The shady side of the glacier—there is usually one more sheltered from the sun—may offer the best chance of finding closed-up crevasses: or of discovering bridges, connecting flakes, etc., such as may help him through any system of open crevasses which extends right across the glacier.
On a curving glacier, the concave side may present to the eye the steeper and more threatening waves of surface ice, since the retardation will be greater on this side. Actually, the ice on this concave side, of slower movement, will be less intricate and the crevasses more contracted. Often the steep crests of ice will be merely surface survivals of crevasses, and their bases will be merged continuously in one another, forming broad glacier dimples easy to negotiate.
On the convex side, the ice crests of swifter movement may appear to be more negotiable, but the crevasses will be open and deep; and on this side there is always more possibility of finding a secondary or tertiary system of edge-crevasses intruding upon the normal side-system. These crevasses are produced by the friction of the glacier against its containing walls, and, like other crevasses, by the unevenness of the rock bed; but being on the convex curve and unsubjected to compression they not only open but stay open. Most edge-crevasses are usually much covered with moraine and stones, on lower glacier, and by many treacherous forms of rotting snow on upper glacier, and they are thoroughly unpleasant to work through. It is best to avoid them by turning out towards the middle of the glacier, until the side-rifts run out in harmless cracks.
On a glacier of approximately straight descent the ordinary systems of crevasses are three in number: these various marginal crevasses, transverse crevasses, and crevasses running lengthways down the fall.
The transverse crevasses are caused by the falls in the rock bed of the descending glacier. They are best crossed on a direct line of ascent or descent of the glacier. According as the glacier is humped in the middle, or flat and even hollow, these crevasses will be found, in the first case, less open nearer the sides of the glacier, and in the second, less open towards the flat or concave centre.
The third system, of lengthways crevasses, is due to the opening out of the containing walls of rock, or to long ‘hog-back’ elevations in the rock bed below, which thrust up the centre of the glacier. This is the more dangerous system to negotiate. Crevasses, especially those whose edges are concealed, must be approached, for crossing, at right angles to their run. A party ascending or descending the glacier lengthways will have constantly to be swinging its whole length on the rope, wheeling until at least two of the party at a time will be approaching the crevasse at right angles to its line and their former line of march. So long as the glacier is clear of snow, the crossings can be thus safely, if tiresomely, managed; but when the crevasses are snow-covered, or where we meet lengthways and transverse crevasses interlinked, it calls for fine icemanship to prevent sometimes all the party finding themselves moving along the same line as the crevasse they intend to cross, or even over its snow-covered length.
Where a glacier, straight or crooked, has one of its sides unsupported by a rock wall, and is therefore falling away at its own edge, or where its marginal and transverse crevasses have joined hands, a disagreeable system of curving crevasses must be expected. This is difficult to prospect, and dangerous to unravel if it is under snow. To locate the direction of a ‘scimitar’ crevasse at one point, and avoid it or cross it correctly, does not protect us against its later loop of unexpected trespass upon our route. The side of the falling away of a glacier should as far as possible be avoided, in spite of the temptation of its easier-looking surface ice.
On large glaciers, where two systems of big crevasses, each equal and still active, have met, we find the worst problem of all: that is, separate ‘cubes’ or islands of ice with crevasses on all their four sides. It leads only to disappointment to attempt to get through a big system of this character. Fortunately it is generally recognizable from a distance by its upstanding squared summits. For a space, also, round the junction of any two glaciers these turbulent effects of cross-pressure must be expected, and the area entered with caution.
On high snow-covered glacier, the huge rotund depressions or mills cannot be disregarded. But they can usually be skirted closely with confidence. They are the result of strains that, fortunately, do not create subsidiary or flanking clefts. They resemble mammoth ice sea urchins, not glacial star-fish.
On glaciers of rapidly changing angle and surface, a disagreeable phenomenon is the formation of an upper crust of ice, or frozen snow, over lesser depressions of the mill type. The crust gets separated from the fall of the surface below it, and while it continues to present a harmless appearance to all but very close inspection, it may surprise a whole party by letting them down simultaneously into a sufficiently startling hollow. The situation is generally more sensational than serious. Similarly, in winter, pools are formed in glacier hollows and frozen over. Then they are covered with snow, and the water drains from under the crust. Thus ice-traps are formed very difficult to detect by the eye, but which sound hollow under the testing tap of the axe point. On covered glacier the axe may never rest idle.