CHAPTER VIII
RIVERS AND LAKES
ON THE DIRECTIONS OF RIVERS
In the last chapter I have alluded to the wanderings of rivers within the limits of their own valleys; we have now to consider the causes which have determined the directions of the valleys themselves.
If a tract of country were raised up in the form of a boss or dome, the rain which fell on it would partly sink in, partly run away to the lower ground. The least inequality in the surface would determine the first directions of the streams, which would carry down any loose material, and thus form little channels, which would be gradually deepened and enlarged. It is as difficult for a river as for a man to get out of a groove.
In such a case the rivers would tend to radiate with more or less regularity from the centre or axis of the dome, as, for instance, in our English lake district (Fig. 37). Derwent Water, Thirlmere, Coniston Water, and Windermere, run approximately N. and S.; Crummock Water, Loweswater, and Buttermere N.W. by S.E.; Waste Water, Ullswater, and Hawes Water N.E. by S.W.; while Ennerdale Water lies nearly E. by W. Can we account in any way, and if so how, for these varied directions?
The mountains of Cumberland and Westmoreland form a more or less oval boss, the axis of which, though not straight, runs practically from E.N.E. to W.S.W., say from Scaw Fell to Shap Fell; and a sketch map shows us almost at a glance that Derwent Water, Thirlmere, Ullswater, Coniston Water, and Windermere run at right angles to this axis; Ennerdale Water is just where the boss ends and the mountains disappear; while Crummock Water and Waste Water lie at the intermediate angles.
Fig. 37.—Map of the Lake District.
So much then for the direction. We have still to consider the situation and origin, and it appears that Ullswater, Coniston Water, the River Dudden, Waste Water, and Crummock Water lie along the lines of old faults, which no doubt in the first instance determined the flow of the water.
Take another case. In the Jura the valleys are obviously (see Fig. 18) in many cases due to the folding of the strata. It seldom happens, however, that the case is so simple. If the elevation is considerable the strata are often fractured, and fissures are produced. Again if the part elevated contains layers of more than one character, this at once establishes differences. Take, for instance, the Weald of Kent (Figs. 38, 39). Here we have (omitting minor layers) four principal strata concerned, namely, the Chalk, Greensand, Weald Clay, and Hastings Sands.
Fig. 38.—a, a, Upper Cretaceous strata, chiefly Chalk, forming the North and South Downs; b, b, Escarpment of Lower Greensand, with a valley between it and the Chalk; c, c, Weald Clay, forming plains; d, Hills formed of Hastings Sand and Clay. The Chalk, etc., once spread across the country, as shown in the dotted lines.
The axis of elevation runs (Fig. 39) from Winchester by Petersfield, Horsham, and Winchelsea to Boulogne, and as shown in the following section, taken from Professor Ramsay, we have on each side of the axis two ridges or "escarpments," one that of the Chalk, the other that of the Greensand, while between the Chalk and the Greensand is a valley, and between the Greensand and the ridge of Hastings Sand an undulating plain, in each case with a gentle slope from about where the London and Brighton railway crosses the Weald towards the east. Under these circumstances we might have expected that the streams draining the Weald would have run in the direction of the axis of elevation, and at the bases of the escarpments, as in fact the Rother does for part of its course, into the sea between the North and South Downs, instead of which as a rule they run north and south, cutting in some cases directly through the escarpments; on the north, for instance, the Wye, the Mole, the Darenth, the Medway, and the Stour; and on the south the Arun, the Addur, the Ouse, and the Cuckmere.
Fig. 39.—Map of the Weald of Kent.
They do not run in faults or cracks, and it is clear that they could not have excavated their present valleys under circumstances such as now exist. They carry us back indeed to a time when the Greensand and Chalk were continued across the Weald in a great dome, as shown by the dotted lines in Fig. 38. They then ran down the slope of the dome, and as the Chalk and Greensand gradually weathered back, a process still in operation, the rivers deepened and deepened their valleys, and thus were enabled to keep their original course.
Other evidence in support of this view is afforded by the presence of gravel beds in some places at the very top of the Chalk escarpment—beds which were doubtless deposited when, what is now the summit of a hill, was part of a continuous slope.
The course of the Thames offers us a somewhat similar instance. It rises on the Oolites near Cirencester, and cuts through the escarpment of the Chalk between Wallingford and Reading. The cutting through the Chalk has evidently been effected by the river itself. But this could not have happened under existing conditions. We must remember, however, that the Chalk escarpment is gradually moving eastwards. The Chalk escarpments indeed are everywhere, though of course only slowly, crumbling away. Between Farnham and Guildford the Chalk is reduced to a narrow ridge known as the Hog's Back. In the same way no doubt the area of the Chalk formerly extended much further west than it does at present, and, indeed, there can be little doubt, somewhat further west than the source of the Thames, almost to the valley of the Severn. At that time the Thames took its origin in a Chalk spring. Gradually, however, the Chalk was worn away by the action of weather, and especially of rain. The river maintained its course while gradually excavating, and sinking deeper and deeper into, the Chalk. At present the river meets the Chalk escarpment near Wallingford, but the escarpment itself is still gradually retreating eastward.
