209 
1914-15.] Meteorological See-Saw over Antarctic Seas. 
during the winter season. For the three months ending January 1912 the 
levelling-up process became more pronounced, pressure being only 1*7 mm. 
lower at Framheim than at Cape Evans, and temperature 2°’5 lower. Pressure 
at Cape Adare was on the mean of the whole period (April to December) 
2'2 mm. higher than at Framheim, temperature 11°‘5 higher, and the wind 
velocity 0*4 m.p.s. greater ; but from October to December the winds were 
lighter than on the barrier. Compared with the temperature at Cape Adare, 
some 380 miles to the north, the Cape Evans temperature is relatively high, 
the thermal gradient being only at the rate of 0°*7 C. per degree of latitude, 
while, as compared with Amundsen’s station, only 53 miles to the south and 
some 400 miles to the east, the excessive warming effect is remarkable. 
This is obviously due (1) to the Fohn effect caused by the suck-out and 
descent of the air from the mountainous Victoria Land, (2) to the effect 
of the lower pressure to the north-eastward over the Ross Sea (most pro- 
nounced in winter), and (3) partly to the greater frequency of calms at 
Amundsen’s station. (See footnote on page 207.) 
The fact that rises of temperature at Cape Evans were frequently 
accompanied by snow does not dispose of our view that dynamic heating 
was responsible for the relative warmth noted under these conditions. 
For it has been shown that under suitable topographical surroundings an 
appreciable warming effect is set up in rainy or snowy weather, due to 
the presence of mountainous areas over which the moisture-laden winds 
had previously passed.* This effect at MAIurdo Sound varies greatly from 
one year to another and was most marked in the period May to August of 
191*2, when the mean pressure was 3‘6 mrn. under the four years’ normal for 
these months, the mean temperature 3°'7 higher, and the mean wind move- 
ment nearly double the average, blowing with a velocity of 12*4 m.p.s. or 
5*0 m.p.s. above the average. 
The normal seasonal variations of pressure, temperature, and wind 
velocity from three regions where the available data are sufficient to give 
fair monthly means are shown in Table II. The means are based on four 
years’ data from M'Murdo Sound, nearly twelve years’ data from the South 
Orkneys, and a composite series from the west coast of Graham’s Land 
derived from Charcot’s two expeditions to that region, supplemented by 
the data from the Belgica records, covering three years in all. The wind 
velocity is not given for the latter series, since the data are not homogeneous. 
Looking at the tabulated values which give for each month the departures 
from the annual mean for each of the three elements, we observe that there 
are marked variations between the three localities. 
* “ The Meteorology of Glen Nevis,” Trans. Roy. Soc. Edin.., vol. xliv, pp. 652, 653. 
VOL. XXXV. 14 
