THE CONTINENTAL GLACIERS OF POLAR REGIONS

The inland ice of Greenland.—In Greenland and in Antarctica the land is almost or quite buried under a cover of snow and ice—the so-called “inland ice”—which always assumes the surface of a very flat dome or shield. In Greenland there is found a marginal ribbon of land generally from five to twenty-five miles in width (Fig. 299), but in Antarctica all the land, with the exception of a few mountain peaks, is inwrapped in a mantle of ice which is also extended upon the sea in a broad shelf of snow and ice. Neither of these vast glaciers has been explored except in its marginal portion, yet such is the symmetry of the profiles along the routes traversed, and such the flatness and monotony of the snow surface within the margins, that there is little reason to doubt that the profile made along Nansen’s route in southern Greenland would, save only for magnitude, fairly represent a section across the middle of the continent (Fig. 300).

The mountain rampart and its portals.—As soon as we examine the coastal belt we observe that the “Great Ice” of Greenland is held in by a wall of mountains and so prevented from spreading out to its natural surface in the marginal portions. Through portals of the inclosing mountain ranges—the outlets—it sends out tongues of ice which in many respects resemble certain types of mountain glaciers.

Such measurements as have been made upon the inland ice of Greenland at points back from, but yet comparatively near to, the outlets, show that it has here a surface rate of motion amounting to less than an inch per day, and it is highly probable that at moderate distances from the margin this amount diminishes to zero. Upon the outlets, on the contrary, surface rates as high as 59 feet per day have been measured, and even 100 feet per day has been reported. We are thus justified in saying that glacier flow within the outlets is from 700 to 1000 times as great as it is upon the near-by inland ice, and that the glacier is in a measure drained through the portals of the inclosing ranges. Back from these outlet streams of ice, or tongues, the surface of the inland ice is depressed to form a dimple or “basin of exudation” as is the surface of a reservoir above the raceway when the water is being rapidly drawn away (Fig. 301).

Fissures in the ice, the so-called crevasses, are the recognized marks of ice movement, and these are always concentrated at the steep slopes of the ice surface in the neighborhood of its margins. Upon the Greenland ice, crevasses are restricted in their distribution to a zone which extends from seven to twenty-five miles within the ice border.

The marginal rock islands.—From its margin the ice surface rises so steeply as to be climbed only with difficulty, but this gradient steadily diminishes until at a distance of between seventy-five and a hundred miles its slope is less than two degrees. Where crossed by Nansen near latitude 64° N. the broad central area of ice was so nearly level as to appear to be a plain.

As we pass across the irregular ice margin in the direction of the interior, larger and larger proportions of the land’s surface are submerged, until only projecting peaks rise above the ice as islands which are described as nunataks (Fig. 302).

Though not a universal observation, it has been often noted that the absorption of the sun’s rays by rock masses projecting through the snow results in a radiation of the heat and a lowering by melting of the surrounding snow and ice. For this reason nunataks are often surrounded by a deep trench due to a melting of the snow. Such a depression in the ice surface about the margin of a nunatak, from its resemblance to a trench about an ancient castle, has been designated a moat (Fig. 303). For the same reason, the outlet tongues of ice which descend in deep fjords between walls of rock are melted away from the walls and a lateral stream of water is sometimes found to flow between ice and rock (pl. 13 B).

Rock fragments which travel with the ice.—Rock surfaces which are exposed to the atmosphere are in high latitudes broken down through the freezing of water within their crevices. The fragments resulting from this rending process fall upon the glacier surface and are carried forward as passengers in the direction of the ice margin. They are either visible as long and narrow ridges or trains following the directions of the steepest slope (Fig. 302), or they become buried under fresh falls of snow and only again become visible where summer melting has lowered the glacier surface in the vicinity of its margin. These longitudinal trains of rock fragments upon the glacier surface always have their starting point at the lower margin of one of the nunataks, and are known as medial moraines (Fig. 301, p. 273, and Fig. 302). Inside the zone of nunataks the glacier surface is, however, clear of rock dÉbris except where dust has been blown on by the wind, and this extends for a few miles only. The material of the medial moraines is a collection of angular blocks whose surfaces are the result of frost rending, for in their travel above the ice they are subjected to no abrading processes.

A contrasted type of surface moraines upon the Greenland glacier, instead of being parallel to the direction of ice movement, is directed transversely or parallel to the margins. The materials of these moraines are more rounded fragments of rock which have come up from the bottom layers, and we shall again refer to the origin of such moraines after the subglacial conditions have been considered.

