The greater or less specific gravity of the different constituents of rock doubtless aids in separating them into distinct masses 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 masses of sand not found in place have been transported and accumulated 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 assuming 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 quantity 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, accumulation, 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 quantity 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 shores of Jutland. Laval estimates the total quantity 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 principally 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 masses 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 quantity 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 mass 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 entitled 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 assertion, 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 space 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 quantity 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 mass, 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 shores 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 shores 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 Massa 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 shore.

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 accumulated by causes which we cannot at present assign. 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 shore. 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 shores 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 shores 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 quantity 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 quantity certainly much smaller than that cast up by the same sea on the shores 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 accumulations 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. xxx., 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 mass 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 masses, 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 illustrated 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.

Very interesting observations, by Blake, on the polishing of hard stones by drifting sand will be found in the Pacific Railroad Report, vol. v., pp. 92, 230, 231. The grinding and polishing power of sand has lately received a new and most ingenious application in America. Jets of sand, and even of small particles of softer substances, thrown with a certain force, are found capable of cutting the hardest minerals and metals. A block of corundum, some inches thick, has been bored through in a few minutes by this process, and it promises to be highly useful in glass-cutting and other similar operations.] So little is it lifted, and so inconsiderable is the quantity yet remaining on the borders of Egypt, that a wall four or five feet high suffices for centuries to check its encroachments. This is obvious to the eye of every observer who prefers the true to the marvellous; but the old-world fable of the overwhelming of caravans by the fearful simoom—which even the Arabs no longer repeat, if indeed they are the authors of it—is so thoroughly rooted in the imagination of Christendom that most desert travellers, of the tourist class, think they shall disappoint the readers of their journals if they do not recount the particulars of their escape from being buried alive by a sand-storm, and the popular demand for a "sensation" must be gratified accordingly. [Footnote: Wilkinson says that, in much experience in the most sandy parts of the Libyan desert, and much inquiry of the best native sources, he never saw or heard of any instance of danger to man or beast from the mere accumulation of sand transported by the wind. Chesney's observations in Arabia, and the testimony of the Bedouins he consulted, are to the same purpose. The dangers of the simoom are of a different character, though they are certainly aggravated by the blinding effects of the light particles of dust and sand borne along by it, and by that of the inhalation of them upon the respiration. ]

Another circumstance is necessary to be considered in estimating the danger to which the arable lands of Egypt are exposed. The prevailing wind in the valley of the Nile and its borders is from the north, and it may be said without exaggeration that the north wind blows for three-quarters of the year. [Footnote: In the narrow valley of the Nile, bounded as it is, above the Delta, by high cliffs, all air-currents from the northern quarter become north winds, though of course varying in partial direction, in conformity with the sinuosities of the valley. Upon the desert plateau they incline westwards, and have already borne into the valley the sands of the eastern banks, and driven those of the western quite out of the Egyptian portion of the Nile basin.] The effect of winds blowing up the valley is to drive the sands of the desert plateau which border it, in a direction parallel with the axis of the valley, not transversely to it; and if it ran in a straight line, the north wind would carry no desert sand into it. There are, however, both curves and angles in its course, and hence, wherever its direction deviates from that of the wind, it might receive sand-drifts from the desert plain through which it runs. But, in the course of ages, the winds have, in a great measure, bared the projecting points of their ancient deposits, and no great accumulations remain in situations from which either a north or a south wind would carry them into the valley. [Footnote: These considerations apply, with equal force, to the supposed danger of the obstruction of the Suez Canal by the drifting of the desert sands. The winds across the isthmus are almost uniformly from the north, and they swept it comparatively clean of flying sands long ages since. The traces of the ancient canal between the Red Sea and the Nile are easily followed for a considerable distance from Suez. Had the drifts upon the isthmus been as formidable as some have feared and others have hoped, those traces would have been obliterated, and Lake Timsah and the Bitter Lakes filled up, many centuries ago. The few particles driven by the rare east and west winds towards the line of the canal, will easily be arrested by plantations or other simple methods, or removed by dredging. The real dangers and difficulties of this magnificent enterprise—and they have been great—consisted in the nature of the soil to be removed in order to form the line, and especially in the constantly increasing accumulation of sea-sand at the southern terminus by the tides of the Red Sea, and of sand and Nile slime at the northern, by the action of the winds and currents. Both seas are shallow for miles from the shore, and the excavation and maintenance of deep channels, and of capacious harbors with easy and secure entrances, in such localities, is doubtless one of the hardest problems offered to modern engineers for practical solution. See post, Geological Importance of Dunes, note.]

The sand let fall in Egypt by the north wind is derived, not from the desert, but from a very different source—the sea. Considerable quantities of sand are thrown up by the Mediterranean, at and between the mouths of the Nile, and indeed along almost the whole southern coast of that sea, and drifted into the interior to distances varying according to the force of the wind and the abundance and quality of the material. The sand so transported contributes to the gradual elevation of the Delta, and of the banks and bed of the river itself. But just in proportion as the bed of the stream is elevated, the height of the water in the annual inundations is increased also, and as the inclination of the channel is diminished, the rapidity of the current is checked, and the deposition of the slime it holds in suspension consequently promoted. Thus the winds and the water, moving in contrary directions, join in producing a common effect.

The sand, blown over the Delta and the cultivated land higher up the stream during the inundation, is covered or mixed with the fertile earth brought down by the river, and no serious injury is sustained from it. That spread over the same ground after the water has subsided, and during the short period when the soil is not stirred by cultivation or covered by the flood, forms a thin pellicle over the surface as far as it extends, and serves to divide and distinguish the successive layers of slime deposited by the annual inundations. The particles taken up by the wind on the sea-beach are borne onward, by a hopping motion, or rolled along the surface, until they are arrested by the temporary cessation of the wind, by vegetation, or by some other obstruction, and they may, in process of time, accumulate in large masses, under the lee of rocky projections, buildings, or other barriers which break the force of the wind.

In these facts we find an important element in the explanation of the sand drifts, which have half buried the Sphinx and so many other ancient monuments in that part of Egypt. These drifts, as I have said, are not wholly from the desert, but in largo proportion from the sea; and, as might be supposed from the distance they have travelled, they have been long in gathering. While Egypt was a great and flourishing kingdom, measures were taken to protect its territory against the encroachment of sand, whether from the desert or from the Mediterranean; but the foreign conquerors, who destroyed so many of its religious monuments, did not spare its public works, and the process of physical degradation undoubtedly began as early as the Persian invasion. The urgent necessity, which has compelled all the successive tyrannies of Egypt to keep up some of the canals and other arrangements for irrigation, was not felt with respect to the advancement of the sands; for their progress was so slow as hardly to be perceptible in the course of a single reign, and long experience has shown that, from the natural effect of the inundations, the cultivable soil of the valley is, on the whole, trenching upon the domain of the desert, not retreating before it.

