the surface, the water easily penetrates the
mass, passing through it more readily where the snow is lightest and
least compact,--in short, where it has not begun its transformation into
ice. A summer's day sends countless rills of water trickling through
such a mass of snow. If the snow be loose and porous throughout, the
water will pass through its whole thickness, accumulating at the bottom,
so that the lower portion of the mass will be damper, more completely
soaked with water, than the upper part; if, on the contrary, in
consequence of the process previously described, alternate melting and
freezing combined with pressure, the mass has assumed the character of
icy snow, it does not admit moisture so readily, and still farther down,
where the snow is actually transformed into pure compact ice, the amount
of surface-water admitted into its structure will, of course, be greatly
diminished. There may, however, be conditions under which even the
looser snow is comparatively impervious to water; as, for instance, when
rain falls upon a snow-field which has been long under a low
temperature, and an ice-crust is formed upon its surface, preventing the
water from penetrating below. Admitting, as I believe we must, that the
water thus introduced into the snow and ice is one of the most powerful
agents to which its motion is due, we must suppose that it has a twofold
influence, since its action when fluid and when frozen would be
different. When fluid, it would contribute to the advance of the mass in
proportion to its quantity; but when frozen, its expansion would produce
a displacement corresponding to the greater volume of ice as compared
with water; add to this that while trickling through the mass it will
loosen and displace the particles of already consolidated ice. I have
already said that I did not intend to trespass on the ground of the
physicist, and I will not enter here upon any discussion as to the
probable action of the laws of hydrostatic pressure and dilatation in
this connection. I will only state, that, so far as my own observation
goes, the movement of the glacier is most rapid where the greatest
amount of moisture is introduced into the mass, and that I believe there
must be a direct relation between these two facts. If I am right in
this, then the motion, so far as it is connected with infiltrated
moisture or with the dilatation caused by the freezing of that moisture,
will, of course, be most rapid where t