Many familiar facts demonstrate the general presence of abundant water beneath the surface of the land. The thousands of wells in regions peopled by civilized races, and the countless springs which issue from the sides of mountains and valleys are a sufficient proof both of the wide distribution of ground-water and of its great abundance.

Certain well-known facts make it clear that ground-water is intimately connected with rainfall. In a dry season the level of the water in wells commonly sinks, and after a heavy rain it rises (p. 71); and the amount of sinking is greater when the drought is long, and the rise is most notable when the rainfall is heavy. Many springs which discharge large quantities of water during a wet season flow with reduced volume, or cease to flow altogether in periods of drought. Furthermore, the water of springs and wells has the properties which rain-water would possess after sinking beneath the surface and dissolving mineral substances. Rain-water is seen to sink beneath the surface with every shower, and since this source seems altogether adequate for ground-water, and since no other source is known whence any considerable amount of ground-water might come, it is concluded that atmospheric precipitation is its chief source.

Water gets beneath the surface by processes which are readily seen. Wherever the soil is porous some of the rain which falls upon it is absorbed. Sinking through the soil to the solid rock it finds cracks and pores through which it descends to great depths. Nowhere are the rocks beneath the mantle rock so compact and so free from cracks, when any considerable area is considered, as to prevent the percolation of water through them.

Conditions influencing descent of rain-water.—There are several conditions which influence not only the amount of water which sinks beneath the surface in a given area, but the proportion of the precipitation which follows this course. These are as follows: (1) Amount of precipitation.—In a general way it is true that the greater the amount of precipitation the greater the amount of water which will sink beneath the surface. Other things being equal, a region of heavy precipitation is a region where wells are easily obtained and springs common. (2) Rate of precipitation.—A given amount of precipitation may be concentrated in a short interval, or distributed through a considerable period of time. In the latter case more of the water sinks beneath the surface; in the former, a larger proportion runs off over the surface. The reason is readily seen. Water passes through small spaces, such as those of soil, slowly, and its rate of passage decreases rapidly with decreasing size of the passageways. When rain falls rapidly on a surface of even moderately close texture, the uppermost layer of soil is promptly filled with water, and since the water passes downward slowly, the uppermost saturated part of the soil becomes virtually impervious. While in this condition, the water which falls on it will run off if there be slope, and stand if there be none. In the latter case it will sink slowly as the water in the soil passes down to lower levels. If precipitation takes place no faster than the water can sink through the soil, all the water may become ground-water. (3) The topography of the surface has much to do with determining the proportion of rainfall which becomes ground-water. If the surface be flat, more will sink in; if it be sloping, more of it will run off before it has time to sink. Other things being equal, the steeper the slope the larger the proportion of the rainfall which will run off over it. (4) The texture of the soil, or other material on which the rain falls, helps to determine what proportion of it sinks beneath the surface. If the surface materials be porous, the water sinks readily; if of close texture, it finds less ready ingress. Other things being equal, the closer the texture of the soil the less the proportion of the rainfall which will enter it. (5) The texture and structure of the rock beneath the surface have some influence on the amount of ground-water. The rock may be stratified or massive; it may be abundantly or sparsely jointed; it may be porous or compact. On the whole, stratified rock is more favorable for the entrance of water than unstratified, partly because of its greater average porosity, and partly because the planes of division between beds often allow the passage of water. If the beds of stratified rock are vertical or inclined, water finds its way into them more readily than if they are horizontal, in so far as it descends along stratification planes. Horizontally bedded rock, or rock which is not bedded at all, may be so much jointed, and the joints so open, as to allow the water to enter readily.

