148 FBIDTJOF NANSEN. 



parison perfectly trustworthy. It seems, however, as if the 

 differences in salinity at the same place and the same depths of 

 the Norwegian sea are less from one year to another than the 

 differences in temperature, and this may be easily explained, for 

 if, for instance, the Gulf Stream one year runs with less velocity 

 than another year, its waters may very nearly keep their sali- 

 nity, while their temperature will be more lowered by emission 

 and conduction of heat. It may therefore be that the tempera- 

 tures will often give more information of the changes in the 

 currents than the salinities. 



From the above discussion we may consequently draw the 

 following general conclusion: The Gulf Stream along the Nor- 

 wegian coast and towards Bear Island had in July, August and 

 September, 1900 less velocity than usual, and its waters were 

 relatively cold. On the other hand the Atlantic northward cur- 

 rents, along the east and west coast of Iceland were more 

 developed than usual and were relatively warm. It may thus 

 be that the regular quantity of warm water was carried daily 

 into the Norwegian East-Greenland Basin from the Atlantic, only 

 that it had to some extent taken a more westerly course than 

 usual, and that the East-Greenland Polar Current was somewhat 

 disturbed in its usual course and the East Icelandic Polar Current 

 was perhaps moved more eastward^, at the same time the Polar 

 waters near Bear Island had a wider distribution than usual, 

 A result of this seems to have been that Iceland, or at any rate 

 its west coast, had an unusually favourable climate in the sum- 

 mer of 1900, while northern Norway had an unusually cold cH- 

 mate^. 



^ It is also a noteworthy fact that the ice-conditions were unusually 

 favorable along the East Greenland coast, north of 66° N. Lat. in the 

 summer and autumn of 1900. 



- I have not yet had an opportunity to examine the serial meteorolo- 

 gical observations made in these regions. 