The grinding mill beneath the ice.—If, now, we examine the front of a glacier tongue which goes out from the inland ice, we find that while the upper portion is white and mainly free from rock dÉbris (plate 13 A), the lower zone is of a dark color and crowded with layers of pebbles and bowlders which have been planed, polished, and scratched in a quite remarkable manner. The ice front is itself subject to forward and retrograde migrations of short period, but it is easily seen that in the main its larger movement has been a retrograde one. The ground from which it has lately withdrawn is generally a hard rock floor unweathered, but smooth, polished, and scratched in the same manner as the bowlders which are imbedded within the ice. It is perfectly apparent that the latter have been derived from some portion of the rock basement upon which the glacier still rests, and that floor and bowlders have alike been ground smooth by mutual contact under pressure.

This erosion beneath the ice is accomplished by two processes; namely, plucking and abrasion. Wherever the rock over which the glacier moves has stood up in projecting masses and is riven by fissure planes of any kind, the ice has found it easy to remove it in larger or smaller fragments by a quarrying process described as plucking. The rock may be said to be torn away in blocks which are largely bounded by the preËxisting fissure planes. Over relatively even surfaces plucking has little importance, but where there are noteworthy inequalities of surface upon the glacier bed, those sides which are away from the oncoming ice (lee side) are degraded by plucking in such a manner as sometimes to leave steep and ragged fracture surfaces. The tools of the ice thus acquired in the process of plucking are quickly frozen into the lowest ice layers, and being now dragged along the floor they abrade in the same manner as does a rasp or file. These tools of the ice are themselves worn away in the process and are thus given their characteristic shapes. Just as the lapidary grinds the surface of a jewel into facets by imbedding the gem in a matrix, first in one and then in another position, each time wearing down the projecting irregularities through contact with the abrading surface; so in like manner the rock fragment is held fast at the bottom of the glacier until “soled” or “shod”, first upon one side and then upon another. Accidental contact with some obstruction upon the floor may suffice to turn the fragment and so expose a new surface to wear upon the abrading floor. Minor obstructions coming in contact with one side of the fragment only, may turn it in its own plane without overturning. Evidence of such interruptions can be later read in the different directions of striÆ upon the same facet (plate 17 A).

The floor beneath the glacier is reduced by the abrading process to a more or less smooth and generally flattened or rounded surface—the so-called glacier pavement (Fig. 304). To accomplish this all former mantle rock due to weathering processes must first be cleared away, and the firm unaltered rock beneath is wherever susceptible of it given a smooth polish although locally scored and scratched by the grinding bowlders. The earlier projections of the surface of the floor, if not entirely planed away, are at least transformed into rounded shoulders of rock, which from their resemblance to closely crowded backs in a flock of sheep have been called “sheep backs” or “roches moutonnÉes.” Thus the effect of the combined action of the processes of plucking and abrasion is to reduce the accent of the relief and to mold the contours of the rock in smoothly flowing curves, generally of large radius.

The lifting of the grinding tools and their incorporation within the ice.—Wherever the ice is locally held in check by the projecting nunataks, relief is found between such obstructions, and there the flow of the ice has a correspondingly increased velocity (Fig. 305 b). If the obstructions are not of large dimensions, the ice which flows around the outer edges is soon joined to that which passes between the obstructions and so normal conditions of flow are restored below the nunataks. The locally rapid flow of the ice is, therefore, restricted to a relatively short distance, the passageway between the nunataks, and the conditions are thus to be likened to the fall of water at a raceway due to the sudden descent of its surface from the level of the reservoir to the level of the stream in the outlet. As is well known, there is under these conditions a prodigious scour upon the bottom which tends to dig a pit just above and below the dam—a scape colk—and carry the materials up to the surface below the pit. Such a tendency was well illustrated by the behavior of the water at the opening of the Neu Haufen dam below the city of Vienna (Fig. 305 a). In the case of ice, material from the bottom may by the upward current be brought up to the surface of the glacier at the lower edge of the colk and thus produce a type of local surface moraine of horseshoe form with its direction generally transverse to the direction of ice movement (Fig. 305 b).

Any obstruction upon the pavement of the glacier apparently exerts a larger or smaller tendency to elevate the bowlders and pebbles and incorporate them within the ice. Rock dÉbris thus incorporated is described as englacial drift. In the case of Greenland glaciers this material seems at the ice front to be largely restricted to the lower 100 feet (plate 13 A).

