[317] The dikes are sometimes founded upon piles, and sometimes protected by one or more rows of piles driven deeply down into the bed of the sea in front of them. "Triple rows of piles of Scandinavian pine," says Wild, "have been driven down along the coast of Friesland, where there are no dunes, for a distance of one hundred and fifty miles. The piles are bound together by strong cross timbers and iron clamps, and the interstices filled with stones. The ground adjacent to the piling is secured with fascines, and at exposed points heavy blocks of stone are heaped up as an additional protection. The earth dike is built behind the mighty bulwark of this breakwater, and its foot also is fortified with stones." * * * "The great Helder dike is about five miles long and forty feet wide at the top, along which runs a good road. It slopes down two hundred feet into the sea, at an angle of forty degrees. The highest waves do not reach the summit, the lowest always cover its base. At certain distances, immense buttresses, of a height and width proportioned to those of the dike, and even more strongly built, run several hundred feet out into the rolling sea. This gigantic artificial coast is entirely composed of Norwegian granite."—Wild, Die Niederlande, i, pp. 61, 62.

[318] The shaking of the ground, even when loaded with large buildings, by the passage of heavy carriages or artillery, or by the march of a body of cavalry or even infantry, shows that such causes may produce important mechanical effects on the condition of the soil. The bogs in the Netherlands, as in most other countries, contain large numbers of fallen trees, buried to a certain depth by earth and vegetable mould. When the bogs are dry enough to serve as pastures, it is observed that trunks of these ancient trees rise of themselves to the surface. Staring ascribes this singular phenomenon to the agitation of the ground by the tread of cattle. "When roadbeds," observes he, "are constructed of gravel and pebbles of different sizes, and these latter are placed at the bottom without being broken and rolled hard together, they are soon brought to the top by the effect of travel on the road. Lying loosely, they undergo some motion from the passage of every wagon wheel and the tread of every horse that passes over them. This motion is an oscillation or partial rolling, and as one side of a pebble is raised, a little fine sand or earth is forced under it, and the frequent repetition of this process by cattle or carriages moving in opposite directions brings it at last to the surface. We may suppose that a similar effect is produced on the stems of trees in the bogs by the tread of animals."—De Bodem van Nederland, i, pp. 75, 76.

It is observed in the Northern United States, that when soils containing pebbles are cleared and cultivated, and the stones removed from the surface, new pebbles, and even bowlders of many pounds weight, continue to show themselves above the ground, every spring, for a long series of years. In clayey soils the fence posts are thrown up in a similar way, and it is not uncommon to see the lower rail of a fence thus gradually raised a foot or even two feet above the ground. This rising of stones and fences is popularly ascribed to the action of the severe frosts of that climate. The expansion of the ground, in freezing, it is said, raises its surface, and, with the surface, objects lying near or connected with it. When the soil thaws in the spring, it settles back again to its former level, while the pebbles and posts are prevented from sinking as low as before by loose earth which has fallen under them. The fact that the elevation spoken of is observed only in the spring, gives countenance to this theory, which is perhaps applicable also to the cases stated by Staring, and it is probable that the two causes above assigned concur in producing the effect.

The question of the subsidence of the Netherlandish coast has been much discussed. Not to mention earlier geologists, Venema, in several essays, and particularly in Het Dalen van de Noordelijke Kuststreken van ons Land, 1854, adduces many facts and arguments to prove a slow sinking of the northern provinces of Holland. Laveleye (Affaissement du sol et envasement des fleuves survenus dans les temps historiques, 1859), upon a still fuller investigation, arrives at the same conclusion. The eminent geologist Staring, however, who briefly refers to the subject in De Bodem van Nederland, i, p. 356 et seqq., does not consider the evidence sufficient to prove anything more than the sinking of the surface of the polders from drying and consolidation.

[319] The elevation of the lands enclosed by dikes—or polders, as they are called in Holland—above low water mark, depends upon the height of the tides, or, in other words, upon, the difference between ebb and flood. The tide cannot deposit earth higher than it flows, and after the ground is once enclosed, the decay of the vegetables grown upon it and the addition of manures do not compensate the depression occasioned by drying and consolidation. On the coast of Zeeland and the islands of South Holland, the tides, and of course the surface of the lands deposited by them, are so high that the polders can be drained by ditching and sluices, but at other points, as in the enclosed grounds of North Holland on the Zuiderzee, where the tide rises but three feet or even less, pumping is necessary from the beginning.—Staring, Voormaals en Thans, p. 152.

