There are other instances, similar to the last in all respects, except that the segregated portion does not take the concretionary form. When gypsum is distributed in small proportion through a formation, there seems very little reason to doubt but that it is, by a molecular action, segregated from the strata in lenticular masses, as at a (Fig. 70). Many of the limestone strata contain irregular aggregations of quartz. It is presumed that the siliceous and calcareous matter was deposited together as sediment, and that the aggregation has resulted from a movement among the particles similar to that by which the concretionary structure is produced.

The columnar structure of basalt seems to have resulted from a peculiar molecular action, at first resembling a concretionary arrangement, while the mass was cooling from a state of fusion. In experimenting to ascertain the cause of this structure, Mr. Watt fused in a furnace seven hundred pounds of basalt. When cooled, he found that “numerous spheroids had been formed, and that when two of them came in contact, they did not penetrate each other, but were mutually compressed and separated by a well-defined plane, invested with a rusty coating. When several met, they formed prisms.” (Fig. 71.

It was shown by the early experiments of Davy, that voltaic currents are capable of taking up mineral substances from their solutions, and removing them from one cup to another. It has been ascertained that in most mineral veins a proper apparatus will detect the existence of electric currents. It may be regarded as certain, that the unequal heating of different parts of the surface at the same time, by the sun, causes a vast current of feeble intensity to circulate around the earth once in twenty-four hours. The unequal distribution of heat below the surface may also produce currents subject to other laws. We should expect that these currents would take up the mineral substances diffused through rocks, and deposit them by themselves. It seems probable, therefore, that the molecular action, from which the segregation of metallic veins has resulted, was that of voltaic currents.

SECTION III.—ORGANIC CAUSES.

Besides these general effects, it is the existence of organic forms that has conferred on all the sedimentary rocks their fossiliferous character. The records of the climate of each geological period, of the physical geography, of the vegetable productions, and of the animal forms by which the earth was peopled, consist in the remains of the living beings of these several periods, imbedded in the contemporaneous rock formations. But in the sediment deposited since the human era there must have been furnished both the remains of human beings and works of art, such as implements of labor and war, pottery, coins, fragments of ships, &c.

Moreover, the quantity of material which has been furnished by organic causes is by no means small. The coal-beds are the product of vegetable growth exclusively. We not unfrequently find strata of great extent consisting almost entirely of the shells of molluscous animals, of the stems of encrinites, or of the shields of microscopic animalcules.

But the most abundant rock which can be regarded as the product of animal organization is the coral formation. It consists of immense walls of coral limestone, separating either an atoll or the land of an island or continent from the open sea. The base of this wall has a width varying from a hundred feet to a mile or more, and the outer edge of it is at such a distance from the shore as to give a depth not much exceeding a hundred feet. Over this area of the bed of the sea, which forms the base of the wall, the coral polyp commenced its work. Attaching itself in immense numbers over this area, it deposits calcareous matter from its under surface, and thus, by degrees, elevates itself towards the surface of the water, till it reaches a level a little above low-water mark. The height of the wall would not, with these conditions, exceed one hundred feet; but some hundreds of the islands surrounded by coral walls are gradually subsiding. The depositions of the polyps keep pace with the subsidence, so that this wall has reached an elevation from its base of a thousand feet, and in one instance of two thousand feet. (See Figs. 61, 62, 63, 64.)

Most of the islands of the torrid zone are thus surrounded with coral reefs, except a few where the cold polar currents reduce the temperature too low to admit of their growth. In one instance, along the north-east coast of New Holland, there is a coral reef, some twenty-five miles from the land, which has a continuous extension, excepting occasional inlets of no great depth, of a thousand miles. The reef along the island of New Caledonia is four hundred miles long. A large number of other reefs have a nearly equal extension. There is thus an area of several thousands of square miles covered to a great depth with this coralline limestone. Some limestone formations of great extent among the older rocks were the work of similar animals. These lower forms of organization have, therefore, always been important geological agents, both in collecting the carbonate of lime from its solution in the waters of the ocean, and in depositing it as solid rock.

SECTION IV.—AQUEOUS CAUSES.

I. The Furnishing of Sediment.

1. Almost all the minerals which occur in the geological formations are, to some slight extent, soluble in water. Hence, rain water, by passing through a stratum of earth or rock and reäppearing as a spring, loses the insipidity which it had as pure water, and becomes palatable. It is then found to hold in solution some small proportion of earthy substances, upon which this change of taste depends. Although the proportion of dissolved matter is very small, yet the surface of earth upon which this distilled water is shed is one-fourth of the surface of the globe, and solution below all that surface is constantly taking place. No inconsiderable amount must thus have been furnished, from the existing rocks of each period, towards the formation of the strata of a later period.

