ANTARCTIC CLIMATE. 47 
type of the climate of the region. This result is of the highest 
importance. If the surface of the Antarctic (ie. the area portion 
bounded by the Antarctic circle) were approximately homogeneous, that 
is if the Pole lay in the centre of a continent, or if there were only 
very small islands, the centre of the South Polar anticyclone would 
lie persistently over the pole, and the system of Polar winds would 
extend itself uniformly on all sides, in winter naturally further than 
in summer, and the data from the Belgian expedition would give 
us a picture the exact opposite of what they do; there would be 
Easterly and Southerly winds in winter, Westerly and Northerly in 
summer. But the anticyclone itself is subject to shiftings in the 
course of the year; in winter it moves further towards the Eastern 
Hemisphere, and for this we can assign no other explanation than 
that a cold centre must lie on that side, towards the Indian ocean. 
But such a centre can only develop itself, apart from the Pole, over 
large areas of land. Accordingly we find in the Belgian winter 
observations a new argument for the idea, that the Antarctic land 
area belongs chiefly to the Eastern Hemisphere, where most topo- 
graphical indications of a continent are presented to us by Victoria 
Land and Wilkes Land. 
In winter therefore the Belgian station lay in the zone of the 
South Polar depression: and this is further proved by the fact that 
the non-periodic barometer oscillations reached their greatest ampli- 
tude at that season. 
On the other hand, the curve of pressure for the year shows a 
character quite different from that of the depression in the Northern 
Hemisphere up to 70° N., for the barometer was lower in summer 
than in winter, so that apparently the Southern edge of the depres- 
sion was not very far distant from the station. Despite the prevalence 
of Equatorial winds the winter was very cold, partly owing to the 
frequency of calms, but principally because the oceanic area from 
which the winds came has no Gulf Stream to warm them. Circum- 
stances as favourable as those which are found in the North Atlantic 
Polar Sea do not occur elsewhere on the surface of the Earth. 
Spring and summer are in fact times of transition. They are 
characterised by the struggle for mastery between the Equatorial and 
Polar winds. This explains the irregularity of the annual curve, the 
abnormal cold of April and September, and the abnormal warmth 
of May. 
The predominance of the Polar winds in summer is even more 
strongly marked than that of the Equatorial in winter. 
The low temperature is not so very surprising for the Antarctic 
