ing the adventurous steps of Saussure, Charpentier, and Agassiz, have, with Mallet, Darwin, Martins, Forbes, and Hopkins, measured, calculated, and imitated the glaciers of many mountains and various latitudes.
Snow is the parent of glaciers; mountains are only their birth-place. Mountain ranges may by their mere narrowness and steepness furnish no cradle; they may be in so dry a region that snows are not abundant, and glaciers grow but feebly, or have such very gradual slopes as to allow of only very slow downward movement. But where the climate favours abundant precipitation of aqueous vapour, on an expanse of high land amidst loftier peaks, from which steep valleys lead down to levels much below the snow-line, the glacier, fed by a perpetual growth from above, and wasted by an eternal corrosion at the lower extremity, is modified by continual transformations of interior substance, and stimulated by a never-ceasing activity of descent.
It is, in fact, a river of ice, slowly winding its way from an inexhaustible upper sea (mer de glace), losing at every instant a part of its substance, and undergoing change in all its features, till, bent, broken, and dissolved, it gives birth to a stormy river, or floats away in icebergs to cool far-distant seas.
The substance of a glacier is not snow, nor is it wholly pure ice; it consists of the peculiar icy compounds, and manifests the peculiar structures which are generated when snow, after partial and interrupted fusion, is re-aggregated by frost. If this fusion be complete, pure ice is the refrigerated result, and this appears in glaciers; but the greater part of the glacial mass is derived from nevé, which is the partially fused and re-aggregated snow. Such being its composition, its parts are not incoherent as snow, nor liquid as water, nor wholly incapable of mutual displacement as solid ice: but it has something of all these properties; for it moves in a coherent mass, which is capable of flexure
