ANTARCTIC OCEANOGRAPHY 



275 



of the Challenger and the Gazelle, hitherto not thus interpreted, have 

 been able to prove that a subsurface current from the North Atlantic 

 subtropics carries warm water to the south. This current we may 

 take to be a feeder of the Antarctic intermediate water and may 

 assume a similar origin in the other oceans, as its occurrence is the 



* Scuftditgs by the Vauss' 



• Other soundings 



Fig. 4 — Bathymetric map of the southernmost Indian Ocean, north of the Antarctic Continent. 

 Scale, 1:44,000,000. Depths in meters. The map is also intended to illustrate the probable separation 

 of the southward-flowing tropical water by the Kerguelen-Gaussberg submarine ridge, the observations 

 of the Gauss expedition having shown that the temperatures of this layer on the western side of the 

 ridge were higher than on the eastern side. The vertical temperature series at the four stations occupied 

 (indicated on the map by soundings 3934 and 3397 on the west and 370S and 31 18 on the east) are 

 published in the second paper cited in footnote 10, from which also the present map is taken. 



same throughout the whole circuit of the Antarctic. I should like to 

 designate this intermediate water "tropical water," because it comes 

 from the tropics or subtropics. 



Origin and Movement of the Polar Water 



This warmer water in the higher southern latitudes lies between 

 two colder layers. This constitutes the mesothermy referred to, 

 which is partly connected with mesosalinity, as the salt content of 

 the warm intermediate water is higher than that of the water above and 

 in places higher than that of the water below. The water above is of 

 Antarctic origin, as could be observed in the Gaussberg region as well 

 as in Weddell Sea. It originates when the walls of the inland ice 

 disintegrate in the flat shelf sea in front of it, and throughout its 

 whole depth it has nearly uniform salinity and temperature, both 

 consideiably lower than those of the intermediate water farther north. 

 For the former we found in the Gaussberg region t = - 1.85°, 5 = 

 34.4/00, and both practically constant down to a depth of 400 meters. 

 Similar conditions have been reported by W. Brennecke, O. Nor- 



