There is no chain of mountains in which the shape of the valleys is more favorable to the formation of glaciers than the Alps.  Contracted at their lower extremity, these valleys widen upward, spreading into deep, broad, trough-like depressions.  Take, for instance, the valley of Hassli, which is not more than half a mile wide where you enter it above Meyringen; it opens gradually upward, till, above the Grimsel, at the foot of the Finster-Aarhorn, it measures several miles across.  These huge mountain-troughs form admirable cradles for the snow, which collects in immense quantities within them, and, as it moves slowly down from the upper ranges, is transformed into ice on its way, and compactly crowded into the narrower space below.  At the lower extremity of the glacier the ice is pure, blue and transparent, but, as we ascend, it appears less compact, more porous and granular, assuming gradually the character of snow, till in the higher regions the snow is as light, as shifting, and incoherent, as the sand of the desert.  A snow-storm on a mountain-summit is very different from a snow-storm on the plain, on account of the different degrees of moisture in the atmosphere.  At great heights, there is never dampness enough to allow the fine snow-crystals to coalesce and form what are called “snow-flakes.”  I have even stood on the summit of the Jungfrau when a frozen cloud filled the air with ice-needles, while I could see the same cloud pouring down sheet of rain upon Lauterbrunnen below.  I remember this spectacle as one of the most impressive I have witnessed in my long experience of Alpine scenery.  The air immediately about me seemed filled with rainbow-dust, for the ice-needles glittered with a thousand hues under the decomposition of light upon them, while the dark storm in the valley below offered a strange contract to the brilliancy of the upper region in which I stood.  One wonder where even so much vapor as may be transformed into the finest snow should come from at such heights.  But the warm winds, creeping up the sides of the valleys, the walls of which become heated during the middle of the day, come laden with moisture which is changed to a dry snow like dust as soon as it comes into contact with the intense cold above.

Currents of warm air affect the extent of the glaciers, and influence also the line of perpetual snow, which is by no means at the same level even in neighboring localities.  The size of glaciers, of course, determines to a great degree the height at which they terminate, simply because a small mass of ice will melt more rapidly, and at a lower temperature, than a larger one.  Thus, the small glaciers, such as those of the Rothhorn or of Trift, above the Grimsel, terminate at a considerable height above the plain, while the Mer de Glace, fed from the great snow-caldrons of Mont Blanc, forces its way down to the bottom of the valley of Chamouni, and the glacier of Grindelwald, constantly renewed from the deep reservoirs where