CHAPTER V.. The Work of Snow and Ice.

THE WORK OF SNOW AND ICE.

A part of the atmospheric precipitation falls as snow, and this, like the rain, does its appropriate work in degrading the land. Over the larger part of the land surface the snow of the winter does not endure through the succeeding summer, and when it melts it follows the same course as the precipitation which falls as rain; but in cold regions where the fall of snow is heavy some of it remains unmelted and constitutes perennial snow-fields.

SNOW- AND ICE-FIELDS.

=Snow-fields.=—Mountain heights and polar lands are the most common habitats of snow-fields, though they are not confined to these situations. In North America there are numerous small snow-fields in the western mountains, from Mexico on the south to Alaska on the north, their number and size increasing in the latter direction. In the United States there are few snow-fields south of the parallel of 36° 30′, and most of the many hundreds north of that latitude (excluding Alaska) are small (Pl. XVIII, Fig. 1, Washington, lat. 48° 5′, long. 121° 5′; Fig. 2, lat. 41° 25′, long. 122° 12′. From Glacier Peak and Shasta Special Quadrangles, U. S. Geol. Surv.). Farther north, especially in Alaska, the snow-fields of the western mountains attain much greater size. In Europe snow-fields comparable to those of the northwestern part of the United States and British Columbia occur in the Alps (Fig. 221), the Pyrenees, the Caucasus, and the Scandinavian mountains. In Asia snow-fields occur in the Himalayas and in many of the high mountains farther north, from Turkestan on the southwest nearly to the coast on the northeast. In South America there are snow-fields of small size even in equatorial latitudes, and farther south in the Chilean Andes there are some of considerable size. Small snow-fields occur on the highest peaks of tropical Africa, and in the mountains of New Zealand. For reasons which will appear later, much of every considerable snow-field is really ice.

In addition to these limited fields of snow in mountain regions, there are fields of much greater extent covering wide expanses of plain and plateau in the polar regions. The greater part of the island of Greenland is covered with a single field of ice and snow, the size of which is variously estimated at 300,000 to 400,000 square miles (Fig. 222)—an area 400 to 600 times as large as the snow-and-ice-covered area of Switzerland. Numerous islands to the west of North Greenland are also partly covered with snow, the areas of the snow-fields far exceeding those of most mountain regions. In Antarctica there is believed to be a still larger field, the largest of the earth. Its area is not even approximately known, but such data as are at hand indicate that it may have an extent of 3,000,000 or 4,000,000 square miles.

The only condition necessary for a snow-field is an excess of snowfall over snow waste. The lower edge of a snow-field, the snow-line, is dependent chiefly on temperature and snowfall. In general it does not depart much from the summer isotherm of 32°, though it may be well above this isotherm where the snowfall is light. That the snow-line is not a function of temperature only is shown by its position in various places. In the equatorial portion of the Andes, for example, the snow-line has an altitude of about 16,000 feet on the east side of the mountains, where the precipitation is heavier, and of about 18,500 feet on the west side, where it is lighter. For the same reason the snow-line in the Himalayas is 3000 or 4000 feet lower on the south side than on the north.

While in equatorial regions the snow-line has an altitude of 15,000 to 18,000 feet, it approaches or even reaches sea-level in the latitude of Antarctica and North Greenland. In intermediate latitudes it has an intermediate position.

While temperature and snowfall are the most important factors controlling the position of the snow-line, both humidity and the movements of the air are of some importance, since both affect the rate of evaporation of snow and ice.

=The passage of snow into névé and ice.=—The snow does not lie on the surface long before it undergoes obvious change. The light flakes soon begin to be transformed into granules, and the snow becomes “coarse-grained.” The granular character, so pronounced in the snow of the last banks which remain in the spring in temperate latitudes, is even more distinct in perennial snow-fields, either at the surface or just beneath it. This granular snow is called névé, or firn. Still deeper beneath the surface, where the thickness of the snow is great, the névé becomes more compact and finally coherent, and grades into porous ice. This gradation is accomplished at no great depth, though the thicknesses of snow and névé are by no means constant.

=Structure of the ice.=—Ice formed beneath a snow-field is in some sense stratified. It is made up of successive falls of snow which tend to retain the form of layers. This follows from two or three conditions. The snow of one season, or of one period of precipitation, may have been considerably changed before the succeeding fall of snow. So also the surface of the snow-field at the end of the melting season is often covered with a visible amount of earthy matter, some of which was blown up and dropped on the surface during the melting season, and some of which was concentrated in that position by the melting of the snow in which it was originally imbedded. The amount of earthy matter is often sufficient to define snows of successive years, or perhaps of minor periods of precipitation, and makes distinct the stratification which would otherwise remain obscure. The snowfall of successive years has been estimated by this means where the snow is exposed in crevices in the snow-field.

In addition to its rude stratification, the ice of the deeper portions often acquires a stratiform structure which may perhaps best be called foliation to distinguish it from the stratification which arises from deposition. The foliation appears to result mainly from the shearing of one part over another in the course of the movements to which the ice is subjected, as will be illustrated presently.

=Texture.=—The ice derived from the snow is formed of interlocking crystalline grains. The crystalline character is present from the beginning, for it is assumed by the snowflakes when they form, and the subsequent changes seem only to modify the original crystals by building up some and destroying others. By the time the snow is converted into névé, the granules have become coarse, and wherever the ice derived from the névé has been examined, the granular crystalline texture is present. The individual crystals in the ice are usually larger than those of the névé, and more closely grown together. In the fresh unexposed ice the crystals are so intimately interlocked that they are not readily seen except under a polarizing microscope, but when the ice has been honeycombed by partial melting, the granules become partially separated and may be easily seen. Fig. 223 shows quantities of them which have been washed down from the surface, and disposed as cones at its base. While a given mass of snow in a great snow- and ice-field cannot be followed consecutively through its whole history, yet since (1) the granular texture is pronounced in the névé stage where the granules show evidences of growth, and since (2) the same texture is also pronounced in the last stages of the ice when it is undergoing dissolution, as well as at all observed intermediate stages, and since (3) the crystals are, on the average, larger in proportion to the lateness of the stage of their history, while (4) experiment has shown that granules grow under the conditions which exist in snow-fields, and (5) that they persist under very considerable pressure, it is legitimate to assume that a granular crystalline condition persists throughout all stages, and is a feature of progressive growth.

=Inauguration of movement.=—Eventually the increase in depth of snow and ice in a snow-field gives rise to motion. The exact nature of the motion has not yet been demonstrated to the satisfaction of all investigators. Brittle and resistant as ice seems, it exhibits, under proper conditions, some of the outward characteristics of a plastic substance. Thus it may be made to change its form, and may even be moulded into almost any desired shape if carefully subjected to sufficient pressure, steadily applied through long intervals of time. These changes may be brought about without visible fracture, and have been thought to point to a viscous condition of the ice. There is much reason, however, as will be seen later, to question this interpretation of the ultimate nature of the movement. Whatever this may be, the mass result of the movement in a field of ice is comparable, in a superficial way at least, to that which would be brought about if the ice were capable of moving like a viscous liquid, the motion taking place with extreme slowness. This slow motion of ice in an ice-field is glacier motion, and ice thus moving is glacier ice.

If both the surface on which the ice-sheet develops and its surroundings be essentially plane, as may happen in high latitudes, and if the snow- and ice-field be symmetrical in shape, the outward movement will be approximately equal in all directions, and the area covered by the spreading ice-field will remain more or less circular. If the ice-field rests on a steeply inclined surface, like a mountain slope, the movement becomes one-sided in conformity to the slope. If the surface, otherwise plane, be affected by valleys parallel to the direction of movement, the ice in the valleys will be deeper than that on the divides between them, and its movement stronger. In the valleys, therefore, the ice will advance farther than elsewhere before being melted, and the outline of the ice will become lobate, the lobes occupying the depressions. These general relations are shown in Figs. 224 and 225. If the depressions be wide and shallow, the lobes will be broad and short (Fig. 226); if the depressions be narrow and deep, the lobes will be relatively narrow and long. If the snow and ice rest on a surface consisting chiefly of steep valleys and sharp ridges, as is common on mountain slopes, the snow and ice are chiefly gathered in the valleys, and take a linear form.

TYPES OF GLACIERS.

These different forms give rise to different terms. The ice which spreads with some approach to equality in all directions from a center is a glacier, is indeed the type of the greatest glaciers, but is commonly called an ice-cap. The same name is applied to any glacier in which there is movement in all directions from the center, even though its shape departs widely from a circle. The glacier covering the larger part of Greenland (Fig. 222) is a good example of a large ice-cap, and the glaciers on some of the flat-topped peninsular promontories of the same island are good examples of small ones (Fig. 224). If ice-caps cover a large part of a continent, as some of those of the past have done, they are often called continental glaciers.

Where ice-caps are developed on plateaus whose borders are trenched by valleys, ice-tongues from the edge of the ice-cap often extend down into the valleys and give rise to one type of valley glacier (Figs. 224 and 227). A second and more familiar type of valley glacier occupies mountain valleys, and is the offspring of mountain snow-fields (Fig. 228). The former are confined chiefly to high latitudes, and are distinguished as polar or high-latitude glaciers (Figs. 227 and 229); the latter are known as alpine glaciers (Figs. 228 and 230). The end and side slopes of polar glaciers are, as a rule, much steeper than those of alpine glaciers. When a valley glacier descends through its valley to the plain beyond, its end deploys, forming a fan (Fig. 231). The deploying ends of adjacent glaciers sometimes merge, and the resulting body of ice constitutes a piedmont glacier (Fig. 232). At the present time, piedmont glaciers are confined to high latitudes. In some cases the snow-field that gives rise to a glacier is restricted to a relatively small depression in the side of a mountain, or in the escarpment of a plateau. In such cases the snow-field and glacier are hardly distinguishable, and the latter descends but little below the snow-line. In many cases it does not even enter the narrow valley which leads out from the depression occupied by the snow-field. Such a glacier is nestled in the face of a cliff, and may therefore be called a cliff glacier (Figs. 233 and 234). The snow-field of a cliff glacier is sometimes no more than a great snowdrift, accumulated through successive years. Cliff glaciers are often as wide as long, and are always small, and between them and valley glaciers there are all gradations (Fig. 235). Occasionally the end of a valley glacier, or the edge of an ice-sheet reaches a precipitous cliff, and the end or edge of the ice breaks off and accumulates like talus below. The ice fragments may then again become a coherent mass by regelation, and the whole may resume motion. Such a glacier is called a reconstructed glacier. The precipitous cliffs of the Greenland coast furnish illustrations.