Near the front of the inland ice the increased slope of the upper surface greatly increases the flow of the upper ice layers in comparison with those nearer the bottom, so that the upper layers override the lower as they would an obstruction. The englacial drift is either for this reason or because of rock obstructions brought to the surface, where it yields parallel ridges corresponding in direction to the glacier margin. Such transverse surface moraines are thus in many respects analogous to those which appear about the lower margins of scape colks. In contrast to the longitudinal or medial surface moraines the materials of the transverse moraines are more faceted and rounded—they have been abraded upon the glacier pavement.

Melting upon the glacier margins in Greenland.—During the short but warm summer season, the margins of the Greenland ice are subject to considerable losses through surface melting. When the uppermost ice layer has attained a temperature of 32° Fahrenheit, melting begins and moves rapidly inward from the glacier margin. In late spring the surface of the outer marginal zone is saturated with water, and this zone of slush advances inward with the season, but apparently never transgresses the inner border of what we have generally referred to as the marginal zone of the ice characterized by relatively steep slopes, crevasses, and nunataks. Upon the ice within this zone are found streams large enough to be designated as rivers and these are connected with pools, lakes, and morasses. The dirt and rock fragments imbedded in the ice are melted out in the lowering of the surface, so that late in the season the ice presents a most dirty aspect. At the front of the great mountain glaciers of Alaska, a more vigorous operation of the same process has yielded a surface soil in which grow such rank forests as entirely to mask the presence of the ice beneath.

In addition to the visible streams upon the surface of the Greenland ice, there are others which flow beneath and can be heard by putting the ear to the surface. All surface streams eventually encounter the marginal crevasses and plunge down in foaming cascades, producing the well known “glacier wells” or “glacier mills.” The progress of the water is now throughout in tunnels within the ice until it again makes its appearance at the glacier margin.

The marginal moraines.—Study of both the Greenland and Antarctic glaciers has shown that if we disregard the smaller and short-period migrations of the ice front, the general later movement has been a retrograde one—we live in a receding hemicycle of glaciation. The earlier Greenland glacier has now receded so as to expose large areas of the former glacier pavement. In places this pavement is largely bare, indicating a relatively rapid retirement of the ice front, but at all points at which the ice margin was halted there is now found a ridge of unassorted rock materials which were dropped by the ice as it melted (Fig. 306). Such ridges, composed of the unassorted materials described as till, come to have a festooned arrangement largely concentric to the ice margin, and are the marginal or terminal moraines (see Fig. 336, p. 312). Marginal moraines, if of large dimensions, usually have a hummocky surface, and are apt to be composed of rock fragments of a wide range of size from rock flour (clay) to large bowlders (plate 17 A), which may represent many types since they have been plucked by the glacier or gathered in at its surface from many widely separated localities.

As the glacier front retires from the moraine which it has built up, the water which emerges from beneath the ice is impounded behind the new dam so as to form a lake of crescentic outline (Fig. 307). Such lakes are particularly short-lived, for the reason that the water finds an outlet over the lowest point in the crest of the moraine and easily cuts a gorge through the loose materials, thus draining the lake (Fig. 308). Thereafter, the escaping water flows in a braided stream across the late lake bottom and thence at the bottom of the gorge through the moraine.

The outwash plain or apron.—The water which descends from the glacier surface in the glacier wells or mills, eventually arrives at the bottom, where it follows a sinuous course within a tunnel melted out in the ice. Much of this water may issue at the ice front beneath the coarse rock materials which are found there, and so be discovered with the ear rather than by the eye. The water within the tunnels not flowing with a free surface but being confined as though it were in a pipe, may, however, reach the glacier margin under a hydrostatic pressure sufficient to carry it up rising grades. Inasmuch as it is heavily charged with rock dÉbris and is suddenly checked upon arriving at the front it deposits its burden about the ice margin so as to build up plains of assorted sands and gravels, and over this surface it flows in ever shifting serpentine channels of braided type (Fig. 308). Such plains of glacier outwash are described as outwash plains or outwash aprons.

Rising as it does under hydrostatic pressure the water issuing at the glacier front may find its way upward in some of the crevasses and so emerge at a level considerably above the glacial floor. It may thus come about that the outwash plain is built up about the nose of the glacier so as partially to bury it from sight. When now the ice front begins a rapid retirement, a depression or fosse (Fig. 309 and Fig. 339, p. 314) is left behind the outwash plain and in front of the moraine which is built up at the next halting place.