The conditions favorable to the sinking of abundant water below the surface are therefore heavy precipitation, falling slowly on a surface with little relief, a soil of open texture underlain by rock which is porous, or affected by vertical or highly inclined planes of cleavage. The annual discharge of water by rivers is estimated to be about 22 percent. of the rainfall on the land.[93]

Supply of ground-water not altogether dependent on local rainfall.—The amount of ground-water in a given region is not always entirely dependent on the local rainfall. Ground-water is in constant movement, and entering the soil or rock at one point it may, after a long subterranean journey, reach a point far from that at which it entered. Thus beneath the Great Plains of the West there is much subterranean water which fell on the eastern slopes of the mountains to the west. It has flowed beneath the surface to the Plains, where some of it is now withdrawn for the purposes of irrigation in regions where rainfall is deficient. The accompanying diagram (Fig. 199) illustrates the flow here described.

The ground-water surface. Water table.—The water table has already been defined (p. 71) as the upper surface of the ground-water. In a flat region of uniform structure the ground-water surface is essentially level, but rises and falls with the rainfall. Where the topography of a region is not flat, the ground-water surface is not level. As a rule it is higher, though farther below the surface, under an elevation than under surrounding lowlands, as illustrated by Fig. 200. The explanation is not far to seek. If a hill of sand be exposed to rainfall, most of the water falling on its porous surface will sink into it. If the precipitation continues long enough, as in a protracted rain, the hill of sand will be filled with water, the water occupying the interstices between the grains. If the sand of the hill could be removed, leaving the water which it contains on the same area, it would constitute a mound perhaps a third or a fourth as high as the hill itself. If unsupported, this mound of water would spread promptly in all directions until its surface was level. While the sand remains, the water in it constitutes a mound, and has a tendency to spread. It does in fact spread, but since the process involves great friction the spreading is slow. With the spreading the surface of the water in the sand sinks, and sinks fastest at the center where it is highest (b, Fig. 201). If the process were not interrupted the surface of the water in the hill would, in time, sink approximately to the level of the water in the surrounding land (d, Fig. 201); but at every stage preceding the last, the surface of the water would be higher beneath the summit of the hill than elsewhere, though farther from the surface. In regions of even moderate precipitation the water surface beneath the hills rarely sinks to the level of that in the lowlands adjacent, before being raised by further rains.

The water-level beneath the lowlands also sinks. Some of it finds its way into valleys, some of it sinks to greater depths, and some of it evaporates; but since the water surface beneath the elevation sinks more rapidly than that beneath the lowland, the two approach a common level. Their difference will be least at the end of a drought, and greatest just after heavy rains.

Depth to which ground-water sinks.—The depth to which ground-water penetrates has not been determined empirically. No borings or excavations of any sort have been made to such depths as to indicate that its limit was being approached, though some of them are a mile or more deep. There is a popular belief that water sinks until it reaches a temperature sufficient to convert it into steam, but except for special localities where hot lava lies near the surface, this belief is not well founded. In the first place, it is not known at what temperature water below the surface would be converted into steam. While water boils at sea-level at a temperature of 212° (Fahr.) a higher temperature would be necessary below that level.

Assuming the temperature of water sinking beneath the surface to be 50° Fahr., its temperature must be raised 162° to bring it to the temperature at which it would boil at sea-level. On the above assumption of initial temperature, the following table shows the depths at which water would reach a temperature of 212° Fahr. under various assumptions as to the rate of increase of temperature. It shows also the pressure in atmospheres which would exist at these several depths if the overlying rock were full of water.

The temperature at which water boils increases with the pressure. A pressure of about 200 atmospheres is the critical pressure for water; that is, the pressure which, if increased, will prevent boiling altogether. The depth at which a pressure of 200 atmospheres would be reached, supposing the upper rock to be full of water, is about 6800 feet. The temperature of the water at this depth, under various assumptions as to initial temperature and rate of increase of heat, is shown in the following table:

Only one of these temperatures reaches the boiling-point of water at sea-level. It is therefore clear that at this depth water has not even closely approached the boiling temperature for this depth, and since this is the depth of the critical pressure, it is clear that it cannot boil at any greater depth. The descent of water is therefore not stopped, under normal conditions of crustal temperature, because it reaches its boiling-point. Locally, as in the vicinity of active or recently extinct volcanoes, the case may be different.

