i8 95 . THE DISTRIBUTION OF FOOD-FISHES. 33 



under-layers consequently greater. The cause of this was probably 

 the great excess of surface temperature over the average in the north- 

 eastern part of the Atlantic basin during the summer of 1893, an d the 

 consequent high level of the sea surface in that region. It may be 

 supposed that the surface current in the Faeroe-Shetland Channel is, 

 in winter at least, chiefly controlled by the winds ; but when winds 

 are weaker than usual, differences of temperature must have increased 

 influence in maintaining currents. 



The stronger the current over the Wyville Thomson ridge, the 

 greater the undertow of cold bottom water. A compensating arrange- 

 ment is accordingly set up, such that the Atlantic stream is always 

 cooled by mixture with the underlying water, and tends to sink 

 below the surface, losing its horizontal motion at the same time. 

 Hence, except when strong south-westerly winds tend to continue the 

 bulk of the current north-eastwards into the Norwegian Sea, there to 

 be absorbed into the general cyclonic circulation of that basin, a 

 mass of Atlantic water collects at the northern end of the Faeroe- 

 Shetland Channel ; and the influence of the earth's rotation, deflecting 

 the current to the right, will make it bank up against the continental 

 plateau to the north-west of the Shetland Islands. We have thus a 

 body of water, partly derived from the Atlantic, and partly from the 

 Norwegian Sea, seeking admission to the German Ocean ; and the 

 volume of this must vary from season to season and from year to year. 

 Further, the relative proportions of Atlantic and Norwegian Sea 

 water composing it must vary ; and since the surface temperatures of 

 the two basins differ widely, the amount of air contained must be 

 different at different seasons and in different years, being smaller the 

 greater the proportion of Atlantic water. 



We may next discuss the mode in which this water is admitted 

 to the North Sea. The "Jackal " observations showed that between the 

 Orkney and Shetland islands tidal influence reigns supreme. It is 

 known that in that region strong tidal currents set north-west and 

 south-east ; and it was accordingly found that water from the Faeroe- 

 Shetland Channel is, as it were, pumped into the North Sea, in 

 a south-easterly direction. This -• probably occurs at all seasons, 

 but the volume of water actually introduced cannot be great, inas- 

 much as it is only the amount substituted by mixture for water 

 already in the North Sea which is withdrawn in its place. Taking 

 next the water introduced by surface currents, we know that the 

 prevailing winds over the whole region are westerly and south-westerly. 

 When the wind influence is strongest, during winter, the tendency 

 will be to drive the surface water either north-eastwards towards the 

 Norwegian coast, out of the North Sea altogether, or eastwards 

 towards the Continental side ; there will be little or no current down 

 the western or British side. In summer again, winds are light and 

 variable, and at the same time the surface of the North Sea is 

 warmer than that of the Faeroe-Shetland Channel, owing to the 



D 