The continental glacier of Antarctica.—In Victoria Land, upon the continent of Antarctica, so far as exploration has yet gone, the continental glacier is held back by a rampart of mountains, as has been shown to be true of the inland ice of Greenland. The same flat dome or shield has likewise been found to characterize its upper surface (Fig. 310).

The most noteworthy differences between the inland ice masses of Greenland and Antarctica are to be ascribed to the greater severity of the Antarctic climate and to the more ample nourishment of the southern glacier measured by the land area which it has submerged. There is here no marginal land ribbon as in Greenland, but the glacier covers all the land and is, moreover, extended upon the sea as a broad floating terrace—the shelf ice (Fig. 311). This barrier at its margin puts a bar to all further navigation, rising as it does in some cases 280 feet above the sea and descending to even greater depths below (plate 15 B).

In that portion of Antarctica which was explored by the German expedition, the inland ice is not as in Victoria Land restrained within walls of rock, but is spread out upon the continent so as to assume its natural ice slopes, which are therefore much flatter than those examined in Greenland and Victoria Land. Here in Kaiser Wilhelm Land the ice rises at its sea margin in a cliff which is from 130 to 165 feet in height, then upon a fairly steeply curving slope to an elevation of perhaps a thousand feet. Here the grades have become relatively level, and on ever flatter slopes the surface appears to continue into the distant interior (plate 14). Near the ice margin numerous fissures betray a motion within the mass which exact measurements indicate to be but one foot per day, and at a distance of a mile and a quarter from the margin even this slight value has diminished by fully one eighth. It can hardly be doubted that at moderate distances only within the ice margin, the glacier is practically without motion.

Rain or general melting conditions being unknown in Antarctica, a striking contrast is offered to the marginal zone of the Greenland continent. This is to a large extent explained by the existence upon the northern land mass of a coastland ribbon which becomes quickly heated in the sun’s rays, and both by warming the air and by radiating heat to the ice it causes melting and produces local air temperatures which in summer may even be described as hot. About Independence Bay in latitude 82° N. and near the northernmost extremity of Greenland, Peary descended from the inland ice into a little valley within which musk oxen were lazily grazing and where bees buzzed from blossom to blossom over a gorgeous carpet of flowers.

Nourishment of continental glaciers.—Explorations upon and about the glaciers of Greenland and Antarctica have shown that the circulation of air above these vast ice shields conforms to a quite simple and symmetrical model subject to spasmodic pulsations of a very pronounced type. Each great ice mass with its atmospheric cover constitutes a sort of refrigerating air engine and plays an important part in the wind system of the globe. (See Fig. 291, p. 263). Both the domed surface and the low temperature of the glacier are essential to the continuation of this pulsating movement within the atmosphere (Fig. 312). The air layer in contact with the ice is during a period of calm cooled, contracted, and rendered heavier, so that it begins to slide downward and outward upon the domed surface in all directions. The extreme flatness of the greater portion of the glacier surface—a fraction only of one degree—makes the engine extremely slow in starting, but like all bodies which slide upon inclined planes, the velocity of its movement is rapidly accelerated, until a blizzard is developed whose vigor is unsurpassed by any elsewhere experienced.

The effect of such centrifugal air currents above the glacier is to suck down the air of the upper currents in order to supply the void which soon tends to develop over the central portion of the glacier dome. This downward vortex, fed as it is by inward-blowing, high-level currents, and drained by outwardly directed surface currents, is what is known as an anticyclone, here fixed in position by the central embossment of the dome.

The air which descends in the central column is warmed by compression, or adiabatically, just as air is warmed which is forced into a rubber tire by the use of a pump. The moisture congealed in the cirrus clouds floating in the uppermost layer of the convective zone, is carried down in this vortex and first melted and in turn evaporated, due to the adiabatic effect. This fusion and evaporation of the ice by its transformation of latent, to sensible, heat, in a measure counteracts, and so retards, the adiabatic elevation of temperature within the column. Eventually the warm air now charged with water vapor reaches the ice surface, is at once chilled, and its burden of moisture precipitated in the form of fine snow needles, the so-called “frost snow”, which in accompaniment to the sudden elevation of temperature is precipitated at the termination of a blizzard.