Of the foregoing types of glaciers, the ice-caps far exceed all others both in size and importance, while the valley glaciers out-rank, in the same respects, the other types; but since the valley glaciers are the most familiar type, the general phenomena of glaciers will be discussed with primary reference to them.

THE GENERAL PHENOMENA OF GLACIERS.

=Dimensions.=—Glaciers which occupy valleys leading down from snow-fields sometimes reach the upper parts of the valleys only, sometimes extend through them, and sometimes push out on the plain beyond. In length they range from a fraction of a mile to many miles, and though their width is usually much less than their length, the reverse is sometimes the case (Figs. 233, 234, and 235). Their thickness is usually measured by hundreds of feet rather than by denominations of other orders, but the variation is great, and exact measurements are almost wholly wanting. The minimum thickness is that necessary to cause movement, and this varies with the slope, the temperature, and other conditions. There is also much variation in the thickness in different parts of a glacier. As a rule, it is thinnest in its terminal portion, and thickest at some point intermediate between this and its source, but nearer the latter than the former. Cliff and reconstructed glaciers are comparable in size to the smaller valley glaciers. Piedmont glaciers may attain greater size.

An ice-cap is theoretically thickest at its center and thins away to its borders, but its actual dimensions are influenced by the topography on which it is developed. The Greenland ice-cap is known to rise about 9000 feet above the sea, and it probably reaches considerably higher than this in the unexplored center of its broad dome. The height of the land surface beneath is unknown, but it is unlikely that it averages half this amount, and hence the ice is probably 5000 feet or more thick in the center. There is reason to think that it is much thicker in Antarctica.

=Limits.=—The ice of a glacier is always moving forward (neglecting temporary halts), but the end of a glacier may be retreating, advancing or remaining stationary, according as the rate of wastage is greater, less, or just equal to the forward movement of the ice. The position of the lower end of the glacier is therefore determined by the ratio of movement to wastage. Its upper end is generally ill-defined. In a superficial sense, it is the point where the ice emerges from the snow-field; but the lower limit of the snow-field is often ill-defined, and in any case is not the true upper limit of the glacier, since there must be movement from the granular mass of ice beneath the snow to make up for the waste below, and the moving ice beneath the snow-field which feeds the tongue of ice in the valley is just as really a part of the glacier as the more consolidated portion in the valley below. If a definite upper limit for an alpine glacier is to be named, it should probably be the Bergschrund, a gaping crevasse, or series of crevasses which sometimes open near the precipitous slope of the peak or cliff where the snow-field lies. The Bergschrund is formed by the moving of the lower part of the snow-field away from the portion above.

The lower end of a glacier is usually free from snow and névé in summer, but, traced toward its source, it first becomes covered with névé, then with snow, and finally merges into the snow-field without having ceased to be a glacier. The term glacier is, however, commonly used to mean merely the more solid portion outside (below) the névé.

=Movement.=—The fact of glacier movement is established in various ways, the most obvious being by the advance of its lower end. Such advance is too slow to be seen from day to day, and is only detected when the lower end of the glacier overrides or overturns objects in front of it, or moves out over ground previously unoccupied. But even when the end of a glacier is not advancing, the movement of the ice may be established by means of stakes or other marks set on the surface. If the positions of these marks relative to fixed points on the sides of the valley be determined, they are found after a time to have moved down the valley. Rows of stakes or lines of stones set across a glacier in the upper, middle and lower portions have revealed many facts concerning the movement of the ice.

Generally speaking, the middle of a valley glacier moves more rapidly than its sides (Fig. 236), but in some cases, especially in large glaciers, there are found to be two or more main lines of movement, with belts of lesser movement between. The top of a glacier moves, on the whole, more rapidly than the bottom, though the observations made do not show that the rate of movement diminishes regularly downward, and it probably does not so diminish in many cases. In Switzerland, where the glaciers have been studied more carefully than elsewhere, the determined rates of movement range from one or two inches to four feet or more per day. Some of the larger glaciers in other regions move more rapidly, but it does not follow that large glaciers always move faster than small ones. The Muir glacier of Alaska has been found to move seven feet or more per day, and some of the glaciers of Greenland have been found to move, in the summer time, 50, 60, or even more feet per day. A single estimate as high as 100 feet per day has been made; but these high rates have been observed only where the ice of a large inland area crowds down into a comparatively narrow fjord, and debouches into the sea, and then only in the summer. In the case of the glacier with the highest recorded rate of summer movement, 100 feet per day, the advance was only 34 feet at about the same place in April.

The average movement of the border of the inland ice of Greenland is very small. Rink says that “between 62° and 68° 30′, the edge of the inland ice is almost stationary for a remarkably long distance.” The observations of the authors between 77° and 78° were of like import. Probably the average movement of the border of the Greenland ice-cap is less than one foot a week.

=Conditions affecting rate of movement.=—The rate of glacier movement appears to depend on (1) the depth of the moving ice; (2) the slope of the surface over which it moves; (3) the slope of the upper surface of the ice; (4) the topography of the bed over which it passes; (5) the temperature; and (6) the amount of water which falls upon it or is carried to it by the drainage of its surroundings, in addition to that produced by the melting of the glacier itself. Great thickness, a steep slope, much water, smoothness of bed, and a high (for ice) temperature favor rapid movement. Since some of these conditions, notably temperature and amount of water, vary with the season, the rate of movement for any given glacier is not constant throughout the year. Other conditions, especially the first of those mentioned above, vary through longer periods of time, and occasion periodic variations in the rate of movement.

Since the volume of ice concerned influences the rate of movement, anything which changes the volume affects the rate. An excess of snowfall with favorable conditions for its preservation for a period of years, increases the volume of ice, and tends to accelerate its movement. A deficiency in snowfall, or in its preservation, as from high average temperature or from aridity, diminishes the quantity of ice, and so retards the movement. An acceleration of velocity causes the ice to move down the valley farther before being melted, that is, causes the end of the glacier to advance, while a decrease of velocity produces the opposite effect. As a matter of fact, the lower ends of glaciers advance for a period of years and then retreat, to advance again at a later time. Observation has shown that the periods of advance follow a succession of years when the snowfall has been heavy and the temperature low, while the periods of retreat follow a succession of years when the snowfall has been light and the temperature above the average. The periods of advance and retreat lag behind the periods of heavy and light snowfall respectively, by some years, and a long glacier responds less promptly than a short one. Present knowledge seems to point to a period of 35 to 40 years as the time within which a cycle of fluctuation, that is, an advance and a retreat, takes place.

A declining upper surface is essential to glacier motion. There are short stretches where this is not the case, and indeed there are particular places where the upper surface slopes backward. This may occur where the ice is pushed up over a swell in its bed, or is crowded up against any considerable obstacle; but such cases are no more than local exceptions, and do not militate against the truth of the general statement that the upper surface of a glacier declines in the direction of motion. A declining lower surface is less necessary. In the case of valley glaciers, the bed does, as a rule, decline in the direction of motion, but that there are local exceptions is shown by the deep basins in rock which such glaciers often leave behind them when they retreat. In the great continental glaciers of recent geologic times, the ice frequently moved up slopes for scores, and even hundreds of miles; but in all such cases, the upper surface must have declined in the direction of movement. With a given thickness of ice, the greater the decline of its lower surface in the direction of motion, the more rapid its progress. A rough bed, or a crooked course retards the motion of a glacier, while a smooth bottom and a straight course facilitate it.

Slope, roughness of bed, and volume affect the movement of glaciers somewhat as they affect the movement of rivers. The temperature of the water, on the other hand, has little effect on the flow of a river so long as it remains unfrozen; but the effect of temperature on the motion of ice is most important. In many cases, indeed, the temperature, together with the water that is incidental to it, seems to be the chief factor in determining the rate of movement. The way in which its effects are felt will be discussed later.

=Likenesses and unlikenesses of glaciers and rivers.=—Many of the characteristics of a valley glacier may be understood from the study of the accompanying figure (Fig. 237) of the White (Alaska) glacier. From this figure it will be seen that the glacier is an elongate river-like body, following the curves of the valley in stream-like fashion. It has its origin in the snows collected on the mountain heights seen in the distance, and it works its way down the valley in a manner which, in the aggregate, is similar to the movement of a stiff liquid. The likeness to a river extends to many details. Not only does the center move faster than the sides, and the upper part faster than the bottom, as in the case of streams, but the movement is more rapid in constricted portions of the valley and slower in the broader parts. These and other likenesses, some of which are apparent rather than real, have given origin to the view that glacier ice moves like a stiff viscous liquid.

But while the points of likeness between glaciers and rivers are several, their differences are at least equally numerous and significant. The trains of débris on the surface (the dark bands in the illustration), like the central currents of streams, pass nearer the projecting points of the valley walls and farther from the receding bends; but beyond this point the analogy fails, for the trains of débris on the ice do not conform in detail to the courses of the currents of a winding stream, nor is there evidence of the rotatory motion that characterizes river water. Furthermore, the glacier is readily fractured, as the numerous gaping crevices on many glaciers show. The crevasses are sometimes longitudinal, sometimes transverse, and sometimes oblique. In the case of Arctic glaciers, longitudinal crevassing is especially conspicuous.

Crevasses appear to be developed wherever there is appreciable tension, and the causes of this tension are many. An obvious cause is an abrupt increase of gradient in the bed (Fig. 238). If the change of gradient be considerable, an ice-fall or cascade results, and the ice may be greatly riven (Fig. 228). Below the cascade, the surface may bristle with wedges and pinnacles of ice (séracs, Fig. 239). Transverse crevices at the margin sometimes appear to be the result of the tension developed on a curve. Oblique crevices on the surface near the sides are commonly ascribed to the tension between the faster-moving center and the slower-moving margins, and in like manner crevasses that rise obliquely from the bottoms are attributed to the tension between the faster-moving portions above and the slower-moving portions below. All these crevasses indicate strains to which a liquid, whose pressures are equal in all directions, does not offer a close analogy. Longitudinal crevasses may affect both the river-like part of a glacier and its deploying end, and are the result of tension developed by movement within the ice itself, to which, again, rivers offer no analogy. Somewhat similar cracks develop in the outer crust of asphalt, when a mass of it is allowed to stand and spread; but in this case there is evaporation of the volatile ingredients, giving to the outer part relative rigidity and brittleness, while the inner part remains more fluent. The analogy is therefore not perfect and probably not really illustrative. The crevices may be narrow or wide, and both narrow and wide may be found in the same glacier. The narrow crevices that never open much are the most significant, as they show that very little stretching is needed to satisfy the tension. The opening of a gaping crevice is sometimes the work of weeks, and, in the slow-moving glaciers of high latitudes, sometimes the work of successive seasons. All this shows that the glacier is a very brittle body, incapable of resisting even very moderate strains brought to bear upon it very slowly. Had the ice even moderate ductility, it would adapt itself to tension brought to bear upon it so slowly as are many of the tensions which produce crevassing. In its behavior under tension therefore a glacier is notably unlike a river.

