The course of events has been the same in Idumæa. The observing traveller discovers everywhere about Petra, particularly if he enters the city by the route of Wadi Ksheibeh, very extensive traces of ancient cultivation, and upon the neighboring ridges are the ruins of numerous cisterns evidently constructed to furnish a supply of water for irrigation.[333] In primitive ages, the precipitation of winter in these hilly countries was, in great part, retained for a time in the superficial soil, first by the vegetable mould of the forests, and then by the artificial arrangements I have described. The water imbibed by the earth was partly taken up by direct evaporation, partly absorbed by vegetation, and partly carried down by infiltration to subjacent strata which gave it out in springs at lower levels, and thus a fertility of soil and a condition of the atmosphere were maintained sufficient to admit of the dense population that once inhabited those now arid wastes. At present, the rain water runs immediately off from the surface and is carried down to the sea, or is drunk up by the sands of the wadis, and the hillsides which once teemed with plenty are bare of vegetation, and seared by the scorching winds of the desert.
In Southern Europe, in the Turkish Empire, and in many other countries, a very large proportion of the surface is, if not absolutely flooded, at least thoroughly moistened by irrigation, a great number of times in the course of every season, and this, especially, at periods when it would otherwise be quite dry, and when, too, the power of the sun and the capacity of the air for absorbing moisture are greatest. Hence it is obvious that the amount of evaporation from the earth in these countries, and, of course, the humidity and the temperature of both the soil and the atmosphere in contact with it, must be much affected by the practice of irrigation. The cultivable area of Egypt, or the space accessible to cultivation, between desert and desert, is more than seven thousand square statute miles. Much of the surface, though not out of the reach of irrigation, lies too high to be economically watered, and irrigation and cultivation are therefore confined to an area of five or six thousand square miles, nearly the whole of which is regularly and constantly watered when not covered by the inundation, except in the short interval between the harvest and the rise of the waters. For nearly half of the year, then, irrigation adds five or six thousand square miles, or more than a square equatorial degree, to the evaporable surface of the Nile valley, or, in other words, more than decuples the area from which an appreciable quantity of moisture would otherwise be evaporated; for after the Nile has retired within its banks, its waters by no means cover one tenth of the space just mentioned.[334] The fresh-water canals now constructing, in connection with the works for the Suez canal, will not only restore the long abandoned fields east of the Nile, but add to the arable soil of Egypt hundreds of square miles of newly reclaimed desert, and thus still further increase the climatic effects of irrigation.[335]
The Nile receives not a single tributary in its course through Egypt; there is not so much as one living spring in the whole land,[336] and, with the exception of a narrow strip of coast, where the annual precipitation is said to amount to six inches, the fall of rain in the territory of the Pharaohs is not two inches in the year. The subsoil of the whole valley is pervaded with moisture by infiltration from the Nile, and water can everywhere be found at the depth of a few feet. Were irrigation suspended, and Egypt abandoned, as in that case it must be, to the operations of nature, there is no doubt that trees, the roots of which penetrate deeply, would in time establish themselves on the deserted soil, fill the valley with verdure, and perhaps at last temper the climate, and even call down abundant rain from the heavens.[337] But the immediate effect of discontinuing irrigation would be, first, an immense reduction of the evaporation from the valley in the dry season, and then a greatly augmented dryness and heat of the atmosphere. Even the almost constant north wind—the strength of which would be increased in consequence of these changes—would little reduce the temperature of the narrow cleft between the burning mountains which hem in the channel of the Nile, so that a single year would transform the most fertile of soils to the most barren of deserts, and render uninhabitable a territory that irrigation makes capable of sustaining as dense a population as has ever existed in any part of the world.[338] Whether man found the valley of the Nile a forest, or such a waste as I have just described, we do not historically know. In either case, he has not simply converted a wilderness into a garden, but has unquestionably produced extensive climatic change.[339]
The fields of Egypt are more regularly watered than those of any other country bordering on the Mediterranean, except the rice grounds in Italy, and perhaps the marcite or winter meadows of Lombardy; but irrigation is more or less employed throughout almost the entire basin of that sea, and is everywhere attended with effects which, if less in degree, are analogous in character to those resulting from it in Egypt. In general, it may be said that the soil is nowhere artificially watered except when it is so dry that little moisture would be evaporated from it, and, consequently, every acre of irrigated ground is so much added to the evaporable surface of the country. When the supply of water is unlimited, it is allowed, after serving its purpose on one field, to run into drains, canals, or rivers. But in most regions where irrigation is regularly employed, it is necessary to economize the water; after passing over or through one parcel of ground, it is conducted to another; no more is withdrawn from the canals at any one point than is absorbed by the soil it irrigates, or evaporated from it, and, consequently, it is not restored to liquid circulation, except by infiltration or precipitation. We are safe, then, in saying that the humidity evaporated from any artificially watered soil is increased by a quantity bearing a large proportion to the whole amount distributed over it; for most even of that which is absorbed by the earth is immediately given out again either by vegetables or by evaporation.