The oases of the Libyan, as well as of many Asiatic deserts, have no such safeguards. The sands are fast encroaching upon them, and threaten soon to engulf them, unless man shall resort to artesian wells and plantations, or to some other efficient means of checking the advance of this formidable enemy, in time to save these islands of the waste from final destruction.

Accumulations of sand are, in certain cases, beneficial as a protection against the ravages of the sea; but, in general, the vicinity, and especially the shifting of bodies of this material, are destructive to human industry, and hence, in civilized countries, measures are taken to prevent its spread. This, however, can be done only where the population is large and enlightened, and the value of the soil, or of the artificial erections and improvements upon it, is considerable. Hence in the deserts of Africa and of Asia, and thee inhabited lands which border on them, no pains are usually taken to check the drifts, and when once the fields, the houses, the springs, or the canals of irrigation are covered or choked, the district is abandoned without a struggle, and surrendered to perpetual desolation. [Footnote: In parts of the Algerian desert, some efforts are made to retard the advance of sand dunes which threaten to overwhelm villages. "At Debila," says Laurent, "the lower parts of the lofty dunes are planted with palms, ... but they are constantly menaced with burial by the sands. The only remedy employed by the natives consists in little dry walls of crystallized gypsum, built on the crests of the dunes, together with hedges of dead palm-leaves. These defensive measures are aided by incessant labor; for every day the people take up in baskets the sand blown over to them the night before and carry it back to the other side of the dune."—Memoires sur le Sahara, p. 14.]

Sand Dunes and Sand Plains.

Two forms of sand deposit are specially important in European and American geography. The one is that of dune or shifting hillock upon the coast, the other that of barren plain in the interior. The coast-dunes are composed of sand washed up from the depths of the sea by the waves, and heaped in more or less rounded knolls and undulating ridges by the winds. The sand with which many plains are covered appears sometimes to have been deposited upon them while they were yet submerged beneath the sea, sometimes to have been drifted from the seacoast, and scattered over them by wind-currents, sometimes to have been washed upon them by running water. In these latter cases, the deposit, though in itself considerable, is comparatively narrow in extent and irregular in distribution, while, in the former, it is often evenly spread over a very wide surface. In all great bodies of either sort, the silicious grains are the principal constituent, though, when not resulting from the disintegration of silicious rock and still remaining in place, they are generally accompanied with a greater or less admixture of other mineral particles, and of animal and vegetable remains, [Footnote: Organic constituents, such as comminuted shells, and silicious and calcareous exuviae of infusorial animals and plants, are sometimes found mingled in considerable quantities with mineral sands. These are usually the remains of aquatic vegetables or animals, but not uniformly so, for the microscopic organisms, whose flinty cases enter so largely into the sand-beds of the Mark of Brandenburg, are still living and prolific in the dry earth. See Wittwer, Physikalische Geographic, p. 142. The desert on both sides of the Nile is inhabited by a land-snail—of which I have counted eighty, in estimation, on a single shrub barely a foot high—and thousands of its shells are swept along and finally buried in the drifts by every wind. Every handful of the sand contains fragments of them. Forchhammer, in Leonhard und Bronn s Jahrbuch, 1841, p. 8, says of the sand-hills of the Danish coast: "It is not rare to find, high in the knolls, marine shells, and especially those of the oyster. They are due to the oyster-eater [Haemalopus ostralegus], which carries his prey to the top of the dunes to devour it." See also Staring, De Bodem van Nederland, i., p. 821.] and they are also, usually somewhat changed in consistence by the ever-varying conditions of temperature and moisture to which they have been exposed since their deposit. Unless the proportion of these latter ingredients is so large as to create a considerable adhesiveness in the mass—in which case it can no longer properly be called sand—it is infertile, and, if not charged with water, partially agglutinated by iron, lime, or other cement, or confined by alluvion resting upon it, it is much inclined to drift, whenever, by any chance, the vegetable network which, in most cases, thinly clothes and at the same time confines it, is broken. Human industry has not only fixed the flying dunes by plantations, but, by mixing clay and other tenacious earths with the superficial stratum of extensive sand plains, and by the application of fertilizing substances, it has made them abundantly productive of vegetable life. These latter processes belong to agriculture and not to geography, and, therefore, are not embraced within the scope of the present subject. But the preliminary steps, whereby wastes of loose, drifting barren sands are transformed into wooded knolls and plains, and finally, through the accummulation of vegetable mould, into arable ground, constitute a conquest over nature which precedes agriculture—a geographical revolution—and, therefore, an account of the means by which the change has been effected belongs properly to the history of man's influence on the great features of terrestrial surface. I proceed, then, to examine the structure of dunes, and to describe the warfare man wages with the sand-hills, striving on the one hand to maintain and even extend them, as a natural barrier against encroachments of the sea, and, on the other, to check their moving and wandering propensities, and prevent them from trespassing upon the fields he has planted and the habitations in which he dwells.

COAST DUNES.

Coast dunes are oblong ridges or round hillocks, formed by the action of the wind upon sands thrown up by the waves on the low beaches of seas, and sometimes of fresh-water lakes. On most coasts, the supply of sand for the formation of dunes is derived from tidal waves. The flow of the tide is more rapid, and consequently its transporting power greater, than that of the ebb; the momentum, acquired by the heavy particles in rolling in with the water, tends to carry them even beyond the flow of the waves; and at the turn of the tide, the water is in a state of repose long enough to allow it to let fall much of the solid matter it holds in suspension. Hence, on all low, tide-washed coasts of seas with sandy bottoms, there exist several conditions favorable to the formation of sand deposits along high-water mark. [Footnote: There are various reasons why the formation of dunes is confined to low shores, and this law is so universal, that when bluffs are surmounted by them, there is always cause to suspect upheaval, or the removal of a sloping beach in front of the bluff, after the dunes were formed. Bold shores are usually without a sufficient beach for the accumulation of large deposits; they are commonly washed by a sea too deep to bring up sand from its bottom; their abrupt elevation, even if moderate in amount, would still be too great to allow ordinary winds to lift the sand above them; and their influence in deadening the wind which blows towards them would even more effectually prevent the raising of sand from the beach at their foot. Forchhammer, describing the coast of Jutland, says that, in high winds, "one can hardly stand upon the dunes, except when they are near the water line and have been cut down perpendicularly by the waves. Then the wind is little or not at all felt—a fact of experience very common on our coasts, observed on all the steep shore bluffs of 200 feet height, and, in the Faroe Islands, on precipices 2,000 feet high. In heavy gales in those islands, the cattle fly to the very edge of the cliffs for shelter, and frequently fall over. The wind, impinging against the vertical wall, creates an ascending current which shoots somewhat past the crest of the rock, and thus the observer or the animal is protected against the tempest by a barrier of air."-Leonhard und Bronn, Jahrbuch, 1841, p. 3. The calming, or rather diversion, of the wind by cliffs extends to a considerable distance in front of them, and no wind would have sufficient force to raise the sand vertically, parallel to the face of a bluff, even to the height of twenty feet.] If the land-winds are of greater frequency, duration, or strength than the sea-winds, the sands left by the retreating wave will be constantly blown back into the water; but if the prevailing air-currents are in the opposite direction, the sands will soon be carried out of the reach of the highest waves, and transported continually farther and farther into the interior of the land, unless obstructed by high grounds, vegetation, or other obstacles.

