In the sixteenth century the elevation of the bed of the valley had become so considerable, that in 1551, at a point about ten miles south of the Arno, it was found to be not less than one hundred and thirty feet above that river; then followed a level of ten miles, and then a continuous descent to the Paglia. Along the level portion of the valley was a boatable channel, and lakes, sometimes a mile or even two miles in breadth, had formed at various points farther south. At this period the drainage of the summit level might easily have been determined in either direction, and the opposite descents of the valley made to culminate at the north or at the south end of the level. In the former case, the watershed would have been ten miles south of the Arno; in the latter, twenty miles, and the division of the valley into two opposite slopes would have been not very unequal.

Various schemes were suggested at this time for drawing off the stagnant waters, as well as for the future regular drainage of the valley, and small operations for those purposes were undertaken with partial success; but it was feared that the discharge of the acc.u.mulated waters into the Tiber would produce a dangerous inundation, while the diversion of the drainage into the Arno would increase the violence of the floods to which that river was very subject, and no decisive steps were taken.

In 1606 an engineer, whose name has not been preserved, proposed, as the only possible method of improvement, the piercing of a tunnel through the hills bounding the valley on the west to convey its waters to the Ombrone, but the expense and other objections prevented the adoption of this scheme. [Footnote: Morozzi, Dello stato dell' Arno, ii., pp. 39, 40.] The fears of the Roman Government for the safety of the basin of the Tiber had induced it to construct embankments across the portion of the valley lying within its territory, and these obstructions, though not specifically intended for that purpose, naturally promoted the deposit of sediment and the elevation of the bed of the valley in their neighborhood. The effect of this measure and of the continued spontaneous action of the torrents was, that the northern slope, which in 1551 had commenced at the distance of ten miles from the Arno, was found in 1605 to begin nearly thirty miles south of that river, and in 1645 it had been removed about six miles farther in the same direction.

[Footnote: Morozzi, Dello stato, etc., dell' Arno, ii., pp. 39, 40.]

In the seventeenth century the Tuscan and Papal Governments consulted Galileo, Torricelli, Castelli, Ca.s.sini, Viviani, and other distinguished philosophers and engineers, on the possibility of reclaiming the valley by a regular artificial drainage. Most of these eminent physicists were of opinion that the measure was impracticable, though not altogether for the same reasons; but they seem to have agreed in thinking that the opening of such channels, in either direction, as would give the current a flow sufficiently rapid to drain the lands properly, would dangerously augment the inundations of the river--whether the Tiber or the Arno--into which the waters should be turned. The general improvement of the valley was now for a long time abandoned, and the waters were allowed to spread and stagnate until carried off by partial drainage, infiltration, and evaporation. Torricelli had contended that the slope of a large part of the valley was too small to allow it to be drained by ordinary methods, and that no practicable depth and width of ca.n.a.l would suffice for that purpose. It could be laid dry, he thought, only by converting its surface into an inclined plane, and he suggested that this might be accomplished by controlling the flow of the numerous torrents which pour into it, so as to force them to deposit their sediment at the pleasure of the engineer, and, consequently, to elevate the level of the area over which it should be spread. [Footnote: Torricelli thus expressed himself on this point: "If we content ourselves with what nature has made practicable to human industry, we shall endeavor to control, as far as possible, the outlets of these streams, which, by raising the bed of the valley with their deposits, will realize the fable of the Tagus and the Pactolus, and truly roll golden sands for him that is wise enough to avail himself of them."--Fos...o...b..oni, Memoris sopra la Val di China, p. 219.] This plan did not meet with immediate general acceptance, but it was soon adopted for local purposes at some points in the southern part of the valley, and it gradually grew in public favor and was extended in application until its final triumph a hundred years later.

In spite of these encouraging successes, however, the fear of danger to the valley of the Arno and the Tiber, and the difficulty of an agreement between Tuscany and Rome--the boundary between which states crossed the Val di Chiana not far from the half-way point between the two rivers--and of reconciling other conflicting interests, prevented the resumption of the projects for the general drainage of the valley until after the middle of the eighteenth century. In the meantime the science of hydraulics had become better understood, and the establishment of the natural law according to which the velocity of a current of water, and of course the proportional quant.i.ty discharged by it in a given time, are increased by increasing its ma.s.s, had diminished if not dissipated the fear of exposing the banks of the Arno to greater danger from inundations by draining the Val di China into it. The suggestion of Torricelli was finally adopted as the basis of a comprehensive system of improvement, and it was decided to continue and extend the inversion of the original flow of the waters, and to turn them into the Arno from a point as far to the south as should be found practicable. The conduct of the works was committed to a succession of able engineers who, for a long series of years, were under the general direction of the celebrated philosopher and statesman Fos...o...b..oni, and the success has fully justified the expectations of the most sanguine advocates of the scheme.

