46 DISCOVERY REPORTS 



the 1933 survey was made immediately after peak conditions of phytoplankton concen- 

 tration, whilst the comparable stations in 1930 were made at the beginning of the 

 upward gradient of phytoplankton growth. Phytoplankton requirements therefore would 

 have removed less phosphate at the time of the survey in 1930 than in 1933, i.e. the 

 residual phosphate should have been greater in 1930 than in 1933. 



The average wind velocity for the dates of the 1930 stations was 18 knots, and 

 for those in 1933 was 12 knots; similarly, the sky was clearer during the 1933 

 observations than during those of 1930. In general, clear, calm conditions favour 

 phytoplankton growth, and therefore the fact that the survey of 1930 was made in more 

 overcast weather and at a time when more turbulence was apparent than in 1933 would 

 tend to favour a greater phytoplankton concentration in 1933 than in 1930; this would 

 involve a smaller residual phosphate content in the surface layer in 1933, which is 

 exactly opposite to what was found. Hydrological conditions were very different in 

 November 1930 from those in 1933. InNovember 1930 vast masses of heavy pack-ice were 

 close to the northern coast of South Georgia, whilst there was no pack-ice visible at the 

 time of the 1933 survey. The presence of pack-ice around South Georgia in early spring 

 is due to a current from the Weddell Sea. Although we have no figures at present for 

 the transport of water out of the Weddell Sea, it is possible that in a year when large 

 masses of ice are found at South Georgia, the Weddell Sea influence predominates over 

 that of the Bellingshausen Sea which is the second source of surface water in this 

 vicinity. At South Georgia, water of Weddell Sea origin appears in spring to have a 

 higher phosphate content than that of Bellingshausen Sea origin. Thus if a greater 

 amount of Weddell water were present in 1930 it would again have been expected that 

 more phosphate would have been present in 1930 than in 1933. The presence of such 

 a large quantity of pack-ice in 1930 lowered the average temperature of the surface layer. 

 The average temperature of the 0-100 m. layer in the two Novembers was as follows: 



November 1930, 49 stations, —0-67° C. 

 November 1930, 23 stations, — o-68° C. 

 November 1933, 12 stations, 0-50° C. 



Thus even if all the stations of 1930 or only those comparable in position with the 

 1933 stations are considered there is a difference of nearly 1-20° C. in the mean 

 integral average of the o- 100 m. layer in the 2 years. It is not known whether phyto- 

 plankton has an optimum temperature for the beginning and duration of the period 

 of increase in concentration. It is possible, however, that such a relatively large differ- 

 ence in temperature of the surface layer in these 2 years is likely to have had a significance, 

 the import of which is at present unknown. 



The above evidence therefore does not explain the presence of more phosphate in 

 the surface layer around South Georgia in November 1933 than there was in the same 

 month of 1930. It is clear that the phosphate content of the surface layer varies very 

 considerably from year to year, and the difference between the results of 1930 and 1933 

 must be due to some annual variation. Until we have much more data, any further 

 discussion on this point must be postponed. 



