346 Glacial Period in the [July, 
that the latter would advance southward, not only by flow- 
ing glacier-like, but by the area of precipitation itself being 
moved southward. A similar accumulation and extension 
must, I think, have taken place in the Antarctic regions. 
We have learnt that there is a great precipitation of snow 
around the edge of the ice-sheet with which the Antarctic 
continent, so far asv/e know, is covered; but an equilibrium 
has been attained between the forces concerned in its forma- 
tion on the one hand, and its liquefaction on the other, so 
that it does not now advance further. 
In the Glacial period that equilibrium must have been 
disturbed in favour of the accumulation of the ice, not 
necessarily by greater cold, which would tend to lessen the 
precipitation, but from some cause favouring a larger amount 
of moisture reaching the polar regions, and falling there in 
the form of snow. If this did occur, I believe that the ice 
might obtain powers of accumulation and growth, where it 
was fed by large evaporating areas, that might enable the 
ice-sheets now nearly confined within the ArCtic and Ant- 
arctic circles to advance upon and invade the temperate 
regions; especially down the ocean depressions. I indicated 
some of the causes of accretion in my last paper. The 
currents of air travelling from the tropical and temperate 
zones towards the Poles, such as the counter trade winds, 
are charged with moisture which is precipitated when they 
reach colder regions or encounter mountain-chains. Con- 
fining ourselves to the southern hemisphere we find that 
much of the moisture falls as rain in low latitudes, yet 
that a certain proportion of it reaches the ArCtic circle, and 
there falls as snow, and recuperates the ever-melting ice- 
sheet. Let us suppose that some change took place by 
which much more of this moisture would reach the southern 
ice-sheet, such, for instance, as would probably be effected 
if the obliquity of the ecliptic was increased. This, by 
raising the temperature of the temperate zone, would cause 
much more moisture to reach the southern ice-cap, and be 
precipitated there so as to add to its extent and height. As 
the edge of the ice-cap moved northward, it would gather to 
itself not only the moisture that would have reached it in 
its old position, but that which belonged to the area it now 
occupied. 
As I remarked in my last paper on the Atlantic Glacier, 
but reversing the direction of its progress, it would be a 
ridge of ice slowly advancing northward, and appropriating 
the precipitation of the regions it invaded to aid in its 
further progress. Its advance would be due to two causes, 