The laws which govern the formation of dunes are substantially these. We have seen that, under certain conditions, sand is accumulated above high-water mark on low sea and lake shores. So long as the sand is kept wet by the spray or by capillary attraction, it is not disturbed by air-currents, but as soon as the waves retire sufficiently to allow it to dry, it becomes the sport of the wind, and is driven up the gently sloping beach until it is arrested by stones, vegetables, or other obstructions, and thus an accumulation is formed which constitutes the foundation of a dune. However slight the elevation thus created, it serves to stop or retard the progress of the sand-grains which are driven against its shoreward face, and to protect from the further influence of the wind the particles which are borne beyond it, or rolled over its crest, and fall down behind it. If the shore above the beach line were perfectly level and straight, the grass or bushes upon it of equal height, the sand thrown up by the waves uniform in size and weight of particles as well as in distribution, and if the action of the wind were steady and regular, a continuous bank would be formed, everywhere alike in height and cross section. But no such constant conditions anywhere exist. The banks are curved, broken, unequal in elevation; they are sometimes bare, sometimes clothed with vegetables of different structure and dimensions; the sand thrown up is variable in quantity and character; and the winds are shifting, gusty, vertical, and often blowing in very narrow currents. From all these causes, instead of uniform hills, there rise irregular rows of sand-heaps, and these, as would naturally be expected, are of a pyramidal, or rather conical shape, and connected at bottom by more or less continuous ridges of the same material.

Elisee Reclus, in describing the coast dunes of Gascony, observes that when, as sometimes happens, the sands are not heaped in a continuous, irregular bulwark, but deposited in isolated hillocks, they have a tendency to assume a crescent shape, the convexity being turned seawards, or towards the direction from which the prevailing winds proceed. This fact, the geological bearing of which is obvious, is not noticed by previous French writers or even by Andresen, though a semi-lunar outline has been long generally ascribed to inland dunes. It is, however evident that such a form would naturally be produced by the action of a wind blowing long in a given direction upon a mass of loose sand with a fixed centre—such as is constituted by the shrub or stone around which the sand is first deposited—and free extremities. On a receding coast, dunes will not attain so great a height as on more secure shores, because they are undermined and carried off before they have time to reach their greatest dimensions. Hence, while at sheltered points in South-western France, there are dunes three hundred feet or more in height, those on the Frisic Islands and the exposed parts of the coast of Schleswig-Holstein range only from twenty to one hundred feet. On the western shores of Africa, it is said that they sometimes attain an elevation of six hundred feet. This is one of the very few points known to geographers where desert sands are advancing seawards, [Footnote: "On the west coast of Africa the dunes are drifting seawards, and always receiving new accessions from the Sahara. They are constantly advancing out into the sea."—Naumann, Geognosie, ii., p.1172.] and here they rise to the greatest altitude to which sand-grains can be carried by the wind. The hillocks, once deposited, are held together and kept in shape, partly by mere gravity, and partly by the slight cohesion of the lime, clay, and organic matter mixed with the sand; and it is observed that, from capillary attraction, evaporation from lower strata, and retention of rain-water, they are always moist a little below the surface. [Footnote: "Dunes are always full of water, from the action of capillary attraction. Upon the summits, one seldom needs to dig more than a foot to find the sand moist, and in the depressions, fresh water is met with near the surface."—Forchhammer, in Leonhard and Bronx, for 1841, p.5, note. On the other hand, Andresen, who has very carefully investigated this as well as all other dune phenomena, maintains that the humidity of the sand ridges cannot be derived from capillary attraction. He found by experiment that a heap of drift-sand was not moistened to a greater height than eight and a half inches, after standing with its base a whole night in water. He states the minimum of water contained by the sand of the dunes, one foot below the surface, after a long drought, at two per cent, the maximum, after a rainy month, at four per cent. At greater depths the quantity is larger. The hygroscopicity of the sand of the coast of Jutland he found to be thirty-three per cent, by measure, or 21.5 by weight. The annual precipitation on that coast is twenty-seven inches, and as the evaporation is about the same, he argues that rain-water does not penetrate far beneath the surface of the dunes, and concludes that their humidity can be explained only by evaporation from below.—Om Klitformationen, pp. 106-110. In the dunes of Algeria, water in so abundant that wells are constantly dug in them at high points on their surface. They are sunk to the depth of three or four inches only, and the water rises to the height of a metre in them.—Laurent, Memoire sur le Sahara, pp. 11, 12, 13. The same writer observes (p. 14) that the 'hollows in the dunes are planted with palms which find moisture enough a little below the surface. It would hence seem that proposal to fix the dunes which are supposed to threaten the Suez Canal, by planting the maratime pine and other trees upon them, is not altogether so absurd as it has been thought to be by some of those disinterested philanthropists of other nations who were distressed with fears that French capitalists would lose the money they had invested in that great undertaking. Ponds of water are often found in the depression between the sand-hills of the dune chains in the North American desert.]

By successive accumulations, they gradually rise to the height of thirty, fifty, sixty, or a hundred feet, and sometimes even much higher. Strong winds, instead of adding to their elevation, sweep off loose particles from their surface, and these, with others blown over or between them, build up a second row of dunes, and so on according to the character of the wind, the supply and consistence of the sand, and the face of the country. In this way is formed a belt of sand-dunes, irregularly dispersed and varying much in height and dimensions, and sometimes many miles in breadth. On the Island of Sylt, in the German Sea, where there are several rows, the width of the belt is from half a mile to a mile. There are similar ranges on the coast of Holland, exceeding two miles in breadth, while at the mouths of the Nile they form a zone not less than ten miles wide. The base of some of the dunes in the Delta of the Nile is reached by the river during the annual inundation, and the infiltration of the water, which contains lime, has converted the lower strata into a silicious limestone, or rather a calcarous sandstone, and thus afforded an opportunity of studying the structure of that rock in a locality where its origin and mode of aggregation and solidification are known.