It is not easy to ascertain precisely either the extent of surface thus watered, or the amount of water supplied, in any given country, because these quantities vary with the character of the season; but there are not many districts in Southern Europe where the management of the arrangements for irrigation is not one of the most important branches of agricultural labor. The eminent engineer Lombardini describes the system of irrigation in Lombardy as, "every day in summer, diffusing over 550,000 hectares of land 45,000,000 cubic mètres of water, which is equal to the entire volume of the Seine, at an ordinary flood, or a rise of three mètres above the hydrometer at the bridge of La Tournelle at Paris."[340] Niel states the quantity of land irrigated in the former kingdom of Sardinia, including Savoy, in 1856, at 240,000 hectares, or not much less than 600,000 acres. This is about four thirteenths of the cultivable soil of the kingdom. According to the same author, the irrigated lands in France did not exceed 100,000 hectares, or 247,000 acres, while those in Lombardy amounted to 450,000 hectares, more than 1,100,000 acres.[341] In these three states alone, then, there were more than three thousand square miles of artificially watered land, and if we add the irrigated soils of the rest of Italy, of the Mediterranean islands, of the Spanish peninsula, of Turkey in Europe and in Asia Minor, of Syria, of Egypt and the remainder of Northern Africa, we shall see that irrigation increases the evaporable surface of the Mediterranean basin by a quantity bearing no inconsiderable proportion to the area naturally covered by water within it. As near as can be ascertained, the amount of water applied to irrigated lands is scarcely anywhere less than the total precipitation during the season of vegetable growth, and in general it much exceeds that quantity. In grass grounds and in field culture it ranges from 27 or 28 to 60 inches, while in smaller crops, tilled by hand labor, it is sometimes carried as high as 300 inches.[342] The rice grounds and the marcite of Lombardy are not included in these estimates of the amount of water applied. Arrangements are concluded, and new plans proposed, for an immense increase of the lands fertilized by irrigation in France and Italy, and there is every reason to believe that the artificially watered soil of the latter country will be doubled, that of France quadrupled, before the end of this century. There can be no doubt that by these operations man is exercising a powerful influence on soil, on vegetable and animal life, and on climate, and hence that in this, as in many other fields of industry, he is truly a geographical agency.[343] The quantity of water artificially withdrawn from running streams for the purpose of irrigation is such as very sensibly to affect their volume, and it is, therefore, an important element in the geography of rivers. Brooks of no trifling current are often wholly diverted from their natural channels to supply the canals, and their entire mass of water completely absorbed, so that it does not reach the river which it naturally feeds, except in such proportion as it is conveyed to it by infiltration. Irrigation, therefore, diminishes great rivers in warm countries by cutting off their sources of supply as well as by direct abstraction of water from their channels. We have just seen that the system of irrigation in Lombardy deprives the Po of a quantity of water equal to the total delivery of the Seine at ordinary flood, or, in other words, of the equivalent of a tributary navigable for hundreds of miles by vessels of considerable burden. The new canals commenced and projected will greatly increase the loss. The water required for irrigation in Egypt is less than would be supposed from the exceeding rapidity of evaporation in that arid climate; for the soil is thoroughly saturated during the inundation, and infiltration from the Nile continues to supply a considerable amount of humidity in the dryest season. Linant Bey computed that twenty-nine cubic mètres per day sufficed to irrigate a hectare in the Delta.[344] This is equivalent to a fall of rain of two millimètres and nine tenths per day, or, if we suppose water to be applied for one hundred and fifty days during the dry season, to a total precipitation of 435 millimètres, about seventeen inches and one third. Taking the area of actually cultivated soil in Egypt at the low estimate of 3,600,000 acres, and the average amount of water daily applied in both Upper and Lower Egypt at twelve hundredths of an inch in depth, we have an abstraction of 61,000,000 cubic yards, which—the mean daily delivery of the Nile being in round numbers 320,000,000 cubic yards—is nearly one fifth of the average quantity of water contributed to the Mediterranean by that river.
Irrigation, as employed for certain special purposes in Europe and America, is productive of very prejudicial climatic effects. I refer particularly to the cultivation of rice in the Slave States of the American Union and in Italy. The climate of the Southern States is not necessarily unhealthy for the white man, but he can scarcely sleep a single night in the vicinity of the rice grounds without being attacked by a dangerous fever.[345] The neighborhood of the rice fields is less pestilential in Lombardy and Piedmont than in South Carolina and Georgia, but still very insalubrious to both man and beast. "Not only does the population decrease where rice is grown," says Escourrou Milliago, "but even the flocks are attacked by typhus. In the rice grounds, the soil is divided into compartments rising in gradual succession to the level of the irrigating canal, in order that the water, after having flowed one field, may be drawn off to another, and thus a single current serve for several compartments, the lowest field, of course, still being higher than the ditch which at last drains both it and the adjacent soil. This arrangement gives a certain force of hydrostatic pressure to the water with which the rice is irrigated, and the infiltration from these fields is said to extend through neighboring grounds, sometimes to the distance of not less than a myriamètre, or six English miles, and to be destructive to crops and even trees reached by it. Land thus affected can no longer be employed for any purpose but growing rice, and when prepared for that crop, it propagates still further the evils under which it had itself suffered, and, of course, the mischief is a growing one."[346]
The attentive traveller in Egypt and Nubia cannot fail to notice many localities, generally of small extent, where the soil is rendered infertile by an excess of saline matter in its composition. In many cases, perhaps in all, these barren spots lie rather above the level usually flooded by the inundations of the Nile, and yet they exhibit traces of former cultivation. Recent observations in India, a notice of which I find in an account of a meeting of the Asiatic Society in the Athenæum of December 20, 1862, No. 1834, suggest a possible explanation of this fact. At this meeting, Professor Medlicott read an essay on "the saline efflorescence called 'Reh' and 'Kuller,'" which is gradually invading many of the most fertile districts of Northern and Western India, and changing them into sterile deserts. It consists principally of sulphate of soda (Glauber's salts), with varying proportions of common salt. Mr. Medlicott pronounces "these salts (which, in small quantities are favorable to fertility of soil) to be the gradual result of concentration by evaporation of river and canal waters, which contain them in very minute quantities, and with which the lands are either irrigated or occasionally overflowed." The river inundations in hot countries usually take place but once in a year, and, though the banks remain submerged for days or even weeks, the water at that period, being derived principally from rains and snows, must be less highly charged with mineral matter than at lower stages, and besides, it is always in motion. The water of irrigation, on the other hand, is applied for many months in succession, it is drawn from rivers at the seasons when their proportion of salts is greatest, and it either sinks into the superficial soil, carrying with it the saline substances it holds in solution, or is evaporated from the surface, leaving them upon it. Hence irrigation must impart to the soil more salts than natural inundation. The sterilized grounds in Egypt and Nubia lying above the reach of the floods, as I have said, we may suppose them to have been first cultivated in that remote antiquity when the Nile valley received its earliest inhabitants. They must have been artificially irrigated from the beginning; they may have been under cultivation many centuries before the soil at a lower level was invaded by man, and hence it is natural that they should be more strongly impregnated with saline matter than fields which are exposed every year, for some weeks, to the action of running water so nearly pure that it would be more likely to dissolve salts than to deposit them.
In pointing out in a former chapter the evils which have resulted from the too extensive destruction of the forests, I dwelt at some length on the increased violence of river inundations, and especially on the devastations of torrents, in countries improvidently deprived of their woods, and I spoke of the replanting of the forests as the only effectual method of preventing the frequent recurrence of disastrous floods. There are many regions where, from the loss of the superficial soil, from financial considerations, and from other causes, the restoration of the woods is not, under present circumstances, to be hoped for. Even where that measure is feasible and in actual process of execution, a great number of years must elapse before the action of the destructive causes in question can be arrested or perhaps even sensibly mitigated by it. Besides this, leaving out of view the objections urged by Belgrand and his followers to the generally received opinions concerning the beneficial influence of the forest as respects river inundations—for no one disputes its importance in preventing the formation and limiting the ravages of mountain torrents—floods will always occur in years of excessive precipitation, whether the surface of the soil be generally cleared or generally wooded.