So, again, the Elbe cuts right across the Erz-Gebirge, the Rhine through the mountains between Bingen and Coblenz, the Potomac, the Susquehannah, and the Delaware through the Alleghanies. The case of the Dranse will be alluded to further on (p. 292). In these cases the rivers preceded the mountains. Indeed as soon as the land rose above the waters, rivers would begin their work, and having done so, unless the rate of elevation of the mountain exceeded the power of erosion of the river, the two would proceed simultaneously, so that the river would not alter its course, but would cut deeper and deeper as the mountain range gradually rose.
Rivers then are in many cases older than mountains. Moreover, the mountains are passive, the rivers active. Since it seems to be well established that in Switzerland a mass, more than equal to what remains, has been removed; and that many of the present mountains are not sites which were originally raised highest, but those which have suffered least, it follows that if in some cases the course of the river is due to the direction of the mountain ridges, on the other hand the direction of some of the present ridges is due to that of the rivers. At any rate it is certain that of the original surface not a trace or a fragment remains in situ. Many of our own English mountains were once valleys, and many of our present valleys occupy the sites of former mountain ridges.
Heim and Rütimeyer point out that of the two factors which have produced the relief of mountain regions, the one, elevation, is temporary and transitory; the other, denudation, is constant, and gains therefore finally the upper hand.
We must not, however, expect too great regularity. The degree of hardness, the texture, and the composition of the rocks cause great differences.
On the other hand, if the alteration of level was too rapid, the result might be greatly to alter the river courses. Mr. Darwin mentions such a case, which, moreover, is perhaps the more interesting as being evidently very recent.
"Mr. Gill," he says, "mentioned to me a most interesting, and as far as I am aware, quite unparalleled case, of a subterranean disturbance having changed the drainage of a country. Travelling from Casma to Huaraz (not very far distant from Lima) he found a plain covered with ruins and marks of ancient cultivation, but now quite barren. Near it was the dry course of a considerable river, whence the water for irrigation had formerly been conducted. There was nothing in the appearance of the water-course to indicate that the river had not flowed there a few years previously; in some parts beds of sand and gravel were spread out; in others, the solid rock had been worn into a broad channel, which in one spot was about 40 yards in breadth and 8 feet deep. It is self-evident that a person following up the course of a stream will always ascend at a greater or less inclination. Mr. Gill therefore, was much astonished when walking up the bed of this ancient river, to find himself suddenly going downhill. He imagined that the downward slope had a fall of about 40 or 50 feet perpendicular. We here have unequivocal evidence that a ridge had been uplifted right across the old bed of a stream. From the moment the river course was thus arched, the water must necessarily have been thrown back, and a new channel formed. From that moment also the neighbouring plain must have lost its fertilising stream, and become a desert."[52]
The strata, moreover, often—indeed generally, as we have seen, for instance, in the case of Switzerland—bear evidence of most violent contortions, and even where the convulsions were less extreme, the valleys thus resulting are sometimes complicated by the existence of older valleys formed under previous conditions.
In the Alps then the present configuration of the surface is mainly the result of denudation. If we look at a map of Switzerland we can trace but little relation between the river courses and the mountain chains.
Fig. 40.—Sketch Map of the Swiss Rivers.
The rivers, as a rule (Fig. 40), run either S.E. by N.W., or, at right angles to this, N.E. and S.W. The Alps themselves follow a somewhat curved line from the Maritime Alps, commencing with the islands of Hyères, by Briancon, Martigny, the Valais, Urseren Thal, Vorder Rhein, Innsbruck, Radstadt, and Rottenmann to the Danube, a little below Vienna,—at first nearly north and south, but gradually curving round until it becomes S.W. by N.E.
The central mountains are mainly composed of Gneiss, Granite, and crystalline Schists: the line of junction between these rocks and the secondary and tertiary strata on the north, runs, speaking roughly, from Hyères to Grenoble, and then by Albertville, Sion, Chur, Inns, bruck, Radstadt, and Hieflau, towards Vienna. It is followed (in some part of their course) by the Isère, the Rhone, the Rhine, the Inn, and the Enns. One of the great folds shortly described in the preceding chapter runs up the Isère, along the Chamouni Valley, up the Rhone, through the Urseren Thal, down the Rhine Valley to Chur, along the Inn nearly to Kufstein, and for some distance along the Enns. Thus, then, five great rivers have taken advantage of this main fold, each of them eventually breaking through into a transverse valley.
The Pusterthal in the Tyrol offers us an interesting case of what is obviously a single valley, which has, however, been slightly raised in the centre, near Toblach, so that from this point the water flows in opposite directions—the Drau eastward, and the Rienz westward. In this case the elevation is single and slight: in the main valley there are several, and they are much loftier, still we may, I think, regard that of the Isère from Chambery to Albertville, of the Rhone from Martigny to its source, of the Urseren Thal, of the Vorder Rhine from its source to Chur, of the Inn from Landeck to below Innsbruck, even perhaps of the Enns from Radstadt to Hieflau as in one sense a single valley, due to one of these longitudinal folds, but interrupted by bosses of gneiss and granite,—one culminating in Mont Blanc, and another in the St. Gotthard,—which have separated the waters of the Isère, the Rhone, the Vorder Rhine, the Inn, and the Enns. That the valley of Chamouni, the Valais, the Urseren Thal, and that of the Vorder Rhine really form part of one great fold is further shown by the presence of a belt of Jurassic strata nipped in, as it were, between the crystalline rocks.