The warming of the air has, however, had the effect of damping as it were, the engine stroke, and, as the process is continued, to start a reverse or upward current within the chimney of the anticyclone. The blizzard is thus suddenly ended in a precipitation of the snow, which by changing the latent heat of condensation to sensible heat tends to increase this counter current.

The glacier broom.—During the calm which succeeds to the blizzard, heat is once more abstracted from the surface air layer, and a new outwardly directed engine stroke is begun. The tempest which later develops acts as a gigantic centrifugal broom which sweeps out to the margins of the glacier all portions of the latest snowfall which have not become firmly attached to the ice surface. The sweepings piled up about the margin of continental glaciers have been described as fringing glaciers, or the glacial fringe. The northern coast of Greenland and Grant Land are bordered by a fringe of this nature (plate 14, and Fig. 315, p. 288). It is by the operation of the glacier broom that the inland ice is given its characteristic shield-like shape (Fig. 312). The granular nature of the snow carried by the wind is well brought out by the little snow deltas about the margins of Greenland ice tongues (Fig. 313). Obviously because of the presence of the vigorous anticyclone, no snows such as nourish mountain glaciers can be precipitated upon continental glaciers except within a narrow marginal zone, and, as shown by Nansen rock dust from the coastland ribbon and from the nunataks of Greenland, is carried by a few miles inside the western margin, and not at all within the eastern.

Field and pack ice.—Within polar regions the surface of the sea freezes during the long winter season, the product being known as sea-ice or field-ice (Fig. 314). This ice cover may reach a thickness by direct freezing of eight or more feet, and by breaking up and being crowded above and below neighboring fragments may increase to a considerably greater thickness. Ice thus crowded together and more or less crushed is described as pack ice or the pack.

The pack does not remain stationary but is continually drifting with the wind and tide, first in one direction and then in another, but with a general drift in the direction of the prevailing winds. Because of the vast dimensions of the pack, the winds over widely separated parts may be contrary in direction, and hence when currents blow toward each other or when the ice is forced against a land area, it is locally crushed under mighty pressures and forced up into lines of hummocks—the so-called pressure ridges. At other times, when the winds of widely separated areas blow away from each other, the pack is parted, with the formation of lanes or leads of open water.

If seen in bird’s-eye view the lines of hummocks would according to Nansen be arranged like the meshes of a net having roughly squared angles and reaching to heights of 15 to 25, rarely 30, feet above the general surface of the pack. The ice within each mesh of the network is a floe, which at the times of pressure is ground against its neighbors and variously shifted in position. At the margin of the pack these floes become separated and float toward lower latitudes until they are melted.

The drift of the pack.—The discovery of the drift in the Arctic pack is a romantic chapter in the history of polar exploration, and has furnished an example of faith in scientific reasoning and judgment which may well be compared with that of Columbus. The great figure in this later discovery is the Norwegian explorer Fridtjof Nansen, and to the final achievement the ill-fated Jeannette expedition contributed an important part.

The Jeannette carrying the American exploring expedition was in 1879 caught in the pack to the northward of Wrangel Island (Fig. 315), and two years later was crushed by the ice and sunk to the northward of the New Siberian Islands. In 1884 various articles, including a list of stores in the handwriting of the commander of the Jeannette, were picked up at Julianehaab near the southern extremity of Greenland but upon the western side of Cape Farewell. Nansen, having carefully verified the facts, concluded that the recovered articles could have found their way to Julianehaab only by drifting in the pack across the polar sea, and that at the longest only five years had been consumed in the transit. After being separated from the pack the articles must have floated in the current which makes southward along the east coast of Greenland and after doubling Cape Farewell flows northward upon the west coast. It was clear that if they had come through Smith Sound they would inevitably have been found upon the other shore of Baffin Bay. In confirmation of this view there was found at Godthaab, a short distance to the northward of Julianehaab (Fig. 315), an ornamented Alaskan “throwing stick” which probably came by the same route. Moreover, large quantities of driftwood reach the shores of Greenland which have clearly come from the Siberian coast, since the Siberian larch has furnished the larger quantity.

Pinning his faith to these indubitable facts, Nansen built the Fram in such a manner as to resist and elude the enormous pressures of the ice pack, stocked her with provisions sufficient for five years, and by allowing the vessel to be frozen into the pack north of the New Siberian Islands, he consigned himself and his companions to the mercy of the elements. The world knows the result as one of the most remarkable achievements in the long history of polar exploration. The track of the Fram, charted in Fig. 315, considered in connection with that of the Jeannette, shows that the Arctic pack drifts from Bering Sea westward until near the northeastern coast of Greenland.

