THE PLEISTOCENE OR GLACIAL PERIOD. 439 



Only a selected portion of this complex process can be further 

 discussed here. The factor that most probably controlled the periods 

 of the glacial oscillations, as it seems to us now, was the reversal in 

 the carbonation of the ocean, and this seems to have bearings of value 

 beyond this immediate problem. 



Let the climatic conditions of the Tertiary period, when figs and 

 magnolias grew in Greenland, be taken as the point of departure. At 

 that time, as apparently at all times, the evaporation and the abso- 

 lute humidity of the air in the low latitudes was greater than in the 

 high latitudes. The general circulation of the atmosphere between 

 the equatorial and polar regions resulted in a loss of humidity in the 

 latter regions, and a gain to the ocean, whose surface was slightly 

 raised and freshened. This gave rise to superficial currents toward 

 the warm zone, to restore the equilibrium. These were gradient cur- 

 rents, for the added waters, though cold, were lighter than the ocean 

 brines. 1 



There was inevitably some mixing of the fresh and salt water, and 

 some of the latter was also carried toward the warm latitudes. In the 

 warm dry latitudes, the excess of evaporation gave rise to increased 

 salinity and density, and the denser salt waters are assumed to 

 have sunk and spread poleward, constituting a counter-current to 

 balance the salt-water element of the equatorward currents. The 

 fresh-water element of the surface circulation had its counterpart 

 in the atmospheric circulation. The flow initiated in the evaporating 

 zone was a density current, due to salinity, notwithstanding its superior 

 warmth. This warm dense water, descending and flowing poleward, 

 must at length have been forced to the surface in high latitudes, and 

 contributed its warmth to them. This is assigned as one reason for 

 the warm temperatures of the high latitudes in those periods when 

 this kind of deep-sea circulation prevailed. 



The validity of this conception of the deep-sea circulation in such 

 periods is based on the conviction that superior evaporation hi the 

 low latitudes was more efficient in inducing high density, than the 

 inferior temperatures in the high latitudes. That this was at least 



1 It is to be borne in mind throughout this discussion that an increase of salinity- 

 is likely to be more effective in increasing density than is a lessening of the tem- 

 perature. Because of the peculiar behavior of water near the freezing-point, the 

 specific gravity at the freezing-point of salt-water is about the same as at 12° C. 