SURFACE FEATURES.

=Topography.=—Many of the minor irregularities of the surface of a glacier are the result of crevassing. After the ice is crevassed, the sun’s rays and the air which has been warmed by them penetrate the openings and melt the ice. The melting is most rapid at the top, and decreases downward. The result is that the sections of ice between adjacent crevasses are narrowed into wedges. If there be cross-crevassing, as is common, points instead of wedges result. As the sort of surface shown in Fig. 239 develops, any débris which was on the ice slides into the crevices, and the upper surface becomes clean.

Where ice is crevassed transversely, and where melting is not rapid, the crevasses may close as the ice moves forward, and the regelation of adjoining faces heals the rents in the surface. Even in this case, however, the surface is likely to be more or less undulating because of the waste on the sides of the crevices before they are closed. After regelation, surface ablation tends to obliterate the protuberances.

The topography of the surface of the ice is affected by other conditions. All parts of the surface of the ice are not equally compact, and the least compact portions melt most rapidly, giving rise to depressions, while the more solid parts occasion protuberances. Both depressions and protuberances may be regular or irregular in form (Figs. 240 and 241). Undulations of the bed often show themselves in the surface of the ice as suggested by Fig. 242. In such cases, ponds or lakelets sometimes accumulate on the surface of the ice. The topography of the ice in such cases seems to show that the ice is forced up slope.

=Surface moraines.=—The surface of a glacier is often affected by débris of one sort or another, and this also influences its topography. The débris is sometimes disposed in the form of belts or moraines (Figs. 237, 243). The surface moraines may be lateral, medial, or terminal. A lateral moraine is any considerable accumulation of débris in a belt on the side of a glacier. A medial moraine is a similar accumulation at some distance from the margins, but not necessarily in or near the middle. There may be several medial moraines on one glacier, in which case some of them may be far from the center. In alpine glaciers, the surface terminal moraine is less well-defined; in polar glaciers it often connects two lateral moraines, making a loop roughly concentric with the terminus of the glacier.

Besides the surface moraines, which represent belted aggregations of débris, there may be scattered bowlders and bits of rock of various sizes on the ice, and, in addition to the coarse material, there is often some dust which has been blown upon the ice.

=Relief due to surface débris.=—The débris on the ice affects its topography by influencing the melting of the subjacent and adjacent ice. The rock débris absorbs heat more readily than the ice. A small and thin piece of stone lying on the ice is warmed through by the sun’s rays, and, melting the ice beneath, sinks, just as a piece of black cloth on snow will sink because of the increased melting beneath it. Though a good absorber of heat, rock is a poor conductor, and so the lower surface of a large mass of stone is not notably warmed. The ice beneath it is protected from the direct rays of the sun, and is therefore melted more slowly than that around it. The result is that the bowlder presently stands on a protuberance of ice (Fig. 244). When its pedestal becomes high, the oblique rays of the sun and the warm air surrounding it cause it to waste away, and the capping bowlder falls. In high latitudes, the great obliquity of the rays sometimes allows them to strike under isolated bowlders. In this case, they are warmed from below, and thus aid rather than hinder the melting of the ice.

The same principles apply to the moraines. A thin bowlder moraine in high latitudes is sometimes sunk below the surface (Fig. 242). Usually, however, a medial moraine protects the ice beneath from melting, and occasions the development of a ridge of ice beneath itself. As the ice on either side is then lowered by ablation, the moraine matter of the medial belt tends to slide down on either hand. The same is true of the lateral moraines. So far does this spreading go, that in some cases the lower end of a glacier is completely covered with the débris which has spread from the medial and lateral moraines. Examples of this may be seen in almost any region of abundant, long, alpine glaciers.

=Dust-wells.=—The wind-blown dust sometimes gives rise to peculiar topographic features of small size. The dust is not distributed by the wind with absolute equality, and the surface drainage of the ice tends to aggregate it. Every dust particle acts like a small stone, and where aggregations of dust occur, they melt their way down into the ice, developing holes or “dust-wells” (Fig. 245). These wells rarely reach a depth of more than a few inches, but they may be so numerous that the pedestrian is obliged to watch his steps. This is especially true near the edge of the large ice-caps. It is evident that the depth of these dust wells must be slight, for so soon as they are deep enough to cut off the sun’s rays from the dust at the bottom, the deepening ceases. Other things being equal, they are deeper in low latitudes than in high.

=Débris below the surface.=—The lower part of a glacier, as well as the upper, carries rock débris. This débris is sometimes so abundant, especially near the ends and edges of the ice, that it is difficult to locate the bottom of the glacier; for between the moving ice which is full of débris, and the stationary débris which is full of ice, there seems to be a nearly complete gradation. The débris in the lower part of arctic glaciers, and to some extent of others, is often disposed in thin sheets sandwiched in between layers of clean ice. These débris sheets are often numerous and usually discontinuous, though groups of such sheets often persist for considerable distances. Débris also occurs to some extent in the ice well above its base. It is sometimes in belts, as seen in section, and sometimes in bunches. These various relations are illustrated by Figs. 227, 229, and 246–249.

Another characteristic of the basal débris-laden part of some glaciers is the foliation of the ice (Figs. 248, 249, etc.). This is especially well shown in the arctic glaciers, the ends and sides of which have steep or vertical faces. The foliation is best developed in the débris zone, though often shown above. The foliation is sometimes minute, consisting of layers of clean ice, an inch or less in thickness, separated by mere films of earthy matter. In extreme cases there are a score or more of laminæ within a foot. Locally, and especially where débris is abundant, the laminæ are much contorted. This is seen both in section (Figs. 248 and 249) and on the surface (Fig. 250).

TEMPERATURE, WASTE, AND DRAINAGE.

The temperature of glacier ice may range downward from the freezing point of water much as other solid portions of the earth’s surface, but it has a fixed upper limit at 32° Fahr. (0° C.) because all the heat it receives tending to raise its temperature above that point, is converted into the latent form by the melting of the ice. The range of temperature is greatest at the surface, where it varies from 32° in the summer, to the coldest temperature of the region where the ice occurs. Beneath the surface the range of temperature is more restricted, and increasingly so with increasing depth.

The variation of temperature at the surface is due primarily to the varying temperature of the air. During the cold season, a wave of low temperature (the winter wave), starting at the surface, penetrates the ice, and during the warm season a wave of higher temperature (the summer wave) takes the same course. The day and night waves and other minor variables are, for present purposes, negligible.

=The winter wave.=—There are but few observations on the internal temperatures of glaciers during the winter season, but it seems certain that the winter wave diminishes rapidly downward and dies out below, much as does the winter wave which affects land surfaces not covered with ice. Conduction alone considered, the temperature of the ice where the cold wave dies out, should correspond, approximately, to the mean annual temperature of the region, provided that temperature is below the melting point of ice.

Assuming that in the high altitudes and high latitudes where glaciers abound, the temperature of the surface may average about -12° Fahr. (about -25° C.) for the winter half of the year, which is about the case for north Greenland, Spitzbergen, and Franz Josef Land, and that the conductivity of the ice in the C. G. S. system is .005, the temperature would be lowered appreciably only about 40 feet below the surface at the close of the winter period, conduction only being considered. How far the internal temperature may be influenced by air forced through the ice by winds and by variations of the barometer is not known and cannot well be estimated. The wave of low temperature descending from the surface in winter would probably become inappreciable before reaching a depth of 60 feet. At this depth the temperature should be about 15° Fahr.—the mean annual temperature of the region.

=The summer wave.=—The warm wave follows the analogy of the summer wave of ice-free land surfaces much less closely. This is because of the low melting temperature of ice as compared with other forms of solid earth-matter. On this account the summer wave is bi-fold. The one part travels downward by conduction, the other by the descent of water; the one has to do primarily with the temperature before the melting-point of ice is reached; the other, with the temperature after that point is reached; the first conforms measurably to the warm wave affecting other solid earth-matter, while the second is governed by special laws. After the surface portion of the ice is brought to the melting temperature, the additional heat which it receives melts the ice and is transformed from sensible into potential heat. Ice charged with water is potentially, but not sensibly, warmer than ice which has just reached the melting temperature.

The warm wave of conduction dies out below like the cold wave. The warm wave descending by the flow of water stops where the freezing temperature of water is reached. In regions where the average temperature is below freezing, the water-wave does not descend so far as the wave of conduction, since the latter descends below the zone where the melting temperature is found.

The foregoing considerations warrant the generalization that glaciers normally consist of two zones (1) an outer or upper zone of fluctuating temperature, and (2) an under zone of nearly constant temperature. The under zone obviously does not exist where the thickness of the ice is less than the thickness of the zone of fluctuating temperature. This may be the case in very thin glaciers in very cold regions, and in the thin ends and edges of all glaciers.

=The temperature of the bottom.=—The internal heat of the earth is slowly conducted to the base of a glacier where it melts the ice at the estimated average rate of about one-fourth of an inch per year. The temperature of melting is a little below 32° Fahr. since pressure lowers the melting-point at the rate of .0133° Fahr. (.0075° C.) for one atmosphere of pressure. At the bottom of a mile of ice therefore the melting temperature is about 30.2° Fahr. (-1° C.) It is probable that in all thick glaciers the temperature of the bottom is constantly maintained at the melting-point. This may be indicated by the streams which issue from beneath glaciers during the winter, though this criterion is hardly decisive since the issuing waters may be derived partly or wholly from the rock beneath. In glaciers or in parts of glaciers so thin as to lie wholly within the zone of fluctuating temperature, the temperature of the bottom is obviously not constant.

=Temperature of the interior of the ice.=—The variation of temperature of the surface of a glacier has already been shown to lie between a maximum of 32° Fahr. and the minimum temperature of the region where the glacier occurs. Lower, in the zone of fluctuating temperature, the variation is less, and where the zone of fluctuating temperature passes into the zone of constant temperature, variation ceases. The thickness of the zone of fluctuating temperature varies with the temperature of the region where the glacier occurs, being greatest where the winters are coldest. In the case of all glaciers except thin ones in very cold regions, the temperatures within the zone of constant temperature range from the mean annual temperature of the region at the top of the zone (provided this is not above the melting-point of ice at this depth) to the melting temperature of the ice at the bottom. Within these limits the range may be great or slight.

