390 T. C. CHAM BERLIN 



greater than precipitation, and the sea waters suffer increase of 

 saHnity. The distribution of these areas of dilution and increase 

 of sahnity are subject to much modification that must be specially 

 considered in any given case. The real cause in the main is the 

 ascent or descent of the atmosphere. 



1 . The source of the upper layer. — There is practically no ground 

 for doubt that the upper layer of water of the Polar Basin is derived 

 from the upper layer of the Atlantic waters with the addition of 

 the excess of precipitation of the region, the inflow from adjacent 

 lands, and the melting of the ice. As the mean dilution of standard 

 sea-water caused by these fresh waters is of the order of 1 5 per cent, 

 it is clear that the main source of salinity is in the Atlantic waters, 

 whose salinity reaches and in some parts passes the mean salinity 

 of the ocean. 



2. The source of the middle layer. — There is no doubt that this 

 layer also comes over the intercontinental ridges from the Atlantic, 

 for that is the only assignable source for its combined salinity and 

 warmth, but there is a rather vital question as to the particular 

 part of the Atlantic column from which it comes. This layer is 

 described by most writers as though it had occupied a surface posi- 

 tion in the Atlantic, as also when it passed over the Southern Inter- 

 continental Ridge, but that it became cooled and sank to its present 

 middle position later. If by this it is intended to affirm that these 

 middle waters of the Polar Basin are those that formed the '' Gulf 

 Stream," or strictly surficial waters of the Atlantic, there appear to 

 be two objections to it: {a) the salinity of the middle layer in the 

 Polar Basin (35.1 to 35.3 per mille — Nansen) seems too high to have 

 suffered surface dilution while it was drifting 3 ,000 miles into a zone 

 in which precipitation is greater than evaporation; and {h) if the 

 Gulf Stream — -or the surface layer under any other name — sinks 

 to form this middle layer of the Polar Basin, what source remains 

 for the upper layer in the Polar Basin ? It seems more consistent 

 with all the considerations that weigh in the case to regard this 

 middle polar layer as a northward extension of the similar middle 

 layer of the Atlantic which will be described later. In a loose sense, 

 this middle polar layer mayl)e regarded as a part of the "Atlantic 



