40 CHAMBERLIN— A POSSIBLE REVERSAL [April i8 



A large part of this cold superficial water flows away in surface cur- 

 rents to lower latitudes. 



In view of these complications, the precise mode by which polar 

 agencies control the deep circulation is much less obvious than it 

 might at first seem. There is ground to suspect that the formation 

 and melting of ice is an important factor. In freezing, the salt and 

 gases of the surface layer are largely forced out into the underlying 

 layer. If the surface layer has an average degree of salinity, the 

 underlayer is super-charged, and being also cold, must tend to sink. 

 On the borders of the ice-covered tracts where the precipitation and 

 melting are considerable and where adjacent polar lands pour in 

 much fresh water, the surface layers are so much fresher than the 

 average sea-water that the concentration of salinity by freezing does 

 not overbalance the original freshness. But in those polar regions 

 where there is no inflowage from the land, where precipitation is 

 slight and almost wholly snow, which accumulates on a previously 

 frozen surface and absorbs most of its own summer melting, and 

 where the ice is borne away to lower latitudes and the waters arising 

 from it do not redilute the concentrated waters, it is believed that a 

 sufficient degree of saline concentration, combined with depression 

 of temperature takes place to cause an effective downward move- 

 ment. This is believed to cooperate with diffusion and conduction 

 in giving the lower body of polar waters the superior gravity which 

 actuates the abysmal circulation. The sea immediately bordering 

 Antarctica and that lying northwest of Greenland seem to furnish 

 these conditions. Moss^ and Krogh- independently have found that 

 at times of northwesterly wind, the air west of Greenland contains 

 about double the usual content of carbon dioxide. This I have sug- 

 gested may come from waters overcharged with it by th*e freezing of 

 the overlying layer. 



It is not to be inferred, however, that the deep-sea waters derived 

 from the polar regions exceed in salinity the waters of the evapo- 

 rating tracts of low latitudes, but merely that by this concentration 

 through freezing conjoined with low temperature and modified by 



^ Moss, "Notes on Arctic Air," Proc. Roy. Dublin Soc, Vol. II, 1880. 

 * Krogh, " Abnormal CO, Percentage in the Air of Greenland," etc., Med- 

 delelser om Gronland, Vol. XXVI, 1804, pp. 409-411. 