[320] The principal engine—called the Leeghwater, from the name of an engineer who had proposed the draining of the lake in 1641—was of 500 horse power, and drove eleven pumps making six strokes per minute. Each pump raised six cubic mètres, or nearly eight cubic yards of water to the stroke, amounting in all to 23,760 cubic mètres, or above 31,000 cubic yards, the hour.—Wild, Die Niederlande, i, p. 87.

[321] In England and New England, where the marshes have been already drained or are of comparatively small extent, the existence of large floating islands seems incredible, and has sometimes been treated as a fable, but no geographical fact is better established. Kohl (Inseln und Marschen Schleswig-Holsteins, iii, p. 309) reminds us that Pliny mentions among the wonders of Germany the floating islands, covered with trees, which met the Roman fleets at the mouths of the Elbe and the Weser. Our author speaks also of having visited, in the territory of Bremen, floating moors, bearing not only houses but whole villages. At low stages of the water these moors rest upon a bed of sand, but are raised from six to ten feet by the high water of spring, and remain afloat until, in the course of the summer, the water beneath is exhausted by evaporation and drainage, when they sink down upon the sand again. See Appendix, No. 40.

Staring explains, in an interesting way, the whole growth, formation, and functions of floating fens or bogs, in his very valuable work, De Bodem van Nederland, i, pp. 36-43. The substance of his account is as follows: The first condition for the growth of the plants which compose the substance of turf and the surface of the fens, is stillness of the water. Hence they are not found in running streams, nor in pools so large as to be subject to frequent agitation by the wind. For example, not a single plant grew in the open part of the Lake of Haarlem, and fens cease to form in all pools as soon as, by the cutting of the turf for fuel or other purposes, their area is sufficiently enlarged to be much acted on by wind. When still water above a yard deep is left undisturbed, aquatic plants of various genera, such us Nuphar, Nymphæa, Limnanthemum, Stratiotes, Polygonum, and Potamogeton, fill the bottom with roots and cover the surface with leaves. Many of the plants die every year, and prepare at the bottom a soil fit for the growth of a higher order of vegetation, Phragmites, Acorus, Sparganium, Rumex, Lythrum, Pedicularis, Spiræa, Polystichum, Comarum, Caltha, &c., &c. In the course of twenty or thirty years the muddy bottom is filled with roots of aquatic and marsh plants, which are lighter than water, and if the depth is great enough to give room for detaching this vegetable network, a couple of yards for example, it rises to the surface, bearing with it, of course, the soil formed above it by decay of stems and leaves. New genera now appear upon the mass, such as Carex, Menyanthes, and others, and soon thickly cover it. The turf has now acquired a thickness of from two to four feet, and is called in Groningen lad; in Friesland, til, tilland, or drijftil; in Overijssel, krag; and in Holland, rietzod. It floats about as driven by the wind, gradually increasing in thickness by the decay of its annual crops of vegetation, and in about half a century reaches the bottom and becomes fixed. If it has not been invaded in the mean time by men or cattle, trees and arborescent plants, Alnus, Salix, Myrica, &c. appear, and these contribute to hasten the attachment of the turf to the bottom, both by their weight and by sending their roots quite through into the ground.

This is the regular method employed by nature for the gradual filling up of shallow lakes and pools, and converting them first into morass and then into dry land. Whenever therefore man removes the peat or turf, he exerts an injurious geographical agency, and, as I have already said, there is no doubt that the immense extension of the inland seas of Holland in modern times is owing to this and other human imprudences. "Hundreds of hectares of floating pastures," says our author, "which have nothing in their appearance to distinguish them from grass lands resting on solid bog, are found in Overijssel, in North Holland and near Utrecht. In short, they occur in all deep bogs, and wherever deep water is left long undisturbed."

In one case, a floating island, which had attached itself to the shore, continued to float about for a long time after it was torn off by a flood, and was solid enough to keep a pond of fresh water upon it sweet, though the water in which it was swimming had become brackish from the irruption of the sea. After the hay is cut, cattle are pastured upon those islands, and they sometimes have large trees growing upon them.

