OCEANOGRAPHICAL RESULTS, MICHAEL SARS, 1900. 153 



course of the current. I consider it probable that the Gulf 

 Stream or Atlantic Current is disturbed in its course and partly 

 deflected towards the north-west (towards Iceland) in some simi- 

 lar way south of the Scotland — Faeroe — Iceland ridge. It is also 

 remarkable how well the isotherms and the isopyknals of the 

 deeper strata in all parts of the sea seem to follow the con- 

 figuration of the sea-bottom (see especially the isopyknal of 

 1-02808 in Sect. II, PI. 13). This proves that the moving deep 

 water is elevated or lowered with the undulations of the bottom 

 of the sea-bed 



If we consider the chart (PI. I) of the surface salinity and 

 temperature it must strike one how almost excactly the most 

 saline surface water follows the deepest channel of the Norwegian 

 Sea, and how the isotherms, especially of 10° G. and 9° C, seem 

 to be deflected in a way similar to the isobaths. It seems to be 

 probable that the central part of the Gulf Stream or Atlantic 

 Current has been deflected towards the north-west by the gra- 

 dual elevation of the sea-bottom off the Norwegian Coast between 

 Latitudes 67 and 69° N. (cf. the shape of the isobaths of 1000 m. 

 and 2000 m. on PI. I). 



But the main portion of the saline, northward flowing current 

 being thus deflected, the coast-water has a corresponding oppor- 

 tunity to extend seaward over this somewhat shallower portion 

 of the sea, and we thus find the less saline surface-water to 

 have a much wider extension in Section II than in Section I. 



Section III (Pis. 8 and 14), from Stat. 7, of Section I, 

 through Stat. 46, of Section II, to Stat. 64 west of Bear Island, 

 demonstrates clearly the northward decrease of the salinity and 

 temperature of the Gulf Stream, which, in the upper strata, are 

 much diminished for every station northward^. The isohaline 



1 The only exception is the difference between Stats. 46 and 68, which 

 is to some extent opposite. But this is evidently the effect of the 

 summer, as the observations at Stat. 46 were taken on Aug. 13th, 1900, 

 while those at Stat. 68 were taken on Sept. 8th, 1900. The effect of 

 the summer comes late. 



