EE PE å 
1906. No. 3. AMUNDSEN'S OCEANOGRAPHIC OBSERVATIONS IN 1901. 13 
has salinities below 338 %,, (33:3—338 9) and densities (o,) below 
27'201, and the cold top layer of Polar water rests on a thick, inter- 
mediate layer of warmer water with much higher salinity. 
It is thus seen that nowhere under the North Polar Current, in re- 
gions hitherto known, can the cold and heavy bottom-water be cooled 
down to its low temperature by direct contact with the cold but much 
lighter top-layer, from which it is everywhere separated by an intermediate 
warmer layer. It is obvious, that the bottom-water of the Norwegian 
Sea cannot originate directly from the East Greenland Polar Current. 
produced down to the greatest depths of the ice, which may possibly be 60 or 70 
metres or perhaps even more; and Prof. O. Pettersson’s suggestion that this 
temperature-minimum may be due to contact between the sea-water and the ice (see 
Geographical Journal, London, vol. XXIV, 1904, pp. 318, 320, and 322), therefore seems 
very probably true, although according to the direct observations made (e. g. measure- 
ments of the height of the same hommock during a long period of many months) 
Pettersson is not right when he assumes that an appreciable melting of the ice 
may be produced by this contact in the North Polar Basin. During the frequent 
ice-pressures the ice-floes are broken and piled up into ridges and hummocks and 
during the winter and spring ice-blocks with very low temperatures, of —20'0 or even 
—30'0 ° C. (vide measurements, the Norw. North Polar Exp. 1893—1896, Scient. Results, 
vol. VI, pp. 544—557), are then pressed down to great depths. This cold ice will 
have a great cooling effect upon the water with which it comes into contact, because 
a considerable quantity of heat is required to raise its temperature to the freezing point 
(—1'84° C.) of sea-water with a salinity of about 33'7 °/,, (not 35'0 °/,, as stated by 
Pettersson, loc. cit. p. 318), and to liquefy the frozen brine in the ice. Much ice will 
thus be formed round the cold ice-blocks, and the blocks of one hummock may be thus 
united more or less into one solid mass before an equilibrium is attained between the 
temperature of the ice and that of the surrounding water which will be cooled down 
to its freezing point of about —1'84 °/,,, at the same time as its salinity is slightly 
raised. Contact between the ice and the sea-water alone will cool the latter, but it is 
clear that after the water-stratum has been thus cooled down to near its freezing-point 
no appreciable melting of ice can be produced, because there is such and extremely 
slow vertical circulation owing to the rapid increase of salinity and density downwards. 
The quantity of heat given off by a water-stratum 10 metres thick which is cooled 
o'ı° C., is hardly sufficient to melt 1°5 centimetre of ice. 
1 Cf. Nansen, op. cit. pp. 246—255, 306, and Pl. XVI. 
Professor Otto Pettersson has propounded the hypothesis (see Geographical Journal, 
vol. XXIV, London, 1904, p. 285) that by the melting of Polar Ice in the western and 
north-western parts of the Norwegian Sea, the Atlantic water should be cooled and 
sink towards the bottom. According to what has been pointed out above, this is not 
possible to any appreciable extent, as long as the Polar ice floats in the water of the 
Polar Current; on account of its low salinity, this water cannot sink, even if it could be 
cooled down by the melting of ice to a lower temperature than it had beforehand. 
The real Polar ice very seldom comes outside the boundaries of the water of the Polar 
Current in the Norwegian or East Greenland Sea. The melting of the Polar Ice, 
cannot therefore have any appreciable direct effect upon the cooling of the underlying 
Atlantic water, from which it is always protected by an layer of cold water with much 
lower salinity. 
To avoid misunderstanding, it may be mentioned here that there are also in the 
Norwegian Sea great masses of ice, which are formed in that same sea during the 
winter, and which occur even outside the region of the Polar Current. This ice, which 
