Ice of this kind may be seen forming every winter on our sidewalks, on the edge of the little ditches which drain them, or on the summits of broad gateposts when capped with snow. Of such ice glaciers are composed; but, in the glacier, another element comes in which we have not considered as yet,--that of immense pressure in consequence of the vast accumulations of snow within circumscribed spaces. We see the same effects produced on a small scale, when snow is transformed into a snowball between the hands. Every boy who balls a mass of snow in his hands illustrates one side of glacial phenomena. Loose snow, light and porous, and pure white from the amount of air contained in it, is in this way presently converted into hard, compact, almost transparent ice. This change will take place sooner, if the snow be damp at first,--but if dry, the action of the hand will presently produce moisture enough to complete the process. In this case, mere pressure produces the same effect which, in the cases we have been considering above, was brought about by alternate thawing and freezing,--only that in the latter the ice is distinctly granular, instead of being uniform throughout, as when formed under pressure. In the glaciers we have the two processes combined. But the investigators of glacial phenomena have considered too exclusively one or the other: some of them attributing glacial motion wholly to the dilatation produced by the freezing of infiltrated moisture in the mass of snow; others accounting for it entirely by weight and pressure. There is yet a third class, who, disregarding the real properties of ice, would have us believe, that, because tar, for instance, is viscid when it moves, therefore ice is viscid because it moves. We shall see hereafter that the phenomena exhibited in the onward movement of glaciers are far more diversified than has generally been supposed.