220 



L. H. N. Cooper 



40* 20* 0* 



Fig. 3. Chart showing position of Marsden squares 217 and 218. 



7f— 



4 - 



3^, 



218 A. 



.. 217 B 



Fig. 4. Decadal mean air temperatures for Marsden 

 square areas 217A and B and 218A and B for the 

 season December to March (after P. R. Brown). 



1|° colder than the three decades 1920-49. That is on a decadal basis the years 1910- 

 19 were more suitable for the formation of oceanic bottom water than the thirty years 

 since. It is probable that in these decadal means the coldness of the years 1920 and 

 1921 was swamped by the eight following warm years. Nothing comparable was 

 experienced for the next thirty years. In these years we had precisely the conditions 

 which we need to explain the fluctuations in the English Channel. In each winter an 

 amount greater than average sank in the Greenland area. In the first winter the excess 

 sinking would have displaced the strata of water upwards by, perhaps, a few tens of 

 metres. This would not have brought much nutrient rich water within reach of the 

 processes of vertical mixing of surface waters which are always operative. The next 

 winter would have been more effective and the third winter most effective of all. By 

 this time the nutrient table would have been lifted by perhaps 100 metres or more, to 

 a level at which processes of vertical mixing would readily and continually draw upon 

 the upwardly displaced water to enrich the surface waters. The deep waters of the 

 ocean move slowly, so that there may be a considerable time lag before events in 

 polar regions come to influence upward displacement in temperate and tropic 

 latitudes. 