Physical improvement in this respect, then, cannot he confined to preventive measures, but, in countries subject to damage by inundation, means must he contrived to obviate dangers and diminish injuries to which human life and all the works of human industry will occasionally be exposed, in spite of every effort to lessen the frequency of their recurrence by acting directly on the causes that produce them. As every civilized country is, in some degree, subject to inundation by the overflow of rivers, the evil is a familiar one, and needs no general description. In discussing this branch of the subject, therefore, I may confine myself chiefly to the means that have been or may be employed to resist the force and limit the ravages of floods, which, left wholly unrestrained, would not only inflict immense injury upon the material interests of man, but produce geographical revolutions of no little magnitude.
The most obvious and doubtless earliest method of preventing the escape of river waters from their natural channels, and the overflow of fields and towns by their spread, is that of raised embankments along their course. The necessity of such embankments usually arises from the gradual elevation of the bed of running streams in consequence of the deposit of the earth and gravel they are charged with in high water; and, as we have seen, this elevation is rapidly accelerated when the highlands around the headwaters of rivers are cleared of their forests. When a river is embanked at a given point, and, consequently, the water of its floods, which would otherwise spread over a wide surface, is confined within narrow limits, the velocity of the current and its transporting power are augmented, and its burden of sand and gravel is deposited at some lower point, where the rapidity of its flow is checked by a diminution in the inclination of the bed, by a wider channel, or finally by a lacustrine or marine basin which receives its waters. Wherever it lets fall solid material, its channel is raised in consequence, and the declivity of the whole bed between the head of the embankment and the slack of the stream is reduced. Hence the current, at first accelerated by confinement, is afterward checked by the mechanical resistance of the matter deposited, and by the diminished inclination of its channel, and then begins again to let fall the earth it holds in suspension, and to raise its bed at the point where its overflow had been before prevented by embankment. The bank must now be raised in proportion, and these processes would be repeated and repeated indefinitely, had not nature provided a remedy in floods, which sweep out recent deposits, burst the bonds of the river and overwhelm the adjacent country with final desolation, or divert the current into a new channel, destined to become, in its turn, the scene of a similar struggle between man and the waters.
Few rivers, like the Nile, more than compensate by the fertilizing properties of their water and their slime for the damage they may do in inundations, and, consequently, there are few whose floods are not an object of dread, few whose encroachments upon their banks are not a source of constant anxiety and expense to the proprietors of the lands through which they flow. River dikes, for confining the spread of currents at high water, are of great antiquity in the East, and those of the Po and its tributaries were begun before we have any trustworthy physical or political annals of the provinces upon their borders. From the earliest ages, the Italian hydraulic engineers have stood in the front rank of their profession, and the Italian literature of this branch of material improvement is exceedingly voluminous. But the countries for which I write have no rivers like the Po, no plains like those of Lombardy, and the dangers to which the inhabitants of English and American river banks are exposed are more nearly analogous to those that threaten the soil and population in the valleys and plains of France, than to the perils and losses of the Lombard. The writings of the Italian hydrographers, too, though rich in professional instruction, are less accessible to foreigners and less adapted to popular use than those of French engineers.[347] For these reasons I shall take my citations principally from French authorities, though I shall occasionally allude to Italian writers on the floods of the Tiber, of the Arno, and some other Italian streams which much resemble those of the rivers of England and the United States.
The floods of mountain streams are attended with greater immediate danger to life and property than those of rivers of less rapid flow, because their currents are more impetuous, and they rise more suddenly and with less previous warning. At the same time, their ravages are confined within narrower limits, the waters retire sooner to their accustomed channel, and the danger is more quickly over, than in the case of inundations of larger rivers. The Ardèche, which has given its name to a department in France, drains a basin of 600,238 acres, or a little less than nine hundred and thirty-eight square miles. Its remotest source is about seventy-five miles, in a straight line, from its junction with the Rhone, and springs at an elevation of four thousand feet above that point. At the lowest stage of the river, the bed of the Chassezac, its largest and longest tributary, is in many places completely dry on the surface—the water being sufficient only to supply the subterranean channels of infiltration—and the Ardèche itself is almost everywhere fordable, even below the mouth of the Chassezac. But in floods, the river has sometimes risen more than sixty feet at the Pont d'Arc, a natural arch of two hundred feet chord, which spans the stream below its junction with all its important affluents. At the height of the inundation of 1827, the quantity of water passing this point—after deducting thirty per cent. for material transported with the current and for irregularity of flow—was estimated at 8,845 cubic yards to the second, and between twelve at noon on the 10th of September of that year and ten o'clock the next morning, the water discharged through the passage in question amounted to more than 450,000,000 cubic yards. This quantity, distributed equally through the basin of the river, would cover its entire area to a depth of more than five inches.
The Ardèche rises so suddenly that, in the inundation of 1846, the women who were washing in the bed of the river had not time to save their linen, and barely escaped with their lives, though they instantly fled upon hearing the roar of the approaching flood. Its waters and those of its affluents fall almost as rapidly, for in less than twenty-four hours after the rain has ceased in the Cévennes, where it rises, the Ardèche returns within its ordinary channel, even at its junction with the Rhone. In the flood of 1772, the water at La Beaume de Ruoms, on the Beaume, a tributary of the Ardèche, rose thirty-five feet above low water, but the stream was again fordable on the evening of the same day. The inundation of 1827 was, in this respect, exceptional, for it continued three days, during which period the Ardèche poured into the Rhone 1,305,000,000 cubic yards of water.
The Nile delivers into the sea 101,000 cubic feet or 3,741 cubic yards per second, on an average of the whole year.[348] This is equal to 323,222,400 cubic yards per day. In a single day of flood, then, the Ardèche, a river too insignificant to be known except in the local topography of France, contributed to the Rhone once and a half, and for three consecutive days once and one third, as much as the average delivery of the Nile during the same periods, though the basin of the latter river contains 500,000 square miles of surface, or more than five hundred times as much as that of the former.