This seems to throw light on the remarkable turns taken by the Rhone at Martigny and the Vorder Rhine at Chur, where they respectively quit the great longitudinal fold, and fall into secondary transverse valleys. The Rhone for the upper part of its course, as far as Martigny, runs in the great longitudinal fold of the Valais; at Martigny it falls into and adopts the transverse valley, which properly belongs to the Dranse; for the Dranse is probably an older river and ran in the present course even before the great fold of the Valais. This would seem to indicate that the Oberland range is not so old as the Pennine, and that its elevation was so gradual that the Dranse was able to wear away a passage as the ridge gradually rose. After leaving the Lake of Geneva the Rhone follows a course curving gradually to the south, until it reaches St. Genix, where it falls into and adopts a transverse valley which properly belongs to the little river Guiers; it subsequently joins the Ain and finally falls into the Saône. If these valleys were attributed to their older occupiers we should therefore confine the name of the Rhone to the portion of its course from the Rhone glacier to Martigny. From Martigny it occupies successively the valleys of the Dranse, Guiers, Ain, and Saône. In fact, the Saône receives the Ain, the Ain the Guiers, the Guiers the Dranse, and the Dranse the Rhone. This is not a mere question of names, but also one of antiquity. The Saône, for instance, flowed past Lyons to the Mediterranean for ages before it was joined by the Rhone. In our nomenclature, however, the Rhone has swallowed up the others. This is the more curious because of the three great rivers which unite to form the lower Rhone, namely, the Saône, the Doubs, and the Rhone itself, the Saône brings for a large part of the year the greatest volume of water, and the Doubs has the longest course. Other similar cases might be mentioned. The Aar, for instance, is a somewhat larger river than the Rhine.
Fig. 41.—Diagram in illustration of Mountain structure.
But why should the rivers, after running for a certain distance in the direction of the main axis, so often break away into lateral valleys? If the elevation of a chain of mountains be due to the causes suggested in p. 214, it is evident, though, so far as I am aware, stress has not hitherto been laid upon this, that the compression and consequent folding of the strata (Fig. 41) would not be in the direction A B only, but also at right angles to it, in the direction A C, though the amount of folding might be much greater in one direction than in the other. Thus in the case of Switzerland, while the main folds run south-west by north-east, there would be others at right angles to the main axis. The complex structure of the Swiss mountains may be partly due to the coexistence of these two directions of pressure at right angles to one another. The presence of a fold so originating would often divert the river to a course more or less nearly at right angles to its original direction.
Switzerland, moreover, slopes northwards from the Alps, so that the lowest part of the great Swiss plain is that along the foot of the Jura. Hence the main drainage runs along the line from Yverdun to Neuchâtel, down the Zihl to Soleure, and then along the Aar to Waldshut: the Upper Aar, the Emmen, the Wiggern, the Suhr, the Wynen, the lower Reuss, the Sihl, and the Limmat, besides several smaller streams, running approximately parallel to one another north-north-east, and at angles to the main axis of elevation, and all joining the Aar from the south, while on the north it does not receive a single contributary of any importance.
On the south side of the Alps again we have the Dora Baltea, the Sesia, the Ticino, the Olonna, the Adda, the Adige, etc., all running south-south-east from the axis of elevation to the Po.
Fig. 42.
Indeed, the general slope of Switzerland, being from the ridge of the Alps towards the north, it will be observed (Fig. 42) that almost all the large affluents of these rivers running in longitudinal valleys fall in on the south, as, for instance, those of the Isère from Albertville to Grenoble, of the Rhone from its source to Martigny, of the Vorder Rhine from its source to Chur, of the Inn from Landeck to Kufstein, of the Enns from its source to near Admont, of the Danube from its source to Vienna, and as just mentioned, of the Aar from Bern to Waldshut. Hence also, whenever the Swiss rivers running east and west break into a transverse valley, as the larger ones all do, and some more than once, they invariably, whether originally running east or westwards, turn towards the north.
But although we thus get a clue to the general structure of Switzerland, the whole question is extremely complex, and the strata have been crumpled and folded in the most complicated manner, sometimes completely reversed, so that older rocks have been folded back on younger strata, and even in some cases these folds again refolded. Moreover, the denudation by aerial action, by glaciers, frosts, and rivers has removed hundreds, or rather thousands, of feet of strata. In fact, the mountain tops are not by any means the spots which have been most elevated, but those which have been least denuded; and hence it is that so many of the peaks stand at about the same altitude.
THE CONFLICTS AND ADVENTURES OF RIVERS
Our ancestors looked upon rivers as being in some sense alive, and in fact in their "struggle for existence" they not only labour to adapt their channel to their own requirements, but in many cases enter into conflict with one another.
In the plain of Bengal, for instance, there are three great rivers, the Brahmapootra coming from the north, the Ganges from the west, and the Megna from the east, each of them with a number of tributary streams. Mr. Fergusson[53] has given us a most interesting and entertaining account of the struggles between these great rivers to occupy the fertile plain of Bengal.