Special casks were for experimental purposes fastened in the ice to the north of Behring Strait by Melville and Bryant, and two of these were afterwards recovered, the one near the North Cape in northern Norway, and the other in northeastern Iceland (see map, Fig. 315). Peary’s trips northward in 1906 and 1909 from the vicinity of Smith Sound have indicated that between the Pole and the shores of Greenland and Grant Land the drift is throughout to the eastward, corresponding to the westerly wind. Upon this border the great area of Arctic drift ice is in contact with great continental glaciers bordered by a glacier fringe. Admiral Peary has shown that instead of consisting of frozen sea ice, the pack is here made up of great floes from 20 to 100 feet in thickness and that these have been derived from the glacier fringe.

Whenever the blizzards blow off the inland ice from the south, leads are opened at the margin of the fringe and may carry strips from the latter northward across the lead. With favorable conditions these leads may be closed by thick sea ice so that with the occurrence of counter winds from the north they do not entirely return to their original position. A continuance of this process may have resulted in the heavy floe ice to the northward of Greenland, which, acting as an obstruction, may have forced the thinner drift ice to keep on the European side of the Arctic pack.

About the Antarctic continent there is a broad girdle of pack ice which, while more indolent in its movements than the Arctic pack, has been shown by the expeditions of the Belgica and the Pourquoi-Pas to possess the same kind of shifting movements. In the southern spring this pack floats northward and is to a large extent broken up and melted on reaching lower latitudes.

The Antarctic shelf ice.—It has been already pointed out that the inland ice of Antarctica is in part at least surrounded by a thick snow and ice terrace floating upon the sea and rising to heights of more than 150 feet above it (plate 15 B and Fig. 316). The visible portions of this shelf-ice are of stratified compact snow, and the areas which have thus far been studied are found in bays from which dislodgment is less easily effected. The origin of the shelf ice is believed to be a sea-ice which because not easily detached at the time of the spring “break-up” is thickened in succeeding seasons chiefly by the deposition of precipitated and drifted snow upon its surface, so that it is bowed down under the weight and sunk to greater and greater depths in the water. To some extent, also, it is fed upon its inner margin by overflow of glacier ice from the inland ice masses.

Icebergs and snowbergs and the manner of their birth.—Greenland reveals in the character of its valleys the marks of a large subsidence of the continent—the serpentine inlets or fjords by which its coast is so deeply indented. Into the heads of these fjords the tongues from the inland ice descend generally to the sea level and below. The glacier ice is thus directly attacked by the waves as well as melted in the water, so that it terminates in the fjords in great cliffs of ice (Fig. 317). It is also believed to extend beneath the water surface as a long toe resting upon the bottom (Fig. 319).

The exposed cliff is notched and undercut by the waves in the same manner as a rock cliff, and the upper portions override the lower so that at frequent intervals small masses of ice from this front separate on crevasses, and toppling over, fall into the water with picturesque splashes. Such small bergs, whose birth may be often seen at the cliff front of both the Greenland and Alaskan glaciers, have little in common with those great floating islands of ice that are drifted by the winds until, wasted to a fraction only of their former proportions, they reach the lanes of transatlantic travel and become a serious menace to navigation (Fig. 318).

Northern icebergs of large dimensions are born either by the lifting of a separated portion of the extended glacier toe lying upon the bottom of the fjord, or else they separate bodily from the cliff itself, apparently where it reaches water sufficiently deep to float it. In either case the buoyancy of the sea water plays a large rÔle in its separation.

If derived from the submerged glacier toe (Fig. 319), a loud noise is heard before any change is visible, and an instant later the great mass of ice rises out of the water some distance away from the cliff, lifting as it does so a great volume of water which pours off on all sides in thundering cascades and exposes at last a berg of the deepest sapphire blue. The commotion produced in the fjord is prodigious, and a vessel in close proximity is placed in jeopardy.

Even larger bergs are sometimes seen to separate from the ice cliff, in this case an instant before or simultaneously, with a loud report, but such bergs float away with comparatively little commotion in the water.