The tide, though a usual, is by no means a necessary condition for the accumulations of sand out of which dunes are formed. The Baltic and the Mediterranean are almost tideless seas, but there are vast ranges of dunes on the Russian and Prussian coasts of the Baltic and at the mouths of the Nile and many other points on the shores of the Mediterranean. The vast shoals in the latter sea, known to the ancients as the Greater and Lesser Syrtis, are of marine origin. They are still filling up with sand, washed up from greater depths, or sometimes drifted from the coast in small quantities, and will probably be converted, at some future period, into dry land covered with sand-hills. There are also extensive ranges of dunes upon the eastern shores of the Caspian, and at the southern, or rather south-eastern, extremity of Lake Michigan. [Footnote: The careful observations of Colonel J. D. Graham, of the United States Army, show a tide of about three inches in Lake Michigan. See "A Lunar Tidal Wave in the North American Lakes," demonstrated by Lieut.-Colonel J. D. Graham, in the fourteenth volume of the Proceedings of the American Association for the Advancement of Science.] There is no doubt that this latter lake formerly extended much farther in that direction, but its southern portion has gradually shoaled and at last been converted into solid land, in consequence of the prevalence of the north-west winds. These blow over the lake a large part of the year, and create a southwardly set of the currents, which wash up sand from the bed of the lake and throw it on shore. Sand is taken up from the beach at Michigan City by every wind from that quarter, and, after a heavy blow of some hours' duration, sand ridges may be observed on the north side of the fences, like the snow wreaths deposited by a drifting wind in winter. Some of the particles are carried back by contrary winds, but most of them lodge on or behind the dunes, or in the moist soil near the lake, or are entangled by vegetables, and tend permanently to elevate the level. Like effects are produced by constant sea-winds, and dunes will generally be formed on all low coasts where such prevail, whether in tideless or in tidal waters.

Jobard thus describes the modus operandi, under ordinary circumstances, at the mouths of the Nile, where a tide can scarcely be detected: "When a wave breaks, it deposits an almost imperceptible line of fine sand. The next wave brings also its contribution, and shoves the preceding line a little higher. As soon as the particles are fairly out of the reach of the water they are dried by the heat of the burning sun, and immediately seized by the wind and rolled or borne farther inland. The gravel is not thrown out by the waves, but rolls backwards and forwards until it is worn down to the state of fine sand, when it, in its turn, is cast upon the land and taken up by the wind." [Footnote: Staring, De Bodun van Nederland, i., p. 327, note.] This description applies only to the common every-day action of wind and water; but just in proportion to the increasing force of the wind and the waves, there is an increase in the quantity of sand, and in the magnitude of the particles carried off from the beach by it, and, of course, every storm in a landward direction adds sensibly to the accumulation upon the shore.

Sand Banks.

Although dunes, properly so called, are found only on dry land and above ordinary high-water mark, and owe their elevation and structure to the action of the wind, yet, upon many shelving coasts, accumulations of sand much resembling dunes are formed under water at some distance from the shore by the oscillations of the waves, and are well known by the name of sand banks. They are usually rather ridges than banks, of moderate inclination, and with the steepest slope seawards, [Footnote: Kohl, Inseln und Marschen Schleswig Holsteins, ii., p. 33. From a drawing in Andresen, Om Klitformationen, p. 24, it would appear that on the Schleswig coast the surf-formed banks have the steepest slope landwards, those farther from the shore, as stated in the text.] and their form differs little from that of dunes except in this last particular and in being lower and more continuous. Upon the western coast of the island of Amrum, for example, there are three rows of such banks, the summits of which are at a distance of perhaps a couple of miles from each other; so that, including the width of the banks themselves, the spaces between them, and the breadth of the zone of dunes upon the land, the belt of moving sands on that coast is probably not less than eight miles wide.

Under ordinary circumstances, sand banks are always rolling, landwards, and they compose the magazine from which the material for the dunes is derived. [Footnote: Sand banks sometimes connect themselves with the coast at both ends, and thus cut off a portion of the sea. In this case, as well as when salt water is enclosed by sea-dikes, the water thus separated from the ocean gradually becomes fresh, or at least brackish. The Haffs, or large expanses of fresh water in Eastern Prussia—which are divided from the Baltic by narrow sand banks called Nehrungen, or, at sheltered points of the coast, by fluviatile deposits called Werders—all have one or more open passages, through which the water of the rivers that supply them at last finds its way to the sea.] The dunes, in fact, are but aquatic sand banks transferred to dry land. The laws of their formation are closely analogous, because the action of the two fluids, by which they are respectively accumulated and built up, is very similar when brought to bear upon loose particles of solid matter. It would, indeed, seem that the slow and comparatively regular movements of the heavy, unelastic water ought to affect such particles very differently from the sudden and fitful impulses of the light and elastic air. But the velocity of the wind currents gives them a mechanical force approximating to that of the slower waves, and, however difficult it may be to explain all the phenomena that characterize the structure of the dunes, observation has proved that it is nearly identical with that of submerged sand banks. [Footnote: Forchhammer ascribes the resemblance between the furrowing of the dune sands and the beach ripples, not to the similarity of the effect of wind and water upon sand, but wholly to the action of the wind; in the first instance, directly, in the latter, through the water. "The wind-ripples on the surface of the dunes precisely resemble the water-ripples of sand flats occasionally overflowed by the sea; and with the closest scrutiny, I have never been able to detect the slightest difference between them. This is easily explained by the fact, that the water-ripples are produced by the action of light wind on the water which only transmits the air-waves to the sand."—Leonhard und Bronn, 1841, pp. 7, 8.] The differences of form are generally ascribable to the greater number and variety of surface accidents of the ground on which the sand hills of the land are built up, and to the more frequent changes, and wider variety of direction, in the courses of the wind.

CHARACTER OF DUNE SAND.

"Dune sand," says Staring, "consists of well-rounded grains of quartz, more or less colored by iron, and often mingled with fragments of shells, small indeed, but still visible to the naked eye. [Footnote: According to the French authorities, the dunes of France are not always composed of quartzose sand. "The dune sands" of different characters, says Bremontier, "partake of the nature of the different materials which compose them. At certain points on the coast of Normandy they are found to be purely calcareous; they are of mixed composition on the shores of Brittany and Saintonge, and generally quartzose between the mouth of the Gironde and that of the Adour."—Memoire sur les Dunes, Annales des Ponts et Chaussees, t. vii., 1833, 1er semestre, p. 146.

In the dunes of Long Island and of Jutland, there are considerable veins composed almost wholly of garnet. For a very full examination of the mechanical and chemical composition of the dune sands of Jutland, see Andresen, Om Klitformationen, p. 110. Fraas informs us, Aus dem Orient, pp. 176, 177, that the dune sands of the Egyptian coast arise from the disintegration of the calcareous sandstone of the same region. This sandstone, composed in a large proportion of detritus of both land and sea shells mingled with quartz sand, appears to have been consolidated under water during an ancient period of subsidence. A later upheaval brought it to or near the surface, when it was more or less disintegrated by the action of the waves and by meteoric influences—a process still going on—and it is now again subsiding with the coast it rests on.