If we consider only the effects of the external seasonal temperatures and the internal heat of the earth, it appears that all the ice in the zone of constant temperature in the lower end of a typical alpine glacier should have a constant melting temperature, for the average temperature of regions where the ends of such glaciers occur is usually above 32° Fahr., and this determines a temperature of 32° Fahr. (or a little less) at the top of the zone, while a melting temperature is maintained at the bottom by the earth’s interior heat. In thin glaciers of very cold regions, where the zone of constant temperature has relatively slight thickness, the low temperature descending from the surface may so far overcome the effect of internal heat as to keep the bottom of the ice at a freezing temperature. In all other cases, the ice at the bottom of the under zone has a melting temperature, while that above is probably colder.

In the higher altitudes and in the polar latitudes where glaciers are chiefly generated, the mean annual temperature of the surface is usually below the melting-point of ice. Here the temperature of the ice between the top and bottom of the zone of constant temperature must, on the average, be below the melting-point, unless heat enough is generated in the interior of the ice to offset the effect of the temperature above. For example, where the mean annual temperature is 20° Fahr. or lower, as in middle Greenland, Spitzbergen, and Franz Josef Land, and at certain high altitudes in more southerly latitudes, the mean temperature in the zone of constant temperature should range from 20° Fahr. at the top to 32° Fahr. (or a little less) below; i.e., it should average about 6° below the melting-point. Under these conditions, all the ice in the zone of constant temperature, except that at its bottom, must be permanently below the melting-point, but it is perhaps worthy of especial note that much of it is but little below. In alpine glaciers the part of the ice affected by this constant low temperature (below freezing) is presumed to be chiefly that which lies beneath the snow-fields. In polar glaciers the low temperature probably prevails beneath the surface, not only throughout the great ice-caps, but also in the marginal glaciers which descend from them.

From these theoretical considerations we may deduce the generalization that in the zone of constant temperature within the area of glacial growth, the temperature of the ice is generally below the melting-point, while within the area of wastage, the temperature of the corresponding zone is generally at the melting-point.

=Compression and friction as causes of heat.=—The foregoing conclusions are somewhat modified by dynamic sources of heat. The compression arising from gravity, and the friction developed where there is motion, are causes of heat. Since friction occurs only when motion takes place, the heat which it generates is secondary and may, for present purposes, be neglected. Compression not only lowers the melting-point slightly, but it produces heat at the point of compression. Where the ice is granular, the compression, due to weight, takes place at the contacts of the grains. At intermediate points the pressure tends to cause them to bulge, and this has the effect of lowering the temperature of the bulging points. If therefore the compression be considerable, the granules may be warmed to the melting-point where they press each other, while at other points their temperature may be lower. In this case melting will take place at the points of compression, and the moisture so produced will be transferred to the adjacent parts of the granule and immediately refrozen. Melting at the points of compression would result in some yielding of the mass, and in some shifting of the pressure to new points where compression and melting would again take place. Thus the melting, the refreezing, and the attendant movement might go on until the limits of the power of gravity in this direction were reached. From considerations already adduced, it appears that the temperature in some parts of every considerable body of ice must be such as to permit these changes. The heat due to depression and friction may modify the theoretical deductions drawn above from atmospheric and internal influences.

=Summary.=—If the foregoing generalizations be correct, (1) the surface of a glacier is likely to be melted during the summer, (2) its immediate bottom is slowly melting all the time (unless the thickness of the ice be less than the thickness of the zone of annual variation or of permanent freezing temperature); (3) its subsurface portion in the zone of waste is generally melting, owing to descending water, compression, and friction; while (4) its subsurface portion in the zone of growth is probably below the melting-point except as locally brought to that temperature by compression, friction, and descending water, and at the bottom by conduction from the rock beneath.

=Movement under low temperature.=—Glacier motion will not be discussed at this point, but one of the bearings of the preceding conclusions on glacier motion may be pointed out. Since there must be motion in the area of growth to supply the loss in the area of waste, the fundamental cause of motion must be operative in bodies of ice the mean temperature of which is below the melting-point, unless the dynamic sources of heat are considerable. This fundamental cause does not exclude the coöperation of causes that work only (1) at the melting temperature, or (2) where the ice is bathed with water, or (3) in the plane of contact between wet ice above and dry ice below. These may be auxiliary causes which abet the fundamental one in producing the more rapid movement of warm seasons, or in bringing about the especially rapid movement in situations where there is abundant water, or in inducing the shearing which is such a remarkable feature of arctic glaciers.

=Evaporation.=—The ice wastes by evaporation as well as by melting, and while the former process is far less important than the latter, its results are probably larger than is commonly apprehended. One of the most remarkable features of some of the deposits of ancient glaciers is the slight evidence they afford of escaping waters. The most plausible explanation seems to lie in the supposition that the ice was largely wasted by evaporation. This conclusion finds support in many places in the presence of a mantle of fine silt over the drift, the silt being apparently composed of dust blown upon the ice. It is supposed to imply aridity in the region about the ice. If a sufficient mantle of dust were spread over the border zone of the ice, and if the air were very dry, nearly all the water melted on the surface of the ice might be held back by the dust-wells until the water was evaporated or absorbed.

=Drainage.=—Some of the water produced by surface melting forms little streams on the ice. Sooner or later they plunge into crevasses or over the sides and ends of the glacier. In the former case, they may melt or wear out well-like passages (moulins) in the ice, and even in the rock beneath. Much of the surface water sinks into the ice. Its ready penetration is aided by the “dust-wells” which mark the surface of many glaciers. In north Greenland wells which contain six or eight inches of water at the end of a warm day are often dry in the morning. The water has leaked out and passed to lower levels. From these and other harmonious observations it is inferred that the superficial part of a glacier at least is readily penetrated by water. The depth to which surface water penetrates is undetermined, but it doubtless varies greatly, not only in different glaciers, but in different parts of the same glacier, and in the same part at different times. Above the line of perennial snow there is little water either from melting or from rain, and hence relatively slight penetration. Below the line of perennial snow there is much melting and much rain, and here it is probable that the water sometimes, perhaps usually, penetrates to the bottom of the ice during the melting season, even independently of crevasses.

Once within the glacier, the course of the water is variable. Exceptionally it follows definite englacial channels, as shown by springs or streams issuing from the ice at some point above its bottom (Fig. 251). Oftener it descends or moves forward through the irregular openings which the accidents of motion have developed. If it reaches a level where the temperature is below its freezing-point, it congeals. Otherwise it remains in cavities or descends to the bottom. The water produced by melting within the glacier probably follows a similar course. So far as these waters descend to the bottom, they join those produced by basal melting and issue from the glacier with them. In alpine glaciers the waters beneath the ice often unite in a common stream in the axis of the valley, and hollow out a tunnel. Thus the Rhone is already a considerable stream where it issues from beneath the Rhone glacier. In the glaciers of high latitudes, subglacial tunnels are less common and the drainage is in streams along the sides of the glaciers or through the débris beneath and about them.

At the end of the glacier, all waters, whether they have been superglacial, englacial or subglacial, unite to bear away the silt, sand, gravel, and even small bowlders set free from the ice, and to spread them in belts along the border of the ice or in trains stretching down the valleys below. These are the most common of the glacio-fluvial deposits.

THE WORK OF GLACIERS.

Erosion and transportation.

The work accomplished by glaciers is distinctive, for while like rivers, they abrade the valleys through which they pass, carry forward the material which they remove from the surface, and wear, grind, and ultimately deposit it, and while their work therefore includes erosion, transportation, and deposition, their method is peculiar.

=Getting load.=—If the surface on which the snow-field which is to become a glacier accumulates be rough and covered with abundant rock débris, as such surfaces usually are, the glacier already has a basal load when its movement begins, for the snow covers, surrounds, and includes such loose blocks of rock as project above the general surface and envelops all projecting points of rock within its field. When the snow becomes ice and the ice begins to move, it carries forward the loose rock already imbedded in it, and tears off the weak points of the enveloped rock-projections. It may perhaps also move some of the soil and mantle rock of the original surface to which it is frozen. In addition to the subglacial load which the glacier thus has at the outset, there may be a surface load which has fallen on the snow or ice from cliffs above. This is especially true of mountain-valley glaciers. If this has been buried by snow and ice it is englacial; if it lies on the surface it is superglacial.

Once in movement, the ice carries away the débris to which it was originally attached, and at the same time gathers new load from the same area. The new load is acquired partly by the rasping effect of the rock-shod ice on its bed, and partly by its rending power which, under favorable circumstances, may quarry out considerable blocks of rock. This “plucking” process is at its best where the ice passes over cliffs of jointed rock or steep-sloped hills.

As the ice advances into new territory it acquires additional basal load, partly by rasping, partly by plucking, partly by freezing to it, and partly in other ways. One of these ways may be illustrated by the sequence of events when the end of a glacier advances on a very large bowlder. As the ice approaches it, the reflection of heat from it melts the adjacent edge of the ice, making a slight reëntrant. With farther advance, the ice closes in against and around the bowlder, and finally carries it along in the bottom of the moving mass. In some cases, especially when its advance is rapid, the ice may push débris in front of itself. Even where this is the case, the amount of material pushed forward is generally slight, partly because the extreme edge of the ice often fails to rest on the land in summer, when the movement is greatest, being melted from below by the heat of the surface over which it is spreading (see Fig. 252), and partly because the earth in front of the glacier is frozen during a large part of the year. In this condition, the earthy matter has greater resistance than the ice, and the latter rides over it. Superglacial material may be acquired during movement by the fall of débris from cliffs, or by the descent of avalanches.

=Conditions influencing rate of erosion.=—An obstructive attitude of the surface toward the movement of the ice is as necessary for effective erosion as the movement of the ice itself. Advancing over a flat surface, ice ordinarily inflicts but little wear, since there is little for it to get hold of. So slight is the abrasive power of ice under these conditions that it frequently overrides and buries the soil with more or less of its herbaceous vegetation. But while a certain measure of roughness of surface is favorable for glacial erosion, the topography may be so uneven as to seriously impede the ice. Erosion is probably at its maximum, so far as influenced by topography, when the roughness of the surface is such as to offer notable catchment for the basal ice, but not such as to impede its motion very seriously. The amount of relief favorable for the greatest erosion increases with increasing thickness of the ice.

