[§ 41.] Longitudinal Crevasses.
276. We have thus unravelled the origin of both transverse and marginal crevasses. But where a glacier issues from a steep and narrow defile upon a comparatively level plain which allows it room to expand laterally, its motion is in part arrested, and the level portion has to bear the thrust of the steeper portions behind. Here the line of thrust is in the direction of the glacier, while the direction at right angles to this is one of tension. Across this latter the glacier breaks, and longitudinal crevasses are formed.
277. Examples of this kind of crevasse are furnished by the lower part of the Glacier of the Rhone, when looked down upon from the Grimsel Pass, or from any commanding point on the flanking mountains.
[§ 42.] Crevasses in relation to Curvature of Glacier.
278. One point in addition remains to be discussed, and your present knowledge will enable you to master it in a moment. You remember at an early period of OUT researches that we crossed the Mer de Glace from the Chapeau side to the Montanvert side. I then desired you to notice that the Chapeau side of the glacier was more fissured than either the centre or the Montanvert side ([75]). Why should this be so? Knowing as we now do that the Chapeau side of the glacier moves more quickly than the other; that the point of maximum motion does not lie on the centre but far east of it, we are prepared to answer this question in a perfectly satisfactory manner.
279. Let A B and C D, in the diagram opposite, represent the two curved sides of the Mer de Glace at the Montanvert, and let m n be a straight line across the glacier. Let o be the point of maximum motion. The mechanical state of the two sides of the glacier may be thus made plain. Supposing the line m n to be a straight elastic string with its ends fixed; let it be grasped firmly at the point o by the finger and thumb, and drawn to o', keeping the distance between o' and the side C D constant. Here the length, n o of the string would have stretched to n o', and the length m o to m o' and you see plainly that the stretching of the short line, in comparison with its length, is greater than that of the long line in comparison with its length. In other words, the strain upon n o' is greater than that upon m o'; so that if one of them were to break under the strain, it would be the short one.
280. These two lines represent the conditions of strain upon the two sides of the glacier. The sides are held back, and the centre tries to move on, a strain being thus set up between the centre and sides. But the displacement of the point of maximum motion through the curvature of the valley makes the strain upon the eastern ice greater than that upon the western. The eastern side of the glacier is therefore more crevassed than the western.
281. Here indeed resides the difficulty of getting along the eastern side of the Mer de Glace: a difficulty which was one reason for our crossing the glacier opposite to the Montanvert. There are two convex sweeps on the eastern side to one on the western side, hence on the whole the eastern side of the Mer de Glace is most riven.
[§ 43.] Moraine-ridges, Glacier Tables, and Sand-Cones.