ought also to be more apparent towards the centre of the glacier than towards the sides. The effect of this progress downward is to produce a very powerful friction between the masses of ice and the earth over which they are pushed, and, consequently, a continual accession to that stream of water which is found issuing from all glaciers.
The result of this continual breaking up is to cause all the water melted by the friction of the blocks of ice which is not retained in the interstices to fall towards the lowest part of the descending valley, and thus increase the stream, and so take away more and more of the support of the central part of the glacier. Hence the advance of the surface of the glacier will be much quicker towards its middle than near the sides.
The consequence of these actions is, that cracks in the ice will occur generally in planes perpendicular to its surface. The rain which falls upon the glacier, the water produced from its surface by the sun's rays and by the effect of the temperature of the atmosphere, as well as the water produced by the friction of its descending fragments, will penetrate through these cracks, and be retained by capillary action on the surfaces, and still more where the distance of the adjacent surfaces is very small. The rest of this unfrozen water will reach the rocky bottom of the glacier, and give up some of its heat to the bed over which it passes, to be again employed in melting away the lowest support of the glacier ice. Although the temperature of the glacier should differ but by a very small quantity from that of the freezing point of water, yet these films will only freeze the more slowly, and therefore become more solid and transparent ice. Their very thinness will enable all the air to be more readily extricated by freezing.
The question of the regelation of pounded ice, if by that
