i84 DISCOVERY REPORTS 



New data may suggest a modification of the loop in 30° E, but some distortion of the convergence and 

 isotherms here is probably normal. Deacon (1937, pp. 34 and 92, and plate xliv) refers to a steep ridge 

 on the sea floor at this point, which may affect conditions at the surface. The slight northward bends 

 shown about 140, 97, 75, 67 and 35° W and 110° E need confirmation, and the whole of the con- 

 vergence between 1 10 and 160° W, and between 50 and 90° E is based on very little material. 



Fig. 2. Positions of the convergence between 10° W and 80° W. Where two or more positions were ascertained within a short 

 period they are joined by a continuous line if the convergence was well defined, and a pecked line if the positions were approxi- 

 mate or uncertain. Single positions are indicated by a short line, continuous or pecked, drawn approximately parallel to the 

 mean line of the convergence. 



Fig. 3. Positions of the convergence between 10° W arid 40° E. See legend to Fig. 2. 



It is noteworthy that the convergence divides the Southern Ocean into two almost equal zones; 

 in fact it lies very nearly half way between the Antarctic coastline and the extremities of each of the 

 southern continents. From Cape Agulhas (South Africa) the distance to the convergence is rather 

 less than half the distance to the nearest part of the Antarctic continent, but from Western Australia, 

 Tasmania, New Zealand and Cape Horn the convergence is very close to the half-way point. It is 

 difficult to say whether this is of any special significance. The fact that the convergence lies in about 