The Megna, though much inferior in size to the Brahmapootra, has one great advantage. It depends mainly on the monsoon rains for its supply, while the Brahmapootra not only has a longer course to run, but relies for its floods, to a great extent, on the melting of the snow, so that, arriving later at the scene of the struggle, it finds the country already occupied by the Megna to such an extent that it has been driven nearly 70 miles northwards, and forced to find a new channel.
Under these circumstances it has attacked the territory of the Ganges, and being in flood earlier than that river, though later than the Megna, it has in its turn a great advantage.
Whatever the ultimate result may be the struggle continues vigorously. At Sooksaghur, says Fergusson, "there was a noble country house, built by Warren Hastings, about a mile from the banks of the Hoogly. When I first knew it in 1830 half the avenue of noble trees, which led from the river to the house, was gone; when I last saw it, some eight years afterwards, the river was close at hand. Since then house, stables, garden, and village are all gone, and the river was on the point of breaking through the narrow neck of high land that remained, and pouring itself into some weak-banded nullahs in the lowlands beyond: and if it had succeeded, the Hoogly would have deserted Calcutta. At this juncture the Eastern Bengal Railway Company intervened. They were carrying their works along the ridge, and they have, for the moment at least, stopped the oscillation in this direction."
This has affected many of the other tributaries of the Ganges, so that the survey made by Rennell in 1780-90 is no longer any evidence as to the present course of the rivers. They may now be anywhere else; in some cases all we can say is that they are certainly not now where they were then.
The association of the three great European rivers, the Rhine, the Rhone, and the Danube, with the past history of our race, invests them with a singular fascination, and their past history is one of much interest. They all three rise in the group of mountains between the Galenstock and the Bernardino, within a space of a few miles; on the east the waters run into the Black Sea, on the north into the German Ocean, and on the west into the Mediterranean. But it has not always been so. Their head-waters have been at one time interwoven together.
At present the waters of the Valais escape from the Lake of Geneva at the western end, and through the remarkable defile of Fort de l'Ecluse and Malpertius, which has a depth of 600 feet, and is at one place not more than 14 feet across. Moreover, at various points round the Lake of Geneva, remains of lake terraces show that the water once stood at a level much higher than the present. One of these is rather more than 250 feet[54] above the lake.
A glance at the map will show that between Lausanne and Yverdun there is a low tract of land, and the Venoge, which falls into the Lake of Geneva between Lausanne and Morges, runs within about half a mile of the Nozon, which falls into the Lake of Neuchâtel at Yverdun, the two being connected by the Canal d'Entreroches, and the height of the watershed being only 76 metres (250 feet), corresponding with the above mentioned lake terrace. It is evident, therefore, that when the Lake of Geneva stood at the level of the 250 feet terrace the waters ran out, not as now at Geneva and by Lyons to the Mediterranean, but near Lausanne by Cissonay and Entreroches to Yverdun, and through the Lake of Neuchâtel into the Aar and the Rhine.
But this is not the whole of the curious history. At present the Aar makes a sharp turn to the west at Waldshut, where it falls into the Rhine, but there is reason to believe that at a former period, before the Rhine had excavated its present bed, the Aar continued its course eastward to the Lake of Constance, by the valley of the Klettgau, as is indicated by the presence of gravel beds containing pebbles which have been brought, not by the Rhine from the Grisons, but by the Aar from the Bernese Oberland, showing that the river which occupied the valley was not the Rhine but the Aar. It would seem also that at an early period the Lake of Constance stood at a considerably higher level, and that the outlet was, perhaps, from Frederichshaven to Ulm, along what are now the valleys of the Schussen and the Ried, into the Danube.
Thus the head-waters of the Rhone appear to have originally run by Lausanne and the Lake of Constance into the Danube, and so to the Black Sea. Then, after the present valley was opened between Waldshut and Basle, they flowed by Basle and the present Rhine, and after joining the Thames, over the plain which now forms the German Sea into the Arctic Ocean between Scotland and Norway. Finally, after the opening of the passage at Fort de l'Ecluse, by Geneva, Lyons, and the Valley of the Saône, to the Mediterranean.
It must not, however, be supposed that these changes in river courses are confined to the lower districts. Mountain streams have also their adventures and vicissitudes, their wars and invasions. Take for instance the Upper Rhine, of which we have a very interesting account by Heim. It is formed of three main branches, the Vorder Rhine, Hinter Rhine, and the Albula. The two latter, after meeting near Thusis, unite with the Vorder Rhine at Reichenau, and run by Chur, Mayenfeld, and Sargans into the Lake of Constance at Rheineck. At some former period, however, the drainage of this district was very different, as is shown in Fig. 43.
The Vorder and Hinter Rhine united then (Fig. 43) as they do now at Reichenau, but at a much higher level, and ran to Mayenfeld, not by Chur, but by the Kunckel Pass to Sargans, and so on, not to the Lake of Constance, but to that of Zurich. The Landwasser at that time rose in the Schlappina Joch, and after receiving as tributaries the Vereina and the Sardasca, joined the Albula, as it does now at Tiefenkasten; but instead of going round to meet the Hinter Rhine near Thusis, the two together travelled parallel with, but at some distance from, the Hinter Rhine, by Heide to Chur, and so to Mayenfeld.