When the turf or peat has been cut, leaving water less than a yard deep, Equisetum limosum grows at once, and is followed by the second class of marsh plants mentioned above. Their roots do not become detached from the bottom in such shallow water, but form ordinary turf or peat. These processes are so rapid that a thickness of from three to six feet of turf is formed in half a century, and many men have lived to mow grass where they had fished in their boyhood, and to cut turf twice in the same spot.

Captain Gilliss says that before Lake Taguataga in Chili was drained, there were in it islands composed of dead plants matted together to a thickness of from four to six feet, and with trees of medium size growing upon them. These islands floated before the wind "with their trees and browsing cattle."—United States Naval Astronomical Expedition to the Southern Hemisphere, i, pp. 16, 17.

[322] A considerable work of this character is mentioned by Captain Gilliss as having been executed in Chili, a country to which we should have hardly looked for an improvement of such a nature. The Lake Taguataga was partially drained by cutting through a narrow ridge of land, not at the natural outlet, but upon one side of the lake, and eight thousand acres of land covered by it were gained for cultivation.—U. S. Naval Astronomical Expedition to the Southern Hemisphere, i, pp. 16, 17.

[323] Économie Rurale de la France, p. 289.

[324] In a note on a former page of this volume I noticed an observation of Jacini, to the effect that the great Italian lakes discharge themselves partly by infiltration beneath the hills which bound them. The amount of such infiltration must depend much upon the hydrostatic pressure on the walls of the lake basins, and, of course, the lowering of the surface of these lakes, by diminishing that pressure, would diminish also the infiltration. It is now proposed to lower the level of the Lake of Como some feet by deepening its outlet. It is possible that the effect of this may manifest itself in a diminution of the water in springs and fontanili or artesian wells in Lombardy. See Appendix, No. 43.

[325] Simonde, speaking of the Tuscan canals, observes: "But inundations are not the only damage caused by the waters to the plains of Tuscany. Raised, as the canals are, above the soil, the water percolates through their banks, penetrates every obstruction, and, in spite of all the efforts of industry, sterilizes and turns to morasses fields which nature and the richness of the soil seemed to have designed for the most abundant harvests. In ground thus pervaded with moisture, or rendered cold, as the Tuscans express it, by the filtration of the canal water, the vines and the mulberries, after having for a few years yielded fruit of a saltish taste, rot and perish. The wheat decays in the ground, or dies as soon as it sprouts. Winter crops are given up, and summer cultivation tried for a time; but the increasing humidity, and the saline matter communicated to the earth—which affects the taste of all its products, even to the grasses, which the cattle refuse to touch—at last compel the husbandman to abandon his fields, and leave uncultivated a soil that no longer repays his labor."—Tableau de l'Agriculture Toscane. pp. 11, 12.

[326] Physikalische Geographie, p. 288. Draining by driving down stakes, mentioned in a note in a chapter on the woods, ante, is a process of the same nature.

[327] "The simplest backwoodsman knows by experience that all cultivation is impossible in the neighborhood of bogs and marshes. Why is a crop near the borders of a marsh cut off by frost, while a field upon a hillock, a few stone's throws from it, is spared?"—Lars Levi Læstadius, Om Uppodlingar i Lappmarken, pp. 69, 74.

[328] Babinet condemns even the general draining of marshes. "Draining," says he, "has been much in fashion for some years. It has been a special object to dry and fertilize marshy grounds. My opinion has always been that excessive dryness is thus produced, and that other soils in the neighborhood are sterilized in proportion."

[329] I ought perhaps to except the Mexicans and the Peruvians, whose arts and institutions are not yet shown to be historically connected with those of any more ancient people. The lamentable destruction of so many memorials of these tribes, by the ignorance and bigotry of the so-called Christian barbarians who conquered them, has left us much in the dark as to many points of their civilization; but they seem to have reached that stage where continued progress in knowledge and in power over nature is secure, and a few more centuries of independence might have brought them to originate for themselves most of the great inventions which the last four centuries have bestowed upon man.