The plan of improvement embraced two branches: the one, the removal of obstructions in the bed of the Arno, and, consequently, the further depression of the channel of that river, in certain places, with the view of increasing the rapidity of its current; the other, the gradual filling up of the ponds and swamps, and raising of the lower grounds of the Val di Chiana, by directing to convenient points the flow of the streams which pour down into it, and there confining their waters by temporary dams until the sediment was deposited where it was needed. The economical result of these operations has been, that in 1835 an area of more than four hundred and fifty square miles of pond, marsh, and damp, sickly low grounds had been converted into fertile, healthy, and well-drained soil, and, consequently, that so much territory has been added to the agricultural domain of Tuscany. But in our present view of the subject, the geographical revolution which has been accomplished is still more interesting. The climatic influence of the elevation and draining of the soil must have been considerable, though I do not know that an increase or a diminution of the mean temperature or precipitation in the valley has been established by meteorological observation. There is, however, in the improvement of the sanitary condition of the Val di Chiana, which was formerly extremely unhealthy, satisfactory proof of a beneficial climatic change. The fevers, which not only decimated the population of the low grounds but infested the adjacent hills, have ceased their ravages, and are now not more frequent than in other parts of Tuscany. The strictly topographical effect of the operations in question, besides the conversion of marsh into dry surface, has been the inversion of the inclination of the valley for a distance of thirty-five miles, so that this great plain which, within a comparatively short period, sloped and drained its waters to the south, now inclines and sends its drainage to the north. The reversal of the currents of the valley has added to the Arno a new tributary equal to the largest of its former affluents, and a most important circ.u.mstance connected with this latter fact is, that the increase of the volume of its waters has accelerated their velocity in a still greater proportion, and, instead of augmenting the danger from its inundations, has almost wholly obviated that source of apprehension. [Footnote: Arrian observes that at the junction of the Hydaspes and the Acesines, both of which are described as wide streams, "one very narrow river is formed of two confluents, and its current is very swift."--Arrian, Alex. Anab., vi., 4.

A like example is observed in the Anapus near Syracuse, which, below the junction of its two branches, is narrower, though swifter than either of them, and such cases are by no means unfrequent. The immediate effect of the confluence of two rivers upon the current below depends upon local circ.u.mstances, and especially upon the angle of incidence. If the two nearly coincide in direction, so as to include a small angle, the join current will have a greater velocity than the slower confluent, perhaps even than either of them. If the two rivers run in transverse, still more if they flow in more or less opposite, directions, the velocity of the princ.i.p.al branch will be r.e.t.a.r.ded both above and below the junction, and at high water it may even set back the current of the affluent.

On the other hand, the diversion of a considerable branch from a river r.e.t.a.r.ds its velocity below the point of separation, and here a deposit of earth in its channel immediately begins, which has a tendency to turn the whole stream into the new bed. "Theory and the authority of all hydrographical writers combine to show that the channels of rivers undergo an elevation of bed below a ca.n.a.l of diversion."--Letter of Fos...o...b..oni, in Salvagnoli, Raccolta di Doc.u.menti, p. 32. See the early authorities and discussions on the principle stated in the text, in Frisi, Del modo di regolare i Fiumi e i Torrenti, libro iii., capit. i., and Mongotti, Idraulica, ii., pp. 88 et seqq., and see p. 498, note, ante.

In my account of these improvements I have chiefly followed Fos...o...b..oni, under whose direction they were princ.i.p.ally executed. Many of Fos...o...b..oni's statements and opinions have been controverted, and in comparatively unimportant particulars they have been shown to be erroneous.--See Lombardini, Guida allo studio dell' Idrologia, cap.

xviii., and same author, Esame degli Studi sul Tevere, Section 33.]

