DEEP CURRENTS: ATLANTIC OCEAN 93 



existence of the Maud Bank, which is not shown in his sections. Section 6 (Plates VIII, IX) 

 passes over the eastern end of the bank in 65 S, 4 E, and the warm water at St. 1 147 

 probably belongs to a branch of the westward current which flows north of the bank. 



Section 5 (Plates VII, IX) crosses the eastern part of the Weddell Sea, between 21 and 

 23 W. In this region the westward current is clearly defined only south of 67 S ; in the 

 central part of the sea between 61 and 67 S the variations in the temperature, salinity, 

 and oxygen content of the deep layer, and the undulations of the isotherms and isohalines 

 suggest that both the surface and deep currents are weak and variable. Earlier sections 

 made across the same region by Brennecke (1921, pis. 4-6) and Mosby (1934, figs. 14-16) 

 have given the impression that the currents are much more regular ; they have suggested 

 that there is only a steady movement towards the west in the southern part of the sea 

 and towards the east in the northern part, the strength of each current decreasing 

 gradually to a narrow boundary region in the middle of the sea. Both sections were, 

 however, based on a much smaller number of observations than section 5, and it is 

 probable that more data would have revealed the same irregularities that we have found. 



Each of the sections across the region shows how the situation of the central part of 

 the sea in a divergence region between the currents towards the west, farther south, and 

 towards the east, farther north, is marked by the upwelling of bottom water. Our 

 observations further emphasize the fact that the divergence region is wider in the 

 Weddell Sea than it is in the eastern half of the Atlantic Ocean. Such a difference and 

 the existence of the irregular movements in the central part of the Weddell Sea may be 

 caused by the falling away of the coast-line towards the south-west ; the expanding of the 

 westward currents in both the surface and deep layers into a wider space in the Weddell 

 Sea may be supposed to make them less regular and more susceptible to wind changes. 



Brennecke 's observations in the western part of the sea show that the isotherms and 

 isohalines bend towards the north along the east coast of Graham Land and back to- 

 wards the east in the northern part of the sea. The deep current must follow the same 

 path, becoming continuously colder and less saline as it mixes with the surface and 

 bottom waters. The Deutschland observations in the southern part of the Weddell Sea 

 show that in 27 1° W the current had a temperature of as much as 0-79° C, but in 44 W 

 it was only o-66° C, and in Nordenskj old's section across the northward current in 

 about 65 S (1917, pi. 2) not more than 0-4° C. In the northern part of the sea and 

 across the northern part of the Atlantic-Antarctic basin the warm deep layer is also 

 weakly developed ; its maximum temperature over a very extensive region (see Fig. 22) 

 is not more than 0-5° C, and its maximum salinity not more than 34-66-34-69 °/ 00 . 

 At St. 638 (Station List, 1932) between the South Shetland Islands and the South 

 Orkney Islands the maximum temperature and salinity in the deep layer were found to 

 be not more than - 0-19° C. and 34-60 % ; the great weakness of the current in this 

 region seems to be caused by the sinking of surface water in a convergence region be- 

 tween the water flowing northwards out of the Weddell Sea and the more easterly drift 

 farther north (see pp. 108, 109). 



The temperature, salinity, and density or specific volume sections across the region of 