There are some substances which are soluble in water, in large quantities. Rock-salt is an example. It is not found in any very large proportion in rocks generally, but a very large aggregate amount has been taken up by the waters which have filtered through the strata. The ocean gathers into itself, by degrees, all the soluble substances which are thus taken up. It receives supplies of water charged with these substances from springs, rivers and lakes. It returns as much water as it receives; but it is always in the form of vapor, and is therefore pure water. Hence the saline properties of the ocean, and of those inland seas which have no outlets. There is thus gathered the materials for the rock-salt deposits.

But many substances which are not considered soluble in water become so by some modification of the water. Water of a high temperature is capable of dissolving silex. In Iceland and other volcanic regions, the hot springs are charged with silex, which is deposited as the water cools. Thus, siliceous formations accumulate around springs of this kind. The various agates may have been deposited from such solutions.

In the decomposition of mica, felspar and volcanic rocks, a considerable amount of potassa is set free. Potassa or soda renders the water in which it is dissolved capable of dissolving silex in large quantity. In these ways water removes, with some degree of rapidity, one of the most insoluble minerals which rocks contain.

In volcanic countries, and in coal districts, carbonic acid is abundant, both in spring-water and in the gaseous form. Water charged with this gas becomes capable of dissolving limestone. Where the water is exposed to the air, the gas gradually escapes, and the calcareous matter is deposited. Many accumulations of this kind are now taking place. Some have already extended several miles in length, and they are often of great thickness, in one instance, in Italy, two hundred feet (Fig. 72). It is also probable that many calcareous springs issue below the surface of lakes and seas, and thus, both fresh-water and marine deposits would now be forming. These formations are distinctly stratified, and are white and crystalline, and become solid at the time of deposition.

2. The abrading action of rivers furnishes considerable detrital matter. The general form of the river courses is determined by other causes than the agency of the river itself, yet a river which has a rapid current is continually deepening its channel. We have proof of this by observing, when the water is low, that irregularity of surface which running water always produces, by wearing away the softer parts of the rock, and leaving the harder in relief. Hence, a river will have its rapids either where the hardest strata occur, and which therefore wear down least rapidly, or where the rock has been hardened by the intrusion or near proximity of dikes.

The abrading power of rivers occasionally becomes greatly increased by waterfalls. The force which the water acquires in its descent is such as to excavate a deep cavity at the foot of the fall, reaching back under the ledge from which the water descends. The ledge is therefore constantly being undermined. The cataract of Niagara is peculiar, in having the rock at its base of a soft and friable texture, so that it is rapidly worn away, while the upper rock is a compact siliceous limestone. If the order of superposition had been the reverse, the falls would have been converted into a series of rapids. It is now preserved as a single fall, and as such it has probably cut the gorge, about two hundred feet deep and seven miles in length, through which its waters now reach Lake Ontario. A few years since, a large mass, perhaps half an acre in area, fell from the centre of the horse-shoe fall. Another mass of equal size has recently fallen from the western extremity of the ledge. Thus the fall is gradually receding.

But the foreign substances, such as drift-wood, ice, sand and gravel, with which the waters of a river are occasionally charged, contribute more than everything else to its abrading power. At such times its volume is generally greatest, and its current the most rapid. Its bed is then sometimes perceptibly deepened and widened in a few hours.

Much the greater part, however, of the earthy matter which rivers convey in such quantity to the ocean, is furnished by other means than the eroding action of the river itself. It is the loose material, the soil and alluvium, to which the solid rocks have been reduced by the imperceptible but incessantly operating atmospheric agencies, from which most of the sediment of rivers is obtained. After a rain, every tributary rivulet is turbid with suspended earthy matter, and it is from these sources that the larger streams receive the most of their sediment.

Some observations have been made for the purpose of ascertaining the quantity of sediment which rivers annually carry into the sea. The Kennebec furnishes materials which, if spread evenly on an area of one mile square, and consolidated into rock of the specific gravity of granite, would have a thickness of six inches. The Merrimac furnishes about two-thirds as much, the Ganges about two hundred and fifty times as much, and the Mississippi two thousand times as much.

