wear away their beds and banks in places where the declivity is very rapid; they hollow out and deepen their channels in mountains composed of rocks of a moderate hardness-they draw along stones, and form accumulations of them in the lower part of their course-and thus their beds are often gradually elevated in the plains, while they are deepened and depres sed in the mountains; but these changes, though continually going on for thousands of years, could only give form to the banks of rivers; they in no wise created the banks themselves.

Declivity of rivers.

It is only the sloping of the land which can at first cause water to flow; but an impulse having been once communicated to the mass, the pressure alone of the water will keep it in motion, even if there were no declivity at all. Many great rivers in fact flow with an almost imperceptible declivity. The river of the Amazons has only ten feet and a half of declivity upon two hundred leagues of extent of water, which makes of an inch for every 1000 feet.† The Seine, between Valvins and Serves, has only one foot declivity out of 6600. The Loire has, between Pouilly and Briare, one foot in 7500; but between Briare and Orleans, only one foot in 13,596. In East Friezland, in

Comp. Saussure, Voyages, 648, 920. Delamétherie, ◊ 1618. (For the action of rivers, see Bourguet, Lettres Philosophiques, 181. Voigt, Mémoires Minéralog. vol. iii. Mémoire sur la Formation des Vallées.)

† Condamine, 1. c. p. 134.

Picart, Traité du Nivellement, p. 152, etc. etc.