58 BJØRN HELLAND-HANSEN AND FRIDTJOF NANSEN. M.-N. Kl. 



winter, and of the intermediate temperature-minimum in the summer, will 

 be higher. 



This process is perfectly analogous to that by which the bottom-water 

 of the Norwegian Sea as also that of the Barents Sea [cf. Nansen, 1906] is 

 formed by the cooling of the surface during the winter, only that in the 

 Ice Fjord the vertical circulation cannot penetrate so deep, and cannot 

 reach the bottom (except in the shallow or inner closed parts of the fjord 

 e.g. at Hjort's Stat. 91 in Sassen Bay, cf. also the Swedish Stat. XXV 

 in Dickson Bay) owing to the too rapid increase of salinity towards the 

 bottom, which is due to the inflow of water along the bottom from outside. 



The intermediate layer with the temperature minimum, observed in 

 the Ice Fjord in the summer, is consequently not due to the melting of 

 ice, as the Swedish authors believe [1911, pp. 13 et seq.], but to the 

 cooling of the sea-surface and the formation of ice during the winter. 



During the summer the surface-layers are gradually heated by the 

 radiation of heat from the sun and their salinities are gradually much 

 reduced by the melting of the ice, and by the river-water from land. 

 Thus the uniform vertical distribution of temperature and salinity existing 

 at the end of the winter, gradually disappears. There are no distinct 

 indications of this uniformity at Hjort's stations of Jul}', 1901; but at the 

 Swedish stations of July, 1908, we still see traces of it in the uniform 

 salinities, while there is very little left at Stat. Ill, on September ist of 

 that year, and at Stat. 41, on September 6th, 1910. 



At the Swedish Stations VIII, of July 30th, and XXIII of Aug. 29th, 

 1908, in the inner part of the Ice Fjord, the heating of the upper layers 

 was farther advanced than at Stats. V — VII nearer the mouth of the fjord, 

 and therefore the temperature minima were situated lower, at 130 and 

 150 metres, but were not so cold (iiiz. — 0.98° and — 0.83" C). In 

 this inner region of the fjord, the cooling during the winter may have 

 been greater, and the vertical circulation more active. This is proved 

 by Hjort's Stat. 89 at the mouth of Sassen Bay, where — 1.8" C. was 

 observed at 100 metres on July 27th, 1901. 



Traces of similar temperature minima are also perceptible in vertical 

 series of observations taken in the sea outside the coast, on the conti- 

 nental shelf. It has already been pointed out that at Hjort's Stat. 91, 

 13 naut. miles outside the mouth of the Ice Fjord there was a minimum 

 of — 0.5^^ C. at 100 metres (with 34.63*^/00). At Isachsen's Stat. 31, there 

 was a minimum of 1.05° C. (with 34.36^/00) at 50 metres, on July 31st, 

 1910, and of 2.23^ C. (with 34.62 "/00) at the same depth on September 6th. 