Thus, the tendency is, to reduce the highest parts of the land, and to fill up the depressions of the sea; and though we have not data enough to form any reliable estimate of the total annual discharge of sediment into the ocean by rivers, yet they are sufficient to show that the effects of this kind are on a large scale, and to relieve us from any impression that existing agencies are inadequate to the production of the stratified rocks.

3. The action of waves is another means by which detrital matter is furnished. Wherever the shore consists of loose materials, and is favorably situated to be acted upon by the waves, there is annually a sensible encroachment of the sea. Such encroachments are rapidly making in many places; and thus a large amount of sediment is delivered to the waters of the ocean.

The waves also encroach upon the coast when it consists of rocks, even of the most indestructible kinds. They continually beat upon it, undermine the cliffs, and precipitate them into the sea. The tides increase the power of the waves, by varying the place of their action, so as to present the same surface of rock alternately to the action of water and of the air, frost and sun. During storms, the waves have sufficient force to break off fragments of rock from the escarpment, sometimes in masses weighing twenty tons or more, and remove them many rods inland.

A bold, rocky coast always exhibits evidence of a great amount of erosion. The steep escarpments and the high rugged shafts of rock (Fig. 73) against which the waves now beat are the remnants of masses of rock which once extended further into the sea, but have been worn away by the waves. It is by such agency that the deep inlets and harbors of the coast of New England and Nova Scotia have been excavated.

For the same reason,—a difference of temperature,—there will be, in the higher regions of the atmosphere, a current of warm and moist air flowing from the equator north and south, while the cold and dry air conies in from the polar regions towards the equator. In this way the equatorial waters are carried, in a state of vapor, towards the poles, where they are condensed, and go to increase the currents of water moving towards the equator.

Such are the general causes of the oceanic movements in a north and south direction; but these currents at once become deflected westward, by the diurnal revolution of the earth, as the trade winds do. Hence there results a Pacific equatorial current, which has a motion of about thirty miles a day, and an Atlantic equatorial current, moving from sixty to seventy miles a day. The principal marine currents are shown in Fig. 74.

The currents moving towards the poles are superficial, and therefore do not produce any marked geological effects. But the polar currents, and those which are produced from them, are of great depth, and there is no reason to suppose that they do not move, from their commencement, along the bed of the ocean. There is also reason to suppose that they exist at great depths, where the opposing superficial currents entirely conceal them.

Wherever these currents come to the surface, their motion is undoubtedly greater than it is at the bottom, where it is retarded by the friction which the moving waters encounter, and by the irregularities of the bed of the ocean. It should, however, be remembered, that they move with the weight of the whole superior body of water; and therefore, though the motion be very slow, it will still possess great power.

Any irregularities in the bed of the ocean beneath such a current must be subject to very rapid abrasion. We shall sea hereafter, that earthquake vibrations often shiver the rocks at the solid surface; and if any of these ridges at the bottom of the ocean were thus acted upon, the loosened portions would be swept away by the current and deposited at lower levels, or where the current subsides. If, in any instance during an earthquake convulsion, a fault should be produced across one of these marine currents, like the great fault of over five hundred feet in England, the abutment thus thrown up would soon be worn down; and if it consisted of unconsolidated matter, it would be swept away almost bodily.

The effect of such currents will be greatest where they are deflected by a continent or island. Thus, a marine current sets from near New Holland in a direct line to the north of the island of Madagascar, where it is arrested by the African coast, and deflected into the narrow Mozambique channel, and there acquires a velocity of four or five miles an hour. It is impossible that any kind of rock should receive the constant force of such a body of water without being rapidly worn away; and, if there should be any difference of texture in this rocky barrier, the softer portions would yield the most rapidly, and thus valleys might be formed.

It is not improbable that the deep indentation on the western coast of Africa may have been due, in a great measure, to the coast current from the Cape of Good Hope; and that the Caribbean Sea and the Gulf of Mexico may have been excavated by the force of the Atlantic equatorial current being thrown into this angle.

We may regard these currents as oceanic rivers; and it is obvious that the volume of the terrestrial rivers would bear no comparison with that of these currents, and their effects would be equally small in the comparison. The Gulf Stream, and the Mozambique and other similar currents, must be wearing down the valleys through which they flow, to such an extent as to furnish an immense amount of detrital matter for the formation of new rocks.

It is principally to the agency of these deep marine currents that we are to refer those extensive denudations, so abundant on the present continents, such as the wearing out of the intermediate masses of rock between the hills already referred to (Fig. 66), the denudation of the Connecticut river sandstone, and, perhaps, the excavations which have formed Lake Erie and Lake Ontario.

