1906. No. 3. AMUNDSEN’S OCEANOGRAPHIC OBSERVATIONS IN 1901. 79 
It ought also to be remembered, that if the heat required to melt, 
for instance, ı metre of ice be taken from the water, it would cool down 
by ı° C. the underlying water-layers to a depth of 79 metres; but this 
is not possible, because on the one hand the underlying layers of Polar 
water are too cold beforehand, and on the other hand no vertical cir- 
culation, necessary to produce such a cooling, can occur in these water- 
strata, where the density too rapidly decreases on rising towards the 
surface. 
Let it, however, be assumed that the vertical circulation in the 
East Greenland Polar Current might reach down to 79 metres (which it 
however does not, as is proved by the vertical series of temperatures 
and salinities). Let it also be supposed that the mean temperature of this 
water-layer, 79 metres thick, be —1'5" C., and its mean salinity be 34'0 °oo 
(which is too high). The freezing-point of water with this salinity is 
about 1:85? C. It is consequently seen that if this bulk of water be 
cooled down to freezing-point from —1:5° C., the quantity of heat thus 
consumed will melt a continuous layer of ice 0'35 metres thick. But in 
the summer, while the ice-melting is going on, the temperature in the 
East Greenland Polar Current rapidly rises towards the surface from the 
temperature-minimum at about 50 or 60 metres. And already by the 
beginning of July, the surface temperature has risen considerably, to 
09°C. (at Nathorst’s Stat. N VII), to —o'ı° C. (at Ryders Stat. À XII), and 
to —o'2° C. (at Amdrup’s Stat. Ag III), whilst in the winter the water was 
cooled down to its freezing-point, about —1'7? C. This rise of temperature 
is due to the heat-wave (caused by direct radiation from the sun) pene- 
strating downwards from above, and it shows that the melting of the ice 
is not able to prevent the heating of the water from above, and conse- 
quently, as long as the ice floats in the diluted cold water-layers of 
the Polar Current, its melting during the summer, both on its upper sur- 
face and on the under-side of the floes, must be due chiefly to this heat 
from above (which may either melt the ice directly, or heat the sur- 
face water in which the ice is floating), and not to heat coming from 
the intermediate layers of warmer water underlying the Polar Current, 
and against which it is well protected by the cold waters of the latter!. 
1 It ought also to be remembered that the melting point of the Polar ice is much above 
the temperature of rhis polar water, which has a layer with a temperature-minimum of 
—ı'4 to —1’9° C. The many observations made of the temperature in the ice down 
to depths of 1°6 metre, during the expedition across the North Polar Basin (see Norw. 
N. Polar Exp. 1893—1896, Scientific Results, vol. VI, pp. 545—557) show that during 
the polar summer, in July and August, the temperature in the ice rises to about 
—0'4? C. even at 1'6 metre below its surface, and this temperature is evidently near 