[330] The necessity of irrigation in the great alluvial plain of Northern Italy is partly explained by the fact that the superficial stratum of fine earth and vegetable mould is very extensively underlaid by beds of pebbles and gravel brought down by mountain torrents at a remote epoch. The water of the surface soil drains rapidly down into these loose beds, and passes off by subterranean channels to some unknown point of discharge; but this circumstance alone is not a sufficient solution. Is it not possible that the habits of vegetables, grown in countries where irrigation has been immemorially employed, have been so changed that they require water under conditions of soil and climate where their congeners, which have not been thus indulgently treated, do not?

There are some atmospheric phenomena in Northern Italy, which an American finds it hard to reconcile with what he has observed in the United States. To an American eye, for instance, the sky of Piedmont, Lombardy, and the northern coast of the Mediterranean, is always whitish and curdled, and it never has the intensity and fathomless depth of the blue of his native heavens. And yet the heat of the sun's rays, as measured by sensation, and, at the same time, the evaporation, are greater than they would be with the thermometer at the same point in America. I have frequently felt in Italy, with the mercury below 60° Fahrenheit, and with a mottled and almost opaque sky, a heat of solar irradiation which I can compare to nothing but the scorching sensation experienced in America at a temperature twenty degrees higher, during the intervals between showers, or before a rain, when the clear blue of the sky seems infinite in depth and transparency. Such circumstances may create a necessity for irrigation where it would otherwise be superfluous, if not absolutely injurious.

In speaking of the superior apparent clearness of the sky in America, I confine myself to the concave vault of the heavens, and do not mean to assert that terrestrial objects are generally visible at greater distances in the United States than in Italy. Indeed I am rather disposed to maintain the contrary; for though I know that the lower strata of the atmosphere in Europe never equal in transparency the air near the earth in New Mexico, Peru, and Chili, yet I think the accidents of the coast line of the Riviera, as, for example, between Nice and La Spezia, and those of the incomparable Alpine panorama seen from Turin, are distinguishable at greater distances than they would be in the United States.

[331] In Egypt, evaporation and absorption by the earth are so rapid, that all annual crops require irrigation during the whole period of their growth. As fast as the water retires by the subsidence of the annual inundation, the seed is sown upon the still moist uncovered soil, and irrigation begins at once. Upon the Nile, you hear the creaking of the water wheels, and sometimes the movement of steam pumps, through the whole night, while the poorer cultivators unceasingly ply the simple shadoof, or bucket-and-sweep, laboriously raising the water from trough to trough by as many as six or seven stages when the river is low. The bucket is of flexible leather, with a stiff rim, and is emptied into the trough, not by inverting it like a wooden bucket, but by putting the hand beneath and pushing the bottom up till the water all runs out over the brim, or, in other words, by turning the vessel inside out.

The quantity of water thus withdrawn from the Nile is enormous. Most of this is evaporated directly from the surface or the superficial strata, but some moisture percolates down and oozes through the banks into the river again, while a larger quantity sinks till it joins the slow current of infiltration by which the Nile water pervades the earth of the valley to the distance, at some points, of not less than fifty miles.

[332] "Forests," "woods," and "groves," are very frequently mentioned in the Old Testament as existing at particular places, and they are often referred to by way of illustration, as familiar objects. "Wood" is twice spoken of as a material in the New Testament, but otherwise—at least according to Cruden—not one of the above words occurs in that volume.

This interesting fact, were other evidence wanting, would go far to prove that a great change had taken place in this respect between the periods when the Old Testament and the New were respectively composed; for the scriptural writers, and the speakers introduced into their narratives, are remarkable for their frequent allusions to the natural objects and the social and industrial habits which characterized their ages and their country. See Appendix, No. 44.

Solomon anticipated Chevandier in the irrigation of forest trees: "I made me pools of water, to water therewith the wood that bringeth forth trees."—Ecclesiastes ii, 6.

[333] One of these, upon Mount Hor, two stories in height, is still in such preservation that I found not less than ten feet of water in it in the month of June, 1851.

The brook Ain Musa, which runs through the city of Petra and finally disappears in the sands of Wadi el Araba, is a considerable river in winter, and the inhabitants of that town were obliged to excavate a tunnel through the rock near the right bank, just above the upper entrance of the Sik, to discharge a part of its swollen current. The sagacity of Dr. Robinson detected the necessity of this measure, though the tunnel, the mouth of which was hidden by brushwood, was not discovered till some time after his visit. I even noticed unequivocal remains of a sluice by which the water was diverted to the tunnel near the arch that crosses the Sik. Immense labor was also expended in widening the natural channel at several points below the town, to prevent the damming up and setting back of the water—a fact I believe not hitherto noticed by travellers.

