86 The Loess of the Rhine and the Danube. [January, 
motion of land ice, and has shown that no slope of the land 
beneath is necessary so long as there is a slope of the upper 
surface of the ice itself down which it moves, and he has 
likened its motion to that of heaped-up pitch— -a simile that 
Prof. James Forbes had also used.* There may be some 
geologists who believe that ice is pushed along as a solid, 
but certainly no such opinion has been advocated by leading 
glacialists, who only claim that if ice be heaped up it will 
flow outwards. The margin of the ice would, however, be 
pushed forward by the ice behind and above it in summer, 
when its fluidity was increased, just as in a railway-cutting 
we often see, in wet weather, clay slip down and push up 
the earth at its base. 
In my theory I assume that when ice in Greenland was 
heaped up high enough to intercept all the moisture of the 
winds blowing northwards, the snow would be heaped up on 
the southern slope, and that, if the frozen precipitation ex- 
ceeded the liquefaction and evaporation, the icy mass would 
move southward, and the highest portion of its crest would 
reach to lower and lower latitudes. It would be a ridge of 
ice and snow gathering to itself the moisture of the atmo- 
sphere, its very growth supplying a basis for future extension, 
like an invading army gathering its supplies from the country 
it was adding to its domain. From this ridge, as from a 
mountain chain, the ice would flow down, chilling the atmo- 
sphere as far as it reached, and carrying with it its own 
wintry climate. There are tree-less districts in America 
where there is not sufficient rain to support forest life. Yet 
if a small extent of country be planted, and sustained until 
the trees grow up, they will cause a greater rainfall. For 
the rain that does fall will not be allowed to run off the land 
to the sea, but much of it will be retained and returned 
again to the atmosphere by evaporation ; and as the amount 
of rain depends primarily upon the amount of moisture in 
the air, precipitation must be increased by increased evapo- 
ration. Thus a small forest may cause sufficient moisture 
to fall and to be retained, so as gradually to increase its 
limits and change a nearly tree-less country into one covered 
with forests. And so at the present time the ice is prevented 
from advancing into the temperate regions by a nice balance 
between the amount of frozen precipitation and of liquefac- 
tion ; but if that balance were only a little shaken in favour 
of the accumulation of the ice, it would obtain powers of 
* See Dana’s Manual of Geology, p. 536, for an excellent account of the 
motion of land ice. 
