I912. Xo. 12. THE SEA WEST OF SPITSBERGEN. 63 



of 28.03 ^^'^^h densities of 27.89 at 240 metres and 27.86 at 150 metres. 

 The latter densities agree perfectly with the densities of 27.90 at 240 

 metres and 27.85 (and 27.87) at 150 metres at Isachsen's Stat. 31, in 

 July and September, 1910. It is unfortunate that the Swedish authors 

 have drawn what they consider to be important conclusions from this 

 high value of 35.14^00 at 200 metres at their Stat. II. without noticing 

 that it must be erroneous. The densities above and below prove that 

 the salinity has been about 34.94^ 00, provided that the temperature (2.95^ C.) 

 observed is correct, and this salinit}' also agrees with the salinitv of 

 34.95^00 (with a temperature of 2.83° C I observed in 200 metres at 

 Isachsen's Stat. 31 on September 6th, 1910. We therefore see no reason 

 why there should be an\- high sill or threshold outside the Swedish Stat. II 

 preventing the free communication of its bottom- water with the waters at 

 the same levels of the coast current. On the contrarv, we consider it 

 probable that a submarine tjord or valley traverses the shelf from the con- 

 tinental slope towards the mouth of the Ice Fjord [cf. our bath3'metrical 

 chart. 1909, PI. I]. 



The Swedish authors pay much attention to the melting of the ice 

 in the Ice Fjord, which they think is chietl}- due to the contact of the ice 

 with the sea-water, and to the influx of warmer water from the Atlantic 

 current outside the tjord. They believe that this supposed process dis- 

 proves the correctness of our view that the polar ice carried b\' the East 

 Greenland Polar Current is chiefly- melted b\' the heat from above, directly 

 due to solar radiation, during the summer, and only to a very insignifi- 

 cant extent by heat coming from the underlving water ^. The Swedish 

 authors base their argument chiefly on the conditions near a wall of 

 glacier-ice descending to a depth of 159 metres^ which is naturally a very 

 different thing. And then they seem to have entirely forgotten that during 

 the winter the Ice Fjord is covered with a thick layer of ice, which is 

 formed there, and this consequenth- does not melt in the winter in spite of 

 anv influx of warmer water. But this ice melts during the summer, and 



' Our view in this respect had led the Swedish oceanographers to the superfluous remark 

 [191 1, p. 17] that ice is melted by contact with sea-water, as long as the latter has 

 „einen Temperaturûberschuss von einem Millionstel Grad über die Gleichgewichts- 

 temperatur des Systems". This, however, is not the question, which is: Hozv much ice 

 can be melted in this manner, c. g. in one year, if the temperature of the sea-water be 

 only some few tenths of a degree above its freezing point? A verj' simple compu- 

 tation will prove that, e. g. in the East Greenland Polar Current, it can only be quite 

 insignificant compared with the quantity of ice which melts there everj' summer, as we 

 have pointed out on several previous occasions. 



