3 8 DISCOVERY REPORTS 



If we consider both the southern and northern curves we find that minima exist as 

 follows : northern region i January, and southern region i November. We know too that 

 in the northern region in November the phytoplankton increases enormously in con- 

 centration. It is obvious that this large increase exerts a tremendous effect on the con- 

 centration of nutrient salts in situ, and it is reasonable to ascribe the minimum in the 

 northern curve at the end of December to this cause. Longitudinal sections of salinity 

 show that the Antarctic surface water has a strong northerly component of movement 

 as it sinks at the Antarctic convergence into sub-Antarctic water. When vertical sections 

 of silicate content are plotted, it is seen that silicate leaves the Antarctic zone in two ways ; 

 the first is in the Antarctic surface current, which sinks at the convergence and 

 ultimately forms part of the Antarctic intermediate water, and the second, the major 

 loss, is in the Antarctic bottom water. Unlike phosphate, which has a maximum concen- 

 tration at about 38°-43° S in the South Atlantic Ocean for example, silicate is found in 

 greatest quantity inside the Antarctic zone. The silicate that leaves the Antarctic in the 

 surface current travels a long way north before it is returned in the upper layers of the 

 warm deep water, whilst the silicate that travels northwards in the Antarctic bottom water 

 begins to return in the lower layers of the warm deep water at a much higher latitude. 

 In the northern region, as we have seen, a sudden and large concentration of phyto- 

 plankton is found in November, which diminishes rapidly during the second half of 

 December. The water which leaves the Antarctic zone and eventually becomes part 

 of the Antarctic intermediate current will thus have a silicate content which varies 

 seasonally, i.e. after the main phytoplankton outburst the surface water contains less 

 nutrient salts, whilst in winter the sinking Antarctic surface water has a high nutrient 

 salt content. It might be argued that after the main outburst, silicate in the form of 

 diatoms also leaves the Antarctic, and therefore the sum total of silicate leaving is the 

 same at all seasons of the year. But this is not correct. Phytoplankton is consumed by 

 zooplankton and mortality occurs from other causes. Consequently a rain of the silicate 

 skeletons of diatoms takes place; this rain will vary in amount depending on the 

 concentration of the plankton, so that a seasonal effect will be found in the sinking 

 silicate particles. The Antarctic intermediate current will have a silicate content which 

 depends on the time of the year when the Antarctic surface water sinks at the convergence. 

 Also, during the rain of silicate skeletons from the surface layer some will pass into the 

 south-going warm deep water and others will sink into the Antarctic bottom current 

 to refresh this layer and to help form the diatom ooze. Re-solution of silicate must take 

 place in all three currents, and since a silicate cycle prevails, the amount of this salt 

 which returns to the Antarctic will vary in different months of the year. It is therefore 

 permissible to argue that a minimum in the silicate content of the surface layer in the 

 northern region at the end of December will be reflected in the silicate content in the 

 southern region, which as we have seen has a minimum at i November. It has been 

 stated earlier in this report that the minimum at the beginning of November in the 

 southern curve bears no relation to the production of phytoplankton in situ, and it is 

 thus possible that this minimum is a reflection of the northern region minimum at 



