eg 2 
1906. No.3. AMUNDSEN'S OCEANOGRAPHIC OBSERVATIONS IN 1901. 81 
below the surface. On the same afternoon even 45? C. was observed on 
the surface, between belts of scattered ice. It was frequently found in 
this region that the surface-temperature was 1° C., or 175? C., or even 
2° C. only short distances, less than a mile, from the edge of the pack-ice, 
whilst amongst the floes the temperature was near or below zero, and 
where there was much ice even as low as —06° C. On many great 
lanes between the floes there was, however, new ice (the socalled "bay- 
ice” of the sealers), and then the surface temperature of the water might 
be about —1'4 or —ı'6°C., indicating that the salinity of the surface- 
water was between 27'0 and 290 %- The temperature of the air was 
at this time between —5° C. and —13°C. The explanation of the low 
surface temperatures and new ice is, that, by the melting of the ice- 
masses, a thin surface-layer of water with a reduced salinity is formed. 
This layer, having a much lower salinity than the underlying water, 
cannot sink by cooling; it may thus be cooled down to freezing point 
by radiation of heat from the surface, and new ice may be formed. 
It is clear that in regions where the conditions are as above, the 
ice-melting process may contribute much to the cooling of the under- 
lying sea-water; but it has of course to be remembered that these ice- 
masses were also formed in the western and north-western parts of the 
same sea, and during the process of formation an equal amount of heat 
was disengaged. The formation and melting of this ice will thus far 
have no effect upon the average temperature of this sea basinl. 
The above considerations show, that the effect which the underlying, 
intermediate, warmer water-layers may have upon the formation or mel- 
ting of ice on the surface, must be an insignificant and negligeable 
quantity for all practical purposes wherever polar conditions prevail, 
2. e. wherever the sea is cövered by a surface layer, 100 or 200 metres 
thick, of less saline water, and temperatures below zero or even below 
—r C.? 
1 In a somewhat different way it will, however, have an effect, for though the forma- 
tion of ice covered by snow duriug the winter will make the temperature of the air 
lower, it will also much reduce the vertical circulation of the underlying sea-water, 
which can then only be cooled by conduction of heat through the overlying ice and snow. 
The formation of ice will thus reduce the cooling of the sea, which is the opposite 
efiect of that attributed to it by Prof. Pettersson. 
= Pettersson’s assertion that ice can only exist in a shallow sea, and over the con- 
tinental shelves, is directly contradictory to facts. There may be mentioned the deep 
North Polar Basin, where the enormous masses of polar ice are formed. But we need 
not go so far, the northern part of the Norwegian Sea itself is a good example. In the 
deep sea between Jan Mayen and Spitsbergen very great masses of ice are formed 
during the winter and spring, as mentioned above. 
/id.-Selsk. Skrifter. I. M.-N. KL 1906. No. 3. 6 
