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



higher temperature than the water; but we do not consider it probable 

 that this cooling will be so great as the cooling due to the emission of 

 heat from the surface and the vertical circulation thus created during the 

 winter; and we think, that this is proved by the observations at Hjort's 

 Stat. 89, at the mouth of Sassen Bay, and at the Swedish Stat. XXV in 

 Dickson Bay. At the mouth of Sassen Bay, Hjort observed a tempe- 

 rature as low as — 1.8'' C. (with 34.53 °/oo) at 100 metres, although the 

 sills of the ba}' inside are not so deep, and there is no glacier descending 

 to that depth. In Dickson Bay the temperatures of the bottom water in 

 August, 1908, were almost as low as in Klaas Billen Bay, although there 

 is no glacier-ice to cool the waters of this fjord. 



The Swedish oceanographers seem to have forgotten that a narrow 

 coast current runs inside the Atlantic Current along the west coast of 

 Spitsbergen, and that its waters are more or less mixed with water from 

 the land and with the water of the Spitsbergen Polar Current coming 

 from the east round South Cape. As we have seen, the water of the 

 coast current has the same temperatures and salinities as the bottom-water 

 in the outer part of the Ice Fjord, according to the observations at Isach- 

 sen's Stat. 41 as also those at Hjort's Stat. 88 and the Swedish Stats. 

 Ill — VII. There is a very good agreement between all these observations^. 



The Swedish oceanographers maintain that the observations at their 

 Stat. II of Aug. 30, 1908 (in 78^ 5' N. and 12^ 50' E. outside the mouth 

 of the Ice Fjord), compared with those of their Stat. I (see Fig. 21 G i), 

 in 77'^ 44' N. and lo*^ E., should prove that there is a sill between these 

 two stations, rising to a level somewhat lower than 100 metres below the 

 surface. We consider the existense of such a high sill to be very impro- 

 bable, and that it is disproved by the vertical series of observations at 

 the Swedish Stat. II, which has evidently been taken in the waters of the 

 coast current, and is very similar to the series of observations taken in 

 July and September, 1910, at Isachsen's Stats. 31 and 30, only that the 

 Swedish salinities are on the whole higher in the surface-layers and lower 

 in the deep layers, except at 200 metres; but the high value of 35.14 °/oo 

 found there, is obviously erroneous, as is proved by the absurd density (ot) 



^ The Swedish authors maintain that the bottom-water of the Ice Fjord should have a 

 salinity of 34.92 ^Jqo, which is verj' like that of the bottom-water of the Norwegian 

 Sea; and they base what they consider to be very important conclusions on this 

 salinity [191 1, pp. 12, 17]. Their own series of observations prove, however, that 

 both the salinities and the temperatures of the bottom-water of the Ice Fjord vary a 

 great deal, and it is only at one depth at one station that they have found the 

 salinity mentioned, which is consequently accidental, as is also proved by the observa- 

 tions at Isachsen's Stat. 41. 