It is conceivable that water may descend until it reaches its critical temperature (somewhere between 610° and 635° Fahr.). The depth at which the critical temperature would be reached, under various assumptions, is shown in the following table:

There is good reason, in the increasing density beneath the surface, for believing that the rate of increase of temperature decreases with depth, and therefore that the rate of 1° for 50 feet for the depths concerned is too high. The greater depths of the table above are therefore believed to more nearly represent the truth than the lesser ones. (See discussion of underground temperature in Chapter XI.)

If descending water attained its critical temperature, the extent to which the resulting water-gas might be absorbed is not known. So far as limited by temperature, therefore, it is not possible to assign a limit to the descent of water under average conditions of crustal temperature.

Other considerations seem to place a limit to the descent of water. Rock, solid and unyielding as it seems, is yet plastic when under sufficiently great pressure. The cracks and cavities affecting it are believed to descend a distance which is but slight in comparison with the radius of the earth. Even if openings were once formed at greater depths, they could not persist, for the adjacent rock, under the pressure which there exists, would “flow” in, in effect (though perhaps not in principle) much as stiff liquid might, and close them. The outer zone of the earth where cavities may exist is known as the zone of fracture.[94] The depth of the zone of fracture differs for different rocks, but is not believed to extend below some such depth as five or six miles, even for the most resistant.[95] It is to be noted that these depths are less than those at which the critical temperature of water would be reached under most of the conditions, including all the more probable ones, specified in the above table.

Let it be assumed that water descends through openings in the rock to a depth of six miles. At this depth it would, under the various assumptions specified in the first and second columns of the following table, have the temperature indicated in the third column:

In two of these cases, namely, those in which the assumed rate of increase of temperature is highest, the temperature of the water at the assumed lower limit of the zone of fracture is above the critical temperature of water. If the assumptions involved in these two cases be correct, water might descend to the point where it would be converted into water-gas, and in this condition it might be occluded by the hot rock. In the other cases, involving the more probable assumptions, the critical temperature is not reached at a depth of six miles. If pores and cracks do not extend to greater depths, liquid water could not; and since the water at this depth has probably not reached its critical temperature, it cannot exist as water-gas. If it does not exist in the form of water-gas, its occlusion by the hot rock substance would not be probable. It would seem, therefore, that the descent of water under ordinary conditions is much more likely to be limited by the zone of fracture, than by temperature.

Movement of ground-water.[96]—Ground-water is in more or less continual movement. If all the water be pumped out of a well it soon fills up again to its normal level by inflow from all sides. Springs and flowing wells also demonstrate the movement of ground-water. Near the surface the movement of ground-water is primarily downward if the medium through which it passes is equally permeable in all directions; but so soon as the descending water reaches the water surface, its descent is checked and its movement is partly lateral.

The commonest sort of movement of ground-water is that exemplified as the water sinks beneath the surface, namely, slow percolation through the pores and cracks of the soil and rock. Ground-water is not generally organized into definite streams, though underground streams, mostly small, are sometimes seen in caves and crevices, and sometimes issue as springs. Most underground streams which issue as springs probably have definite channels for short distances only before they issue. It is probable that ground-water frequently flows in considerable quantity along somewhat definite planes, without having open channels. Thus every porous bed of rock is likely to serve as the pathway along which subterranean drainage passes. This is especially true where the porous bed is underlain by an impervious one. The “reservoirs” from which artesian wells draw their supply are not usually streams or lakes, but porous beds of rock through which abundant water passes. As the supply is drawn off at one point, it is renewed by water entering elsewhere. Since the freedom of movement of ground-water is notably influenced by the porosity of the rock, and since the rock is, on the average, most porous and the pores largest near the surface, the movement of ground-water is, on the average, greatest near the surface, and least at its lower limit. In general the decrease of movement is much more rapid than the decrease in the size of the pores. It follows that while the upper part of the ground-water, especially that above ground-water level, moves somewhat freely, the lower part moves much more slowly. It is probable, indeed, that the movement in the lower part of the subterranean hydrosphere is extremely slight.