The calcareous sand arising from the comminution of corals forms dunes on some of the West India Islands.—Agassiz, Bulletin of the Museum of Comparative Zoology, vol. i.] These fragments are not constant constituents of dune sand. They are sometimes found at the very summits of the hillocks, as at Overveen; in the King's Dune, near Egmond, they form a coarse, calcareous gravel very largely distributed through the sand, while the interior dunes between Haarlem and Warmond exhibit no trace of them. It is yet undecided whether the presence or absence of these fragments is determined by the period of the formation of the dunes, or whether it depends on a difference in the process by which different dunes have been accumulated. Land shells, such as snails, for example, are found on the surface of the dunes in abundance, and many of the shelly fragments in the interior of the hillocks may be derived from the same source." [Footnote: De Bodem van Nederland, i., p. 323.]

Sand concretions form within the dunes and especially in the depressions between them. These are sometimes so extensive and impervious as to retain a sufficient supply of water to feed perennial springs, and to form small permanent ponds, and they are a great impediment to the penetration of roots, and consequently to the growth of trees planted, or germinating from self-sown seeds, upon the dunes. [Footnote: Staring, De Bodem van Nederland, i., p.317. See also Bergsoe, Reventlov's Virksomhed, ii., p. 11.

"In the sand-hill ponds mentioned in the text, there is a vigourous growth of bog plants accompanied with the formation of peat, which goes on regularly as long as the dune sand does not drift. But if the surface of the dunes is broken, the sand blows into the ponds, covers the peat, and puts and end to its formation. When, in the course of time, marine currents cut away the coast, the dunes move landwards and fill up the ponds and thus are formed the remarkable strata of fossile peat called Martorv, which appears to be unknown to the geologists of other parts of Europe." — Forchhammer, in Leonhard und Bronn, 1841, p. 18. Martorv has a specific gravity thrice as great as that of ordinary peat in consequence of the pressure of the sand.—Asbjornsen, Torv og Torvdrift, p.26.]

Interior Structure of Dunes.

The interior structure of the dunes, the arrangement of their particles, is not, as might be expected, that of an unorganized, confused heap, but they show a strong tendency to stratification. This is a point of much geological interest, because it indicates that sandstone may owe its stratified character to the action of other forces as well as of water. The origin and peculiar character of these layers are due to a variety of causes.

For example, a south-west wind and current may deposit upon a dune a stratum of a given color and mineral composition, and this may be succeeded by a north-west wind and current, bringing with them particles of a different hue, constitution, and origin. Again, if we suppose a violent tempest to strew the beach with sand-grains very different in magnitude and specific gravity, and, after the sand is dry, to be succeeded by a gentle breeze, it is evident that only the lighter particles will be taken up and carried to the dunes. If, after some time, the wind freshens, heavier grains will be transported and deposited on the former, and a still stronger succeeding gale will roll up yet larger kernels. Each of these deposits will form a stratum. If we suppose the tempest to be followed, after the sand is dry, not by a gentle breeze, but by a wind powerful enough to lift at the same time particles of very various magnitudes and weights, the heaviest will often lodge on the dune while the lighter will be carried farther. This would produce a stratum of coarse sand, and the same effect might result from the blowing away of light particles out of a mixed layer, while the heavier remained undisturbed. [Footnote: The lower strata must be older than the superficial layers, and the particles which compose them may in time become more disintegrated, and therefore finer than those deposited later and above them.

Hull ingeniously suggests that, besides other changes, fine sand intermixed with or deposited above a coarser stratum, as well as the minute particles resulting from the disintegration of the grains of the latter, may be carried by rain in the case of dunes, or by the ordinary action or sea-water in that of sand-banks, down through the interstices in the coarser layer, and thus the relative position of sand and gravel may be changed.—Oorsprong der Hollandsche Duinen, p. 103.]

Still another cause of apparent stratification may be found in the occasional interposition of a thin layer of leaves or other vegetable remains between successive deposits, and this I imagine to be more frequent than has been generally supposed. Some geologists have thought that the sand strata of dunes are of annual formation; [Footnote: Schomann, Geologische Wanderungen durch die Preussischen Ost-See Provinzen, 1869, p. 81.] but the autumnal deposit of foliage from neighboring trees and shrubs furnishes a more probable explanation of the division of the sand-heaps into regular layers.

A late distinguished American admiral communicated to me an interesting observation made by him at San Francisco, which has an important bearing on the arrangement of the particles of sand in dunes and other irregular accumulations of that substance. In laying out a navy-yard at that port, a large quantity of earthy material was removed from the dunes and other hillocks and carted to a low piece of ground which required filling up. Sand of various characters, fine and coarse gravel, and common earth were dropped promiscuously by the carts as accident or convenience dictated, and of course they were all confusedly intermixed. Some time after, when the new ground was consolidated, various excavations were made in it, and the different materials of which the filling was composed were found to be stratified with considerable regularity, according to their specific gravity.

Two explanations of this remarkable fact suggest themselves to me, which, however, do not perhaps exclude others. San Francisco is subject to earthquakes, and though violent or even sensible shocks are not very frequent, it is highly probable that, as is shown to be the case in many other countries, by late seismological observations, there are, in the course of the year, a great number of slight shocks which escape unscientific observation. A frequent repetition of slight tremblings of the earth would, like any other moderate mechanical agitation, probably produce the separation of a miscellaneous mass, like that described, into distinct layers. Again, the Pacific coast, like all others upon an open sea, is exposed to incessant concussion from the shock of the waves, which is repeated many thousand times a day. This concussion is often sensibly felt by the observer, and it seems not in the least improbable that the agitation may have tended to produce a stratified arrangement in the case at San Francisco, as well as in all coast dunes and other accumulations of loose mineral material in similar situations. Kohl observes that the shore on the landward side of the files of dunes often trembles from the shock of the waves on the beach, [Footnote: Inseln und Marschen, etc., ii., p. 34.] and Villeneuve established by careful experiment that at Dunkerque the ground is sensibly agitated by the same cause, in stormy weather, to a distance of more than a mile from the sea.

The eddies of strong winds between the hillocks must also occasion disturbances and re-arrangements of the sand layers, and it seems possible that the irregular thickness and the strange contortions of the strata of the sandstone at Petra may be due to some such cause. A curious observation of Professor Forchhammer suggests an explanation of another peculiarity in the structure of the sandstone of Mount Seir. He describes dunes in Jutland, composed of yellow quartzose sand intermixed with black titanian iron. When the wind blows over the surface of the dunes, it furrows the sand with alternate ridges and depressions, ripples, in short, like those of water. The swells, the dividing ridges of the system of sand ripples, are composed of the light grains of quartz, while the heavier iron rolls into the depressions between, and thus the whole surface of the dune appears as if covered with a fine black network.