Between the beginning of the fifteenth century and the year 1761, thirty-one destructive floods of the Arno are recorded; between 1761, when the princ.i.p.al streams of the Val di Chiana were diverted into that river, and 1835, not one. [Footnote: Fos...o...b..oni, Memorie Idraulico-storiche, Introduzione, p. xvi. Between the years 1700 and 1799 the chroniclers record seventeen floods of the Arno, and twenty between 1800 and 1870, but none of these were of a properly destructive character except those in 1844, 1864, and 1870, and the ravages of this latter were chiefly confined to Pisa, and were occasioned by the bursting of a dike or wall. They are all three generally ascribed to extraordinary, if not unprecedented, rains and snows, but many inquirers attribute them to the felling of the woods in the valleys of the upper tributaries of the Arno since 1835. See a paper by Griffini, in the Italia Nuova, 18 Marzo, 1871.]

Results of Operations.

It is now a hundred years since the commencement of the improvements in the Val di Chiana, and those of the Maremma have been in more or less continued operation for above a generation. They have, as we have seen, produced important geographical changes in the surface of the earth and in the flow of considerable rivers, and their effects have been not less conspicuous in preventing other changes, of a more or less deleterious character, which would infallibly have taken place if they had not been arrested by the improvements in question.

The sediment washed into the marshes of the Maremma is not less than 12,000,000 cubic yards per annum. The escape of this quant.i.ty into the sea, which, is now almost wholly prevented, would be sufficient to advance the coast-line fourteen yards per year, for a distance of forty miles, computing the mean depth of the sea near the sh.o.r.e at twelve yards. It is true that in this case, as well as in that of other rivers, the sedimentary matter would not be distributed equally along the sh.o.r.e, and much of it would be carried out into deep water, or perhaps transported by the currents to distant coasts. The immediate effects of the deposit in the sea, therefore, would not be so palpable as they appear in this numerical form, but they would be equally certain, and would infallibly manifest themselves, first, perhaps, at some remote point, and afterwards more energetically at or near the outlets of the rivers which produced them. The elevation of the bottom of the sea would diminish the inclination of the beds of the rivers discharging themselves into it on that coast, and of course their tendency to overflow their banks and extend still further the domain of the marshes which border them would be increased in proportion.

It has been already stated that, in order to prevent the overflow of the valley of the Tiber by freely draining the Val di Chiana into it, the Papal authorities, long before the commencement of the Tuscan works, constructed strong barriers near the southern end of the valley, which detained the waters of the wet season until they could be gradually drawn off into the Paglia. They consequently deposited most of their sediment in the Val di Chiana and carried down comparatively little earth to the Tiber. The lateral streams contributing the largest quant.i.ties of sedimentary matter to the Val di Chiana originally flowed into that valley near its northern end; and the change of their channels and outlets in a southern direction, so as to raise that part of the valley by their deposits and thereby reverse its drainage, was one of the princ.i.p.al steps in the process of improvement.

We have seen that the north end of the Val di Chiana near the Arno had been raised by spontaneous deposit of sediment to such a height as to interpose a sufficient obstacle to all flow in that direction. If, then, the Roman dam had not been erected, or the works of the Tuscan Government undertaken, the whole of the earth, which has been arrested by those works and employed to raise the bed and reverse the declivity of the valley, would have been carried down to the Tiber and thence into the sea. The deposit thus created would, of course, have contributed to increase the advance of the sh.o.r.e at the mouth of that river, which has long been going on at the rate of three metres and nine-tenths (twelve feet and nine inches) per annum. [Footnote: See the careful estimates of Rozet, Moyens de forcer les Torrents, etc., pp. 42, 44.] It is evident that a quant.i.ty of earth, sufficient to effect the immense changes I have described in a wide valley more than thirty miles long, if deposited at the outlet of the Tiber, would have very considerably modified the outline of the coast, and have exerted no unimportant influence on the flow of that river, by raising its point of discharge and lengthening its channel.