Other conditions which influence erosion by ice are (1) the amount of loose or slightly attached débris on the surface, (2) the resistance of the rock, (3) the slope of the surface, (4) the thickness of the ice, (5) the rate of movement, and (6) the abundance and character of the débris which the ice has to work with. The effect of the first five of these conditions is evident. The effect of the last is less simple. Clean ice passing over a smooth surface of solid rock has little effect upon it; but a rock-shod glacier will abrade the same surface notably. The effect of this abrasion is shown in the grooves and scratches (striæ) which the stones in the bottom of the ice inflict on the surface of the rock over which they pass (Figs. 253, 255, and 256). At the same time the stones in the ice are themselves worn both by abrasion with the bottom, and with one another (Fig. 254). It does not follow, however, that the more material in the bottom of the ice the greater the erosion it effects; for with increase of débris there may be decrease of motion and, beyond a certain point, the decrease of motion seriously interferes with the efficiency of erosion. When any considerable thickness of ice at the bottom of the glacier is full of débris, this loaded basal portion may approach stagnancy, and the lower limit of considerable movement may lie between the loaded ice below and the relatively clean ice above. A moderate, but not an excessive load of débris is, therefore, favorable for great erosion. Something depends, too, on the character of the load. Coarse, hard, and angular débris is a more effective instrument of erosion than fine, soft, or rounded material. The adverse influence of the overloading of the ice on its motion has been likened to the stiffening of a viscous liquid by the addition of foreign matter, but it may better perhaps be referred to the destruction of the granular-crystalline continuity on which glacier motion probably depends.

From the preceding statement, it is evident that erosion is not equally effective at all points beneath a glacier. So far as concerns the ice itself, erosion is not most effective at the end of a valley glacier, or at the edge of an ice sheet, for here the strength of movement is too slight and the load too great; nor is the most effective erosion at the source or near it, for though the ice may here be thick, the movement is slow and the load likely to be slight. Ice conditions only being considered, erosion is most effective somewhere between the source and the terminus, and probably much nearer the latter than the former. The conditions of the surface over which the ice passes may be such as to vary the place of greatest erosion widely. Thus in an Alpine glacier, erosion may be most effective at the Bergschrund because the slope here favors “plucking.” Here, notable amphitheatres (cirques) are sometimes excavated. After the glacier disappears, the bottom of the cirque is often seen to contain rock basins (Fig. 257). Glacial cirques abound in mountains where glaciers once existed, but from which they have now disappeared. The cirques of the Bighorn mountains of Wyoming (Pl. XIX) are examples.

=Summary.=—In summary it may be said that rapidly moving ice of sufficient thickness to be working under goodly pressure, shod with a sufficient but not excessive quantity of hard-rock material, passing over incoherent or soft formations possessing a topography of sufficient relief to offer some resistance, and yet too little to retard seriously the progress of the ice, will erode most effectively.

=Varied nature of glacial débris.=—From its mode of erosion it will readily be seen that the bottom of a glacier may be charged with various sorts of material. There may be (1) bowlders which the ice has picked up from the surface, or which it has broken off from projecting points of rock over which it has passed; (2) smaller pieces of rock of the size of cobbles, pebbles, etc., either picked up by the ice from its bed or broken off from larger masses; (3) the fine products (rock-flour) produced by the grinding of the débris in the ice on the rock-bed over which it passes, and similar products resulting from the rubbing of stones in the ice against one another; and (4) sand, clay, soil, vegetation, etc., derived from the surface overridden. Thus the materials which the ice carries (drift) are of all grades of coarseness and fineness, from large bowlders to fine clay. The coarser material may be angular or round at the outset, and its form may be changed and its surface striated as it is moved forward. Whether one sort of material or another predominates, depends primarily on the nature of the surface overridden.

PART OF THE BIGHORN MOUNTAIN RANGE, WYOMING. U. S. Geol. Surv.]

A SECTION OF THE CALIFORNIA COAST NEAR SAN MATEO, CALIFORNIA. U. S. Geol. Surv.]

=The topographic effects of glacial erosion.=—In passing through its valley, an alpine glacier deepens and widens its bottom and smooths its slopes up to the upper limit of the ice. It tends to change a V-shaped valley (Fig. 258) into a U-shaped one (Fig. 259). The change in topography at the upper limit of glaciation is often marked (Figs. 260 and 261).

The deepening of a valley by glacial erosion may throw its tributaries out of topographic adjustment. Thus if a main valley is lowered 100 feet by glacial erosion while its tributary is not deepened, the lower end of the latter will be 100 feet above the former when the ice disappears. Such a valley is called a hanging valley (Figs. 262 and 263). Such valleys are of common occurrence in regions recently glaciated, but now ice-free. Examples are common in the western mountains of North America and elsewhere.

Ice-caps which overspread the surface irrespective of valleys and hills, tend to reduce the angularities of the surface. Hills and ridges are cut down and smoothed (Figs. 264 and 265); but since valleys parallel to the direction of movement are deepened at the same time, it is doubtful if the relief of the surface is commonly reduced by the erosion of an ice-cap.

=Fiords.=—A glacier descending into the head of a narrow bay may gouge out the bay to a very considerable depth, causing its head to recede. When the ice finally melts, the bay may be a fiord. Thus have arisen the glacial features of many of the fiords of high-latitude coasts, and many of the glaciers of those coasts are now making fiords (Fig. 266). Fiords also arise in other ways. Coasts indented by fiords are likely to be bordered by islands.

=The positions in which débris is carried.=—As a result of the methods by which a glacier becomes loaded, there are three positions in which the débris is carried: (1) the basal or subglacial, (2) the englacial, and (3) the superglacial. The material picked up or rubbed off from the surface over which the ice moves is normally carried forward in the base of the ice; while that which falls on the surface is usually carried in the form of surface moraines. In the former position the drift is basal; in the latter, superglacial. It is doubtful if much débris is moved along beneath (that is, strictly below the bottom of) the ice, though the movement of the latter would have a tendency to drag or urge along with it the loose material of its bed. If drift were carried forward in such positions, it would be strictly subglacial.

The basal load of a glacier is constantly being mixed with new accessions derived from ground over which the ice is passing, and this admixture tells the story of the work done by the bottom of the ice. The englacial and superglacial material, on the other hand, is normally borne from the place of origin to the place of deposition without such intermixture. It is a case of “local” versus “through” transportation.

=Transfers of load.=—While the origin of the load usually determines its position, exceptions and complications arise from the transfer of load from one position to another, and from the gradation of one horizon into another.

Most of the débris gathered by ice is acquired at its bottom. While such material is basal at the outset, some of it may find itself above the bottom a little later. Thus when ice passes over a hill (Fig. 267) the bottom of the ice rends débris from the top of the hill. When it descends from one level to another there is a similar result (Fig. 268). To the lee of the hill the ice from either side may close in under that which came over the top, in which case the débris derived from the top of the hill by the bottom of the overriding ice will be well up in the ice. It has passed from an initial basal to a subsequent englacial position. The change does not usually involve an actual rise of the material, but rather a decline. If carried upward at all, the upward movement is temporary only, and incident to the passage of the ice over the hill, or to other local causes. The englacial débris may be little or much above the basal zone according to the height of the elevation overridden.

Superglacial débris may obviously become englacial by falling into crevasses or by being carried down by descending waters. Either superglacial or englacial débris may become basal by the same means.

From their form and position, there is less ice-free land in immediate association with ice-caps than with valley glaciers. Furthermore, the ice-free land about the borders of an ice-cap is less likely to be in the form of cliffs above it. As a result, the surfaces of ice-caps are comparatively clean, except at their edges where the ice is thin.

Englacial material may become superglacial by surface ablation. In this case the drift does not rise, but melting brings the surface of the ice down to its level. This occurs chiefly at the end or edge of the ice, where the surface melting is greatest. Englacial débris, especially that near the bottom, may also become basal by the melting of the bottom of the ice.

Englacial material plucked or rasped from an elevation over which the ice has passed is liable to be disposed in a longitudinal belt in the ice in the lee of the elevation itself. By surface ablation this material may reach the surface at some point below its source, and be disposed as a medial moraine. Such a moraine has an origin very different from that of a medial moraine formed by the junction of two lateral moraines of superglacial origin.

Much less in the natural order of things is the transfer of material from a basal to an englacial and from an englacial to a superglacial position by upward movement of the débris itself. Such transfer is remarkable because the specific gravity of rock is from two and a half to three times as great as that of ice, so that its normal tendency is to sink.

In arctic glaciers, and probably in others, some material which has been basal becomes englacial by being sheared forward over ice in front of it. So far as observed this takes place chiefly where the ice in front of the plane of shearing lies at a lower level than that behind, as where the surface of an upland falls off into a valley, or where a boss of rock shelters the ice in its lee from the thrust of the overriding ice (Fig. 268).

At the borders of arctic glaciers the lower layers are not infrequently upturned, as shown in Figs. 269 to 272. Where the layers turn up at the end of a glacier (Figs. 269 and 270), basal and englacial débris is carried to the surface by actual upward movement, and a terminal moraine or a series of terminal moraines sometimes aggregated where the upturned layers of ice outcrop at the surface (Fig. 271). That the material of these moraines was originally basal is abundantly demonstrated by the bruised and scratched condition of the bowlders and pebbles, and sometimes by the nature of the material itself. For example, in two cases in North Greenland where glaciers descend into the heads of shallow bays and move forward on their bottoms, moraines formed by the upturning of the layers were seen to contain abundant molluscan shells derived from the bottom of the bay. The upturning sometimes affects the side-edges of ice-tongues (Fig. 272) as well as their ends, and the material thus brought to the surface gives origin to lateral moraines altogether different in origin from the lateral moraines formed by the falling of débris upon the glaciers. Sometimes also there is an upturning of the ice along a longitudinal zone well back from the lateral margins (Fig. 273), and the material so borne to the surface in such a zone gives rise to a moraine resembling the medial moraine formed by the union of lateral moraines, but of wholly different origin.

The phenomenon of upturning here referred to has been observed only at or near the terminus of the ice, and is perhaps due in most part to the resistance of frozen morainic or other material beneath and in front of the edge. To this should probably be added the effect of the increased rigidity of the ice at its borders, due to the low external temperature during the larger part of the year, while the interior, with its higher temperature, remains more fluent. But even this probably leaves the explanation inadequate. In not a few instances the upturning is associated with a notable thickening of the layers toward their edges (Fig. 274). This suggests that perhaps there is an exceptional growth of the granular crystals of the ice near the edge of the layers, owing to the penetration of the surface-waters which are much more abundant at the borders than elsewhere, and which in the arctic glaciers probably do not penetrate deeply before they reach a freezing temperature.

