I912. No. 12. THE SEA WEST OF SPITSBERGEN. 4I 



In our memoir on the Norwegian Sea [1909] we put forth as an ex- 

 periment the suposition that the water of the Atlantic Current may spend 

 approximately one year in travelling from the region of our Sognefjord- 

 section to the region of our Lofoten-section. 



Our calculations and conclusions based upon this supposition gave 

 very reasonable results. It seems probable, however, that the greater 

 part of the water of the current may travel somewhat farther than this 

 distance in the period of one year. We do not, however, consider it 

 probable that water passing the Sognefjord-section in one year will on the 

 average reach the region of Isachsen's Sect. VI, or say 79 '' N. Lat., much 

 earlier than two years later, leaving out of consideration the possibly 

 more rapid drift of surface-water. These same slowly-moving water-masses 

 must naturally reach the region of Isachsen's Sections I and II in a shorter 

 period, perhaps nearly half a year less. If it be assumed that they spend 

 two years in travelling from about 62 " N. Lat. (the Sognefjord-section) to 

 79'' N. Lat., it would follow that they would need about 5 months less 

 to reach the region of Isachsen's Sect. I, in 75'' N. Lat. 



We found [1909, pp. 193 et seq.] that the variations in the mean tem- 

 perature of the Atlantic water of the Sognefjord-section in May, each year, 

 was to some extent followed by similar variations in the mean temperature 

 of the atmosphere in Norway in the winters following. If this be correct, 

 and if the variations in the mean temperature of the Spitsbergen Atlantic 

 Current, near 79*^ N. Lat., be approximately similar to those of the tem- 

 perature of the Sognefjord-section two years earlier, we may expect some 

 kind of agreement between these variations of the Spitsbergen Current, 

 and the variations in the winter-temperature of the preceding winters. 



The following table gives the mean anomalies of the temperatures for 

 the period from December ist to May 31st, of various years, for all 22 

 meteorological stations of Norway, and the corresponding anomalies for 

 the five northern stations, Tromso, Andenes, Alten, \'ardo and Sydvar- 

 anger, compared with the possible anomalies of the mean temperature of 

 the Spitsbergen Current in the summer a year and a half later: 