The icebergs of the south polar region are usually built upon a far grander scale than those of the Arctic regions, and are, further, both distinctly tabular in form and bounded by rectangular outlines (Fig. 321). Whereas the large bergs of Greenlandic origin are of ice and blue in color, the tabular bergs of Antarctica might better be described as snowbergs, since they are of a blinding whiteness and their visible portions are either compacted snow or alternating thick layers of compact snow and thin ribbons of blue ice, the latter thicker and more abundant toward the base. All such bergs have been derived from the shelf ice and not from the inland ice itself. Blue icebergs which have been derived from the inland ice have been described from the one Antarctic land that has been explored in which that ice descends directly to the sea.

In both the northern and southern hemispheres those bergs which have floated into lower latitudes have suffered profound transformations. Their exposed surfaces have been melted in the sun, washed by the rain, and battered by the waves, so that they lose their relatively simple forms but acquire rounded surfaces in place of the early angular ones (Fig. 318, p.291). Sir John Murray, who had such extended opportunities of studying the southern icebergs from the deck of the Challenger, has thus described their beauties:

“Waves dash, against the vertical faces of the floating ice island as against a rocky shore, so that at the sea level they are first cut into ledges and gullies, and then into caves and caverns of the most heavenly blue, from out of which there comes the resounding roar of the ocean, and into which the snow-white and other petrels may be seen to wing their way through guards of soldier-like penguins stationed at the entrances. As these ice islands are slowly drifted by wind and current to the north, they tilt, turn and sometimes capsize, and then submerged prongs and spits are thrown high into the air, producing irregular pinnacled bergs higher, possibly, than the original table-shaped mass.”

Reading References for Chapters XX and XXI

General:—

I. C. Russell. Glaciers of North America. Ginn, Boston, 1897, pp. 210, pls. 22.

Chamberlin and Salisbury. Geology, vol. 1, pp. 232-308.

H. Hess. Die Gletscher, Braunschweig, 1904, pp. 426 (illustrated).

William H. Hobbs. Characteristics of Existing Glaciers. Macmillan, 1911, pp. 301, pls. 34.

Special districts of mountain glaciers:—

James D. Forbes. Travels Through the Alps of Savoy and other Parts of the Pennine Chain with Observations on the Phenomena of Glaciers. Edinburgh, 1845, pp. 456, pls. 9, maps 2.

A. Penck, E. BrÜckner, et L. du Pasquier. Le systÈme glaciare des alpes, etc., Bull. Soc. Sc. Nat. NeuchÂtel, vol. 22, 1894, pp. 86.

E. Richter. Die Gletscher der Ostalpen. Stuttgart, 1888, pp. 306, 7 maps.

James D. Forbes. Norway and Its Glaciers, etc. Edinburgh, 1853, pp. 349, pls. 10, map.

I. C. Russell. Existing Glaciers of the United States, 5th Ann. Rept. U. S. Geol. Surv., 1885, pp. 307-355, pls. 32-55; Glaciers of Mt. Ranier, 18th Ann. Rept. U. S. Geol. Surv., 1898, pp. 349-423, pls. 65-82.

W. H. Sherzer. Glaciers of the Canadian Rockies and Selkirks, Smith. Cont. to Knowl. No. 1692, Washington, 1907, pp. 135, pls. 42.

H. F. Reid. Studies of Muir Glacier, Alaska, Nat. Geogr. Mag., vol. 4, 1892, pp. 19-84, pls. 1-16.

I. C. Russell. Malaspina Glacier, Jour. Geol., vol. 1, 1893, pp. 219-245.

G. K. Gilbert. Harriman Alaska Expedition, vol. 3, Glaciers, 1904, pp. 231, pls. 37.

W. M. Conway. Climbing and Exploration in the Karakoram Himalayas, Maps and Scientific Reports, 1894, map sheets I-II.

Fanny Bullock Workman and William Hunter Workman. The Hispar Glacier, Geogr. Jour., vol. 35, 1910, pp. 105-132, 7 pls. and map.

The cycle of glaciation:—

William H. Hobbs. The Cycle of Mountain Glaciation, Geogr. Jour., vol. 36, 1910, pp. 146-163, 268-284.

Upper and lower cloud zones of the atmosphere:—

R. Assmann, A. Berson, and H. Gross. Wissenschaftliche Luftfahrten ausgefÜhrt vom deutschen Verein zur FÖrderung der Luftschiffahrt in Berlin, 1899-1900, 3 vols.

E. Gold and W. A. Harwood. The Present State of our Knowledge of the Upper Atmosphere as Obtained by the Use of Kites, Balloons, and Pilot-balloons, Rept. Brit. Assoc. Adv. Sci., 1909, pp. 1-55.

