AMELIORATIONS OF PRESENT ARCTIC CLIMATES 387 



continent. This ridge lies near the edge of the two continents and 

 not far back from the abyssal portion of the Polar Basin. The 

 bordering flat between the ridge and the basin has about the usual 

 width, depth, and aspect of a continental shelf. The deepest sub- 

 merged notch in this ridge seems to Ue between Spitzbergen and 

 Greenland, but it has not as yet been positively defined by sound- 

 ings. The difference between the abyssal waters of the Norwegian 

 Sea and those of the Polar Basin, however, make the existence of a 

 defining weir of this kind practically certain, and fix the maximum 

 depth of water over it at about 3,000 meters. As well brought out 

 by Nansen and others, these two ridges, acting as weirs, determine 

 the upper boundary of the abyssal waters in the Polar Basin and in 

 the Norwegian Basin, respectively, and through this determination 

 they indirectly fix the depths attained by the currents or layers of 

 water formed by the circulation above the abyssal waters. 



So far as known, the inlets through the American Archipelago 

 into Baffin Bay and Hudson Bay are too shallow to be occupied by 

 any but the upper currents flowing out of the Polar Basin; this 

 statement may, however, be found to need revision. It is wise to 

 bear in mind that canyon-hke channels across barriers like these 

 partially submerged ridges that serve as weirs may easily excape the 

 sounding-lead, especially when soundings are few. And more than 

 that, such canyons may once have existed but were filled afterward 

 and may now be quite indetectable. The shallow Behring Strait 

 plays only a small part relatively in the exchanges between high and 

 low latitude and need not be considered in this discussion. 



One of the notable effects of the Southern Intercontinental 

 Ridge is that it holds the surface of the cold saline abyssal waters 

 of the Polar Basin a thousand meters or more above the surface of 

 similar abyssal water in the mid-Atlantic as now determined. As 

 the less dense layers of water developed above these heaviest 

 abyssal waters yield to them, and must conform to them in both 

 basins, this great difference is a radical feature. It implies that in 

 seeking an interpretation we should proceed on the assumption that 

 the superabyssal water-strata of the mid-Atlantic, taken together, are 

 much thicker than the corresponding layers of the Polar Basin. This 

 greater depth is well established for the tracts of the mid-Atlantic, 