The amount of ground-water.—The porosity of surface rocks varies widely, and the porosity of but few has been determined.[97] Such determinations as have been made are chiefly on building stones, in which the range of porosity varies from a fraction of one percent., in the case of granite, to nearly 30 percent. in the case of some sandstones. Building stone is perhaps more dense than the average surface rock. Furthermore, such tests as have been made do not take account of the larger cracks and openings of rock, for these would not appear in the specimens tested; nor of the mantle rock, which generally contains a large amount of water. From such determinations as have been made it is estimated that the average porosity of the outer part of the lithosphere is somewhere between 5 and 10 percent. If the porosity diminishes regularly to a depth of six miles, where it becomes zero, the average porosity to this depth would be half the surface porosity.[98] An average porosity of two and one-half percent. would mean that the rock contains enough water to form a layer nearly 800 feet deep. With an average porosity of 5 percent., this figure would be doubled.[99] While these figures are not to be regarded as measurements, they perhaps give some idea of the amount of ground-water. It is this sphere of ground-water which justifies the term hydrosphere, as applied to the waters of the earth.

Fate of ground-water.—Most of the water which sinks into the earth reaches the surface again after a longer or shorter journey. Some of it is evaporated from the surface directly; some of it is taken up by plants and is passed by them into the atmosphere; some of it issues in the form of springs; some of it seeps out; some of it is drawn out through wells; and much of the remainder finds its way underground to the sea or to lakes, issuing as springs beneath them. A small portion of the descending waters enters into permanent combination with mineral matter. Many minerals are known to take up water, being changed thereby from an anhydrous to a hydrous condition. It does not necessarily follow, however, that the total supply of water is thereby decreasing. Minerals once hydrated may be dehydrated subsequently, the water being set free. Furthermore, considerable quantities of water in the form of vapor issue from volcanoes, and volcanic vents often continue to steam long after volcanic action proper has ceased. It is probable that some, and perhaps much of the water issuing from these vents has never been at the surface before, and it is not now possible to affirm that the supply from this source does not offset, or even surpass, the depletion of the hydrosphere resulting from mineral hydration.

THE WORK OF GROUND-WATER.

Ground-water effects very considerable results in the course of its history. These results are partly chemical and partly mechanical, the former being far more important than the latter.

Chemical Work.

The results of the chemical and chemico-physical action of water may be grouped in several more or less distinct categories.

1. The simplest effect is the subtraction of soluble mineral matter. Pure water is in itself a solvent of certain minerals; but the carbonic-acid gas extracted from the atmosphere, and the products of organic decay extracted from the soil make ground-water a much more efficient solvent. Something of the results which it achieves is shown by its composition. All ground-water, whether issuing as springs or drawn out through wells, contains much more mineral matter than the water which falls as rain, and the excess is acquired in its underground course.

The subtraction of soluble matter from rock renders it porous. The amount of material dissolved from a given place may be trivial or considerable, according to the character of the rock, the readiness with which water has access to it, and the character of the water. Locally, the subtraction of mineral matter may be the chief, or even the only appreciable, effect of the ground-water.

2. It sometimes happens that ground-water with certain mineral substances in solution exchanges them for other substances extracted from the rock. Thus the process of substitution is effected. By this process the lime carbonate of a shell imbedded in rock may be removed, molecule by molecule, and some other substance, such as silica, left in its place. When the process is complete, the substance of the shell has been completely removed, though its form and structure are still preserved in the new material which has taken the place of the old. Buried logs are sometimes converted into stone by the substitution of mineral matter for the vegetable tissue. This is petrification. Petrification is altogether distinct from incrustation, which simply means the coating of an object with mineral matter. A bird’s nest may be incrusted with lime carbonate, but it is not thereby petrified. Solution is a necessary antecedent of substitution.

3. The materials which are subtracted from the rock at one point may be added to other rock elsewhere. Thus a third type of change, addition, is effected. Rock may at one time and place be rendered porous by the subtraction of some of its substance, and the openings thus formed may subsequently become the receptacles of deposits from solution. This is exemplified in the stalactitic deposits of many caves. Not uncommonly cracks and fissures are filled with mineral matter deposited by the waters which pass through them. Thus arise veins which, for the most part, are nothing more than cracks and crevices filled by mineral matter brought to them in solution, and precipitated on their walls. Most veins of metallic ores have originated in this way.