W. H. Moore. Descriptive Meteorology, Appleton, New York, 1910, pp. 95-136.

William H. Hobbs. The Pleistocene Glaciation of North America Viewed in the Light of our Knowledge of Existing Continental Glaciers, Bull. Am. Geogr. Soc., vol. 42, 1911, pp. 647-650.

The continental glacier of Greenland:—

F. Nansen. The First Crossing of Greenland, 2 vols, Longmans, London, 1890 (the scientific results are contained in an appendix to volume 2, pp. 443-497).

R. E. Peary. A Reconnaissance of the Greenland Inland Ice, Jour. Am. Geogr. Soc., vol. 19, 1887, pp. 261-289; Journeys in North Greenland, Geogr. Jour., vol. 11, 1898, pp. 213-240.

T. C. Chamberlin. Glacier Studies in Greenland, Jour. Geol., vol. 2, 1894, pp. 649-668, 768-788, vol. 3, pp. 61-69, 198-218, 469-480, 565-582, 668-681, 833-843, vol. 4, pp. 582-592, 769-810, vol. 5, pp. 229-245; Recent glacial studies in Greenland (Presidential address), Bull. Geol. Soc. Am., vol. 6, 1895, pp. 199-220, pls. 3-10.

R. S. Tarr. The Margin of the Cornell Glacier, Am. Geol., vol. 20, 1897, pp. 139-156, pls. 6-12.

R. D. Salisbury. The Greenland Expedition of 1895, Jour. Geol., vol. 3, 1895, pp. 875-902.

E. v. Drygalski. GrÖnland Expedition der Gesellschaft fÜr Erdkunde zu Berlin 1891-1893, Berlin, 1897, 2 vols., pp. 551 and 571, pls. 53, maps 10.

William H. Hobbs. Characteristics of the Inland Ice of the Arctic Regions, Proc. Am. Phil. Soc., vol. 49, 1910, pp. 57-129, pls. 26-30.

The Antarctic continental glacier:—

R. F. Scott. The Voyage of the Discovery. London, 2 vols., 1905.

E. H. Shackleton. The Heart of the Antarctic. London, 2 vols., 1910.

E. von Drygalski. Zum Kontinent des eisigen SÜdens, Deutsche SÜdpolar-Expedition, Fahrten und Forschungen des “Gauss”, 1901-1903, Berlin, 1904, pp. 668, pls. 21.

Otto NordenskiÖld and J. S. Andersson. Antarctica or Two Years Amongst the Ice of the South Pole. London, 1905, pp. 608, illustrated.

E. Philippi. Ueber die fÜnf Landeis-Expeditionen, etc., Zeit. f. Gletscherk., vol. 2, 1907, pp. 1-21.

Nourishment of continental glaciers:—

William H. Hobbs. Characteristics of the Inland Ice of the Arctic Regions, Proc. Am. Phil. Soc., vol. 49, 1910, pp. 96-110; The Ice Masses on and about the Antarctic Continent, Zeit. f. Gletscherk., vol. 5, 1910, pp. 107-120; Characteristics of Existing Glaciers. New York, 1911, pp. 143-161, 261-289. Pleistocene Glaciation of North America Viewed in the Light of our Knowledge of Existing Continental Glaciers, Bull. Am. Geogr. Soc., vol. 43, 1911, pp. 641-659.

Field and pack ice:—

Emma de Long. The Voyage of the Jeannette, the ship and ice journals of George W. de Long, etc. Berlin, 1884, 2 vols., chart in back of vol. 1.

Robert E. Peary. The Discovery of the North Pole (for further references on both sea and pack ice and Antarctic shelf ice, consult Hobbs’s Characteristics of Existing Glaciers, pp. 210-213, 242-244).

Icebergs:—

Wyville Thomson. Challenger Report, Narrative, vol. 1, 1865, Pt. i, pp. 431-432, pls. B-D.

I. C. Russell. An Expedition to Mt. St. Elias, Nat. Geogr. Mag., vol. 3, 1891, pp. 101-102, fig. 1.

H. F. Reid. Studies of Muir Glacier, Alaska, ibid., vol. 4, 1892, pp. 47-48.

E. von Drygalski. GrÖnland-Expedition, etc., vol. 1, pp. 367-404.

M. C. Engell. Ueber die Entstehung der Eisberge, Zeit. f. Gletscherk., vol. 5, 1910, pp. 112-132.

---