The average annual precipitation in the basin of the Ardèche is not greater than in many other parts of Europe, but excessive quantities of rain frequently fall in that valley in the autumn. On the 9th of October, 1827, there fell at Joyeuse, on the Beaume, no less than thirty-one inches between three o'clock in the morning and midnight. Such facts as this explain the extraordinary suddenness and violence of the floods of the Ardèche, and the basins of many other tributaries of the Rhone exhibit meteorological phenomena not less remarkable.[349] The inundation of the 10th September, 1857, was accompanied with a terrific hurricane, which passed along the eastern slope of the high grounds where the Ardèche and several other western affluents of the Rhone take their rise. The wind tore up all the trees in its path, and the rushing torrents bore their trunks down to the larger streams, which again transported them to the Rhone in such rafts that one might almost have crossed that river by stepping from trunk to trunk.[350] The Rhone, therefore, is naturally subject to great and sudden inundations, and the same remark may be applied to most of the principal rivers of France, because the geographical character of all of them is approximately the same.
The height and violence of the inundations of most great rivers are determined by the degree in which the floods of the different tributaries are coincident in time. Were all the affluents of the Rhone to pour their highest annual floods into its channel at once, were a dozen Niles to empty themselves into its bed at the same moment, its water would rise to a height and rush with an impetus that would sweep into the Mediterranean the entire population of its banks, and all the works that man has erected upon the plains which border it. But such a coincidence can never happen. The tributaries of this river run in very different directions, and some of them are swollen principally by the melting of the snows about their sources, others almost exclusively by heavy rains. When a damp southeast wind blows up the valley of the Ardèche, its moisture is condensed, and precipitated in a deluge upon the mountains which embosom the headwaters of that stream, thus producing a flood, while a neighboring basin, the axis of which lies transversely or obliquely to that of the Ardèche, is not at all affected.[351]
It is easy to see that the damage occasioned by such floods as I have described must be almost incalculable, and it is by no means confined to the effects produced by overflow and the mechanical force of the superficial currents. In treating of the devastations of torrents in a former chapter, I confined myself principally to the erosion of surface and the transportation of mineral matter to lower grounds by them. The general action of torrents, as there shown, tends to the ultimate elevation of their beds by the deposit of the earth, gravel, and stone conveyed by them; but until they have thus raised their outlets so as sensibly to diminish the inclination of their channels—and sometimes when extraordinary floods give the torrents momentum enough to sweep away the accumulations which they have themselves heaped up—the swift flow of their currents, aided by the abrasion of the rolling rocks and gravel, scoops their beds constantly deeper, and they consequently not only undermine their banks, but frequently sap the most solid foundations which the art of man can build for the support of bridges and hydraulic structures.[352]
In the inundation of 1857, the Ardèche destroyed a stone bridge near La Beaume, which had been built about eighty years before. The resistance of the piers, which were erected on piles, the channel at that point being of gravel, produced an eddying current that washed away the bed of the river above them, and the foundation, thus deprived of lateral support, yielded to the weight of the bridge, and the piles and piers fell up stream.
By a curious law of compensation, the stream which, at flood, scoops out cavities in its bed, often fills them up again as soon as the diminished velocity of the current allows it to let fall the sand and gravel with which it is charged, so that when the waters return to their usual channel, the bottom shows no sign of having been disturbed. In a flood of the Escontay, a tributary of the Rhone, in 1846, piles driven sixteen feet into its gravelly bed for the foundation of a pier were torn up and carried off, and yet, when the river had fallen to low-water mark, the bottom at that point appeared to have been raised higher than it was before the flood, by new deposits of sand and gravel, while the cut stones of the half-built pier were found buried to a great depth, in the excavation which the water had first washed out. The gravel with which rivers thus restore the level of their beds is principally derived from the crushing of the rocks brought down by the mountain torrents, and the destructive effects of inundations are immensely diminished by this reduction of large stones to minute fragments. If the blocks hurled down from the cliffs were transported unbroken to the channels of large rivers, the mechanical force of their movement would be irresistible. They would overthrow the strongest barriers, spread themselves over a surface as wide as the flow of the waters, and convert the most smiling valleys into scenes of the wildest desolation.
There are few operations of nature where the effect seems more disproportioned to the cause than in the comminution of rock in the channel of swift waters. Igneous rocks are generally so hard as to be wrought with great difficulty, and they bear the weight of enormous superstructures without yielding to the pressure; but to the torrent they are as wheat to the millstone. The streams which pour down the southern scarp of the Mediterranean Alps along the Riviera di Ponente, near Genoa, have short courses, and a brisk walk of a couple of hours or even less takes you from the sea beach to the headspring of many of them. In their heaviest floods, they bring rounded masses of serpentine quite down to the sea, but at ordinary high water their lower course is charged only with finely divided particles of that rock. Hence, while, near their sources, their channels are filled with pebbles and angular fragments, intermixed with a little gravel, the proportions are reversed near their mouths, and, just above the points where their outlets are partially choked by the rolling shingle of the beach, their beds are composed of sand and gravel to the almost total exclusion of pebbles. The greatest depth of the basin of the Ardèche is seventy-five miles, but most of its tributaries have a much shorter course. "These affluents," says Mardigny, "hurl into the bed of the Ardèche enormous blocks of rock, which this river, in its turn, bears onward, and grinds down, at high water, so that its current rolls only gravel at its confluence with the Rhone."[353]
Guglielmini argued that the gravel and sand of the beds of running streams were derived from the trituration of rocks by the action of the currents, and inferred that this action was generally sufficient to reduce hard rock to sand in its passage from the source to the outlet of rivers. Frisi controverted this opinion, and maintained that river sand was of more ancient origin, and he inferred from experiments in artificially grinding stones that the concussion, friction, and attrition of rock in the channel of running waters were inadequate to its comminution, though he admitted that these same causes might reduce silicious sand to a fine powder capable of transportation to the sea by the currents.[354] Frisi's experiments were tried upon rounded and polished river pebbles, and prove nothing with regard to the action of torrents upon the irregular, more or less weathered, and often cracked and shattered rocks which lie loose in the ground at the head of mountain valleys. The fury of the waters and of the wind which accompanies them in the floods of the French Alpine torrents is such, that large blocks of stone are hurled out of the bed of the stream to the height of twelve or thirteen feet. The impulse of masses driven with such force overthrows the most solid masonry, and their concussion cannot fail to be attended with the crushing of the rocks themselves.[355]
