3 8 DISCOVERY REPORTS 



more steeply; it is only marked by a change in surface temperature of 1-5° C. in 

 15 miles. 



Farther east, between 140 and 150 W, the isotherms and the Antarctic convergence 

 advance towards the north. The bottom current is strengthened by a large volume of 

 water which is turned northwards at the Cape Adare-Easter Island ridge, and the effect of 

 the ridge and the greater bottom current is to cause the easterly current in the deep and 

 surface layers to be deflected northwards. The greater northward movement at the 

 surface is shown by the northward advance of the isotherms (Fig. 8) and of the limit 

 of pack-ice (Fig. 9). The convergence was found to be very well defined, and there 

 was a sudden increase of 3 C. in 55 35' S ; it must bend still farther towards the north 

 in 140 W because a further tongue of Antarctic water was crossed between 53 ° 43' S 

 and 53 05' S. In the deep basin east of 140 W (see Plate XLIV) the convergence 

 retreats to a very high latitude ; it makes a small bend towards the north near the 

 longitude of Peter 1st Island, but the main trend, as far as 8o° W, is towards the south. 

 In accordance with the general rule, the temperature and salinity distribution indicates 

 that the retreat of the convergence at the surface is associated with a decline of the 

 northward current at the bottom. 



In 120 20' W the position of the convergence is probably indicated by some observa- 

 tions of the Ellsworth Antarctic Expedition (1934-5) which showed that in 59 20' S 

 the temperature of the sea varied as much as 3 within a run of a few miles. 1 The south- 

 ward bend of the convergence to this point from 53 to 54 S in 140 W is very sharp. 

 It is probably caused by a southward movement of the deep and bottom currents; 

 these currents, which are deflected northwards as they approach the Cape Adare-Easter 

 Island ridge, appear to turn back to the south in the deep water farther east. 



Between 120 20' W and no° 12' W the convergence remains in approximately the 

 same latitude ; it was crossed in 59 15' S by section 16 (Plates XXXVI I-XXXIX). East 

 of 1 io° W it once more bends southwards. In section 17 (Plates XL, XLII) the surface 

 temperature record showed a first increase of 2° C. in 61 ° 06' S, 93 10' W, and then 

 after an irregular decrease, a second increase of 3 C. in 6o° S, 90 32' W. These 

 observations suggest that the convergence bends roughly north-east between 93 and 

 90 W and indicate that it may lie as far south as 61-62 S between no and 93 W. 

 East of 93 W there are several indications that the main drift of Antarctic water towards 

 the east is deflected northwards ; a northward set was experienced near Peter 1st Island 

 by the 'Pourquoi Pas?' and the 'Odd I' (see p. 16), and the surface isotherms bend 

 northwards. The bending of the convergence is probably associated with a stronger 

 bottom current, and it gives some indication that the Antarctic shelf may extend 

 farther north in this region. 2 



East of 90 W the convergence recedes once more to the south, until it reaches its 

 farthest south in 8o° W. In this longitude (section 18, Plates XLI, XLII) it was marked 



1 See note on Ellsworth Antarctic Expedition in Polar Record, No. 9, 1935, p. 67. 



2 Such an extension is shown tentatively in the American Geographical Society's Bathymetric Chart 

 of the Antarctic, December 1929, revised November 193 1. 



