POSSIBILITIES OF REVERSAL OF DEEP-SEA CIRCULATION 435 



As one of the particularly interesting problems connected with glacia- 

 tion is the question as to what agency or agencies brought the several 

 Glacial epochs to a close, it is worth while to consider briefly the data 

 bearing on the question of reversal of deep-sea circulation. Such a re- 

 versal may never have occurred, and it is at least very doubtful if it was 

 the chief cause for the retreat of the glaciers. However, the possibilities 

 should not be igiiored. The close balance between the forces maintain- 

 ing the present type of circulation and those tending to produce a re- 

 versal is therefore worth pointing out. 



At present the average salinity of the ocean is 3.5 per cent, giving a 

 density of about 1.026, fresh water being 1.000. The most saline sam- 

 ples of the open ocean taken by the Challenger Expedition had a salinity 

 of 3.7-i per cent, giving a density of about 1.028. The somewhat less salty 

 surface waters of the open ocean in high latitudes are reported to have a 

 density of 1.024. Thus the observed range in density due to salinity is 

 -OO-t in favor of the tropics. A change in temperature from 4 degrees 

 ceiitigrade, the temperature of greatest density of fresh water, to 30 de- 

 grees centigrade (86 degrees Fahrenheit) produces a change of density of 

 exactly the same amount, .004, according to the "Smithsonian Physical 

 Tables." Hence the most extreme cooling now experienced has no more 

 influence on density than the observed difference in density due to 

 salinity. 



If the contrast in density between low and high latitudes were in- 

 creased slightly, the tropical waters would become distinctly more dense 

 than the most dense subpolar waters, and henc3 a rex'ersal would be 

 probable. Such an increased contrast might result either if tr()])ieal 

 waters were more dense than now, as they would l)e if the average salinity 

 were increased, or if the rate of tropical evaporation were increased. It 

 would also result if the subpolar surface waters were less dense than now, 

 as they would be if subpolar regions were not so cold as now or if there 

 were more precipitation in high latitudes. 



The prolmble increased steadiness and strength of the southwesterly 

 winds, suggested by Huntington*^ as probable during times when cyclonic 

 storms are few, would liave warmed high latitudes and would also have 

 reduced evaporation, tbereby increasing fogginess and cloudiness. Pre- 

 cipitation would probably increase with cloudiness and fog. 



If at any time in the past the deep-sea circulation was of the reversed 

 type during a long period, the presence of magnolias, cycads, and sequoia 

 found fossilized in nortliern Greeidaiid in Miocene rocks, the tree-ferns 



lOHswoi-di Iliintinf4(<m : Cliniatic chaiis^cs ( in pi-css). 