The month of May, 1856, was remarkable for violent and almost uninterrupted rains, and most of the river basins of France were inundated to an extraordinary height. In the valleys of the Loire and its affluents, about a million of acres, including many towns and villages, were laid under water, and the amount of pecuniary damage was almost incalculable.[356] The flood was not less destructive in the valley of the Rhone, and in fact an invasion by a hostile army could hardly have been more disastrous to the inhabitants of the plains than was this terrible deluge. There had been a flood of this latter river in the year 1840, which, for height and quantity of water, was almost as remarkable as that of 1856, but it took place in the month of November, when the crops had all been harvested, and the injury inflicted by it upon agriculturists was, therefore, of a character to be less severely and less immediately felt than the consequences of the inundation of 1856.[357]
In the fifteen years between these two great floods, the population and the rural improvements of the river valleys had much increased, common roads, bridges, and railways had been multiplied and extended, telegraph lines had been constructed, all of which shared in the general ruin, and hence greater and more diversified interests were affected by the catastrophe of 1856 than by any former like calamity. The great flood of 1840 had excited the attention and roused the sympathies of the French people, and the subject was invested with new interest by the still more formidable character of the inundations of 1856. It was felt that these scourges had ceased to be a matter of merely local concern, for, although they bore most heavily on those whose homes and fields were situated within the immediate reach of the swelling waters, yet they frequently destroyed harvests valuable enough to be a matter of national interest, endangered the personal security of the population of important political centres, interrupted communication for days and even weeks together on great lines of traffic and travel—thus severing as it were all Southwestern France from the rest of the empire—and finally threatened to produce great and permanent geographical changes. The well-being of the whole commonwealth was seen to be involved in preventing the recurrence, and in limiting the range of such devastations. The Government encouraged scientific investigation of the phenomena and their laws. Their causes, their history, their immediate and remote consequences, and the possible safeguards to be employed against them, have been carefully studied by the most eminent physicists, as well as by the ablest theoretical and practical engineers of France. Many hitherto unobserved facts have been collected, many new hypotheses suggested, and many plans, more or less original in character, have been devised for combating the evil; but thus far, the most competent judges are not well agreed as to the mode, or even the possibility, of applying a remedy.
Perhaps no one point has been more prominent in the discussions than the influence of the forest in equalizing and regulating the flow of the water of precipitation. As we have already seen, opinion is still somewhat divided on this subject, but the conservative action of the woods in this respect has been generally recognized by the public of France, and the Government of the empire has made this principle the basis of important legislation for the protection of existing forests, and for the formation of new. The clearing of woodland, and the organization and functions of a police for its protection, are regulated by a law bearing date June 18th, 1859, and provision was made for promoting the restoration of private woods by a statute adopted on the 28th of July, 1860. The former of these laws passed the legislative body by a vote of 246 against 4, the latter with but a single negative voice. The influence of the government, in a country where the throne is so potent as in France, would account for a large majority, but when it is considered that both laws, the former especially, interfere very materially with the rights of private domain, the almost entire unanimity with which they were adopted is proof of a very general popular conviction, that the protection and extension of the forests is a measure more likely than any other to check the violence, if not to prevent the recurrence, of destructive inundations. The law of July 28th, 1860, appropriated 10,000,000 francs, to be expended, at the rate of 1,000,000 francs per year, in executing or aiding the replanting of woods. It is computed that this appropriation will secure the creation of new forest to the extent of about 250,000 acres, or one eleventh part of the soil where the restoration of the forest is thought feasible and, at the same time, specially important as a security against the evils ascribed in a great measure to its destruction.
The provisions of the laws in question are preventive rather than remedial; but some immediate effect may be expected to result from them, particularly if they are accompanied with certain other measures, the suggestion of which has been favorably received. The strong repugnance of the mountaineers to the application of a system which deprives them of a part of their pasturage—for the absolute exclusion of domestic animals is indispensable to the maintenance of an existing forest and to the formation of a new—is the most formidable obstacle to the execution of the laws of 1859-'60. It is proposed to compensate this loss by a cheap system of irrigation of lower pasture grounds, consisting in little more than in running horizontal furrows along the hillsides, thus converting the scarp of the hills into a succession of small terraces which, when once turfed over, are very permanent. Experience is said to have demonstrated that this simple process suffices to retain the water of rains, of snows, and of small springs and rivulets, long enough for the irrigation of the soil, thus increasing its product of herbage in a fivefold proportion, and that it partially checks the too rapid flow of surface water into the valleys, and, consequently, in some measure obviates one of the most prominent causes of inundations.[358] It is evident that, if such results are produced by this method, its introduction upon an extensive scale must also have the same climatic effects as other systems of irrigation.
Whatever may be the ultimate advantages of reclothing a large extent of the territory of France with wood, or of so shaping its surface as to prevent the too rapid flow of water over it, the results to be obtained by such processes can be realized in an adequate measure only after a long succession of years. Other steps must be taken, both for the immediate security of the lives and property of the present generation, and for the prevention of yet greater and remoter evils which are inevitable unless means to obviate them are found before it is forever too late. The frequent recurrence of inundations like those of 1856, for a single score of years, in the basins of the Rhone and the Loire, with only the present securities against them, would almost depopulate the valleys of those rivers, and produce physical revolutions in them, which, like revolutions in the political world, could never be made to "go backward."
Destructive inundations are seldom, if ever, produced by precipitation within the limits of the principal valley, but almost uniformly by sudden thaws or excessive rains on the mountain ranges where the tributaries take their rise. It is therefore plain that any measures which shall check the flow of surface waters into the channels of the affluents, or which shall retard the delivery of such waters into the principal stream by its tributaries, will diminish in the same proportion the dangers and the evils of inundation by great rivers. The retention of the surface waters upon or in the soil can hardly be accomplished except by the methods already mentioned, replanting of forests, and furrowing or terracing. The current of mountain streams can be checked by various methods, among which the most familiar and obvious is the erection of barriers or dams across their channels, at points convenient for forming reservoirs large enough to retain the superfluous waters of great rains and thaws. Besides the utility of such basins in preventing floods, the construction of them is recommended by very strong considerations, such as the meteorological effects of increased evaporable surface, the furnishing of a constant supply of water for agricultural and mechanical purposes, and, finally, their value as ponds for breeding and rearing fish, and, perhaps, for cultivating aquatic vegetables.