The sea side of dunes, being more exposed to the caprices of the wind, is more irregular in form than the lee or land side, where the arrangement of the particles is affected by fewer disturbing and conflicting influences. Hence, the stratification of the windward slope is somewhat confused, while the sand on the lee side is found to be disposed in more regular beds, inclining landwards, and with the largest particles lowest, where their greater weight would naturally carry them. The lee side of the dunes, being thus formed of sand deposited according to the laws of gravity, is very uniform in its slope, which, according to Forchhammer, varies little from an angle of 30 degrees with the horizon, while the more exposed and irregular weather side lies at an inclination of from 5 degrees to 10 degrees. When, however, the outer tier of dunes is formed so near the waterline as to be exposed to the immediate action of the waves, it is undermined, and the face of the hill is very steep and sometimes nearly perpendicular.

Geological Importance of Dunes.

These observations, and other facts which a more attentive study on the spot would detect, might furnish the means of determining interesting and important questions concerning geological formations in localities very unlike those where dunes are now thrown up. For example, Studer supposes that the drifting sand-hills of the African desert were originally coast dunes, and that they have been transported to their present position far in the interior, by the rolling and shifting leeward movement to which all dunes not covered with vegetation are subject. The present general drift of the sands of that desert appears to be to the south-west and west, the prevailing winds blowing from the north-east and east; but it has been doubted whether the shoals of the western coast of Northern Africa, and the sands upon that shore, are derived from the bottom of the Atlantic, in the usual manner, or, by an inverse process, from those of the Sahara. The latter, as has been before remarked, is probably the truth, though observations are wanting to decide the question. [Footnote: "The North African desert falls into two divisions: the Sahel, or western, and the Sahar, or eastern. The sands of the Sahar were, at a remote period, drifted to the west. In the Sahel, the prevailing east winds drive the sand-ocean with a progressive westward motion. The eastern half of the desert is swept clean."—Naumann, Geognosie, ii., p. 1173.] There would be nothing violently improbable in the a priori supposition that they may have been in part first thrown up by the Mediterranean on its Libyan coast, and thence blown south and west over the vast space they now cover. But inasmuch as it is now geologically certain that the Sahara is an uplifted bed of an ancient sea, we may suppose that, while submerged, it was, like other sea-bottoms, strewn with sand, and that its present supply of that material was, in great proportion, brought up with it. Laurent observed, some years ago, that marine shells of still extinct species were found in the Sahara, far from the sea, and even at considerable depths below the surface. [Footnote: Memoires sur le Sahara Oriental p. 62] These observations have been confirmed past all question by Desor, Martins, and others, and the facts and the obvious conclusion they suggest are at present not disputed. But whatever has been the source and movement of these sands, they can hardly fail to have left on their route some sandstone monuments to mark their progress, such, for example, as we have seen are formed from the dune sand at the mouth of the Nile; and it is conceivable that the character of the drifting sands themselves, and of the conglomerates and sandstones to whose formation they have contributed, might furnish satisfactory evidence as to their origin, their starting-point, and the course by which they have wandered so far from the sea. [Footnote: Forchhammer, after pointing out the coincidence between the inclined stratification of dunes and the structure of ancient tilted rocks, says: "But I am not able to point out a sandstone formation corresponding to the dunes. Probably most ancient dunes have been destroyed by submersion before the loose sand became cemented to solid stone, but we may suppose that circumstances have existed somewhere which have preserved the characteristics of this formation."—Leonhard und Bronn, 1841, p. 8, 9. Such formations, however, certainly exist. Laurent (Memoire sur le Sahara, etc., p. 12) tells us that in the Algerian desert there are "sandstone formation" not only "corresponding to the dunes," but, actually consolidated within them. "A place called El-Mouia-Tadjer presents a repetition of what we saw at El-Baya; one of the funnels formed in the middle of the dunes contains wells from two metres to two and a half in depth, dug in a sand which pressure, and probably the presence of certain salts, have cemented so as to form true sandstone, soft indeed, but which does not yield except to the pickaxe. These sandstones exhibit an inclination which seems to be the effect of wind; for they conform to the direction of the sands which roll down a scarp occasioned by the primitive obstacle."

"At New Quay the dune sands are converted to stone by an oxide of iron held in solution by the water which pervades them. This stone, which is formed, so to speak, under our eye, has been found solid enough to be employed for building."-Esquiros, L'Angleterre, etc., in Revue des Deux Mondes, 1864, pp. 44, 45.

The dunes near the mouth of the Nile, the lower sands of which have been cemented together by the infiltration of Nile water, would probably show a similar stratification in the sandstone which now forms their base.

Dana describes a laminated rook often formed by the infiltration of water into the sand dunes on the Hawaian islands.—Corals and Coral Islands, 1872, p.155.]

If the sand of coast dunes is, as Staring describes it, composed chiefly of well-rounded, quartzose grains, fragments of shells, and other constant ingredients, it would often be recognizable as coast sand, in its agglutinate state of sandstone. The texture of this rock varies from an almost imperceptible fineness of grain to great coarseness, and affords good facilities for microscopic observation of its structure. There are sandstones, such, for example, as are used for grindstones, where the grit, as it is called, is of exceeding sharpness; others where the angles of the grains are so obtuse that they scarcely act at all on hard metals. The former may be composed of grains of rock, disintegrated indeed, and re-cemented together, but not, in the meanwhile, much rolled; the latter, of sands long washed by the sea, and drifted by land-winds. There is, indeed, so much resemblance between the effects of driving winds and of rolling water upon light bodies, that there might be difficulty in distinguishing them; but after all, it is not probable that sandstone, composed of grains thrown up from the salt sea, and long tossed by the winds, would be identical in its structure with that formed from fragments of rock crushed by mechanical force, or disintegrated by heat, and again agglutinated without much exposure to the action of moving water.

Dunes of American Coasts.

Upon the Atlantic coast of the United States, the prevalence of western or off-shore winds is unfavorable to the formation of dunes, and, though marine currents lodge vast quantities of sand, in the form of banks, on that coast, its shores are proportionally more free from sand-hills than some others of lesser extent. There are, however, very important exceptions. The action of the tide throws much sand upon some points of the New England coast, as well as upon the Beaches of Long Island and other more southern shores, and here dunes resembling those of Europe are formed. There are also extensive ranges of dunes on the Pacific coast of the United States, and at San Francisco they border some of the streets of the city.

The dunes of America are far older than her civilization, and the soil they threaten or protect possesses, in general, too little value to justify any great expenditure in measures for arresting their progress or preventing their destruction. Hence, great as is their extent and their geographical importance, they have, at present, no such intimate relations to human life as to render them objects of special interest in the point of view I am taking, and I do not know that the laws of their formation and motion have been made a subject of original investigation by any American observer.

Dunes of Western Europe.

