On the Glaciers and Climate of Iceland. 291 



mentioned observations on the tropical and arctic currents 

 on the one hand, with the isothermal lines on the other, it 

 appears at once that the positions of the isothermal lines are 

 mainly determined by the directions of the currents. The 

 mean temperature of the sea at Reykjavik is about a degree 

 centigrade higher than the mean temperature of the air. 

 Now, in continental regions at the same latitude of 64°, the 

 mean temperature of the air is usually zero, or even less. The 

 Gulf Stream, therefore, produces a mean increase of tempera- 

 ture of four degrees on the air, while the water itself still re- 

 tains a higher temperature than that which the air thus re- 

 ceives. 



It has been already remarked that the Gulf Stream spreads 

 itself out from Feroe towards the Scandinavian coast, and 

 that, from the 63d degree of latitude, it begins to scatter it- 

 self out towards the north ; so that in Finmark its ultimate 

 ramifications alone are perceived. From the paper by L. 

 von Buch, already cited, it appears to be indubitable that the 

 influence of this cuiTent reaches even to the Bear Island, in 

 latitude 75°. South-westerly winds are, in that island, as 

 well as in Iceland, accompanied by mild weather. The 

 months of November and December usually bring rain, but 

 no snow, and the taking of the walrus can be continued even 

 till Christmas. The most severe cold commences in spring, 

 accompanying the approach of the drift-ice from the north- 

 east ; but, even then, the island enjoys the beneficial wann- 

 ing influence of the Atlantic Ocean, and is in climate very 

 difi^erent from the neighbouring icy Spitzbergen. All along 

 the course of the current, so far as to this island, we per- 

 ceive the isothermal lines sharply bent towards the north, 

 whilst they suddenly turn towards the south, at the places 

 where the influence of the current ceases, and the action of 

 the continental climate begins to prevail. 



Finally, it appears that the retrograde polar current pro- 

 duces efffccts on the climate of Jan Mayen, of Northern Ice- 

 land, and of Greenland, in the same way as the Gulf Stream 

 does, but that these eff'ects are of the opposite kind. Under 

 such circumstances, it is rather striking that the mean tem- 

 perature of Akureyre is not still less than it is actually 

 found to be ; and the influence of the tropical heat, in miti- 