=Wear of drift in transit.=—Drift carried at the bottom of the ice is subject to notable wear. The materials in transportation abrade one another and are abraded by the bed over which they pass. Englacial drift is subject to less wear because it is commonly more scattered. Superglacial drift is worn little or none while it lies on the surface of the ice; but in so far as superglacial or englacial drift is derived from the basal load, it may show the same evidences of wear as the basal drift itself. Superglacial drift often reveals its history in this way.

Deposition of the Drift.

1. =Beneath the body of the ice.=—During the advance of a glacier, deposition may take place both beneath the body of the ice and beneath its end and edges. Deposition beneath the body of the ice is liable to take place wherever the topography favors lodgment, or wherever the ice is overloaded. The topography favoring deposition is much the same as that favoring erosion, but the two processes are not favored at the same point. Erosion is greatest on the “stoss” side of an obstruction (the side against which the ice advances), and deposition on the lee side. The ice is likely to be overloaded (1) just beyond a place where conditions have favored the gathering of a heavy load, and (2) where the ice is rapidly thinning. On the whole, however, the deposition of material beneath the main body of a glacier is much more than balanced by erosion in the same position.

2. =At ends and edges of glaciers.=—At and near the end of a glacier the conditions of deposition are somewhat different. Here deposition beneath the ice goes on faster than elsewhere, chiefly because of the more rapid melting and the more rapid thinning and weakening of the ice. If the end of the glacier be stationary in position, drift is being continually brought to it and left there, for though the end is stationary, the ice continues to move. If the glacier moves forward 500 feet per year, and if its end is melted at the same rate, all the débris in the 500 feet of ice which has been melted has been deposited, and all except that which has been washed away has been deposited at and beneath the end of the glacier. If the end of the glacier is retreating, the retreat means that the waste at the end exceeds the forward movement. If the ice advances 300 feet per year, and is melted back 500 feet in the same time, all the débris carried by the 500 feet which has been melted has been deposited, and largely in the narrow zone (200 feet) from which the ice has receded. Even in this case, therefore, there is a notable tendency to marginal accumulation. If the end of the glacier is advancing 500 feet per year while it is being melted but 300 feet, all the drift in the 300 feet melted has been deposited, and chiefly at or beneath the immediate margin of the ice. To the marginal and sub-marginal accumulations made in this way, the material carried on the ice is added whenever the ice is melted from beneath it. This addition is sometimes considerable and sometimes meagre. If the edge of the ice is without much fluctuation in position, the material dumped over its end may take the form of a narrow ridge or bowlder-wall (Geschiebe-wall). If a glacier pushes material in front of it, this, too, becomes a part of the general terminal aggregation of drift.

TYPES OF MORAINES.

=The terminal moraine.=—The thick accumulation of drift made at the end of a glacier or at the edge of an ice sheet, especially where its end or edge is stationary, or nearly stationary, for a considerable time, is the terminal moraine. That part of the aggregation deposited beneath the ice is sometimes called the lodge moraine (Figs. 275 and 276; see also Fig. 235); that carried on the ice and dropped at its edge, the dump moraine; and that pushed before the ice, the push moraine. Many moraines marginal to the ice appear to be push moraines, when they are really lodge moraines from which the ice has withdrawn (Fig. 277). The push moraine can rarely be distinguished, and the dump moraine by no means always, after the disappearance of the ice.

=The ground moraine.=—When a glacier disappears by melting, all its débris is deposited. All the drift deposited beneath the advancing ice and all deposited from the base of the ice during its dissolution constitutes the ground moraine. The thickness of the ground moraine is notably unequal. In general, it is thicker toward the terminus of the glacier and thinner toward its source, but considerable portions of a glacier’s bed are often left without débris when the ice melts. In general, the terminal moraine is not only thicker, but more irregularly disposed than the ground moraine.

=The lateral moraines.=—The surface lateral moraines of valley glaciers are let down on the surface beneath when the ice melts out from under them; but the lateral moraines in a valley from which the ice has melted are not merely the lateral moraines which were on the glacier at a given time. They are often far more massive than any which ever existed on the ice itself at any one time (Fig. 278). As a glacier retreats, its lateral moraine material is more or less bunched. Thus if the ice advances 200 feet while its end is being melted back 300 feet, the lateral moraines on the 300 feet melted are concentrated into 100 feet, as they are delivered on to the land by the melting of the ice from beneath. If the retreat of the end of a glacier be very slow, the bunching may be great. But even this cannot explain the massiveness of some lateral moraines. Furthermore, the materials of which many lateral moraines are composed are nearly as well worn as those of the ground moraine. The massive lateral moraines of which this is true are often made up chiefly of the drift accumulated beneath the lateral margins of the glaciers. This accumulation is the result of the lateral motion of the ice from center to side. Such sublateral accumulations are akin to terminal moraines. Some of the lateral moraines of ancient valley glaciers, such as those of the Uinta, Wasatch, and Bighorn mountains are several hundred feet high, and in one case about 1000 feet. In northern Italy lateral moraines are said to be 1500 to 2000 feet high.

Most of the material which was englacial during the transportation becomes either subglacial or superglacial before deposition, for it ordinarily reaches the bottom or the top of the ice before being deposited. Where the ends or edges of a glacier are vertical or nearly so, as in the high arctic regions, deposition may take place from the englacial position directly.

=Distinctive nature of glacial deposits.=—The deposits made by glaciers are distinctive. In the first place the ice does not assort its material, and bowlders, cobbles, pebbles, sand, and clay are confusedly commingled (Fig. 279). In this respect, the deposits of ice differ notably from those of water. Furthermore, many stones of the drift show the peculiar type of wear which glaciers inflict. They are not rounded as the stones carried by rivers, though they are notably worn. Many of them have subangular forms with planed and beveled faces, the planes being striated and bruised (Fig. 254). The absence of stratification, the physical heterogeneity, and the striation of at least a part of the stones are among the most distinctive characteristics of glacial drift. A not less real though less obvious characteristic is the constitution of the fine material, for it is in general not the product of rock decay, but of rock grinding. The fine material handled by streams (except glacial streams) on the other hand, is usually the product of rock decay.

=Glaciated rock surfaces.=—Another distinctive mark which a glacier leaves behind it is the character of the surface of the rock on which the drift rests. This is generally smoothed by the severe abrasion to which it has been subjected, and the smoothed surfaces are marked by grooves and striæ, similar to those on the stones of the drift (Figs. 255 and 256). Other distinctive features of a glaciated area are the rounded bosses of rock (roches moutonnées, Fig. 280; see also surface about the lakes, Fig. 261), the rock basins, the lakes (Fig. 261), ponds, and marshes, and the peculiar topographies resulting from the unequal erosion, and the still more unequal deposition of the drift. Surface bowlders, often unlike the underlying formations of rock, and sometimes in peculiar and apparently unstable positions, are still another mark of a glaciated area.

GLACIO-FLUVIAL WORK.

The constant but unequal waste of glaciers has already been referred to. The streams to which this gives rise are usually laden with gravel, sand and silt derived from the ice. Since the mud is often light-colored, the streams are sometimes described as “milky.” Where the amount of material carried is great, much of it is dropped at a slight distance from the ice, the coarsest being dropped first. Glacial streams are, as a rule, aggrading streams, and therefore develop alluvial plains, called valley trains (Fig. 281 and 282), or where they enter lakes (Fig. 283), bays, or other streams, deltas. In its transportation, the river-borne drift is assorted; after its deposition, it is stratified. True glacial deposits in the upper part of a mountain valley are, therefore, often continued below by glacio-fluvial deposits derived from the same source.

The most common form of such deposit is a valley train (Fig. 281) of glacial wash stretching indefinitely down the valley. The silt, sand, and gravel of such trains can usually be distinguished from valley deposits of non-glacial origin by the character of the material, as much of it is the product of grinding, crushing, and fracture, rather than of ordinary surface decay. Its materials are, therefore, fresh and often include rock material which, if long exposed at the surface, would be decomposed or dissolved.

Where an ice sheet ends in a broad face, as did the ancient continental glaciers, numerous streams flow from it and spread their débris in front of the terminal moraine, forming a broad fringing sheet or “apron” (outwash plain) along it. Where streams of considerable size form tunnels under or in the ice, these may become more or less filled with wash, and when the ice melts the aggraded channels appear as long ridges of gravel and sand known as eskers (osars and serpentine kames and kames of authors. See chapter on glacial period). It has been thought that similar ridges are sometimes formed in valleys cut in the ice from top to bottom, and even that they arise from gravel and sand lodged in superglacial channels. The latter at least is probably rare, as the surface streams usually have high gradients, swift currents, and smooth bottoms, and hence give little opportunity for lodgment. In the case of ice-sheets, too, in connection with which eskers are chiefly developed, there is usually no surface material except at the immediate edge, where the ice is thin and its layers upturned.

At the mouths of ice-tunnels or ice-channels, especially where they end against terminal moraines, sands and gravels are liable to be bunched in quantity, giving rise, after the adjacent ice has melted, to peculiar hills and hollows of the knob-and-basin type. The hills and short ridges are known as kames (see glacial period). Subglacial streams may leave washed and assorted material in their tracks under the ice, and this is sometimes buried under deposits made by the ice itself, so that glacio-fluvial and glacial deposits are interbedded.

ICEBERGS.

When glaciers advance into water, the depth of which approaches their thickness, their ends are broken off (Fig. 284), and the detached masses float away as icebergs (Fig. 285). Many of the bergs are overturned, or at least tilted, as they set sail. If this does not happen at the outset, it is likely to occur later as the result of the melting and wave-cutting which disturb their equilibrium. The great majority of bergs do not travel far before losing all trace of stony and earthy débris, but the finding of glacial material in dredgings far south of all glaciers shows that they occasionally carry stones far from land.

THE INTIMATE STRUCTURE AND THE MOVEMENT OF GLACIERS.

With the preceding account of glaciers in mind, we may return to a closer study of their origin, their intimate structure, and their mode of motion. The key to this study is the thesis that a glacier is a mass of crystalline rock—the purest and simplest type of crystalline rock known—since it is made up of a single mineral of simple composition and rare purity, which never appears in a solid state except in the crystalline form.