II. The Transportation of Sediment.

The detrital matter obtained in these several ways is swept away by running water. The specific gravity of rocks does not, in general, exceed two and a half. Hence, to keep them suspended in water, will require a force of only three-fifths of what would be necessary to suspend them in the atmosphere. In the case of river currents, the velocity and irregularity of motion are generally sufficient to keep all the finer sediment equally distributed.

There will, however, be a division of the sediment according to the strength of the current. Hence, the bed of a mountain stream, if there is any loose material, always consists of pebbles. As it approaches the alluvial region, the bed is sandy; and when the current becomes very sluggish, it consists of a fine mud.

Rivers never deposit all their sediment, some of them none of it, along their course. Large rivers continue partially distinct from the ocean water to a considerable distance beyond their mouths. The waters of the Amazon have been recognized at a distance of three hundred miles. This depends in part upon the volume and velocity of the river; more, however, upon the fact that river water is lighter than sea water. This extension of a river will, in most cases, be sufficient to deliver a part of its sediment into a marine current. When such a current sweeps very near the mouth of a river, as it does to that of the Niger, the Amazon, or the Mississippi, it is probable that most of its sediment is carried away by it.

The transporting power of a marine current is greater than that of a river, in consequence of the greater specific gravity of its water; but it has scarcely any of that irregular motion of rapid rivers, upon which their transporting power in a great degree depends. The force of the current alone, when it reaches the bottom, is, however, sufficient to remove every form of loose earthy matter. Thus it may be presumed that the Gulf Stream sweeps all the sediment from its bed until it reaches the latitude of Cape Hatteras, where the cold waters from the north begin to underlie it, and it takes the character of a surface stream.

But the transporting power of marine currents depends mostly upon the depth of water. It is found, by experiment, that ordinary river sediment will sink in water about one foot in an hour. A current, therefore, of a thousand feet in depth, which moves a mile in an hour, would carry its sediment a thousand miles. It is obvious, then, that there is no part of the bed of the sea which may not be receiving sediment.

III. The Deposition of Sediment.

Of the extent of these deposits we can form no estimate. Those of rivers and lakes are comparatively unimportant, as they are in the older formations. Some of the delta deposits are already of great extent. That of the Ganges contains an area of twenty-six thousand square miles, that of the Niger twenty-five thousand, and that of the Nile twelve thousand. The delta of the Rhone has increased its area by three hundred square miles in the last thousand years. The Po has encroached upon the Adriatic two thousand square miles in the last two thousand years, and the Mississippi has enlarged its delta by one hundred square miles in the last hundred years. In the deep valleys of the ocean accumulations may be taking place on as large a scale as they ever have been in former times.

IV. Character of the Formations thus produced.

The occurrence of layers of different composition, as one way in which the stratification is indicated, is produced by local and frequently recurring causes. There are, however, other alternations of much greater extent; those, for example, nearly twenty in number, distinguished by striking differences in lithological character, into which the New York system of rocks is divided. These alternations have resulted from more general causes. The physical geography of a wide region must have been so different, at the different periods during which these several formations were deposited, as to change, at each period, the kind of sediment furnished to the forming currents, and modify the types of animal life.

We have seen that the same causes that determined the stratified arrangement will determine the alternations of strata of coarse and fine materials.

It is obvious that the stratification of the marine deposits will be nearly horizontal. If the surface were very irregular upon which the deposition commenced, the irregularity would constantly diminish; for the movement of the water over this surface, however slow, would tend to remove the accumulations from the highest points, and leave them at the lowest (Fig. 75). Delta and lake deposits will, however, dip somewhat, though never at a high angle, towards the deep water. In certain situations, where a river and a tidal wave, coming in conflict, cause, in succession, eddies and currents in opposite directions, we should expect to find the stratification very irregular (Fig. 76); sometimes false stratifications (a b), sometimes the strata cut off abruptly, and at other times contorted or dipping in opposite directions within short distances.

Wherever sediment is deposited, it will entomb whatever of the remains of animal or vegetable life may be mingled with it. They will be at once protected against the influence of all the ordinary decomposing agencies, and will continue for ages to retain their peculiar markings, and even their colors. They will thus constitute, in all future time, a record of the present condition of the organic world. The lacustrine deposits can contain only fresh-water species of animals, marine deposits only marine animals, while deltas may contain the remains of marine life mingled with those which have been washed down by rivers. The remains of birds, insects, and terrestrial animals, may occasionally occur, in every kind of deposit. Sediment deposited in deep water will never contain fossils in abundance, the deep parts of the ocean being almost wholly destitute of animal or vegetable life. It is only in water of a few fathoms that the greater number of species and of individuals occur. In all these particulars the deposits now forming sustain a close resemblance to the older formations.