In the meanwhile, however, the Landquart was stealthily creeping up the valley, attacked the ridge which then united the Casanna and the Madrishorn, and gradually forcing the passage, invaded (Fig. 44) the valleys of the Schlappina, Vereina, and Sardasca, absorbed them as tributaries, and, detaching them from their allegiance to the Landwasser, annexed the whole of the upper province which had formerly belonged to that river.
Fig. 43.—River system round Chur, as it used to be.
The Schyn also gradually worked its way upwards from Thusis till it succeeded in sapping the Albula, and carried it down the valley to join the Vorder Rhine near Thusis. In what is now the main valley of the Rhine above Chur another stream ate its way back, and eventually tapped the main river at Reichenau, thus diverting it from the Kunckel, and carrying it round by Chur.
Fig. 44.—River system round Chur, as it is.
At Sargans a somewhat similar process was repeated, with the addition that the material brought down by the Weisstannen, or perhaps a rockfall, deflected the Rhine, just as we see in Fig. 30 that the Rhone was pushed on one side by the Borgne. The Rhone, however, had no choice, it was obliged to force, and has forced its way over the cone deposited by the Borgne. The Rhine, on the contrary, had the option of running down by Vaduz to Rheinach, and has adopted this course. The watershed between it and the Weisstannen is, however, only about 20 feet in height, and the people of Zurich watch it carefully, lest any slight change should enable the river to return to its old bed. The result of all these changes is that the rivers have changed their courses from those shown in Fig. 43 to their present beds as shown in Fig. 44.
Another interesting case is that of the Upper Engadine (Fig. 45), to which attention has been called by Bonney and Heim. The fall of the Val Bregaglia is much steeper than that of the Inn, and the Maira has carried off the head-waters of that river away into Italy. The Col was formerly perhaps as far south as Stampa: the Albegna, the Upper Maira, and the stream from the Forgno Glacier, originally belonged to the Inn, but have been captured by the Lower Maira. Their direction still indicates this; they seem as if they regretted the unwelcome change, and yearned to rejoin their old companions.
Fig. 45.—River system of the Maloya.
Moreover, as rivers are continually cutting back their valleys they must of course sometimes meet. In these cases when the valleys are at different levels the lower rivers have drained the upper ones, and left dry, deserted valleys. In other cases, especially in flatter districts, we have bifurcations, as, for instance, at Sargans, and several of the Italian lakes. Every one must have been struck by the peculiar bifurcation of the Lakes of Como and Lugano, while a very slight depression would connect the Lake Varese with the Maggiore, and give it also a double southern end.
ON LAKES
The problem of the origin of Lakes is by no means identical with that of Valleys. The latter are due, primarily as a rule to geological causes, but so far as their present condition is concerned, mainly to the action of rain and rivers. Flowing water, however, cannot give rise to lakes.
It is of course possible to have valleys without lakes, and in fact the latter are, now at least, exceptional. There can be no lakes if the slope of the valley is uniform. To what then are lakes due?
Professor Ramsay divides Lakes into three classes:—
1. Those due to irregular accumulations of drift, and which are generally quite shallow.
2. Those formed by moraines.
3. Those which occupy true basins scooped by glacier ice out of the solid rock.
To these must, however, I think be added at least one other great class and several minor ones, namely,—
4. Those due to inequalities of elevation or depression.
5. Lakes in craters of extinct volcanoes, for instance, Lake Avernus.
6. Those caused by subsidence due to the removal of underlying soluble rocks, such as some of the Cheshire Meres.
7. Loop lakes in deserted river courses, of which there are many along the course of the Rhine.
8. Those due to rockfalls, landslips, or lava currents, damming up the course of a river.
9. Those caused by the advance of a glacier across a lateral valley, such as the Mergelen See, or the ancient lake whose margins form the celebrated "Parallel Roads of Glen Roy."
As regards the first class we find here and there on the earth's surface districts sprinkled with innumerable shallow lakes of all sizes, down to mere pools. Such, for instance, occur in the district of Le Doubs between the Rhone and the Saône, that of La Sologne near Orleans, in parts of North America, and in Finland. Such lakes are, as a rule, quite shallow. Some geologists, Geikie, for instance, ascribe them to the fact of these regions having been covered by sheets of ice which strewed the land with irregular masses of clay, gravel, and sand, lying on a stratum impervious to water, either of hard rock such as granite or gneiss, or of clay, so that the rain cannot percolate through it, and without sufficient inclination to throw it off.
2. To Ramsay's second class of Lakes belong those formed by moraines. The materials forming moraines being, however, comparatively loose, are easily cut through by streams. There are in Switzerland many cases of valleys crossed by old moraines, but they have generally been long ago worn through by the rivers.
3. Ramsay and Tyndall attribute most of the great Swiss and Italian lakes to the action of glaciers, and regard them as rock basins. It is of course obvious that rivers cannot make basin-shaped hollows surrounded by rock on all sides. The Lake of Geneva, 1230 feet above the sea, is over 1000 feet deep; the Lake of Brienz is 1850 feet above the sea, and 2000 feet deep, so that its bottom is really below the sea level. The Italian Lakes are even more remarkable. The Lake of Como, 700 feet above the sea, is 1929 feet deep. Lago Maggiore, 685 feet above the sea, is no less than 2625 feet deep.
