RECENT HYDROBIOLOGICAL INVESTIGATIONS 485 



that the periodicity is also the same. But the dates of 

 occurrence of the maxima and minima differ. The maximum 

 at Sognefjord occurred in 1902, at Lofoten in 1903, and at 

 the Barentz Sea stations in 1904. The conclusion is that the 

 water passing the more southerly of the three lines — the 

 Sognefjord one — in 1902 traversed the Lofoten line in 1903, 

 and arrived in the Barentz Sea in 1904. Thus the observations 

 give us a rough measure of the velocity of the Gulf Stream 

 drift in these latitudes. It is impossible from the existing 

 data to speak more precisely than this. The observations 

 are made only at intervals of three months, and determinations 

 which are strictly comparable with each other are available 

 for the period of 1902-8. It is also difficult at the present 

 time to allow accurately for the influence of the Polar Stream. 

 One must not expect that an exact value for the velocity can 

 be obtained, nor that the correspondence in the case of the 

 three curves would be very minute. Nevertheless, it seems 

 to be close enough to justify the conclusions. 



Climate and Hydrography 



The area of ice-free surface in the Barentz Sea is shown 

 graphically in fig. 3. The estimations have been made by 

 Helland-Hansen and Nansen from planimetric measurements 

 on the charts published by the Danish Meteorological Office, 

 for the years 1900-8. The correlation between the curves 

 of sea-temperature and extent of sea covered by ice is 

 extraordinarily good. This is as might be expected, for the 

 amount of ice in the sea must obviously vary as the 

 temperature, allowing, of course, for disturbances due to a 

 variable drift from the Polar seas. But, admitting the incom- 

 pleteness of the data, it seems also that there is a close 

 correspondence between the quantity of ice in the Barentz 

 Sea and the amount of heat in the water of the Gulf Stream 

 drift in the latitudes of Southern and Middle Norway two 

 and three years previously. That is, a strong Gulf Stream 

 flow corresponds to a minimum ice year in the Barentz Sea, 

 and vice versa. Not only so, but it can also be shown that 

 there is an intimate correspondence between the volume of 

 the Gulf Stream and the air-temperature in Norway. Fig. 4 

 is taken from Helland-Hansen's and Nansen's memoir (p. 345) 