=The growth and constitution of a glacier.=—The origin and history of a glacier is little more than the origin and aggregate history of the crystals that compose it. The fundamental conception of a glacier is therefore best obtained by tracing the growth of its constituent crystals. A basal fact ever to be kept in mind is that water in the solid form is always controlled by crystalline forces. When it solidifies from the vapor of the atmosphere it takes the form of separate crystals (Figs. 286–291). Perfect forms are developed only when the flakes fall quietly through a saturated atmosphere which allows them to grow as they descend. Under other conditions, the crystals are imperfect in growth and are mutilated by impact. But however modified, they are always crystals. The molecules are arranged on the hexagonal plan, and, as the expansive power of freezing water shows, the arrangement is controlled by a strong force. Once the definite crystalline arrangement is established, the molecules can be displaced only by relatively great force.

Snow crystals often continue to grow so long as they are in the atmosphere; but if they pass through an under-saturated stratum of air or a stratum whose temperature is above 32° Fahr., they suffer from evaporation or melting. When they reach the ground, the processes of growth and decadence continue, and the crystals grow or diminish according to circumstances.

A glacier is a colossal aggregation of crystals grown from snowflakes to granules of much greater sizes. The microscopic study of new-fallen snow reveals the mode of change from flakes to granules. The slender points and angles of the former yield to melting and evaporation more than the more massive central portions, and this change probably illustrates a law of vital importance. It may often be seen that the water melted from the periphery of a flake gathers about its center, and if the temperature be right, it freezes there. This is a first step toward the pronounced granulation of snow which has lain for some time on the ground. If measured systematically from day to day, the larger granules taken from beneath the surface of this coarse-grained snow are found to be growing. In a series of experiments to determine the law of growth it was found that when the temperature of the atmosphere was above the melting-point the growth was appreciably more rapid than when the air was colder, but there was, on the average, an increase under all conditions of temperature. A portion of this average increase of the larger granules appears to come from the diminution and destruction of the smaller ones, for the total number of granules steadily diminishes. A portion of the growth doubtless comes from the moisture of the atmosphere which penetrates the snow and another portion from the moisture derived from surface melting; but beneath the surface of a large body of snow the growth of the large granules is probably chiefly at the expense of the small ones. To follow the process it should be noted that the free surface of every granule is constantly throwing off particles of water-vapor (evaporation); that the rate at which the particles are thrown off is dependent, among other things, on the curvature of the surface, being greater the sharper the curve; that the surfaces of the granules are at the same time liable to receive and retain molecules thrown from other granules, and that, other things being equal, the retention of particles also depends on the curvature of the surface, the less curved surface retaining more than the sharply curved one. Under these laws, it is obvious that the larger granules of smaller curvature will lose less and gain more, on the average, than the smaller granules of greater curvature. It follows that the larger granules will grow at the expense of the smaller. It is also to be noted that, other things being equal, small granules melt more readily than large ones, and that where the temperature is nicely adjusted between melting and freezing, the smaller may lose while the larger gain.

Another factor that enters into the process is that of pressure and tension. The granules are compressed at the points of contact and put under tension at points not in contact, and the pressure and tension are, on the average, likely to be relatively greatest for the smallest granules. Tension increases the tendency to evaporation and adds its effects to curvature, and the capillary spaces adjoining the points of contact probably favor condensation. Ice expands in crystallizing and pressure reduces the melting-point, while tension raises it. The effect of this is slight (p. 276), and it probably plays little part in glacial action, but it is to be correlated with the much more important fact that compression produces heat which may raise the temperature of the ice to the melting-point, while tension may reduce the temperature to or below freezing. There is therefore a tendency for the ice to melt at the points of contact and compression, and for the water so produced to refreeze at adjacent points where the surface is under tension. This process becomes effective beneath a considerable body of snow, and here the granules gradually lose the spheroidal form assumed in the early stages of granulation and become irregular polyhedrons interlocked into a more or less solid mass.

A third factor is also to be recognized, though its effectiveness is unknown. Under severe wind pressure, air penetrates porous bodies with appreciable facility. The “breathing” of soils and the curious phenomena of “blowing-wells” and “blowing-caves” teach us of the effective penetration and extrusion of the air under variations of barometric pressure. In the snow-fields, and in the more granular portions of glaciers near their heads, the porosity is doubtless sufficient to allow of the appreciable penetration of the atmosphere. During a part of the time, the probable effect is the condensation within the ice of moisture from the air, and during another part, evaporation from the ice. These alternating processes are attended by oscillations of temperature. While the balance between loss and gain of substance may be immaterial, the oscillating nature of the process and the fluctuations of temperature are probably favorable to granular change.

Whether these processes furnish an adequate explanation of the changes or not, the observed fact is that there are all gradations from snowflakes and pellets into granular névé, and thence into glacier granules (Gletscherkörner), varying in size up to that of filberts and walnuts, and even beyond. In coherence, these aggregations may vary from the early slightly coherent granular stage, where the grains are small and spheroidal, to the ice stage, where the cohesion has become strong through the interlocking growths of the large granules. Even when the mass has become seemingly solid ice, sufficient space is usually left between the granules to give the dispersive reflection to light which imparts to glacier ice its distinctive whitish color.

=The arrangement of the crystal axes.=—The most radical difference between glacier ice and ice formed directly from water is in the arrangement (orientation) of the crystals. In the ice formed on undisturbed water, the bases of the crystals are at the surface and their principal axes are vertical, as shown by Mügge. As they grow, the crystal prisms extend downwards. This gives a columnar or prismatic structure to the ice, well seen when it is “honeycombed” by partial melting. In the glacier, on the other hand, the crystals, starting from snowflakes, have their axes turned in various directions according to the accidents of their fall; and as the snow develops into ice, the principal axes of the crystals continue to lie in all directions. Hence glacier ice, unlike pond ice, cannot usually be split along definite planes, except where cleavage planes are subsequently developed by extraneous agencies.

Figures to illustrate the method of deformation of ice crystals.]

While the crystals of a glacier usually have their principal axes in various directions, there appears to be a tendency for them to approach parallelism in certain positions, especially in the basal parts of a glacier near its terminus. Observations on this point are not so full and critical as could be desired, but it is probable that the parallel orientation is partly general, and due to the vertical pressure of the ice, and partly special and local, and connected with the shearing planes and foliation.

The bearing of this partial parallelism of the crystals on shearing and foliation is supposed to reside in the fact that a crystal of ice is made up of a series of plates arranged at right angles to the principal axis of the crystal. These plates may be likened to a pile of cards, the principal axis being represented by a line vertical to them. If a cube be cut from a large crystal of ice, it will behave much like a cube cut from the pile of cards. If the cube be so placed that its plates are horizontal (Fig. 291a), and if it be rested on supports at two edges and heavily weighted in the middle, it will sag, the plates sliding slightly over one another so as to give oblique ends, but in this case the cube offers considerable resistance to deformation. If the cube be so placed that the plates stand on edge, each reaching from support to support (Fig. 291b), it will offer very great resistance to deformation; but if the plates be vertical and transverse to the line joining the supports, as in Fig. 291c, the middle portion will sag under very moderate weighting by the sliding of the plates on one another, and in a comparatively short time the middle portion may be pushed entirely out, dividing the cube. These properties have been demonstrated by McConnel and Mügge, and they appear to throw light on certain phases of the action of glaciers that are most pronounced in their basal parts, and are best illustrated in arctic glaciers.

The Probable Fundamental Element in Glacial Motion.

=Melting and freezing.=—It has already been shown (p. 279) that the initial or fundamental cause of glacial motion must be operative at the heads of glaciers where the temperature is lowest and the material most loosely granular. In this condition, there is reason to believe that motion takes place between the grains, rather than by their distortion through the displacement of their laminæ. The fact that the granular structure is not destroyed, as it would be by the indefinite sliding of the crystal plates over each other, sustains this view. The inference is that the gliding planes play a notable rôle in glacial movement only in the basal parts of the lower ends of glaciers, where the greatest thrusts are developed, and where the granules have become largest and most completely interlocked. At the heads of glaciers, where motion is initiated, there may be great downward pressure, but not vigorous thrusts from behind, and probably only moderate thrusts developed within the body itself. There seems therefore no escape from the conclusion that the primal cause of glacial motion is one which may operate even under the relatively low temperatures, the relatively dry conditions, and the relatively granular textures which affect the heads of glaciers. These considerations lead to the view that movement takes place by the minute individual movements of the grains upon one another. While they are in the spheroidal form, as in the névé, this would not seem to be at all difficult. They may rotate and slide over each other as the weight of the snow increases; but as they become interlocked by growth, both rotation and sliding must apparently encounter more resistance. The amount of rotary motion required of an individual granule is, however, surprisingly small, and the meltings and refreezings incident to shifting pressures and tensions, and to the growth of the granules, seem adequate to meet the requirements. In order to account for a movement of three feet per day in a glacier six miles long, the mean motion of the average granule relative to its neighbor would be, roundly, ¹⁄₁₀₀₀₀ of its own diameter per day, or one diameter in 10,000 days; in other words, it would change its relations to its neighbors to the extent of its diameter in about thirty years. A change of so great slowness under the conditions of granular alteration can scarcely be thought incredible, or even improbable, in spite of the interlocking which the granules may develop. The movement is supposed to be permitted chiefly by the temporary passage of minute portions of the granules into the fluid form at the points of greatest compression, the transfer of the moisture to adjoining points, and its resolidification. The points of greatest compression are obviously just those whose yielding most promotes motion, and a successive yielding of the points that come in succession to oppose motion most (and thus to receive the greatest stresses) permits continuous motion. It is merely necessary to assume that the gravity of the accumulated mass is sufficient to produce the minute temporary liquefaction at the points of greatest stress, the result being accomplished not so much by the lowering of the melting-point as by the development of heat by pressure.

This conception of glacial “flowage” involves only the momentary liquefaction of minute portions of the mass, while the ice as a whole remains rigid, as its crystalline nature requires. Instead of assigning a slow viscous fluidity like that of asphalt to the whole mass, which seems inconsistent with its crystalline character, it assigns a free fluidity to a succession of particles that form only a minute fraction of the whole at any instant.

This conception is consistent with the retention of the granular condition of the ice, with the heterogeneous (in the main) orientation of the crystals, with the rigidity and brittleness of the ice, and with its strictly crystalline character, a character which a viscous liquid does not possess however much its high viscosity may make it resemble a rigid body.