The Fellahheen above Petra still employ the waters of Ain Musa for irrigation, and in summer the superficial current is wholly diverted from its natural channel for that purpose. At this season, the bed of the brook, which is composed of pebbles, gravel, and sand, is dry in the Sik and through the town; but the infiltration is such that water is generally found by digging to a small depth in the channel. Observing these facts in a visit to Petra in the summer, I was curious to know whether the subterranean waters escaped again to daylight, and I followed the ravine below the town for a long distance. Not very far from the upper entrance of the ravine, arborescent vegetation appeared upon its bottom, and as soon as the ground was well shaded, a thread of water burst out. This was joined by others a little lower down, and, at the distance of a mile from the town, a strong current was formed and ran down toward Wadi el Araba.

[334] The authorities differ as to the extent of the cultivable and the cultivated soil of Egypt. Lippincott's, or rather Thomas and Baldwin's, Gazetteer—a work of careful research—estimates "the whole area comprised in the valley [below the first cataract] and delta," at 11,000 square miles. Smith's Dictionary of the Bible, article "Egypt," says: "Egypt has a superficies of about 9,582 square geographical miles of soil, which the Nile either does or can water and fertilize. This computation includes the river and lakes as well as sundry tracts which can be inundated, and the whole space either cultivated or fit for cultivation is no more than about 5,626 square miles." By geographical mile is here meant, I suppose, the nautical mile of sixty to an equatorial degree, or about 2,025 yards. The whole area, then, by this estimate, is 12,682 square statute or English miles, that of the space "cultivated or fit for cultivation," 7,447. Smith's Dictionary of Greek and Roman Geography, article "Ægyptus," gives 2,255 square miles as the area of the valley between Syene and the bifurcation of the Nile, exclusive of the Fayoom, which is estimated at 340. The area of the Delta is stated at 1,976 square miles between the main branches of the river, and, including the irrigated lands east and west of those branches, at 4,500 square miles. This latter work does not inform us whether these are statute or nautical miles, but nautical miles must be intended.

Other writers give estimates differing considerably from those just cited. The latest computations I have seen are those in the first volume of Kremer's Ægypten, 1863. This author (pp. 6, 7) assigns to the Delta an area of 200 square German geographical miles (fifteen to the degree); to all Lower Egypt, including, of course, the Delta, 400 such miles. These numbers are equal, respectively, to 4,239 and 8,478 square statute miles, and the great lagoons are embraced in the areas computed. Upper Egypt (above Cairo) is said (p. 11) to contain 4,000,000 feddan of culturfläche, or cultivable land. The feddan is stated (p. 37) to contain 7,333 square piks, the pik being 75 centimètres, and it therefore corresponds almost exactly to the English acre. Hence, according to Kremer, the cultivable soil of Upper Egypt is 6,250 square statute miles, or twice as much as the whole area of the valley between Syene and the bifurcation of the Nile, according to Smith's Dictionary of Greek and Roman Geography. I suspect that 4,000,000 feddan is erroneously given as the cultivable area of Upper Egypt alone, when in fact it should be taken for the arable surface of both Lower and Upper Egypt; for from the statistical tables in the same volume, it appears that 3,317,125 feddan, or 5,253 square statute miles, were cultivated, in both geographical divisions, in the year referred to in the tables, the date of which is not stated.

The area which the Nile would now cover at high water, if left to itself, is greater than in ancient times, because the bed of the river has been elevated, and consequently the lateral spread of the inundation increased. See SMITH'S Dictionary of Geography, article "Ægyptus." But the industry of the Egyptians in the days of the Pharaohs and the Ptolomies carried the Nile-water to large provinces which have now been long abandoned and have relapsed into the condition of a desert. "Anciently," observes the writer of the article "Egypt" in Smith's Dictionary of the Bible, "2,735 square miles more [about 3,700 square statute miles] may have been cultivated. In the best days of Egypt, probably all the land was cultivated that could be made available for agricultural purposes, and hence we may estimate the ancient arable area of that country at not less than 11,000 square statute miles, or fully double its present extent."