If the mind is at first staggered at the magnitude of the scale, we must remember that the ice which is supposed to have scooped out the valley in which the Lake of Geneva now reposes, was once at least 4000 feet thick; while the moraines were also of gigantic magnitude, that of Ivrea, for instance, being no less than 1500 feet above the river, and several miles long.
Indeed it is obvious that a glacier many hundred, or in some cases several thousand, feet in thickness, must exercise great pressure on the bed over which it travels. We see this from the striæ and grooves on the solid rocks, and the fine mud which is carried down by glacial streams. The deposit of glacial rivers, the "loess" of the Rhine itself, is mainly the result of this ice-waste, and that is why it is so fine, so impalpable. That glaciers do deepen their beds seems therefore unquestionable.
Moreover, though the depth of some of these lakes is great, the true slope is very slight.
Tyndall and Ramsay do not deny that the original direction of valleys, and consequently of lakes, is due to cosmical causes and geological structure, while even those who have most strenuously opposed the theory which attributes lakes to glacial erosion do not altogether deny the action of glaciers. Favre himself admits that "it is impossible to deny that valleys, after their formation, have been swept out and perhaps enlarged by rivers and glaciers."
Even Ruskin admits "that a glacier may be considered as a vast instrument of friction, a white sand-paper applied slowly but irresistibly to all the roughness of the hill which it covers."
It is obvious that sand-paper applied "irresistibly" and long enough, must gradually wear away and lower the surface. I cannot therefore resist the conclusion that glaciers have taken an important part in the formation of lakes.
The question has sometimes been discussed as if the point at issue were whether rivers or glaciers were the most effective as excavators. But this is not so. Those who believe that lakes are in many cases due to glaciers might yet admit that rivers have greater power of erosion. There is, however, an essential difference in the mode of action. Rivers tend to regularise their beds; they drain, rather than form lakes. Their tendency is to cut through any projections so that finally their course assumes some such curve as that below, from the source (a) to its entrance into the sea (b).
Fig. 46.—Final Slope of a River.
Glaciers, however, have in addition a scooping power, so that if similarly a d b in Fig. 47 represent the course of a glacier, starting at a and gradually thinning out to e, it may scoop out the rock to a certain extent at d; in that case if it subsequently retires say to c, there would be a lake lying in the basin thus formed between c and e.
Fig. 47.
On the other hand I am not disposed to attribute the Swiss lakes altogether to the action of glaciers. In the first place it does not seem clear that they occupy true rock basins. On this point more evidence is required. That some lakes are due to unequal changes of level will hardly be denied. No one, for instance, as Bonney justly observes,[55] would attribute the Dead Sea to glacial erosion.
The Alps, as we have seen, are a succession of great folds, and there is reason to regard the central one as the oldest. If then the same process continued, and the outer fold was still further raised, or a new one formed, more quickly than the rivers could cut it back, they would be dammed up, and lakes would result.
Moreover, if the formation of a mountain region be due to subsidence, and consequent crumpling, as indicated on p. 217, so that the strata which originally occupied the area A B C D are compressed into A' B' C' D', it is evident, as already mentioned, that while the line of least resistance, and, consequently, the principal folds might be in the direction A' B', there must also be a tendency to the formation of similar folds at right angles, or in the direction A' C'. Thus, in the case of Switzerland, while the main folds run south-west by north-east there would also be others at right angles, though the amount of folding might be much greater in the one direction than in the other. To this cause the bosses, for instance—at Martigny, the Furca, and the Ober Alp,—which intersect the great longitudinal valley of Switzerland, are perhaps due.
The great American lakes also are probably due to differences of elevation. Round Lake Ontario, for instance, there is a raised beach which at the western end of the lake is 363 feet above the sea level, but rises towards the East and North until near Fine it reaches an elevation of 972 feet. As this terrace must have been originally horizontal we have here a lake barrier, due to a difference of elevation, amounting to over 600 feet.
In the same way we get a clue to the curious cruciform shape of the Lake of Lucerne as contrasted with the simple outline of such lakes as those of Neuchâtel or Zurich. That of Lucerne is a complex lake. Soundings have shown that the bottom of the Urner See is quite flat. It is in fact the old bed of the Reuss, which originally ran, not as now by Lucerne, but by Schwytz and through the Lake of Zug. In the same way the Alpnach See is the old bed of the Aa, which likewise ran through the Lake of Zug. The old river terraces of the Reuss can be traced in places between Brunnen and Goldau. Now these terraces must have originally sloped from the upper part downwards, from Brunnen towards Goldau. But at present the slope is the other way, i.e. from Goldau towards Brunnen. From this and other evidence we conclude that in the direction from Lucerne towards Rapperschwyl there has been an elevation of the land, which has dammed up the valleys and thus turned parts of the Aa and the Reuss into lakes—the two branches of the Lake of Lucerne known as the Alpnach See and Urner See.