The Coast of the Netherlands. It has been shown in a former section that the dikes of the Netherlands and the adjacent states have protected a considerable extent of coast from the encroachments of the sea, an have won a large tract of cultivable land from the dominion of the ocean waters. The immense results obtained from the operations of the Tuscan engineers in the Val di Chiana, and the Maremma have suggested the question, whether a different method of accomplishing these objects might not have been adopted with advantage. It has been argued, as in the case of the Po, that a system of transverse inland dikes and ca.n.a.ls, upon the principle of those which have been so successfully employed in the Val di Chiana and in Egypt, might have elevated the low grounds above the ocean tides, by spreading over them the sediment brought down by the Rhine, the Maes, and the Scheld. If this process had been introduced in the Middle Ages, and constantly pursued to our times, the superficial and coast geography, as well as the hydrography of the countries in question, would undoubtedly have presented an aspect very different from their present condition; and by combining the process with a system of maritime dikes, which would have been necessary, both to resist the advance of the sea and to retain the slime deposited by river overflows, it is, indeed, possible that the territory of those states would have been as extensive as it now is, and, at the same time, somewhat elevated above its natural level.

The argument in favor of that method rests on the a.s.sumption that all the sea-washed earth, which the tides have let fall upon the shallow coast of the Netherlands, has been brought down by the rivers which empty upon those sh.o.r.es, and could have been secured by allowing those rivers to spread over the flats and deposit their sediment in still-water pools formed by cross-dikes like those of Egypt.

But we are ignorant of the proportions in which the marine deposits that form the soil of the polders have been derived from materials brought down by these rivers, or from other more remote sources. Much of the river slime has, no doubt, been transported by marine currents quite beyond the reach of returning streams, and it is uncertain how far this loss has been balanced by earth washed by the sea from distant sh.o.r.es and let fall on the coasts of the Netherlands and other neighboring countries.

We know little or nothing of the quant.i.ty of solid matter brought down by the rivers of Western Europe in early ages, but, as the banks of those rivers are now generally better secured against wash and abrasion than in former centuries, the sediment transported by them must be less than at periods nearer the removal of the primitive forests of their valleys, though certainly greater than it was before those forests were felled. Kladen informs us that the sedimentary matter transported to the sea by the Rhine would amount to a cubic geographical mile in five thousand years. [Footnote: Erdhunde, vol. i, p. 384. The Mississippi--a river "undercharged with sediment"--with a mean discharge of about ten times that of the Rhine, deposits a cubic geographical mile in thirty-three years.] The proportion of this suspended matter which, with our present means, could be arrested and precipitated upon the ground, is almost infinitesimal, for only the surface-water, which carries much less sediment than that at the bottom of the channel, would flow over the banks, and as the movement of this water, if not checked altogether, would be greatly r.e.t.a.r.ded by the proposed cross-dikes, the quant.i.ty of solid matter which would be conveyed to a given portion of land during a single inundation would be extremely small. Inundations of the Rhine occur but once or twice a year, and high water continues but a few days, or even hours; the flood-tide of the sea happens seven hundred times in a year, and at the turn of the tide the water is brought to almost absolute rest. Hence, small as is the proportion of suspended matter in the tide-water, the deposit probably amounts to far more in a year than would be let fall upon the same area by the Rhine.

This argument, except as to the comparison between river and tide water, applies to the Mississippi, the Po, and most other great rivers. Hence, until that distant day when man shall devise means of extracting from rivers at flood, the whole volume of their suspended material and of depositing it at the same time on their banks, the system of cross-dikes and COLMATAGE must be limited to torrential streams transporting large proportions of sediment, and to the rivers of hot countries, like the Nile, where the saturation of the soil with water, and the securing of a supply for irrigation afterwards, are the main objects, while raising the level of the banks is a secondary consideration.

CHAPTER V.

THE SANDS.

Origin of Sand--Sand now Carried to the Sea--Beach Sands of Northern Africa--Sands of Egypt--Sand Dunes and Sand Plains--Coast Dunes--Sand Banks--Character of Dune Sand--Interior Structure of Dunes--Geological Importance of Dunes--Dunes on American Coasts--Dunes of Western Europe--Age, Character, and Permanence of Dunes--Dunes as a Barrier against the Sea--Encroachments of the Sea--Liimfjord--Coasts of Schleswig-Holstein, Netherlands, and France--Movement of Dunes--Control of Dunes by Man--Inland Dunes--Inland Sand Plains.

Origin of Sand.