[335] A canal has been constructed, and new ones are in progress, to convey water from the Nile to the city of Suez, and to various points on the line of the ship canal, with the double purpose of supplying fresh water to the inhabitants and laborers, and of irrigating the adjacent soil. The area of land which may be thus reclaimed and fertilized is very large, but the actual quantity which it will be found economically expedient to bring under cultivation cannot now be determined.

[336] The so-called spring at Heliopolis is only a thread of water infiltrated from the Nile or the canals.

[337] The date and the doum palm, the sont and many other acacias, the caroub, the sycamore, and other trees, grow well in Egypt without irrigation, and would doubtless spread through the entire valley in a few years.

[338] Wilkinson has shown that the cultivable soil of Egypt has not been diminished by encroachment of the desert sands, or otherwise, but that, on the contrary, it must have been increased since the age of the Pharaohs. The Gotha Almanac for 1862 states the population of Egypt in 1859 at 5,125,000 souls; but this must be a great exaggeration, even supposing the estimate to include the inhabitants of Nubia, and of much other territory not geographically belonging to Egypt. In general, the population of that country has been estimated at something more than three millions, or about six hundred to the square mile; but with a better government and better social institutions, the soil would sustain a much greater number, and in fact it is believed that in ancient times its inhabitants were twice, perhaps even thrice, as numerous as at present.

Wilkinson (Handbook for Travellers in Egypt, p. 10) observes that the total population, which two hundred years ago was estimated at 4,000,000, amounted till lately only to about 1,800,000 souls, having been reduced since 1800 from 2,500,000 to that number.

[339] Ritter supposes Egypt to have been a sandy desert when it was first occupied by man. "The first inhabitant of the sandy valley of the Nile was a desert dweller, as his neighbors right and left, the Libyan, the nomade Arab, still are. But the civilized people of Egypt transformed, by canals, the waste into the richest granary of the world; they liberated themselves from the shackles of the rock and sand desert, in the midst of which, by a wise distribution of the fluid through the solid geographical form, by irrigation in short, they created a region of culture most rich in historical monuments."—Einleitung zur allgemeinen vergleichenden Geographie, pp. 165, 166.

This view seems to me highly improbable; for though, by canals and embankments, man has done much to modify the natural distribution of the waters of the Nile, and possibly has even transferred its channel from one side of the valley to the other, yet the annual inundation is not his work, and the river must have overflowed its banks and carried spontaneous vegetation with its waters, as well before as since Egypt was first occupied by the human family. There is, indeed, some reason to suppose that man lived upon the banks of the Nile when its channel was much lower, and the spread of its inundations much narrower than at present; but wherever its flood reached, there the forest would propagate itself, and its shores are much more likely to have been morasses than sands.

[340] Memorie sui progetti per l'estensione dell' Irrigazione, etc., il Politecnico, for January, 1863, p. 6.

[341] Niel, L'Agriculture des États Sardes, p. 232.

[342] Niel, Agriculture des États Sardes, p. 237. Lombardini's computation just given allows eighty-one cubic mètres per day to the hectare, which, supposing the season of irrigation to be one hundred days, is equal to a precipitation of thirty-two inches. But in Lombardy, water is applied to some crops during a longer period than one hundred days; and in the marcite it flows over the ground even in winter.

According to Boussingault (Économie Rurale, ii, p. 246) grass grounds ought to receive, in Germany, twenty-one centimètres of water per week, and with less than half that quantity it is not advisable to incur the expense of supplying it. The ground is irrigated twenty-five or thirty times, and if the full quantity of twenty-one centimètres is applied, it receives about two hundred inches of water, or six times the total amount of precipitation. Puvis, quoted by Boussingault, after much research comes to the conclusion that a proper quantity is twenty centimètres applied twenty-five or thirty times, which corresponds with the estimate just stated. Puvis adds—and, as our author thinks, with reason—that this amount might be doubled without disadvantage.

Boussingault observes that rain water is vastly more fertilizing than the water of irrigating canals, and therefore the supply of the latter must be greater. This is explained partly by the different character of the substances held in solution or suspension by the waters of the earth and of the sky, partly by the higher temperature of the latter, and, possibly, partly also by the mode of application—the rain being finely divided in its fall or by striking plants on the ground, river water flowing in a continuous sheet.

