I9I2. No. 12. 



THE SEA WEST OF SPITSBERGEN, 



45 



Fig. 35- Curve I: Temperature- Anomaly of 

 Spitsbergen Atlantic Current (scale to the left). 

 Curve II: The Area (in hundred-thousands of 

 square kilometres) of open water in the Barents 

 Sea in May (scale to the right). 



l^OQ BOl 1905 1908 1910 



-/' 



1910 



Considering the very inaccurate 

 methods, and the insufficient obser- 

 vation-material at our disposal, a bet- 

 ter agreement cannot be expected. 

 The variations seem to have gone 

 in the same direction, as is demon- 

 strated by the curves in Fig. 35. 



It might seem probable that the 

 distribution of the ice in the sea west 

 of Spitsbergen is also more or less ^T 

 dependent on the variations in the 

 Spitsbergen Atlantic Current. But in 

 this deep sea it is evidently less de- 

 pendent on the temperature of the 

 water than in the shallow Barents 

 Sea, and is more dependent on the 

 distribution of the currents, and on 

 the prevailing winds. 



We have previously [1909, pp. 204, ei seq.] pointed out that there is 

 probably a close relation between the annual variations in the temperature 

 of the Norwegian Adantic Current and the growth and spawning of the 

 food-fishes; and we found especially that the Lofoten fisheries had corre- 

 sponding variations. When the Atlantic Current had a comparatively high 

 temperature in the Sognefjord-section (northeast of the Færoe-Shetland 

 Channel) in May of one year the fisheries began late at Lofoten in the 

 following winter, and the cod was in a comparatively poor condition, e. g. 

 with small liver; and vice versa if the current was cold. We might 

 consequently expect that there should be a somewhat similar agreement 

 between the annual variations of the temperature of the Spitsbergen At- 

 lantic Current and the variations of the Lofoten fisheries of each prece- 

 ding year. 



The curves of Fig. 36 represent the variations in the temperature of 

 the Spitsbergen Atlantic Current in different years, in the northern region 

 (Curve I) as found above ^, the variations in the mean anomaly of the 

 winter temperature (December — May) a year and a half earlier at the five 

 northern meteorological Stations of Norway (Curve II), and the variations 

 in the quantity of cod-liver in hectolitres per 1000 fish, fished at Lofoten 

 in each previous year (Curve III). 



1 For 1900 and 1901 we had to use values from the southern region. 



