THE COLORADO PLATEAU.
We had a suspicion that the area between Foreland Sound and Ice Fjord was not covered by an icesheet, but we still thought it probable that one would be found in the region north-east of Ice Fjord. The result of our present expedition was to prove this not to be the case. We traversed a great deal of glacier and snowfield, but none belonging to a true icesheet. The whole of this region, which I have named Garwood Land, after my excellent companion, is a glaciated mountain and valley system. Each glacier in it is a clearly-marked unit, with its evident watersheds dividing it from its neighbours. North of the Chydenius range, by which Garwood Land is bounded, there does come a true icesheet covering the whole of New Friesland and flowing down to the sea on all sides. North-East Land, too, is buried under an icesheet. These are the only ones in the Spitsbergen archipelago.
The mountains of Garwood Land are remains of a denuded plateau, resembling those of the Sassendal region. They have been carved out by a denuding agent eating a series of valleys back into the plateau. Readers of my former book, “The First Crossing of Spitsbergen,” will remember how many examples of the rapid formation and extension of valleys by the eating back of the head-waters are there recorded. The Colorado Berg north of the Sassendal was the best example of the process. That plateau, now bare of ice, is being rapidly cut up into separate hills by the excavation of a series of deep, narrow cañons, which will widen and creep further back year by year. Now, the hills of Garwood Land are of a similar type. The wide, deep valley, into the head of which we looked down from our farthest point, sends back into the plateau (or remnant of a plateau) a number of tributary valleys, all of the same deep, gently sloping, steep-headed type. From many indications we concluded that a series of similar valley-heads and cliffs lay to the eastward of our whole route from where we turned back as far as the Terrier. This row of cliffs and bluffs probably flanks the eastern watershed of the Nordenskiöld Glacier. The bad weather that prevented our ascent of the Terrier prevented also the verification of this hypothesis.
If we could assume that Garwood Land was at any time considerably less glacier-covered than it now is, so that its valleys were bog-bottomed like the Sassendal, and its uplands resembled the Colorado Berg, it would be easy to account for the present configuration of the land surface. We should say that it was formed by aqueous denudation, and subsequently covered up by the increase of the ice. It is certain that there has been a great increase in the ice-supply on the land hereabouts during the last two centuries, for in that time the Negri Glacier has advanced at least fifteen, probably twenty, miles into the sea along a front fifteen miles in width. This fact, however, does not suffice as foundation for so great an assumption. It is rather to the steady elevation of the land that we must look for a solution. Everywhere in Franz Josef Land and Spitsbergen the land is known to be rising. The western belt of the island has been longer exposed to denudation than the east belt. The latter, therefore, has perhaps been later elevated. It came up from the sea as relatively flat ground. As its elevation continued this flat ground was raised into a plateau. At first it did not reach the level of perpetual snow, so that whilst rising it was being cut down into valleys and cañons by the action of water, pouring off from the plateau over its edge, and hurrying down a frost-split rock-face. The bed of such a valley has of necessity a very gentle slope. The head is steep, almost a cliff, the whole face of which is being continually stripped off, so that the valley, once begun by a waterfall over the edge of the horizontally stratified plateau, penetrates steadily backward.
These valleys once formed, with their steep heads and sides, would maintain themselves even after the remains of the plateau were covered with an icesheet and the valleys filled with glaciers. There is no need to predicate for the glaciers any power of erosion; that is not the way arctic icesheets act, for the upper layers of ice flow over the lower at a far greater speed than is the case in glaciers under lower latitudes. Given an existing cliff, however, with a glacier below it, and the denuding agencies of frost and water at work upon it, that cliff tends to maintain itself and to eat its way back into the mountain mass behind, for its débris fall upon the glacier below and are carried away; they do not pile themselves up into a protecting slope at the base of the cliff. This eating-back process will go forward with unequal speed according to the varying qualities of the rocks. Bays will thus be formed and will eat back into the plateau, just as the gullies eat back in the Sassendal region, only the bays will tend to grow wider in proportion to their depth in a glaciated country than in a region mainly bare of snow and ice.
For this process to begin it is necessary that somewhere a rock-face should be exposed to the air. The exposure may be produced by a fault, or by a denuding process begun before the land was much glaciated. We are in no position yet to assert how the process commenced in Garwood Land, but that the bays, valleys, and cliffs now existing are being maintained in the manner above described is certain. If the ice were again to cover up the Colorado Berg and the hills opposite, and were to flow into and down the Sassendal to Sassen Bay, the aspect of that region would resemble that of Garwood Land to-day. It is only in the case of a country like Greenland, entirely buried under an icecap thousands of feet thick, through which, save along the coast, no rock appears and no cliff is exposed—it is only in such a country that the conservative action of ice is complete and the modelling of an elevated land-mass into hills is practically arrested. Hence the scientific importance of distinguishing between a proper icesheet (in the Greenland sense) and a mere assemblage of separate glaciers, however large in volume and intimate in their connexion with one another. An icesheet, or inland-ice, operates in a totally different manner from a series of glaciers. Save in North-East Land and in the part of Spitsbergen called New Friesland, there is no proper icesheet in Spitsbergen, and the phrase “inland-ice” should be expunged from maps and descriptions of regions to which it is not applicable. A chief and no unimportant result of our explorations in the interior of Spitsbergen is this discovery that the parts supposed to be enveloped in an icesheet are in fact merely glacier regions.