The temperature of the water is thought even more important than its composition. The sources which irrigate the marcite of Lombardy—meadows so fertile that less than an acre furnishes grass for a cow the whole year—are very warm. The ground watered by them never freezes, and a first crop, for soiling, is cut from it in January or February. The Canal Cavour, just now commenced—which is to take its supply from the Po at Chivasso, fourteen or fifteen miles below Turin—will furnish water of much higher fertilizing power than that derived from the Dora Baltea and the Sesia, both because it is warmer, and because it transports a more abundant and a richer sediment than the latter streams, which are fed by Alpine icefields and melting snows, and which flow, for long distances, in channels ground smooth and bare by ancient glaciers, and not now contributing much vegetable mould or fine slime to their waters.

[343] It belongs rather to agriculture than to geography to discuss the quality of the crops obtained by irrigation, or the permanent effects produced by it on the productiveness of the soil. There is no doubt, however, that all crops which can be raised without watering are superior in flavor and in nutritive power to those grown by the aid of irrigation. Garden vegetables, particularly, profusely watered, are so insipid as to be hardly eatable. Wherever irrigation is practised, there is an almost irresistible tendency, especially among ignorant cultivators, to carry it to excess; and in Piedmont and Lombardy, if the supply of water is abundant, it is so liberally applied as sometimes not only to injure the quality of the product, but to drown the plants and diminish the actual weight of the crop.

Professor Liebig, in his Modern Agriculture, says: "There is not to be found in chemistry a more wonderful phenomenon, one which more confounds all human wisdom, than is presented by the soil of a garden or field. By the simplest experiment, any one may satisfy himself that rain water filtered through field or garden soil does not dissolve out a trace of potash, silicic acid, ammonia, or phosphoric acid. The soil does not give up to the water one particle of the food of plants which it contains. The most continuous rains cannot remove from the field, except mechanically, any of the essential constituents of its fertility."

"The soil not only retains firmly all the food of plants which is actually in it, but its power to preserve all that may be useful to them extends much farther. If rain or other water holding in solution ammonia, potash, and phosphoric and silicic acids, be brought in contact with soil, these substances disappear almost immediately from the solution; the soil withdraws them from the water. Only such substances are completely withdrawn by the soil as are indispensable articles of food for plants; all others remain wholly or in part in solution."

The first of the paragraphs just quoted is not in accordance with the alleged experience of agriculturists in those parts of Italy where irrigation is most successfully applied. They believe that the constituents of vegetable growth are washed out of the soil by excessive and long-continued watering. They consider it also established as a fact of observation, that water which has flowed through or over rich ground is far more valuable for irrigation than water from the same source, which has not been impregnated with fertilizing substances by passing through soils containing them; and, on the other hand, that water, rich in the elements of vegetation, parts with them in serving to irrigate a poor soil, and is therefore less valuable as a fertilizer of lower grounds to which it may afterward be conducted.

The practice of irrigation—except in mountainous countries where springs and rivulets are numerous—is attended with very serious economical, social, and political evils. The construction of canals and their immensely ramified branches, and the grading and scarping of the ground to be watered, are always expensive operations, and they very often require an amount of capital which can be commanded only by the state, by moneyed corporations, or by very wealthy proprietors; the capacity of the canals must be calculated with reference to the area intended to be irrigated, and when they and their branches are once constructed, it is very difficult to extend them, or to accommodate any of their original arrangements to changes in the condition of the soil, or in the modes or objects of cultivation; the flow of the water being limited by the abundance of the source or the capacity of the canals, the individual proprietor cannot be allowed to withdraw water at will, according to his own private interest or convenience, but both the time and the quantity of supply must be regulated by a general system applicable, as far as may be, to the whole area irrigated by the same canal, and every cultivator must conform his industry to a plan which may be quite at variance with his special objects or with his views of good husbandry. The clashing interests and the jealousies of proprietors depending on the same means of supply are a source of incessant contention and litigation, and the caprices or partialities of the officers who control, or of contractors who farm the canals, lead not unfrequently to ruinous injustice toward individual landholders. These circumstances discourage the division of the soil into small properties, and there is a constant tendency to the accumulation of large estates of irrigated land in the hands of great capitalists, and consequently to the dispossession of the small cultivators, who pass from the condition of owners of the land to that of hireling tillers. The farmers are no longer yeomen, but peasants. Having no interest in the soil which composes their country, they are virtually expatriated, and the middle class, which ought to constitute the real physical and moral strength of the land, ceases to exist as a rural estate, and is found only among the professional, the mercantile, and the industrial population of the cities.

