I877J 
Southern Hemisphere. 
349 
Most of this snow is again liquefied before it reaches the 
earth by the higher temperature of the lower portions of 
the atmosphere it has to pass through. But if it be inter- 
cepted above, on a mountain range, it may accumulate there 
and feed glaciers that descend for thousands of feet below 
the snow-line. In its descent, in this form, only its upper 
surface is exposed to the warm air, and the lower portions 
of the mass are preserved as in an ice-house. Thus, gla- 
ciers may be said to bring down the cold of the upper 
regions of the atmosphere, and the advance of a ridge of 
ice from the north and south into the temperate zones, in 
the Glacial period, would give us the machinery for doing 
this on a scale commensurate with the extent of the glacia- 
tion of the two hemispheres. 
According to this theory, the ice that reached the coasts 
of New Zealand and South America was not due to a great 
lowering of the mean temperature of these countries, but 
flowed down upon them, bringing with it the cold of regions 
thousands of feet above the sea-level and hundreds of miles 
to the southward. The theories of local glaciation require 
a much greater change of climate than this does, just as in 
the valley of Chamoinix a very considerable lowering of the 
present mean temperature would be necessary to cause ice 
to accumulate there, instead of flowing down to it from 
Mont Blanc. In all the treatises on the Glacial period 
that I have seen, the question sought to be answered is — 
“ What causes would bring about changes of climate to 
allow perennial snow to accumulate on lands in temperate 
regions. Instead of this, I think the problem to be solved 
is — What are the conditions that would cause ice to be 
piled up around the Ardtic and Antarctic circles, and gra- 
dually invade the temperate zones, bringing down the cold 
of the upper regions of the atmosphere as it progressed ? 
I shall not now attempt to answer this question, but occupy 
what space I have left in showing how such an advance of 
ice explains the fadts we have been considering in the 
southern hemisphere and some others in this part of the 
world. 
The advance of the ice from the Antardtic regions would 
not be diredtly southward. The moist winds that fed it 
blew from the north-west, and its progress would be to the 
south-east towards its source of supply. And as the quan- 
tity of moisture in the air would be greater in some regions 
than in others, owing to the irregular distribution of land 
and water to the northward, the progress of the ice would 
not be equal all around the Antardtic, but some parts would 
