ANTARCTIC CLIMATE. 49 
that the surrounding land must contribute to raising the temperature 
over the inner part of the Arctic Sea, but that ought to manifest itself 
by great contrasts in the thermal wind-rose, but according to Dr. 
Mohn’s lecture at the Geographical Conference at Berlin, no such 
contrasts are discoverable. The summer temperatures from the Fram 
are therefore the correct ones for the North Polar regions, and are 
not seriously affected by winds, and we must look for another cause 
for the very low temperatures of the Antarctic summer. This cause 
Number of Days with 
: valias. 
STORRS. Light Winds Clear for 
f all 7 O t 
Observations mat above or foe ie Snow. Rain. 
1898 
March aa ee cas 9-1 Sie 6 15 13 we 
April iat atl. Seis 9-1 2 10 14 22 ae 
May seo se) a) ee 9-1 3 15 8 30 4 
JUNC! ss ee ak is 72 3 5 16 24 = 
JULY ee. aie: Cele a8 25°3 15 7 22 14 a 
August .. we 8:9 3 9 15 26 1 
September oe) ae 19°3 7 9 14 19 ste 
October .. .. 10°1 4 16 12 25 2 
November... .. .. 16°8 8 18 10 25 ee 
December.. .. .. 14°2 4 9 13 18 ae 
1899 
January .. .. 5:9 5 17 6 19 4 
February .. .. .. 3:1 1 21 1 22 
Autumn .. 0... 971 5 31 37 65 4 
Winter .. ow... 13°9 21 21 53 64 1 
Spring .. .. 15°3 19 38 36 69 2 
Summer .. .. .. 79 10 47 20 59 7 
Year oe. we) cass ae 11°6 55 137 146 257 14 
must be the Antarctic continent, which must be entirely covered 
with ice, the low summer temperature of which is carried. far and 
wide over the ocean by the Polar winds. Land-ice and sea-ice behave 
themselves, contrary to the usual ideas, very differently as regards 
their summer conditions. The sea-ice breaks up into floes, while the 
land-ice remains always a compact mass. 
Accordingly we find in the summer observations a new confirma- 
tion of the supposed Antarctic continent, and we may fairly expect 
that, owing to the very slight rise of temperature in summer, we 
E 
