28o DISCOVERY REPORTS 



Upon phytoplankton production: light, and the distribution of pack-ice. The duration 

 and intensity of the Hght will vary more or less directly with the distance one proceeds 

 to the south, so long as the latitude of the convergence remains fairly constant, since it 

 is of extra-terrestrial origin. The distribution of the pack-ice, on the other hand, can 

 be extremely erratic as climatic conditions fluctuate. Our knowledge of it is now suffi- 

 cient, however, to make the following subdivision, based on the gradient of these two 

 factors, reasonably satisfactory in the open oceans.^ We divide these parts of the Antarctic 

 zone into Northern, Intermediate and Southern Regions. 



The Northern Region extends 330 sea miles south of the Antarctic convergence, all the 

 way round the world, with the exclusion of the special areas to be described later. It is 

 never covered by continuous pack-ice and only invaded by loose pack- and drift-ice in 

 spring on rare occasions. 



The Intermediate Region extends from the southern boundary of the Northern Region 

 to the Antarctic circle — an unavoidably arbitrary boundary. It is largely covered by 

 pack-ice in winter and spring, and mainly free during summer and early autumn. Here 

 again it is necessary to exclude the 'special areas'. 



The Southern Region lies between the Antarctic circle and the Antarctic continent, 

 excluding the immediate coastal areas. It is largely covered by pack-ice throughout the 

 year and free only in summer. New ice frequently forms in March. 



To the south-west of South America and south of New Zealand the Antarctic con- 

 vergence lies far to the south of its mean latitude, and the gradient in the conditions of 

 existence is consequently ' telescoped ' so that three clearly defined regions can no longer 

 be distinguished. Hence the need for separate treatment of these ' special areas ', iiorth 

 of the Ross Sea and the eastern south Pacific. These are oceanic, but cannot be divided 

 into Northern and Intermediate Regions on the same basis as those previously described. 

 To the south of them, however, it appears that no serious anomaly is introduced by re- 

 garding the Ross Sea and Bellingshausen Sea as comparable with the Southern Region. 



To the south and south-east of South America conditions are extremely com- 

 plicated. These are the only localities where considerable land masses and a sharp 

 rise in the sea floor — the Graham Land Peninsula, the Scotia arc with island groups — 

 intrude upon the northern part of the Antarctic zone. The complications clearly exert a 

 profound influence upon the phytoplankton development. For present purposes they 

 may be somewhat loosely summarized as neritic influences, and in the light of our 

 observations it is possible to distinguish further ' special areas ' based partly on latitude 

 but mainly on ' degree of neritic influences'. Chief among them are the South Georgia 

 area and the Scotia Sea. To make the scheme of subdivision complete, one would need 

 to consider as special areas the Bransfield Strait, the central portion of the Weddell 

 Sea, and other areas around isolated islands with local neritic conditions. Little of the 

 work considered here falls in these regions however, so that they may be treated under 

 the general heading of 'other special areas'. It may be noted that conditions around 



1 See Mackintosh and Herdman, Distribution of the Pack-ice in the Southern Ocean, Discovery Repts., xix, 

 pp. 285-96, plates LXix-xcv, published since the above was written. 



