AMELIORATIONS OF PRESENT ARCTIC CLIMATES 391 



Drift," if that term is understood to include all the warm waters 

 that flow north over the intercontinental ridge, whether in strictly 

 surficial or deeper levels, but the middle layer of the Polar Basin 

 can scarcely be assigned to the Gulf Stream if that term is confined, 

 as it apparently should be, to such surface waters as were scarcely 

 normal in saHnity when they left the Gulf of Mexico and afterward 

 traveled 3,000 miles in zones where precipitation is generally greater 

 than evaporation. These considerations will take on more force 

 when compared with the middle Atlantic layer discussed later. 

 3. The source of the bottom layer of the Polar Sea. — The thick 

 abysmal mass of cold saline water which forms the bottom layer in 

 the Polar Basin is assigned by Nansen to the cooHng of the middle 

 layer by contact with colder water. That the bottom layer is 

 derived in some way from the waters above seems quite beyond 

 question, for the Southern Intercontinental Ridge cuts off the cold 

 abysmal waters of the Atlantic and there seems no other source for 

 the low temperature than the surface effects of the Polar climate. 

 But the view that the middle warm layer was cooled to the requisite 

 degree by any form of simple contact with the upper waters seems 

 incompatible with the observed salinities. The mean salinity of the 

 abysmal mass is given by Nansen as 35.29 parts per thousand. 

 The mean saHnity of the warm middle layer is slightly less than this, 

 and the salinity of the cold waters of the upper layer is notably less 

 than this. Any simple commingling of the two upper layers should 

 give too low sahnity to the lowest layer. Nansen holds with good 

 reason that the bottom waters should be heated rather than cooled 

 from below, and gives observational evidence in support of this. 

 He also holds that the lower part of the upper layer, which is colder 

 and more saHne than the uppermost portion, arises from freezing 

 at the surface. Freezing, as is well known, forces out of the forming 

 ice the larger part of the salts and gases previously held in the con- 

 geahng water. This salt added to that in the layer of water next 

 below gives it higher saHnity. Sometimes it rises to the quaHty Of 

 a brine. At the same time it gives coldness, which is sometimes 

 intense. Nansen, however, is inclined to limit these effects of freez- 

 ing to his second layer — ^that is, to the lower part of what is here 

 called the upper layer. He thinks that a small and slowly formed 