The objections to the general adoption of the system of reservoirs are these: the expense of their construction and maintenance; the reduction of cultivable area by the amount of surface they must cover; the interruption they would occasion to free communication; the probability that they would soon be filled up with sediment, and the obvious fact that when full of earth or even water, they would no longer serve their principal purpose; the great danger to which they would expose the country below them in case of the bursting of their barriers;[359] the evil consequences they would occasion by prolonging the flow of inundations in proportion as they diminished their height; the injurious effects it is supposed they would produce upon the salubrity of the neighboring districts; and, lastly, the alleged impossibility of constructing artificial basins sufficient in capacity to prevent, or in any considerable measure to mitigate, the evils they are intended to guard against.
The last argument is more easily reduced to a numerical question than the others. The mean and extreme annual precipitation of all the basins where the construction of such works would be seriously proposed is already approximately known by meteorological tables, and the quantity of water, delivered by the greatest floods which have occurred within the memory of man, may be roughly estimated from their visible traces. From these elements, or from recorded observations, the capacity of the necessary reservoirs can be calculated. Let us take the case of the Ardèche. In the inundation of 1857, that river poured into the Rhone 1,305,000,000 cubic yards of water in three days. If we suppose that half this quantity might have been suffered to flow down its channel without inconvenience, we shall have about 650,000,000 cubic yards to provide for by reservoirs. The Ardèche and its principal affluent, the Chassezac, have, together, about twelve considerable tributaries rising near the crest of the mountains which bound the basin. If reservoirs of equal capacity were constructed upon all of them, each reservoir must be able to contain 54,000,000 cubic yards, or, in other words, must be equal to a lake 3,000 yards long, 1,000 yards wide, and 18 yards deep, and besides, in order to render any effectual service, the reservoirs must all have been empty at the commencement of the rains which produced the inundation.
Thus far, I have supposed the swelling of the waters to be uniform throughout the whole basin; but such was by no means the fact in the inundation of 1857, for the rise of the Chassezac, which is as large as the Ardèche proper, did not exceed the limits of ordinary floods, and the dangerous excess came solely from the headwaters of the latter stream. Hence reservoirs of double the capacity I have supposed would have been necessary upon the tributaries of that river, to prevent the injurious effects of the inundation. It is evident that the construction of reservoirs of such magnitude for such a purpose is financially, if not physically, impracticable, and when we take into account a point I have just suggested, namely, that the reservoirs must be empty at all times of apprehended flood, and, of course, their utility limited almost solely to the single object of preventing inundations, the total inapplicability of such a measure in this particular case becomes still more glaringly manifest.
Another not less conclusive fact is that the valleys of all the upland tributaries of the Ardèche descend so rapidly, and have so little lateral expansion, as to render the construction of capacious reservoirs in them quite impracticable. Indeed, engineers have found but two points in the whole basin suitable for that purpose, and the reservoirs admissible at these would have only a joint capacity of about 70,000,000 cubic yards, or less than one ninth part of what I suppose to be required. The case of the Ardèche is no doubt an extreme one, both in the topographical character of its basin and in its exposure to excessive rains; but all destructive inundations are, in a certain sense, extreme cases also, and this of the Ardèche serves to show that the construction of reservoirs is not by any means to be regarded as a universal panacea against floods.
Nor, on the other hand, is this measure to be summarily rejected. Nature has adopted it on a great scale, on both flanks of the Alps, and on a smaller, on those of the Adirondacks and lower chains, and in this as in many other instances, her processes may often be imitated with advantage. The validity of the remaining objections to the system under discussion depends on the topography, geology, and special climate of the regions where it is proposed to establish such reservoirs. Many upland streams present numerous points where none of these objections, except those of expense and of danger from the breaking of dams, could have any application. Reservoirs may be so constructed as to retain the entire precipitation of the heaviest thaws and rains, leaving only the ordinary quantity to flow along the channel; they may be raised to such a height as only partially to obstruct the surface drainage; or they may be provided with sluices by means of which their whole contents can be discharged in the dry season and a summer crop be grown upon the ground they cover at high water. The expediency of employing them and the mode of construction depend on local conditions, and no rules of universal applicability can be laid down on the subject.
It is remarkable that nations which we, in the false pride of our modern civilization, so generally regard as little less than barbarian, should have long preceded Christian Europe in the systematic employment of great artificial basins for the various purposes they are calculated to subserve. The ancient Peruvians built strong walls, of excellent workmanship, across the channels of the mountain sources of important streams, and the Arabs executed immense works of similar description, both in the great Arabian peninsula and in all the provinces of Spain which had the good fortune to fall under their sway. The Spaniards of the fifteenth and sixteenth centuries, who, in many points of true civilization and culture, were far inferior to the races they subdued, wantonly destroyed these noble monuments of social and political wisdom, or suffered them to perish, because they were too ignorant to appreciate their value, or too unskilful as practical engineers to be able to maintain them, and some of their most important territories were soon reduced to sterility and poverty in consequence.
Another method of preventing or diminishing the evils of inundation by torrents and mountain rivers, analogous to that employed for the drainage of lakes, consists in the permanent or occasional diversion of their surplus waters, or of their entire currents, from their natural courses, by tunnels or open channels cut through their banks. Nature, in many cases, resorts to a similar process. Most great rivers divide themselves into several arms in their lower course, and enter the sea by different mouths. There are also cases where rivers send off lateral branches to convey a part of their waters into the channel of other streams.[360] The most remarkable of these is the junction between the Amazon and the Orinoco by the natural canal of the Cassiquiare and the Rio Negro. In India, the Cambodja and the Menam are connected by the Anam; the Saluen and the Irawaddi by the Panlaun. There are similar examples, though on a much smaller scale, in Europe. The Torneå and the Calix rivers in Lapland communicate by the Tarando, and in Westphalia, the Else, an arm of the Haase, falls into the Weser.