Sand, which is found in beds or strata at the bottom of the sea or in the channels of rivers, as well as in extensive deposits upon or beneath the surface of the dry land, appears to consist essentially of the detritus of rocks. It is not always by any means clear through what agency the solid rock has been reduced to a granular condition; for there are beds of quartzose sand, where the sharp, angular shape of the particles renders it highly improbable that they have been formed by gradual abrasion and attrition, and where the supposition of a crushing mechanical force seems equally inadmissible. In common sand, the quartz grains are the most numerous; but this is not a proof that the rocks from which these particles were derived were wholly, or even chiefly, quartzose in character; for, in many composite rocks, as, for example, in the granitic group, the mica, feldspar, and hornblende are more easily decomposed by chemical action, or disintegrated, comminuted, and reduced to an impalpable state by mechanical force, than the quartz. In the destruction of such rocks, therefore, the quartz would survive the other ingredients, and remain unmixed, when they had been decomposed and recomposed into new mineralogical or chemical combinations, or been ground to slime and washed away by water currents.

The greater or less specific gravity of the different const.i.tuents of rock doubtless aids in separating them into distinct ma.s.ses when once disintegrated, though there are veined and stratified beds of sand where the difference between the upper and lower layers, in this respect, is too slight to be supposed capable of effecting a complete separation. [Footnote: In the curiously variegated sandstone of Arabia Petraea--which is certainly a reaggregation of loose sand derived from disaggregation of older rocks--the continuous veins frequently differ very widely in color, but not sensibly in specific gravity or in texture; and the singular way in which they are now alternated, now confusedly intermixed, must be explained otherwise than by the weight of the respective grains which compose them. They seem, in fact, to have been let fall by water in violent ebullition or tumultuous mechanical agitation, or deposited by a succession of sudden aquatic or aerial currents flowing in different directions and charged with differently colored matter.] In cases where rock has been reduced to sandy fragments by heat, or by obscure chemical and other molecular forces, the sand-beds may remain undisturbed, and represent, in the series of geological strata, the solid formations from which they were derived.

The large ma.s.ses of sand not found in place have been transported and acc.u.mulated by water or by wind, the former being generally considered the most important of these agencies; for the extensive deposits of the Sahara, of the Arabian peninsulas, of the Llano Estacado and other North and South American deserts, of the deserts of Persia, and of that of Gobi, are supposed to have been swept together or distributed by marine currents, and to have been elevated above the ocean by the same means as other upheaved strata. Meteoric and mechanical influences are still active in the reduction of rocks to a fragmentary state; [Footnote: A good account of the agencies now operative in the reduction of rock to sand will be found in Winkler, Zand en Duinen, Dockarm, 1865, pp. 4-20.

I take this occasion to acknowledge my obligations to this author for a.s.suming the responsibility of many of the errors I may have committed in this chapter, by translating a large part of it from a former edition of the present work and publishing it as his own.] but the quant.i.ty of sand now transported to the sea seems to be comparatively inconsiderable, because--not to speak of the absence of diluvial action--the number of torrents emptying directly into the sea is much less than it was at earlier periods. The formation of alluvial plains in maritime bays, by the sedimentary matter brought down from the mountains, has lengthened the flow of such streams and converted them very generally into rivers, or rather affluents of rivers of later geographical origin than themselves. The filling up of the estuaries has so reduced the slope of all large and many small rivers, and, consequently, so checked the current of what the Germans call their Unterlauf, or lower course, that they are much less able to transport heavy material than at earlier epochs. The slime deposited by rivers at their junction with the sea, is usually found to be composed of material too finely ground and too light to be denominated sand, and it can be abundantly shown that the sand-banks at the outlet of most large streams are of tidal, not of fluviatile, acc.u.mulation, or, in lakes and tideless seas, a result of the concurrent action of waves and of wind. Large deposits of sand, therefore, must in general be considered as of ancient, not of recent formation, and many eminent geologists ascribe them to diluvial action. Staring has discussed this question very fully, with special reference to the sands of the North Sea, the Zuiderzee, and the bays and channels of the Dutch coast. [Footnote: De Bodem van Nederland, i., pp. 243, 246-377, et seqq. See also the arguments of Bremontier as to the origin of the dune-sands of Gascony, Annales des Ponts et Chaussees, 1833, 1er semestre, pp. 158, 161. Bremontier estimates the sand anually thrown up on that coast at five cubic toises and two feet to the running toise (ubi supra, p. 162), or rather more than two hundred and twenty cubic feet to the running foot. Laval, upon observations continued through seven years, found the quant.i.ty to be twenty-five metres per running metre, which is equal to two hundred and sixty-eight cubic feet to the running foot.--Annales des Ponts et Chaussees, 1842, 2me semestre, p. 229. These computations make the proportion of sand deposited on the coast of Gascony three or four times as great as that observed by Andresen on the sh.o.r.es of Jutland. Laval estimates the total quant.i.ty of sand annually thrown up on the coast of Gascony at 6,000,000 cubic metres, or more than 7,800,000 cubic yards.]

