CH. Il] THE OCEANOGRAPHY OF THE NORTH-WESTERN OCEAN 51 



1902 we will see how hopeless it is at the present time to attempt 

 to describe the nature and extent of these perturbations, or un- 

 periodic variations, much less to explain their cause by reference 

 to general laws. For this purpose the accumulated observations 

 of many years will be necessary. What we do know at the present 

 time only gives us elusive glimpses of the working of great causes, 

 probably cosmic in their nature. The maximum flow of Atlantic 

 water into the Norwegian Sea ought to take place at the same 

 time every year if we had to deal only with such a variation 

 in the extent of the Gulf Stream eddy as would be caused only by 

 the revolution of the earth and the inclination of the axis of the 

 latter to the plane of the orbit. But we know that the time of 

 maximum flow of the European Stream through the Faeroe 

 Channel may vary by a month or more. Thus the study of the 

 temperature variations of the sea at the Faeroe Islands shows 

 that the Atlantic flow culminated in the year 1902 in September 

 instead of August, which is the month in which this should 

 normally take place. There is in fact no absolute fixity in the 

 time of appearance of the height of the European Stream. Tem- 

 perature observations taken on the coast of Norway show that 

 there is a two-yearly period in the variation of the temperature of 

 the sea. In the even years the temperature is above the normal, 

 while in the odd years it is below the normal. These variations 

 can only be due to the fact that the flow of comparatively warm 

 water into this area from the Atlantic is less in the odd years 

 than in the even ; that is, the Atlantic Stream has a two-yearly 

 periodicity. What is the cause of this ? Another striking fact 

 is mentioned by Pettersson in the paper quoted. It is now beyond 

 question that the migration of the herring into the various fishing 

 grounds of the North Sea depends on the presence of a layer of 

 water of a certain density. That is, herrings are only present 

 in the sea in sufficient numbers to form a fishery if the hydro- 

 graphical condition of the water is favourable to them. But the 

 temperature and salinity of the sea depend mainly on the strength 

 of flow of the European Stream, in our latitudes of course. Now 

 since the year 895 the winter herring fishery in the Skagerak has 

 been recorded, and it is found that it has returned at intervals of, 

 on the whole. 111 j^ears. This is one of those isolated observations 



4—2 