The change of bed in rivers by gradual erosion of their banks is familiar to all, but instances of the sudden abandonment of a primitive channel are by no means wanting. At a period of unknown antiquity, the Ardèche pierced a tunnel 200 feet wide and 100 high, through a rock, and sent its whole current through it, deserting its former bed, which gradually filled up, though its course remained traceable. In the great inundation of 1827, the tunnel proved insufficient for the discharge of the water, and the river burst through the obstructions which had now choked up its ancient channel, and resumed its original course.[361]
It was probably such facts as these that suggested to ancient engineers the possibility of like artificial operations, and there are numerous instances of the execution of works for this purpose in very remote ages. The Bahr Jusef, the great stream which supplies the Fayoum with water from the Nile, has been supposed, by some writers, to be a natural channel; but both it and the Bahr el Wady are almost certainly artificial canals constructed to water that basin, to regulate the level of Lake Moeris, and possibly, also, to diminish the dangers resulting from excessive inundations of the Nile, by serving as waste-weirs to discharge a part of its surplus waters. Several of the seven ancient mouths of the Nile are believed to be artificial channels, and Herodotus even asserts that King Menes diverted the entire course of that river from the Libyan to the Arabian side of the valley. There are traces of an ancient river bed along the western mountains, which give some countenance to this statement. But it is much more probable that the works of Menes were designed rather to prevent a natural, than to produce an artificial, change in the channel of the river.
Two of the most celebrated cascades in Europe, those of the Teverone at Tivoli and of the Velino at Terni, owe, if not their existence, at least their position and character, to the diversion of their waters from their natural beds into new channels, in order to obviate the evils produced by their frequent floods. Remarkable works of the same sort have been executed in Switzerland, in very recent times. Until the year 1714, the Kander, which drains several large Alpine valleys, ran, for a considerable distance, parallel with the Lake of Thun, and a few miles below the city of that name emptied into the river Aar. It frequently flooded the flats along the lower part of its course, and it was determined to divert it into the Lake of Thun. For this purpose, two parallel tunnels were cut through the intervening rock, and the river turned into them. The violence of the current burst up the roof of the tunnels, and, in a very short time, wore the new channel down not less than one hundred feet, and even deepened the former bed at least fifty feet, for a distance of two or three miles above the tunnel. The lake was two hundred feet deep at the point where the river was conducted into it, but the gravel and sand carried down by the Kander has formed at its mouth a delta containing more than a hundred acres, which is still advancing at the rate of several yards a year. The Linth, which formerly sent its waters directly to the Lake of Zurich, and often produced very destructive inundations, was turned into the Wallensee about forty years ago, and in both these cases a great quantity of valuable land was rescued both from flood and from insalubrity.
In Switzerland, the most terrible inundations often result from the damming up of deep valleys by ice slips or by the gradual advance of glaciers, and the accumulation of great masses of water above the obstructions. The ice is finally dissolved by the heat of summer or the flow of warm waters, and when it bursts, the lake formed above is discharged almost in an instant, and all below is swept down to certain destruction. In 1595, about a hundred and fifty lives and a great amount of property were lost by the eruption of a lake formed by the descent of a glacier into the valley of the Drance, and a similar calamity laid waste a considerable extent of soil in the year 1818. On this latter occasion, the barrier of ice and snow was 3,000 feet long, 600 thick, and 400 high, and the lake which had formed above it contained not less than 800,000,000 cubic feet. A tunnel was driven through the ice, and about 300,000,000 cubic feet of water safely drawn off by it, but the thawing of the walls of the tunnel rapidly enlarged it, and before the lake was half drained, the barrier gave way and the remaining 500,000,000 cubic feet of water were discharged in half an hour. The recurrence of these floods has since been prevented by directing streams of water, warmed by the sun, upon the ice in the bed of the valley, and thus thawing it before it accumulates in sufficient mass to threaten serious danger.
In the cases of diversion of streams above mentioned, important geographical changes have been directly produced by those operations. By the rarer process of draining glacier lakes, natural eruptions of water, which would have occasioned not less important changes in the face of the earth, have been prevented by human agency.
The principal means hitherto relied upon for defence against river inundations has been the construction of dikes along the banks of the streams, parallel to the channel and generally separated from each other by a distance not much greater than the natural width of the bed.[362] If such walls are high enough to confine the water and strong enough to resist its pressure, they secure the lands behind them from all the evils of inundation except those resulting from infiltration; but such ramparts are enormously costly in original construction and maintenance, and, as we have already seen, the filling up of the bed of the river in its lower course, by sand and gravel, involves the necessity of occasionally incurring new expenditures in increasing the height of the banks.[363] They are attended, too, with some collateral disadvantages. They deprive the earth of the fertilizing deposits of the waters, which are powerful natural restoratives of soils exhausted by cultivation; they accelerate the rapidity and transporting power of the current at high water by confining it to a narrower channel, and it consequently conveys to the sea the earthy matter it holds in suspension, and chokes up harbors with a deposit which it would otherwise have spread over a wide surface; they interfere with roads and the convenience of river navigation, and no amount of cost or care can secure them from occasional rupture, in case of which the rush of the waters through the breach is more destructive than the natural flow of the highest inundation.[364]
For these reasons, many experienced engineers are of opinion that the system of longitudinal dikes ought to be abandoned, or, where that cannot be done without involving too great a sacrifice of existing constructions, their elevation should be much reduced, so as to present no obstruction to the lateral spread of extraordinary floods, and they should be provided with sluices to admit the water without violence whenever they are likely to be overflowed. Where dikes have not been erected, and where they have been reduced in height, it is proposed to construct, at convenient intervals, transverse embankments of moderate height running from the banks of the river across the plains to the hills which bound them. These measures, it is argued, will diminish the violence of inundations by permitting the waters to extend themselves over a greater surface and thus retarding the flow of the river currents, and will, at the same time, secure the deposit of fertilizing slime upon all the soil covered by the flood.
Rozet, an eminent French engineer, has proposed a method of diminishing the ravages of inundations, which aims to combine the advantages of all other systems, and at the same time to obviate the objections to which they are all more or less liable.[365] The plan of Rozet is recommended by its simplicity and cheapness as well as its facility and rapidity of execution, and is looked upon with favor by many persons very competent to judge in such matters. He proposes to commence with the amphitheatres in which mountain torrents so often rise, by covering their slopes and filling their beds with loose blocks of rock, and by constructing at their outlets, and at other narrow points in the channels of the torrents, permeable barriers of the same material promiscuously heaped up, much according to the method employed by the ancient Romans in their northern provinces for a similar purpose. By this means, he supposes, the rapidity of the current would be checked, and the quantity of transported pebbles and gravel much diminished.