His general conclusion is, that the rivers of the Netherlands "move sand only by a very slow displacement of sand-banks, and do not carry it with them as a suspended or floating material." The sands of the German Ocean he holds to be a product of the "great North German drift," deposited where they now lie before the commencement of the present geological period, and he maintains similar opinions with regard to the sands thrown up by the Mediterranean at the mouths of the Nile and on the Barbary coast. [Footnote: De Bodem van Nederland, i., p. 339.]

Sand now carried to the Sea.

There are, however, cases where mountain streams still bear to the sea perhaps relatively small, but certainly absolutely large, amounts of disintegrated rock. [Footnote: The conditions favorable to the production of sand from disintegrated rock, by causes now in action, are perhaps nowhere more perfectly realized than in the Sinaitic Peninsula.

The mountains are steep and lofty, unprotected by vegetation or even by a coating of earth, and the rocks which compose them are in a shattered and fragmentary condition. They are furrowed by deep and precipitous ravines, with beds sufficiently inclined for the rapid flow of water, and generally without basins in which the larger blocks of stone rolled by the torrents can be dropped and left in repose; there are severe frosts and much snow on the higher summits and ridges, and the winter rains are abundant and heavy. The mountains are princ.i.p.ally of igneous formation, but many of the less elevated peaks are capped with sandstone, and on the eastern slope of the peninsula you may sometimes see, at a single glance, several lofty pyramids of granite, separated by considerable intervals, and all surmounted by horizontally stratified deposits of sandstone often only a few yards square, which correspond to each other in height, are evidently contemporaneous in origin, and were once connected in continuous beds. The degradation of the rock on which this formation rests is constantly bringing down ma.s.ses of it, and mingling them with the basaltic, porphyritic, granitic, and calcareous fragments which the torrents carry down to the valleys, and, through them, in a state of greater or less disintegration, to the sea. The quant.i.ty of sand annually washed into the Red Sea by the larger torrents of the Lesser Peninsula, is probably at least equal to that contributed to the ocean by any streams draining basins of no greater extent.

Absolutely considered, then, the ma.s.s may be said to be large, but it is apparently very small as compared with the sand thrown up by the German Ocean and the Atlantic on the coasts of Denmark and of France. There are, indeed, in Arabia Petraea, many torrents with very short courses, for the sea-waves in many parts of the peninsular coast wash the base of the mountains. In these cases, the debris of the rocks do not reach the sea in a sufficiently comminuted condition to be ent.i.tled to the appellation of sand, or even in the form of well-rounded pebbles. The fragments retain their annular shape, and, at some points on the coast, they become cemented together by lime or other binding substances held in solution or mechanical suspension in the sea-water, and are so rapidly converted into a singularly heterogeneous conglomerate, that one deposit seems to be consolidated into a breccia before the next winter's torrents cover it with another.

In the northern part of the peninsula there are extensive deposits of sand intermingled with agate pebbles and petrified wood, but these are evidently neither derived from the Sinaitic group, nor products of local causes known to be now in action.

I may here notice the often repeated but mistaken a.s.sertion, that the petrified wood of the Western Arabian desert consists wholly of the stems of palms, or at least of endogenous vegetables. This is an error.

I have myself picked up in that desert, within the s.p.a.ce of a very few square yards, fragments apparently of fossil palms, and of at least two petrified trees distinctly marked as of exogenous growth both by annular structure and by knots. In ligneous character, one of these almost precisely resembles the grain of the extant beech, and this specimen was worm-eaten before it was converted into silex.]