[344] Boussingault, Économie Rurale, ii, pp. 248, 249.

[345] The cultivation of rice is so prejudicial to health everywhere that nothing but the necessities of a dense population can justify the sacrifice of life it costs in countries where it is pursued.

It has been demonstrated by actual experiment, that even in Mississippi, cotton can be advantageously raised by the white man without danger to health; and in fact, a great deal of the cotton brought to the Vicksburg market for some years past has been grown exclusively by white labor. There is no reason why the cultivation of cotton should be a more unhealthy occupation in America than it is in other countries where it was never dreamed of as dangerous, and no well-informed American, in the Slave States or out of them, believes that the abolition of slavery in the South would permanently diminish the cotton crop of those States.

[346] L'Italie à propos de l'Exposition de Paris, p. 92.

[347] The very valuable memoirs of Lombardini, Cenni idrografi sulla Lombardia, Intorno al sistema idraulico del Po, and other papers on similar subjects, were published in periodicals little known out of Italy; and the Idraulica Pratica of Mari has not, I believe, been translated into French or English. These works, and other sources of information equally inaccessible out of Italy, have been freely used by Baumgarten, in a memoir entitled Notice sur les Rivières de la Lombardie, in the Annales des Ponts et Chaussées, 1847, 1er sémestre, pp. 129 et seqq., and by Dumont, Des Travaux Publics dans leurs Rapports avec l'Agriculture, note, viii, pp. 269 et seqq. For the convenience of my readers, I shall use these two articles instead of the original authorities on which they are founded.

[348] Sir John F. W. Herschel, citing Talabot as his authority, Physical Geography (24).

In an elaborate paper on "Irrigation," printed in the United States Patent Report for 1860, p. 169, it is stated that the volume of water poured into the Mediterranean by the Nile in twenty-four hours, at low water, is 150,566,392,368 cubic mètres; at high water, 705,514,667,440 cubic mètres. Taking the mean of these two numbers, the average daily delivery of the Nile would be 428,081,059,808 cubic mètres, or more than 550,000,000,000 cubic yards. There is some enormous mistake, probably a typographical error, in this statement, which makes the delivery of the Nile seventeen hundred times as great as computed by Talabot, and many times more than any physical geographer has ever estimated the quantity supplied by all the rivers on the face of the globe.

[349] The Drac, a torrent emptying into the Isère a little below Grenoble, has discharged 5,200, the Isère, which receives it, 7,800 cubic yards, and the Durance an equal quantity, per second.—Montluisant, Note sur les Desséchements, etc., Annales des Ponts et Chaussées, 1833, 2me sémestre, p. 288.

The floods of some other French rivers scarcely fall behind those of the Rhone. The Loire, above Roanne, has a basin of 2,471 square miles, or about twice and a half the area of that of the Ardèche. In some of its inundations it has delivered above 9,500 cubic yards per second.—Belgrand, De l'Influence des Forêts, etc., Annales des Ponts et Chaussées, 1854, 1er sémestre, p. 15, note.

[350] The original forests in which the basin of the Ardèche was rich have been rapidly disappearing, for many years, and the terrific violence of the inundations which are now laying it waste is ascribed, by the ablest investigators, to that cause. In an article inserted in the Annales Forestières for 1843, quoted by Hohenstein, Der Wald, p. 177, it is said that about one third of the area of the department had already become absolutely barren, in consequence of clearing, and that the destruction of the woods was still going on with great rapidity. New torrents were constantly forming, and they were estimated to have covered more than 70,000 acres of good land, or one eighth of the surface of the department, with sand and gravel.

[351] "There is no example of a coincidence between great floods of the Ardèche and of the Rhone, all the known inundations of the former having taken place when the latter was very low."—Mardigny, Mémoire sur les Inondations des Rivières de l'Ardèche, p. 26.

I take this occasion to acknowledge myself indebted to the interesting memoir just quoted for all the statements I make respecting the floods of the Ardèche, except the comparison of the volume of its waters with that of the Nile, and the computation with respect to the capacity required for reservoirs to be constructed in its basin.