=Accumulated motion in the terminal part of a glacier.=—However slight the relative motion of one granule on its neighbor, the granules in any part of a glacier partake in the accumulated motion of all parts nearer the source, and hence all are thrust forward. Herein appears to lie the distinctive nature of glacial movement. Each part of a stream of water feels the hydrostatic pressure of neighboring parts (theoretically equal in all directions) and the momentum of motion, but not the rigid thrust of the mass behind. Lava streams are good types of viscous fluids flowing in masses comparable to those of glaciers and on similar slopes, and, in their last stages, at similar rates, but their special modes of flow and their effects on the sides and bottoms of their paths are radically different from those of glaciers. Forceful abrasion, and particularly the rigid holding of imbedded stones while they score and groove the rock beneath, is unknown in lava streams and is scarcely conceivable. There is, so far as we know, no experimental or natural evidence that any typical viscous body in flowing over a rugose bottom detaches and picks up fragments and holds them as graving tools in its base so fixedly as to cut deep, long, straight grooves in the hard bottom over which it flows. It would seem that competency to do this peculiar class of work, which is distinctive of glaciers, should be demonstrated before the viscous theory of glacial movement is accepted as even a good working hypothesis. Somewhat in contrast with viscous movement, it is conceived that a glacier is thrust forward rigidly by internal elongation, shears forcibly over its sides and bottoms, and leaves its distinctive marks upon them.

Auxiliary Elements.

=Shearing.=—In the lower portion of a glacier where normally the thrusts are greatest, the granules fewest, and their interlocking most intimate, shearing takes place within the ice itself. This is illustrated by the accompanying Figs., 292–295. The shearing results in the foliation of the ice and in the forcing of débris between the sheared layers. Thus the ice becomes loaded in a special englacial or baso-englacial fashion, as previously mentioned and illustrated in Fig. 268.

Within the zone of shearing, it is probable that the gliding planes of the crystals come into effective function. It is thought that the combined effect of the vertical pressure, the forward thrust, and the basal drag of the ice, may be to increase the number of granules whose gliding planes are parallel to the glacier’s bottom. At any rate, Drygalski reports that there is a tendency to such an arrangement in the basal portion of the Greenland glaciers at their borders. It is conceived that where strong thrusts are brought to bear upon such a mass of granules, those whose gliding planes are parallel to the direction of thrust are strained with sufficient intensity to cause the plates to slide over each other, while those which are not parallel to the direction of thrust are either rotated into parallelism—when they also yield—or are pressed aside out of the plane of shear. As previously noted, shearing is observed to occur chiefly where the ice below the plane of shearing is protected more or less from the force of the thrust. It perhaps also occurs where the basal ice becomes so overloaded with débris that it is incapable of ready movement.

It is also probable that sharp differential strain and shearing are developed at the level where the surface-water of the warm season, descending into the ice, reaches the zone of freezing. The expanding of the freezing water at the upper limit of the cold zone may cause the layer expanded by it to shear over that below. As the level of freezing is lowered with the advance of the warm season, the zone of shearing also sinks. This may be regarded as an auxiliary agency of shearing, of application to a special horizon.

=High temperature and water.=—In the zone of waste, a higher temperature and more water lend their aid to the fundamental agencies of movement, and there is need for these aids to promote a proportionate movement, for here the granules are more intimately interlocked and the ice more compact and inherently more solid and rigid. The average temperature is, however, near the melting-point (p. 276), and during the warm season the ice is bathed in water so that the necessary changes in the crystals are facilitated, and movement apparently takes place even more readily than in the more open granular portion of lower temperature and dryer state. The extraordinary movements of certain tongues of ice in some of the great fiords of Greenland are probably due to the convergence of very thick slow-moving ice from the interior into basins leading down to the fiords. Into the same basins a large amount of surface-water is concentrated at the same time, with the result that the thick ice, bathed with water and having a high gradient, develops unusual velocity during the warm season.

=Applications.=—By a studious consideration of the coöperation of the auxiliary agencies with the fundamental ones, the peculiarities of glacial movement may apparently be explained. In regions of intense cold, where a dry state and low temperature prevail, as in the heart of Greenland, the snow-ice mass may accumulate to extraordinary thicknesses, for the burden of movement seems to be thrown almost wholly upon compression, with the slight aid of molecular changes due to internal evaporation and allied inefficient processes. Since the temperature in the upper part of the ice is very adverse (see p. 277), the compression must be great before it becomes effective in melting the ice, and hence the great thickness of the mass antecedent to much motion. Similar conditions more or less affect the heads of alpine glaciers, though here the high gradients favor motion with lesser thicknesses of ice; but in the lower reaches of alpine glaciers, where the temperatures are near the melting-point, and the ice is bathed in water, movement may take place in ice which is thin and compact.

If the views here presented are correct, there is also, near the end or edge of a glacier, the coöperation of rigid thrust from behind with the tendency of the mass to move on its own account. The latter is controlled by gravity, and conforms in its results to laws of liquid flow. The former is a derived factor, and is a mechanical thrust. This thrust is different from the pressure of the upper part of a liquid stream on the lower part, because it is transmitted through a body whose rigidity is effective, while the latter is transmitted on the hydrostatic principle of equal pressure in all directions.

Corroborative Phenomena.

The conception of the glacier and its movement here presented explains some of the anomalies that otherwise seem paradoxical. While a glacier in a sense flows over a surface, it often cuts long, deep furrows in firm rock. It is difficult to explain this if the ice be so yielding as to flow under its own weight on a surface which is almost flat. If the mass is really viscous, its hold on its imbedded débris should also be viscous, and a bowlder in the bottom should be rotated in the yielding mass when its lower point catches on the rock beneath, instead of being firmly held while a deep groove is cut. This is more to the point since viscous fluids flow by a partially rotary movement. If, on the other hand, the ice is always a rigid body which yields only as its interlocking granules change their form by loss and gain, a rigid hold on the imbedded rock at some times, and a yielding hold at others, is intelligible, for on this view the nature of its hold is dependent on the temperature and dryness of the ice. Stones in the base of a glacier may be held with very great rigidity when the ice is dry, scoring the bottom with much force, while they may be rotated with relative ease when the ice is wet. In short, the relation of the ice to the bowlders in its bottom varies radically according to its dryness and temperature. A dry glacier is a rigid glacier. A dry glacier is necessarily cold, and a cold glacier is necessarily dry.

On the view here presented, a glacier should be more rigid in winter than in summer, and the whole thickness of the glacier should experience this rigidity chiefly at the ends and edges, where the relative thinness of the ice permits the low temperature to reach its bottom. The motion in these parts during the winter is, therefore, very small.

In this view may also be found an explanation of the movement of glaciers for considerable distances on upward slopes, even when the surface as well as the base is inclined backwards. So far does this go that superglacial streams sometimes run for some distance backwards, i.e. toward the heads of the glaciers, while in other places surface-waters are collected into ponds and lakelets. Such a slope of the surface of ice is not difficult to understand if the movement be due to thrust from behind, or if it be occasioned by internal crystalline changes acting upon a rigid body; but it must be regarded as very remarkable if the movement be that of a fluid body, no matter how viscous, for the length of the acclivity is sometimes several times the thickness of the ice. Crevassing and other evidences of brittleness and rigidity find a ready elucidation under the view that the ice is a really solid body at all times, and that its apparent fluency is due to the momentary fluidity of small portions of the mass assumed in succession as compression demands.

In addition to the considerations already adduced, it may be urged that a glacier does not flow as a stiff liquid because its granules are not habitually drawn out into elongated forms, as are cavities in lavas and plastic lumps in viscous bodies. Flowage lines comparable to those in lavas are unknown in glaciers.

All this is strictly consistent with our primary thesis, that a glacier is a crystalline rock of the purest and simplest type, and that it never has other than the crystalline state. This strictly crystalline character is incompatible with viscous liquidity.

Other Views of Glacier Motion.

While these views of glacial motion seem to us to best accord with the known facts, they are not to be regarded as established in scientific opinion, or as the views most commonly held. The mode of glacial motion has long been a mooted question, and is still so regarded. The main alternative interpretations that have been entertained are the following:

(1) In the early days of glacial studies De Saussure thought that glaciers slid bodily on their beds;

(2) Charpentier and Agassiz referred the movement to the expansion of descending water freezing within the glacier;

(3) Rendu and Forbes, followed by many, perhaps most, modern writers, believed ice to be viscous, and that in sufficiently large masses it flows under the influence of its own weight, like pitch or asphalt;

(4) Others, realizing the fundamental difference between crystalline ice and a true viscous body, have fallen back on a vague notion of plasticity which scarcely amounts to a definite hypothesis at all;

(5) Tyndall urged that the movement was accomplished by minute repeated fracturing and regelation, appealing to the fact that broken pieces of ice slightly pressed together at melting temperatures freeze together, but neglecting the fact that this would destroy the integrity of the crystals;

(6) Moseley assigned the movement to a bodily expansion and contraction of the glacier, analogous to the creeping of a mass of lead on a roof;

(7) James Thompson demonstrated that pressure lowers the melting-point, and while this effect is so small as probably to be ineffectual, it is correlated with the very important fact that compression may cause melting, which is not the case in most other rocks. He recognized that under pressure partial liquefaction took place, that the water so liberated might be refrozen as it escaped from pressure, and appears to have regarded this as a vital factor;

(8) Croll held that the movement was due to a consecutive series of molecular changes somewhat like the chain of chemical combinations in electrolysis;

(9) Hugi, Eli de Beaumont, Bertin, Forel, and others thought that the growth of the granules was the leading factor in the ice movement;

(10) McConnel and Mügge have made the gliding planes of the ice crystals serve an important function in glacial movement.

It will be seen that the principle of partial liquefaction for which Thompson laid the basis, the crystallization of descending water, urged by Charpentier and Agassiz, and the granular growth on which Hugi, Beaumont, Forel, and others founded their hypotheses, are incorporated in the view already presented. Probably the agencies on which some of the other views are based may also be participants in producing glacial motion, sometimes as incidental factors, and sometimes perhaps as important ones, for under different conditions, different agencies may play rôles of varying importance. For example, in going over the brinks of precipices of sufficient height, glaciers break into fragments which are re-cemented below, and the “reconstructed” glacier moves on as before. Here fracture and regelation are evident. The movement of the gliding planes of the ice crystals over each other, which has been looked upon as a special kind of viscoid movement, probably plays a large part in the shearing movements in certain cases. But neither of these is probably a large factor in ordinary glacial movement, and it seems highly improbable that any of them are essential factors in the primary movements in the snow-fields where glacial action begins.