AMELIORATIONS OF PRESENT ARCTIC CLIMATES 403 



the northern regions to account for the fairly close approximation of 

 the middle and bottom layers of even the Polar Basin to the mean 

 salinity of the whole ocean. The concentration of salts by excess 

 of evaporation over precipitation is the sole dependence of any 

 moment for offsetting the dilution of the tracts where precipitation 

 is greater than evaporation. The high latitudes belong to this 

 latter class of areas, and adequate saline waters must flow into these 

 regions to maintain the observed salinity of their waters. 



As the weir formed by the Southern Intercontinental Ridge, 

 though not yet fully explored, has a water depth of about 600 meters 

 at least, there is room for a warm, saline middle current as well as 

 the somewhat diluted and cooled surficial current — the Gulf Stream 

 factor — to flow over it poleward. Both classes of water are recog- 

 nizable in the Norwegian Sea beyond. The only point of present 

 divergence from current opinion — if indeed there is really any at 

 all — Hes in the assignment of the partially diluted and somewhat 

 cooled waters flowing over the weir northward to the Gulf Stream 

 group, and in the assignment of the more saline and warmer class 

 of waters going over the weir to the middle layer which better pre- 

 serves its warmth and salinity because it flows under cover. 



The bottom Atlantic layer. — ^ Where the Mediterranean layer 

 of warm sahne water pours out in delta form into the Atlantic, it 

 forms a well-defined middle layer. This gives place below at about 

 2,000 meters' depth to a still heavier layer which is both saline and 

 cold. This fills the abyssal depths of the whole Atlantic Basin, and 

 similar waters occupy the depths of all the oceans. These abyssal 

 waters of the Atlantic are of the same type as the bottom waters 

 of the Polar Basin. They are thought to be derived from them in 

 part at least, for this seems to be the only adequate line of outflow, 

 for the abyssal waters of the Polar Basin and, as noted before, their 

 nature seems to show that they are a part of a persistent and effec- 

 tive circulation. The abyssal waters of the North Atlantic seem, 

 as present data stand, to be more saline and heavier than those of 

 the more southerly oceanic basins. They should hence tend to 

 flow outward toward these in part also; this would tend, in turn, 

 to offset the surficial waters shunted into the North Atlantic by 

 Cape St. Roque. This interchange would help toward the equaliza- 