There are certain formations, as that of the coal, which required conditions for their formation different from those of ordinary sedimentary deposits. Coal consists of mineralized vegetable matter. Its vegetable origin is proved by the uniform occurrence of vegetable fossils almost exclusively in the coal measures. When reduced to thin slices and examined under a high magnifying power, a structure very similar to the ligneous tissue of existing coniferæ is sometimes found to exist. There are probably vegetable deposits now taking place not altogether unlike those which produced the coal measures.

We know that many rivers—the Mississippi, for example—now carry into the sea great quantities of ligneous matter. Before the country was inhabited by man, the quantity was undoubtedly much greater than it now is. It floats for a time; but the ligneous tissue itself is heavier than water, and as soon as the air is excluded from the pores, and they are filled with water, it will sink. The woody and earthy matters are swept into the sea together; but, as they sink under different circumstances, they will be deposited separately. Thus wood may continue to accumulate in particular places in the sea for long periods, with but little intermixture of earthy substances.

It is, however, to be expected that, in the progress of geological changes, the places which at one time receive deposits of wood will at another receive detrital matter, and thus the wood will become deeply buried beneath sedimentary strata.

Wood thus situated will become converted into coal. Trees which had been covered to considerable depth with earth have been found near the Mississippi river changed to lignite, a substance resembling charcoal. In this case, the wood had been exposed to no greater heat than is common to the crust of the earth at the depth where it was found; and yet it had undergone this change since the country has been known to Europeans, as it retained the marks of the axe when it was discovered. It has also been found by experiment that vegetable matter, by long submersion in water, passes into the state of lignite. This is the first step in the conversion of wood into mineral coal.

When lignite is exposed to moderate heat and great pressure, it loses the characters of lignite, and becomes mineral coal. This is shown by facts observed in Germany, Ireland and Iceland, where beds of lignite have been overspread by basalt. The upper portions of the lignite are changed to mineral coal. The lower portions, which the heat did not reach, retain the characters of lignite.

Beds of vegetable matter, with a great thickness of rock deposited above them, would therefore be subject to all the conditions necessary to convert them into coal, namely, pressure from the superincumbent mass, and the heat which the strata uniformly assume at great depths.

It is not improbable, therefore, that coal-beds are now forming, and that they have been formed at every geological period since an abundant terrestrial vegetation commenced. Accordingly, there occurs in Virginia an extensive coal-field in the oölite formation. Coal-fields also occur in England, of less extent, in the same formation. In France, and other parts of Europe, there are extensive beds of lignite in the tertiary formation.

We have therefore no difficulty in accounting, in a general way, for the formations of the carboniferous period. The vegetables were probably less woody than those of the present time of equal size, and were therefore more easily prostrated and committed to the waters. They grew rapidly in moist ground, and perhaps in shoal-water, and required an atmosphere charged with moisture and of a high temperature. Thus much is inferred from the conditions most favorable for the growth of recent species analogous to the coal-plants. These recent species are tropical plants, and grow in moist insular situations, conditions which would have existed at the carboniferous period, if the present coal-fields were then an archipelago dotted with low islands.

Such being regarded as the origin of the coal-beds, the alternations of the earthy and carbonaceous strata may be referred, provisionally, to those great changes in physical geography upon which the other alternations of strata on a large scale depend. But the regularity with which the coal-seams and sandstone succeed each other presents some difficulties which, in the present state of knowledge, we cannot satisfactorily account for.

Beds of salt occur, interstratified with other rocks, in nearly all countries. Still, it is not a sedimentary deposit, and its formation must depend upon peculiar circumstances. In New York, saline, together with earthy matter, constitutes the Onondaga limestone, one of the formations of the New York system. In Kentucky, the strata of rock-salt are in the coal formation; in England, they are in the new red sandstone; in Spain, they are in the greensand, and in Poland they are in tertiary strata. The conditions of its formation have therefore existed in connection with the deposition of every fossiliferous rock.

