SOME RESULTS OF THE INTERNATIONAL FISHERY INVESTIGATIONS. 443 



I 



of the Gulf Stream eddy were investigated.* The water of 36 and 

 37 salinity touched the Azores in March of that year, but not the 

 coasts of Africa or Europe. Between these coasts and the stream 

 was water of less salinity than 36. In November a great exten- 

 sion of the eddy had taken place, and in tlmt month it had actually 

 touched the coasts of Africa and Southern Europe. With this extension 

 of heavy and relatively warm sub-tropical water had also occurred an 

 extension of the area containing sub-tropical microscopic drifting 

 organisms. In March of the following year the limits of the Gulf 

 Stream eddy had again contracted. 



The Gulf Stream circulation, though it never actually reaches our 

 northern latitudes, thus undergoes a periodic expansion and contraction. 

 Now analogous to these gigantic annual pulsations there occur hydro- 

 graphic events in the seas of Northern Europe. A periodic flooding 

 of the North Sea, the Skagerak, the Norwegian sea, and even the 

 remote Barentz Sea, with water of Atlantic origin, occurs annually in 

 such a manner as to render it an undoubted fact that the oceanic 

 circulation in these regions is dependent on that of the Gulf Stream, 

 and ultimately on the equatorial current. In some way or other a 

 great stream or drift takes origin in the Gulf Stream eddy and invades 

 our northern seas. This is the " European stream." It is sometimes 

 said that it is the result of the propulsion of surface water by the 

 prevailing south-westerly cyclonic storms which reach our latitudes. 

 This may be so, but the cause of the Norwegian stream is more prob- 

 ably a complex thermo-dynamical one. Anyhow, there is a continual 

 drift of relatively warm and dense water from the south-west towards 

 Northern Europe. Just as the Gulf Stream eddy pulsates, so does this 

 drift of water become augmented or contracted. And with these 

 augmentations and contractions of the European stream are correlated 

 changes in the barometric pressure and temperature of the atmosphere, 

 and in the prevailing fisheries of the regions into which it penetrates. 



The chartf (Fig. 2) reproduced on page 444 illustrates the distribution 

 of the European stream in August, 1890. This chart was constructed from 

 observations made prior to the beginning of the International Investiga- 

 tions, and the results obtained since 1902 indicate that the distrilmtion 

 of the stream in 1S9G was rather abnormal. A glance at the chart, 

 however, will illustrate what may perhaps be regarded as the maximum 

 flooding of the European seas by Atlantic water. The stream has 

 invaded the Icelandic coastal regions, and has penetrated into the 

 Denmark strait between Iceland and Greenland. Impinging on the 



* Cleve, Ekman, and Pettcrsson, Variations annuellea (h Vcau dc surface dc Vocean 

 atlnntiqxie. 



t I'eteiiiianii's MitUicilungen, 1000, Ilcft i. u. ii. ; .sec also Rajtpts. et Proc.-rcrh., 

 vol. iii., l^Oft, p. 4. 