During the earthquakes of 1819 while part of the Runn of Cutch, 2000 square miles in area, sunk several feet, a ridge of land, called by the natives the Ulla-Bund or "the wall of God," thirty miles long, and in parts sixteen miles wide, was raised across an ancient arm of the Indus, and turned it temporarily into a lake.
In considering the great Italian lakes, which descend far below the sea level, we must remember that the Valley of the Po is a continuation of the Adriatic, now filled up and converted into land, by the materials brought down from the Alps. Hence we are tempted to ask whether the lakes may not be remains of the ancient sea which once occupied the whole plain. Moreover just as the Seals of Lake Baikal in Siberia carry us back to the time when that great sheet of fresh water was in connection with the Arctic Ocean, so there is in the character of the Fauna of the Italian lakes, and especially the presence of a Crab in the Lake of Garda, some confirmation of such an idea. Further evidence, however, is necessary before these interesting questions can be definitely answered.
Lastly, some lakes and inland seas seem to be due to even greater cosmical causes. Thus a line inclined ten degrees to the pole beginning at Gibraltar would pass through a great chain of inland waters—the Mediterranean, Black Sea, Caspian, Aral, Baikal, and back again through the great American lakes.
But though many causes have contributed to the original formation and direction of Valleys, their present condition is mainly due to the action of water. When we contemplate such a valley, for example, as that which is called par excellence the "Valais," we can at first hardly bring ourselves to realise this; but we can trace up valleys, from the little water-course made by last night's rains up to the greatest valleys of all.
These considerations, however, do not of course apply to such depressions as those of the great oceans. These were probably formed when the surface of the globe began to solidify, and, though with many modifications, have maintained their main features ever since.
ON THE CONFIGURATION OF VALLEYS
The conditions thus briefly described repeat themselves in river after river, valley after valley, and it adds, I think, very much to the interest with which we regard them if, by studying the general causes to which they are due, we can explain their origin, and thus to some extent understand the story they have to tell us, and the history they record.
What, then, has that history been? The same valley may be of a very different character, and due to very different causes, in different parts of its course. Some valleys are due to folds (see Fig. 41) caused by subterranean changes, but by far the greater number are, in their present features, mainly the result of erosion. As soon as any tract of land rose out of the sea, the rain which fell on the surface would trickle downwards in a thousand rills, forming pools here and there (see Fig. 37), and gradually collecting into larger and larger streams. Wherever the slope was sufficient the water would begin cutting into the soil and carrying it off to the sea. This action would be the same in any case, but, of course, would differ in rapidity according to the hardness of the ground. On the other hand, the character of the valley would depend greatly on the character of the strata, being narrow where they were hard and tough; broader, on the contrary, where they were soft, so that they crumbled readily into the stream, or where they were easily split by the weather. Gradually the stream would eat into its bed until it reached a certain slope, the steepness of which would depend on the volume of water. The erosive action would then cease, but the weathering of the sides and consequent widening would continue, and the river would wander from one part of its valley to another, spreading the materials and forming a river plain. At length, as the rapidity still further diminished, it would no longer have sufficient power even to carry off the materials brought down. It would form, therefore, a cone or delta, and instead of meandering, would tend to divide into different branches. These three stages, we may call those of—
1. Deepening and widening;
2. Widening and levelling;
3. Filling up;
and every place in the second stage has passed through the first; every one in the third has passed through the second.
A velocity of 6 inches per second will lift fine sand, 8 inches will move sand as coarse as linseed, 12 inches will sweep along fine gravel, 24 inches will roll along rounded pebbles an inch diameter, and it requires 3 feet per second at the bottom to sweep along angular stones of the size of an egg.
When a river has so adjusted its slope that it neither deepens its bed in the upper portion of its course, nor deposits materials, it is said to have acquired its "regimen," and in such a case if the character of the soil remains the same, the velocity must also be uniform. The enlargement of the bed of a river is not, however, in proportion to the increase of its waters as it approaches the sea. If, therefore, the slope did not diminish, the regimen would be destroyed, and the river would again commence to eat out its bed. Hence as rivers enlarge, the slope diminishes, and consequently every river tends to assume some such "regimen" as that shown in Fig. 46.
Now, suppose that the fall of the river is again increased, either by a fresh elevation, or locally by the removal of a barrier. Then once more the river regains its energy. Again it cuts into its old bed, deepening the valley, and leaving the old plain as a terrace high above its new course. In many valleys several such terraces may be seen, one above the other. In the case of a river running in a transverse valley, that is to say of a valley lying at right angles to the "strike" or direction of the strata (such, for instance, as the Reuss), the water acts more effectively than in longitudinal valleys running along the strike. Hence the lateral valleys have been less deeply excavated than that of the Reuss itself, and the streams from them enter the main valley by rapids or cascades. Again, rivers running in transverse valleys cross rocks which in many cases differ in hardness, and of course they cut down the softer strata more rapidly than the harder ones; each ridge of harder rock will therefore form a dam and give rise to a rapid, or cataract. We often as we ascend a river, after a comparatively flat plain, find ourselves in a narrow defile, down which the water rushes in an impetuous torrent, but at the summit of which, to our surprise, we find another broad flat valley.