When the stream has reached that part of its course where it is bordered by soil capable of cultivation, and worth the expense of protection, he proposes to place along one or both sides of the stream, according to circumstances, a line of cubical blocks of stone or pillars of masonry three or four feet high and wide, and at the distance of about eleven yards from each other. The space between the two lines, or between a line and the opposite high bank, would, of course, be determined by observation of the width of the swift-water current at high floods. As an auxiliary measure, small ditches and banks, or low walls of pebbles, should be constructed from the line of blocks across the grounds to be protected, nearly at right angles to the current, but slightly inclining downward, and at convenient distances from each other. Rozet thinks the proper interval would be 300 yards, and it is evident that, if he is right in his main principle, hedges, rows of trees, or even common fences, would in many cases answer as good a purpose as banks and trenches or low walls. The blocks or pillars of stone would, he contends, check the lateral currents so as to compel them to let fall all their pebbles and gravel in the main channel—where they would be rolled along until ground down to sand or silt—and the transverse obstructions would detain the water upon the soil long enough to secure the deposit of its fertilizing slime. Numerous facts are cited in support of the author's views, and I imagine there are few residents of rural districts whose own observation will not furnish testimony confirmatory of their soundness.[366]
The deposit of slime by rivers upon the flats along their banks not only contributes greatly to the fertility of the soil thus flowed, but it subserves a still more important purpose in the general economy of nature. All running streams begin with excavating channels for themselves, or deepening the natural depressions in which they flow;[367] but in proportion as their outlets are raised by the solid material transported by their currents, their velocity is diminished, they deposit gravel and sand at constantly higher and higher points, and so at last elevate, in the middle and lower part of their course, the beds they had previously scooped out.[368] The raising of the channels is compensated in part by the simultaneous elevation of their banks and the flats adjoining them, from the deposit of the finer particles of earth and vegetable mould brought down from the mountains, without which elevation the low grounds bordering all rivers would be, as in many cases they in fact are, mere morasses.
All arrangements which tend to obstruct this process of raising the flats adjacent to the channel, whether consisting in dikes which confine the waters, and, at the same time, augment the velocity of the current, or in other means of producing the last-mentioned effect, interfere with the restorative economy of nature, and at last occasion the formation of marshes where, if left to herself, she would have accumulated inexhaustible stores of the richest soil, and spread them out in plains above the reach of ordinary floods.[369]
If a system of continuous lateral dikes, like those of the Po, had been adopted in Egypt in the early dynasties, when the power and the will to undertake the most stupendous material enterprises were so eminently characteristic of the government of that country, and the waters of the annual inundation consequently prevented from flooding the land, it is conceivable that the productiveness of the small area of cultivable soil in the Nile valley might have been long kept up by artificial irrigation and the application of manures. But nature would have rebelled at last, and centuries before our time the mighty river would have burst the fetters by which impotent man had vainly striven to bind his swelling floods, the fertile fields of Egypt would have been converted into dank morasses, and then, perhaps, in some distant future, when the expulsion of man should have allowed the gradual restoration of the primitive equilibrium, would be again transformed into luxuriant garden and plough land. Fortunately, the "wisdom of Egypt" taught her children better things. They invited and welcomed, not repulsed, the slimy embraces of Nilus, and his favors have been, from the hoariest antiquity, the greatest material blessing ever bestowed upon a people.[370]
The valley of the Po has probably not been cultivated or inhabited so long as that of the Nile, but embankments have been employed on its lower course for at least two thousand years, and for many centuries they have been connected in a continuous chain. I have pointed out in a former chapter the effects produced on the geography of the Adriatic by the deposit of river sediment in the sea at the mouths of the Po, the Adige, and the Brenta. If these rivers had been left unconfined, like the Nile, and allowed to spread their muddy waters at will, according to the laws of nature, the slime they have carried to the coast would have been chiefly distributed over the plains of Lombardy. Their banks would have risen as fast as their beds, the coast line would not have been extended so far into the Adriatic, and, the current of the streams being consequently shorter, the inclination of their channel and the rapidity of their flow would not have been so greatly diminished. Had man spared a reasonable proportion of the forests of the Alps, and not attempted to control the natural drainage of the surface, the Po would resemble the Nile in all its essential characteristics, and, in spite of the difference of climate, perhaps be regarded as the friend and ally, not the enemy and the invader, of the population which dwells upon its banks.[371]
The Nile is larger than all the rivers of Lombardy together,[372] it drains a basin twenty times as extensive, its banks have been occupied by man probably twice as long. But its geographical character has not been much changed in the whole period of recorded history, and, though its outlets have somewhat fluctuated in number and position, its historically known encroachments upon the sea are trifling compared with those of the Po and the neighboring streams. The deposits of the Nile are naturally greater in Upper than in Lower Egypt. They are found to have raised the soil at Thebes about seven feet within the last seventeen hundred years, and in the Delta the rise has been certainly more than half as great.
We shall, therefore, not exceed the truth if we suppose the annually inundated surface of Egypt to have been elevated, upon an average, ten feet, within the last 5,000 years, or twice and a half the period during which the history of the Po is known to us.[373]
We may estimate the present actually cultivated area of Egypt at about 5,500 square statute miles. As I have computed in a note on page 372, that area is not more than half as extensive as under the dynasties of the Pharaohs and the Ptolemies; for—though, in consequence of the elevation of the river bed, the inundations now have a wider natural spread—the industry of the ancient Egyptians conducted the Nile water over a great extent of soil it does not now reach. We may, then, adopt a mean between the two quantities, and we shall probably come near the truth if we assume the convenient number of 7,920 square statute miles as the average measure of the inundated land during the historical period. Taking the deposit on this surface at ten feet, the river sediment let fall on the soil of Egypt within the last fifty centuries would amount to fifteen cubic miles.
Had the Nile been banked in, like the Po, all this deposit, except that contained in the water diverted by canals or otherwise drawn from the river for irrigation and other purposes, would have been carried out to sea.[374] This would have been a considerable quantity; for the Nile holds earth in suspension even at low water, a much larger proportion during the flood, and irrigation must have been carried on during the whole year. The precise amount which would have been thus distributed over the soil is matter of conjecture, but three cubic miles is certainly a liberal estimate. This would leave twelve cubic miles as the quantity which embankments would have compelled the Nile to transport to the Mediterranean over and above what it has actually deposited in that sea. The Mediterranean is shoal for some miles out to sea along the whole coast of the Delta, and the large bays or lagoons within the coast line, which communicate both with the river and the sea, have little depth of water. These lagoons the river deposits would have filled up, and there would still have been surplus earth enough to extend the Delta far into the Mediterranean.[375]