Upon the western coast of Europe, on the contrary, the ravages occasioned by the movement of sand dunes, and the serious consequences often resulting from the destruction of them, have long engaged the earnest attention of Governments and of scientific men, and for nearly a century persevering and systematic effort has been made to bring them under human control. The subject has been carefully studied in Denmark and the adjacent duchies, in Western Prussia, in the Netherlands, and in France; and the experiments in the way of arresting the drifting of the dunes, and of securing them, and the lands they shelter, from the encroachments of the sea, have resulted in the adoption of a system of coast improvement substantially the same in all these countries. The sands, like the forests, have now their special literature, and the volumes and memoirs, which describe them and the processes employed to subdue them, are full of scientific interest and of practical instruction.

Dunes of Gascony.

In the small kingdom of Denmark, inclusive of the duchies of Schleswig and Holstein, the dunes cover an area of more than two hundred and sixty square miles. The breadth of the chain is very various, and in some places it consists only of a single row of sand-hills, while in others, it is more than six miles wide. [Footnote: Andersen, Om Klitformationen, pp. 78, 202, 275.] The dunes of the Prussian coast are vaguely estimated to cover from eighty-five to one hundred and ten thousand acres; those of Holland one hundred and forty thousand acres; and those of Gascony more than two hundred thousand acres. I do not find any estimate of their extent in other provinces of France, or in the Baltic provinces of Russia, but it is probable that the entire quantity of dune land upon the Atlantic and Baltic shores of Europe does not fall much short of a million of acres. [Footnote: In an article on the dunes of Europe, in vol. 29 (1864) of Aus der Natur, p. 590, the dunes are estimated to cover, on the islands and coasts of Schleswig Holstein, in North-west Germany, Denmark, Holland, and France, one hundred and eighty-one German, or nearly four thousand English square miles; in Scotland, about ten German, or two hundred and ten English miles; in Ireland, twenty German, or four hundred and twenty English miles; and in England, one hundred and twenty German, or more than twenty-five hundred English miles. Pannewitz (Anleitung zum Anbau der Sandfluchen), as cited by Andresen (Om Klitformationen, p. 45), states that the drifting sands of Europe, including, of course, sand plains as well as dunes, cover an extent of 21,000 square miles. This is, perhaps, an exaggeration, though there is, undoubtedly, much more desert-land of this description on the European continent than has been generally supposed. There is no question that most of this waste is capable of reclamation by simple planting, and no mode of physical improvement is better worth the attention of civilized Governments than this.

There are often serious objections to extensive forest planting on soils capable of being otherwise made productive, but they do not apply to sand wastes, which, until covered by woods, are not only a useless incumbrance, but a source of serious danger to all human improvements in the neighborhood of them.] This vast deposit of sea-sand extends along the coasts for a distance of several hundred miles, and from the time of the destruction of the forests which covered it, to the year 1789, the whole line was rolling inwards and burying the soil beneath it, or rendering the fields unproductive by the sand which drifted from it. At the same time, as the sand-hills moved landwards, the ocean was closely following their retreat and swallowing up the ground they had covered, as fast as their movement left it bare.

Age, Character, and Permanence of Dunes.

The origin of most great lines of dunes goes back past all history. There are on many coasts several distinct ranges of sand-hills which seem to be of very different ages, and to have been formed under different relative conditions of land and water. [Footnote: Krause, speaking of the dunes on the coast of Prussia, says: "Their origin belongs to three different periods, in which important changes in the relative level of sea and land have unquestionably taken place.... Except in the deep depressions between them, the dunes are everywhere sprinkled, to a considerable height, with brown oxydulated iron, which has penetrated into the sand to the depth of from three to eighteen inches, and colored it red. ... Above the iron is a stratum of sand differing in composition from ordinary sea-sand, and on this, growing woods are always found.... The gradually accumulated forest soil occurs in beds of from one to three feet thick, and changes, proceeding upward, from gray sand to black humus."

Even on the third or seaward range, the sand grasses appear and thrive luxuriantly, at least on the west coast, though Krause doubts whether the dunes of the east coast were ever thus protected.—Der Dunenbau, pp. 8, 11.] In some cases there has been an upheaval of the coast line since the formation of the oldest hillocks, and these have become inland dunes, while younger rows have been thrown up on the new beach laid bare by elevation of the sea-bed. Our knowledge of the mode of their first accumulation is derived from observation of the action of wind and water in the few instances where, with or without the aid of man, new coast dunes have been accumulated, and of the influence of wind alone in elevating new sand-heaps inland of the coast tier, when the outer rows are destroyed by the sea, as also when the sodded surface of ancient sands has been broken, and the subjacent strata laid open to the air.

It is a question of much interest, in what degree the naked condition of most dunes is to be ascribed to the improvidence and indiscretion of man. There are, in Western France, extensive ranges of dunes covered with ancient and dense forests, while the recently formed sand-hills between them and the sea are bare of vegetation, and in some cases are rapidly advancing upon the wooded dunes, which they threaten to bury beneath their drifts. Between the old dunes and the new there is no discoverable difference in material or in structure; but the modern sand-hills are naked and shifting, the ancient, clothed with vegetation and fixed. It has been conjectured that artificial methods of confinement and plantation were employed by the primitive inhabitants of Gaul; and Laval, basing his calculations on the rate of annual movement of the shifting dunes, assigns the fifth century of the Christian era as the period when those processes wore abandoned. [Footnote: Laval, Memoire sur les Dunes de Gascogne, Annales des Ponts et Chaussees, 1847, 2me semestre, p. 231. The same opinion had been expressed by Bremontier, Annales des Ponts et Chaussees, 1833, 1er semestre, p. 185.]

There is no historical evidence that the Gauls were acquainted with artificial methods of fixing the sands of the coast, and we have little reason to suppose that they were advanced enough in civilization to be likely to resort to such processes, especially at a period when land could have had but a moderate value.