It has been shown that the ocean is the principal reservoir of the saline matters which are taken up whenever water percolates through rocks. It must happen not unfrequently, in the course of submarine elevations, that a basin of sea-water will be cut off from its communication with the sea; and from this basin the evaporation might be more rapid than the supply of water. The great salt-lake of Utah is undoubtedly a basin of this kind. The Mediterranean Sea is another such basin, not yet wholly separated from the ocean. The evaporation exceeds the supply of water from the rivers, and a powerful stream is therefore continually thrown in from the ocean, through the Strait of Gibraltar. The waters of the Mediterranean are already more highly charged with salt than ordinary sea-water. This sea may ultimately become a saturated solution, and begin to deposit salt. But whether it does, or not, it indicates the way in which salt-beds may be formed.

V. Solidification of Aqueous Deposits.

Some rocks are composed of such materials that they set, like hydraulic cement, when they are deposited. Other rocks become solid simply by drying. Thus a deposit now forming in Lake Superior becomes, by drying, nearly as hard as granite. Such a deposit will therefore become solid whenever it shall be elevated above the water.

The pressure to which all but the upper layers are subjected is probably sufficient to reduce most rocks to the solid state. Dry and pulverized clay is reduced by artificial pressure, for a moment, almost to stone. The pressure upon the deep-seated rocks is constant, and greater than any artificial pressure can be.

In addition to these causes, all the older rocks have been subjected to a high temperature, some of them nearly to that of fusion. By this means the solidification of every kind of rock would be promoted, and probably some may have been reduced by it to the solid state, which would otherwise have remained as an incoherent mass.

SECTION V.—AQUEO-GLACIAL ACTION.

1. Glaciers.—A glacier is a mass of ice occupying the bed of a mountain valley, having a slow progressive motion, and reaching somewhat lower in the valley than the line of constant snow. (Fig. 77. The Glacier des Bois, which may be regarded as a specimen of the Alpine glaciers, covers an area of about seventeen square miles. In its lowest portion, when all its branches have become united into one stream, it has an average width of half a mile, and is five miles long. It is estimated that the glaciers of the Alps cover an area of fourteen hundred square miles. These have been the most carefully studied, though glaciers are found in the valleys of various other ranges of mountains.

In the higher valleys, the snow, which falls at all seasons of the year, accumulates in immense quantities, and the steep mountain sides contribute, by frequent avalanches, to this accumulation. The snow, when thus increased, does not become a compact, adhesive mass; but, changing into particles of solid ice, it resembles sand rather than snow. It is this névé which constitutes the upper part of every glacier, and which, in a modified form, constitutes the lower part.

The valleys descend rapidly towards the base of the mountains; and this snow-ice, having no cohesion between its particles, moves slowly down the slope of the valley, like a very imperfect liquid. After descending below the line of perpetual snow, the surface will melt during the day; and the water, sinking into the porous mass, becomes frozen, and converts the whole into more or less compact ice, yet never into a rigid mass. Influenced by its own weight, and by the pressure of the snow-ice behind, it still continues its motion, and conforms itself to the shape and curves of the valley through which it passes. The average movement per annum may be stated at about five hundred feet.

The temperature of the rocky bed of the valley will be a little, and but a little, higher than thirty-two degrees. There will therefore be but little melting at the bed of this river of ice. As it receives continual accessions from the atmosphere, it will therefore increase in volume till it descends to the level of perpetual snow. Below this line the waste exceeds the addition; and as it approaches the lower and cultivated portions of the valley, it rapidly diminishes, till it finally loses the solid form, and becomes a rivulet. The terminus of the glacier is determined principally by the general climate of the country. Any considerable variation of climate will cause it to recede, or descend lower down the valley. The terminus varies, however, somewhat with the seasons, being lower in winter than in summer, though the motion is much less in the cold season than in the warm; and it descends many rods further some seasons than it does others.

The glacier consists principally of snow, more or less modified in structure; but it also contains whatever else may have been thrown upon its surface, or into the snows by which it is fed. Tributary glaciers extend up through all the gorges into which the irregular surface of the mountain-top is divided. On these rough peaks there are always fragments of rock, varying in size from fine sand to masses weighing many tons; some of them loosened when the mountain was upheaved, some by subsequent earthquake vibrations, and others still by tempests, lightnings, and changes of temperature. When the snow has accumulated to a certain extent on the steep slopes, it falls in avalanches into the valleys, carrying with it loosened masses of rock, and often breaking off large fragments from the rocky escarpments against which it strikes. These avalanches are almost constantly descending, and hence a glacier always contains considerable earthy matter distributed through it.