4. A further series of changes is effected by ground-water when it, or the mineral matter it contains, enters into combination with the mineral matter through which it passes. One of the commonest processes of this sort, hydration, has already been referred to (pp. 43 and 428); but in the development of many of the commoner hydrous minerals changes other than hydration are involved. These changes result in new mineral combinations, the new minerals being developed out of the old, usually with some additions or subtractions. In the long course of time changes of this sort may be very great, so great indeed that large bodies of rock are radically changed, both chemically and physically. Much of the old substance may remain, but it has entered into new and more stable combinations with the materials which the water has brought to it.

Quantitative importance of solution.—In general, solution is probably most effective at a relatively slight distance below the surface. In the outermost zone of mantle rock the materials are usually less soluble than below, for they often represent the residuum after the soluble parts of the formation from which they originated were dissolved out. Below this zone the rock contains more soluble matter, and the water, charged with organic matter in its descent through the soil, is in condition to dissolve it. At greater depths the water has become saturated to some extent, and, so far forth, less active. Here, too, the movement is less free. The increased pressure at considerable depths, on the other hand, facilitates solution, which must be understood to take place under proper circumstances in any zone reached by the water.

Calculations have been made which illustrate the quantitative importance of the solution effected by ground-water. The springs of Leuk (Switzerland) bring to the surface annually more than 2000 tons of calcium sulphate (gypsum) in solution, and in the same time the springs of Bath (England) bring up an amount of mineral matter in solution sufficient to make a column 9 feet in diameter, and 140 feet high.[100]

The amount of mineral matter in solution in streams is also significant, for while stream-water is not all derived from ground-water, much of it had such an origin. In the case of several streams, among them the Thames and the Elbe, careful estimates of the amount of dissolved mineral matter have been made. Though the Thames drains an area only about one-tenth as large as the State of New York, it is estimated to carry about 1500 tons of mineral matter in solution to the sea daily.[101]

From the uppermost 20,000 square miles of its drainage basin the Elbe is estimated to carry yearly about 1,370,000 tons of mineral matter in solution. Estimates of the amounts of material carried to the sea in solution by several rivers are given on pp. 102 and 103. Much of this matter was brought to the rivers by waters which had been underground before reaching the streams.

From these figures it is clear that we have to reckon here with a very considerable factor in the lowering of land surfaces. From the amount of lime carbonate carried by the Thames it has been estimated that the average amount of this material dissolved from the limestone area drained by this stream is 143[102] tons per square mile per year. It is estimated that, on the average, something like one-third as much matter is carried to the sea in solution as in the form of sediment, and that by this process alone land areas would be lowered something like one foot in 13,000 years.[103]

Deposition of mineral matter from solution.—The deposition of material from solution is effected in several ways. (1) It is sometimes deposited by evaporation. This is well shown where water seeps out on arid lands. The same process is illustrated when water is boiled. (2) Reduction of temperature often occasions deposition. In general, hot water is a better solvent of mineral matter than cold,[104] and if it issues with abundant mineral matter in solution the precipitation of some of it is likely to take place. (3) Plants sometimes cause the precipitation of mineral matter from solution. About some hot springs, even where the temperature of the water is very high small plants of low type (algæ) grow in profusion. In ways which are not perfectly understood these algæ extract the mineral matter from the hot water. They are now thought to be a chief factor in the deposits about the hot springs of the Yellowstone Park.[105] The influence of organisms on precipitation from solution is not confined to the waters of hot springs. (4) A fourth factor involved in the deposition of mineral matter from solution is pressure. Pressure increases the solvent power of water with respect to minerals directly; it produces the same effect indirectly by its effect on the solution of gases. As water charged with gas comes to the surface, the pressure is relieved and some of the gas escapes. Such mineral matter as was held in solution by the help of the gas which escapes is then precipitated. (5) Precipitation is also sometimes effected by the mingling of waters containing different mineral substances in solution. Such mingling of solutions would be most common along lines of ready subterranean flow, and while each portion of the water entering a crevice or porous bed may be able to keep its own mineral matter in solution, their mingling may involve chemical changes resulting in the formation of insoluble compounds, and therefore in deposition. This principle has probably been involved in the filling of many fissures and crevices, converting them into veins. (6) The escape of gases from water, whether from increase of temperature or by the disturbance of water, sometimes causes the deposition of mineral matter held in solution.