Another lesson which we learn from the study of river valleys, is that, just as geological structure was shown by Sir C. Lyell to be no evidence of cataclysms, but the result of slow action; so also the excavation of valleys is due mainly to the regular flow of rivers; and floods, though their effects are more sudden and striking, have had, after all, comparatively little part in the result.
The mouths of rivers fall into two principal classes. If we look at any map we cannot but be struck by the fact that some rivers terminate in a delta, some in an estuary. The Thames, for instance, ends in a noble estuary, to which London owes much of its wealth and power. It is obvious that the Thames could not have excavated this estuary while the coast was at its present level. But we know that formerly the land stood higher, that the German Ocean was once dry land, and the Thames, after joining the Rhine, ran northwards, and fell eventually into the Arctic Ocean. The estuary of the Thames, then, dates back to a period when the south-east of England stood at a higher level than the present, and even now the ancient course of the river can be traced by soundings under what is now sea. The sites of present deltas, say of the Nile, were also once under water, and have been gradually reclaimed by the deposits of the river.
It would indeed be a great mistake to suppose that rivers always tend to deepen their valleys. This is only the case when the slope exceeds a certain angle. When the fall is but slight they tend on the contrary to raise their beds by depositing sand and mud brought down from higher levels. Hence in the lower part of their course many of the most celebrated rivers—the Nile, the Po, the Mississippi, the Thames, etc.—run upon embankments, partly of their own creation.
Fig. 48.—Diagrammatic section of a valley (exaggerated) R R, rocky basis of valley; A A, sedimentary strata; B, ordinary level of river; C, flood level.
The Reno, the most dangerous of all the Apennine rivers, is in some places as much as 30 feet above the adjoining country. Rivers under such conditions, when not interfered with by Man, sooner or later break through their banks, and leaving their former bed, take a new course along the lowest part of their valley, which again they gradually raise above the rest. Hence, unless they are kept in their own channels by human agency, such rivers are continually changing their course.
If we imagine a river running down a regularly inclined plane in a more or less straight line; any inequality or obstruction would produce an oscillation, which when once started would go on increasing until the force of gravity drawing the water in a straight line downwards equals that of the force tending to divert its course. Hence the radius of the curves will follow a regular law depending on the volume of water and the angle of inclination of the bed. If the fall is 10 feet per mile and the soil homogeneous, the curves would be so much extended that the course would appear almost straight. With a fall of 1 foot per mile the length of the curve is, according to Fergusson, about six times the width of the river, so that a river 1000 feet wide would oscillate once in 6000 feet. This is an important consideration, and much labour has been lost in trying to prevent rivers from following their natural law of oscillation. But rivers are very true to their own laws, and a change at any part is continued both upwards and downwards, so that a new oscillation in any place cuts its way through the whole plain of the river both above and below.
The curves of the Mississippi are, for instance, for a considerable part of its course so regular that they are said to have been used by the Indians as a measure of distance.
If the country is flat a river gradually raises the level on each side, the water which overflows during floods being retarded by reeds, bushes, trees, and a thousand other obstacles, gradually deposits the solid matter which it contains, and thus raising the surface, becomes at length suspended, as it were, above the general level. When this elevation has reached a certain point, the river during some flood bursts its banks, and deserting its old bed takes a new course along the lowest accessible level. This then it gradually fills up, and so on; coming back from time to time if permitted, after a long cycle of years, to its first course.
In evidence of the vast quantity of sediment which rivers deposit, I may mention that the river-deposits at Calcutta are more than 400 feet in thickness.
In addition to temporary "spates," due to heavy rain, most rivers are fuller at one time of year than another, our rivers, for instance, in winter, those of Switzerland, from the melting of the snow, in summer. The Nile commences to rise towards the beginning of July; from August to October it floods all the low lands, and early in November it sinks again. At its greatest height the volume of water sometimes reaches twenty times that when it is lowest, and yet perhaps not a drop of rain may have fallen. Though we now know that this annual variation is due to the melting of the snow and the fall of rain on the high lands of Central Africa, still when we consider that the phenomenon has been repeated annually for thousands of years it is impossible not to regard it with wonder. In fact Egypt itself may be said to be the bed of the Nile in flood time.
Some rivers, on the other hand, offer no such periodical differences. The lower Rhone, for instance, below the junction with the Saône, is nearly equal all through the year, and yet we know that the upper portion is greatly derived from the melting of the Swiss snows. In this case, however, while the Rhone itself is on this account highest in summer and lowest in winter, the Saône, on the contrary, is swollen by the winter's rain, and falls during the fine weather of summer. Hence the two tend to counterbalance one another.
Periodical differences are of course comparatively easy to deal with. It is very different with floods due to irregular rainfall. Here also, however, the mere quantity of rain is by no means the only matter to be considered. For instance a heavy rain in the watershed of the Seine, unless very prolonged, causes less difference in the flow of the river, say at Paris, than might at first have been expected, because the height of the flood in the nearer affluents has passed down the river before that from the more distant streams has arrived. The highest level is reached when the rain in the districts drained by the various affluents happens to be so timed that the different floods coincide in their arrival at Paris.
FOOTNOTES:
[52] Darwin's Voyage of a Naturalist.
[53] Geol. Jour., 1863.
[54] Favre, Rech. Geol. de la Savoie.
[55] Growth and Structure of the Alps.