In other countries, dunes have spontaneously clothed themselves with forests, and the rapidity with which their surface is covered by various species of sand-plants, and finally by trees, where man and cattle and burrowing animals are excluded from them, renders it highly probable that they would, as a general rule, protect themselves, if left to the undisturbed action of natural causes. The sand-hills of the Frische Nehrung, on the coast of Prussia, were formerly wooded down to the water's edge, and it was only in the last century that, in consequence of the destruction of their forests, they became moving sands. [Footnote: "In the Middle Ages," says Willibald Alexis, as quoted by Muller, Das Buch der Pflanzenwelt, i., p. 16, "the Nebrung was extending itself further, and the narrow opening near Lochstadt had filled itself up with sand. A great pine forest bound with its roots the dune sand and the heath uninterruptedly from Danzig to Pillau. King Frederick William I. was once in want of money. A certain Herr von Korff promised to procure it for him, without loan or taxes, if he could be allowed to remove something quite useless. He thinned out the forests of Prussia, which then, indeed, possessed little pecuniary value; but he felled the entire woods of the Frische Nebrung, so far as they lay within the Prussian territory. The financial operation was a success. The king had money, but in the material effects which resulted from it, the state received irreparable injury. The sea-winds rush over the bared hills; the Frische Haff is half-choked with sand; the channel between Elbing, the sea, and Konigsberg is endangered, and the fisheries in the Haff injured. The operation of Herr von Korff brought the king 200,000 thalers. The state would now willingly expend millions to restore the forests again."] There is every reason to believe that the dunes of the Netherlands were clothed with trees until after the Roman invasion. The old geographers, in describing these countries, speak of vast forests extending to the very brink of the sea; but drifting coast dunes are first mentioned by the chroniclers of the Middle Ages, and so far as we know they have assumed a destructive character in consequence of the improvidence of man. [Footnote: Staring, Voormaals en Thans, p. 231. Had the dunes of the Netherlandish and French coasts, at the period of the Roman invasion, resembled the moving sand-hills of the present day, it is inconceivable that they could have escaped the notice of so acute a physical geographer as Strabo; and the absolute silence of Caesar, Ptolemy, and the encyclopaedic Pliny, respecting them, would be not less inexplicable.] The history of the dunes of Michigan, so far as I have been able to learn from my own observation, or that of others, is the same. Thirty years ago, when that region was scarcely inhabited, they were generally covered with a thick growth of trees, chiefly pines, and underwood, and there was little appearance of undermining and wash on the lake side, or of shifting of the sands, except where the trees had been cut or turned up by the roots. [Footnote: The sands of Cape Cod were partially, if not completely, covered with vegetation by nature. Dr. Dwight, describing the dunes as they were in 1800, says: "Some of them are covered with beach grass; some fringed with whortleberry bushes; and some tufted with a small and singular growth of oaks. ... The parts of this barrier which are covered with whortleberry bushes and with oaks, have been either not at all or very little blown. The oaks, particularly, appear to be the continuation of the forests originally formed on this spot. ... They wore all the marks of extreme age; were, in some instances, already decayed, and in others decaying; were hoary with moss and were deformed by branches, broken and wasted, not by violence, but by time."—Travels, iii., p. 91]

Nature, as she builds up dunes for the protection of the seashore, provides, with similar conservatism, for the preservation of the dunes themselves; so that, without the interference of man, these hillocks would be, not perhaps absolutely perpetual, but very lasting in duration, and very slowly altered in form or position. When once covered with the trees, shrubs, and herbaceous growths adapted to such localities, dunes undergo no apparent change, except the slow occasional undermining of the outer tier, and accidental destruction by the exposure of the interior, from the burrowing of animals, or the upturning of trees with their roots, and all these causes of displacement are very much less destructive when a vegetable covering exists in the immediate neighborhood of the breach.

Protection of Dunes.

Before the occupation of the coasts by man, dunes, at all points where they have been observed, seem to have been protected in their rear by forests, which served to break the force of the winds in both directions, [Footnote: Bergsoe (Reventlovs Virksomhed, ii., 3) states that the dunes on the west coast of Jutland were stationary before the destruction of the forests to the east of them. The felling of the tall trees removed the resistance to the lower currents of the westerly winds, and the sands have since buried a great extent of fertile soil. See also same work, ii., p. 124.] and to have spontaneously clothed themselves with a dense growth of the various plants, grasses, shrubs, and trees, which nature has assigned to such soils. It is observed in Europe that dunes, though now without the shelter of a forest country behind them, begin to protect themselves as soon as human trespassers are excluded, and grazing animals denied access to them. Herbaceous and arborescent plants spring up almost at once, first in the depressions, and then upon the surface of the sand-hills. Every seed that sprouts, binds together a certain amount of sand by its roots, shades a little ground with its leaves, and furnishes food and shelter for still younger or smaller growths. A succession of a very few favorable seasons suffices to bind the whole surface together with a vegetable network, and the power of resistance possessed by the dunes themselves, and the protection they afford to the fields behind them, are just in proportion to the abundance and density of the plants they support.

The growth of the vegetable covering can, of course, be much accelerated by judicious planting and watchful care, and this species of improvement is now carried on upon a vast scale on the sandy coasts of Western Europe, wherever the value of land is considerable and the population dense.

Use of Dunes as a Barrier against the Sea.

Although the sea throws up large quantities of sand on flat lee-shores, there are many cases where it continually encroaches on those same shores and washes them away. At all points of the shallow North Sea where the agitation of the waves extends to the bottom, banks are forming and rolling eastwards. Hence the sea-sand tends to accumulate upon the coast of Schleswig-Holstein and Jutland, and were there no conflicting influences, the shore would rapidly extend itself westwards. But the same waves which wash the sand to the coast undermine the beach they cover, and still more rapidly degrade the shore at points where it is too high to receive partial protection by the formation of dunes upon it. The earth of the coast is generally composed of particles finer, lighter, and more transportable by water than the sea-sand. While, therefore, the billows raised by a heavy west wind may roll up and deposit along the beach thousands of tons of sand, the same waves may swallow up even a larger quantity of fine shore-earth. This earth, with a portion of the sand, is swept off by northwardly and southwardly currents, and let fall at other points of the coast, or carried off, altogether, out of the reach of causes which might bring it back to its former position.

Although, then, the eastern shore of the German Ocean here and there advances into the sea, it in general retreats before it, and but for the protection afforded it by natural arrangements seconded by the art and industry of man, whole provinces would soon be engulfed by the waters. This protection consists in an almost unbroken chain of sand banks and dunes, extending from the northernmost point of Jutland to the Elbe, a distance of not much less than three hundred miles, and from the Elbe again, though with more frequent and wider interruptions, to the Atlantic borders of France and Spain. So long as the dunes are maintained by nature or by human art, they serve, like any other embankment or dike, as a partial or a complete protection against the encroachments of the sea; and on the other hand, when their drifts are not checked by natural processes, or by the industry of man, they become a cause of as certain, if not of as sudden, destruction as the ocean itself whose advance they retard. On the whole, the dunes on the coast of the German Sea, notwithstanding the great quantity of often fertile land they cover, and the evils which result from their movement, are a protective and beneficial agent, and their maintenance is an object of solicitude with the Governments and people of the shores they defend. [Footnote: "We must, therefore, not be surprised to see the people here deal as gingerly with their dunes as if treading among eggs. He who is lucky enough to own a molehill of dune pets it affectionately, and spends his substance in cherishing and fattening it. That fair, fertile, rich province, the peninsula of Eiderstadt in the south of Friesland, has, on the point towards the sea, only a tiny row of dunes, some six miles long or so; but the people talk of their fringe of sand hills, as if it were a border set with pearls. They look upon it as their best defence against Neptune. They have connected it with their system of dikes, and for years have kept sentries posted to protect it against wanton injury."—J. G. Kohl, Die Inseln u. Marschen Schleswig-Holsteins, ii., p. 115.]