The quant.i.ty of sand and gravel carried into the Mediterranean by the torrents of the Maritime Alps, the Ligurian Apennines, the islands of Corsica, Sardinia, and Sicily, and the mountains of Calabria, is apparently great. In mere ma.s.s, it is possible, if not probable, that as much rocky material, more or less comminuted, is contributed to the basin of the Mediterranean by Europe, even excluding the sh.o.r.es of the Adriatic and the Euxine, as is washed up from it upon the coasts of Northern Africa and Syria. A great part of this material is thrown out again by the waves on the European sh.o.r.es of that sea. The harbors of Luni, Albenga, San Remo, and Savona west of Genoa, and of Porto Fino on the other side, are filling up, and the coast near Carrara and Ma.s.sa is said to have advanced upon the sea to a distance of 475 feet in thirty-three years. [Footnote: Bottger, Das Mittelmeer, p. 128.] Besides this, we have no evidence of the existence of deep-water currents in the Mediterranean, extensive enough and strong enough to transport quartzose sand across the sea. It may be added that much of the rock from which the torrent sands of Southern Europe are derived contains little quartz, and hence the general character of these sands is such that they must be decomposed or ground down to an impalpable slime, long before they could be swept over to the African sh.o.r.e.

Sands of Northern Africa.

The torrents of Europe, then, do not at present furnish the material which composes the beach sands of Northern Africa, and it is equally certain that those sands are not brought down by the rivers of the latter continent. They belong to a remote geological period, and have been acc.u.mulated by causes which we cannot at present a.s.sign. The wind does not stir water to great depths with sufficient force to disturb the bottom, [Footnote: The testimony of divers and of other observers on this point is conflicting, as might be expected from the infinite variety of conditions by which the movement of water is affected. It is generally believed that the action of the wind upon the water is not perceptible at greater depths than from fifteen feet in ordinary to eighty or ninety in extreme cases; but these estimates are probably very considerably below the truth. Andresen quotes Bremontier as stating that the movement of the waves sometimes extends to the depth of five hundred feet, and he adds that others think it may reach to six or even seven hundred feet below the surface.--Andresen, Om Klitformationen, p. 20.

Many physicists now suppose that the undulations of great bodies of water reach even deeper. But a movement of undulation is not necessarily a movement of translation, and besides, there is very frequently an undertow, which tends to carry suspended bodies out to sea as powerfully as the superficial waves to throw them on sh.o.r.e. Sand-banks sometimes recede from the coast, instead of rolling towards it. Reclus informs us that the Mauvaise, a sand-bank near the Point de Grave, on the Atlantic coast of France, has moved five miles to the west in less than a century.--Revue des Deux Mondes for December, 1862, p. 905.

The action of currents may, in some cases, have been confounded with that of the waves. Sea-currents, strong enough, possibly, to transport sand for some distance, flow far below the surface in parts of the open ocean, and in narrow straits they have great force and velocity. The divers employed at Constantinople in 1853 found in the Bosphorus, at the depth of twenty-five fathoms and at a point much exposed to the wash from Galata and Pera, a number of bronze guns supposed to have belonged to a ship-of-war blown up about a hundred and fifty years before. These guns were not covered by sand or slime, though a crust of earthy matter, an inch in thickness, adhered to their upper surfaces, and the bottom of the strait appeared to be wholly free from sediment. The current was so powerful at this depth that the divers were hardly able to stand, and a keg of nails, purposely dropped into the water, in order that its movements might serve as a guide in the search for a bag of coin accidentally lost overboard from a ship in the harbor, was rolled by the stream several hundred yards before it stopped.] and the sand thrown upon the coast in question must be derived from a narrow belt of sea. It must hence, in time, become exhausted, and the formation of new sand-banks and dunes upon the southern sh.o.r.es of the Mediterranean will cease at last for want of material. [Footnote: Few seas have thrown up so much sand as the shallow German Ocean; but there is some reason to think that the amount of this material now cast upon its northern sh.o.r.es is less than at some former periods, though no extensive series of observations on this subject has been recorded. On the Spit of Agger, at the present outlet of the Liimfjord, Andresen found the quant.i.ty during ten years, on a beach about five hundred and seventy feet broad, equal to an annual deposit of an inch and a half over the whole surface.--Om Klitformationen, p. 56. This gives seventy-one and a quarter cubic feet to the running foot--a quant.i.ty certainly much smaller than that cast up by the same sea on the sh.o.r.es of the Dano-German duchies and of Holland, and, as we have seen, scarcely one-fourth of that deposited by the Atlantic on the coast of Gascony.]