The deposition of material held in solution is most notable at two zones, one below that of most active solution, and the other at the surface, where evaporation is active. Under proper conditions, however, deposition may take place at any level reached by water.

Mechanical Work.

The mechanical work of ground-water is relatively unimportant. Wherever it is organized into definite streams, the channels through which it flows are likely to be increased by mechanical erosion as well as by solution. Either beneath the surface, or after the streams issue, the mechanical sediment carried will be deposited.

RESULTS OF THE WORK OF GROUND-WATER.

Weathering.—Where the solution effected by ground-water in any locality is slight and equally distributed, the result is to make the rock porous. If, for example, some of the cement of sandstone is dissolved, the texture of the rock becomes more open; but if all the cement be removed the rock is changed from sandstone to sand. If a complex crystalline rock contains among its many minerals some one which is more soluble than the others, that one may be dissolved. This has the effect of breaking up the rock, since each mineral acts as a binder for the rest. It might happen that no one of the minerals is dissolved completely, but that some one of them is decomposed by water, and certain of its constituents removed. Such change would be likely to cause the mineral so affected to crumble, and with its crumbling, if it be an important constituent of the rock, the integrity of the rock is destroyed. Where considerable chemical changes, especially subtractions, are going on, the rock is likely to crumble. The increase in volume attendant on hydration, etc., sometimes leads to the disruption of rock. These are phases of weathering. (For other phases of weathering see pp. 54 and 110.)

Caverns.[106]—Where local solution is very great results of another sort may be effected. In formations like limestone, which are relatively soluble, considerable quantities of material are frequently dissolved from a given place. Instead of making the rock porous, in the usual sense of the term, large caverns may be developed (Fig. 202). In their production, solution may be abetted by the mechanical action of the water passing through the openings which solution has developed. Considerable caves are found chiefly in limestone. They were probably developed when the surface relief was slight, and surface drainage therefore poor. Regions where caves were developed under these conditions may subsequently acquire relief, so that caves are not now confined to flat regions.

One of the best known regions of caves is in the basin of the Ohio in Kentucky and southern Indiana, where the number of caves is large, and the size of some of them, such as Mammoth and Wyandotte, very great. A ground-plan of Wyandotte (Ind.) Cave is shown in Fig. 203. The aggregate length of the passageways is about 23½ miles.

Deposition often takes place in caves after they are formed (Figs. 204 and 205). It may even go on at the same time that the cave is being excavated. Here are formed the well-known stalactites and stalagmites. A stalactite may start from a drop of water leaking through the roof of the cave. Evaporation, or the escape of some of the carbonic gas in solution, results in the deposition of some of the lime carbonate about the margin of the drop, in the form of a ring. Successive drops make successive deposits on the lower edge of the ring, which grows downward into a hollow tube through which descending water passes, making its chief deposits at the end. Deposition in the tube may ultimately close it, while deposition on the outside, due to water trickling down in that position, may greatly enlarge it.

Underground caves sometimes give rise to topographic features which are of local importance. When the solution of material in a cavern has gone so far that its roof becomes thin and weak, it may collapse, giving rise to a sink or depression in the surface over the site of the original cave. This is so common that regions of limestone caves are often affected by frequent sinks formed in this way. They are a conspicuous feature of the landscape in the cave region of Kentucky, and are well known in many other limestone districts. They are known as limestone sinks. (Fig. 206 and Fig. 2, Pl. XVII.)