But even in the cases where the acc.u.mulations of sand in extensive deserts appear to be of marine formation, or rather aggregation, and to have been brought to their present position by upheaval, they are not wholly composed of material collected or distributed by the currents of the sea; for, in all such regions, they continue to receive some small contributions from the disintegration of the rocks which underlie, or crop out through, the superficial deposits. [Footnote: See, on this subject, an article in Aus der Natur, vol. x.x.x., p. 590.

The Florentine Frescobaldi, who visited the Sinaitic peninsula five hundred years ago, observed the powerful action of the solar heat in the disintegration of the desert rocks. "This place," says he, "was a ridge of rocks burnt to powder by the sun, and this powder is blown away from the rock by the wind and is the sand of the desert; and there be many hills which are pure bare rock, and when the sun parcheth them, the wind carries off the dust, and other sand is there none in that land,"--Viaggio, pp. 69, 70. In Arabia Petraea, when a wind, powerful enough to scour down below the ordinary surface of the desert and lay bare a fresh bed of stones, is followed by a sudden burst of sunshine, the dark agate pebbles are often cracked and broken by the heat; and this is the true explanation of the occurrence of the fragments in situations where the action of fire is not probable. If the fragments are small enough to be rolled by the winds, they are in time ground down to sand and contribute to the stock of that material which covers the face of the desert, though the sand thus formed is but an infinitesimal proportion of the whole.] In some instances, too, as in Northern Africa, additions are constantly made to the ma.s.s by the prevalence of sea-winds, which transport, or, to speak more precisely, roll the finer beach-sand to considerable distances into the interior. But this is a very slow process, and the exaggerations of travellers have diffused a vast deal of popular error on the subject.

Sands of Egypt.

In the narrow valley of the Nile--which, above its bifurcation near Cairo, is, throughout Egypt and Nubia, generally bounded by precipitous cliffs--wherever a ravine or other considerable depression occurs in the wall of rock, one sees what seems a stream of desert sand pouring down, and common observers have hence concluded that the whole valley is in danger of being buried under a stratum of infertile soil. The ancient Egyptians apprehended this, and erected walls, often of unburnt brick, across the outlet of gorges and lateral valleys, to check the flow of the sand-streams. In later ages, these walls have mostly fallen into decay, and no preventive measures against such encroachments are now resorted to. But the extent of the mischief to the soil of Egypt, and the future danger from this source, have been much overrated. The sand on the borders of the Nile is neither elevated so high by the wind, nor transported by that agency in so great ma.s.ses, as is popularly supposed; and of that which is actually lifted or rolled and finally deposited by air-currents, a considerable proportion is either calcareous, and, therefore, readily decomposable, or in the state of a very fine dust, and so, in neither case, injurious to the soil. There are, indeed, both in Africa and in Arabia, considerable tracts of fine, silicious sand, which may be carried far by high winds, but these are exceptional cases, and in general the progress of the desert sand is by a rolling motion along the surface. [Footnote: Sand heaps, three and even six hundred feet high, are indeed formed by the wind, but this is effected by driving the particles up an inclined plane, not by lifting them.

Bremontier, speaking of the sand-hills on the western coast of France, says: "The particles of sand composing them are not large enough to resist wind of a certain force, nor small enough to be taken up by it, like dust; they only roll along the surface from which they are detached, and, though moving with great velocity, they rarely rise to a greater height than three or four inches."--Memoirs sur les Dunes, Annales des Ponts et Chaussecs, 1833, ler semestre, p, 148.

Andresen says that a wind, having a velocity of forty feet per second, is strong enough to raise particles of sand as high as the face and eyes of a man, but that, in general, it rolls along the ground, and is scarcely ever thrown more than to the height of a couple of yards from the surface. Even in these cases, it is carried forward by a hopping, not a continuous, motion; for a very narrow sheet or channel of water stops the drift entirely, all the sand dropping into it until it is filled up.

Blake observes, Pacific Railroad Report, vol. v., p. 242, that the sand of the Colorado desert does not rise high in the air, but bounds along on the surface or only a few inches above it.

The character of the motion of sand drifts is well ill.u.s.trated by an interesting fact not much noticed hitherto by travellers in the East. In situations where the sand is driven through depressions in rock-beds, or over deposits of silicious pebbles, the surface of the stone is worn and smoothed much more effectually than it could be by running water, and I have picked up, in such localities, rounded, irregularly broken fragments of agate, which had received from the attrition of the sand as fine a polish as could be given them by the wheel of the lapidary.