ANTARCTIC SURFACE WATER 45 



magnitude of these values is correct, phosphate is withdrawn at South Georgia by the 

 plankton to an extent of about 45 per cent of the available content in the surface. The 

 amount withdrawn from the 0-100 m. layer is a little less. 



At Grytviken, South Georgia, during part of 1925-6, a daily surface sample was taken 

 at high tide and analysed for phosphate content. Although such factors as dilution by 

 land drainage, snow water, glacier ice, and varying contamination during the whaling 

 season diminished the value of these estimations, it is considered that the results furnish 

 some indications of the seasonal variation of surface phosphate at one position at South 

 Georgia. The results were plotted as monthly means and are given in graphical form 

 in Fig. 10. 



izo- 



m 

 O 



2 

 O 



80 



60 



* 



-Q MONTHLY AVERAGES MAY i92S • APRIL 1926 



-X EXTRAPOLATED VALUES JAN.UAR Y - AP R I L '9S5 



N FEB MAR APR MAY JUN JUL AUG SEPT OCT NOV DEC , . JAN FEB MAR APR_ 



' \9ziT~ '926 



Fig. io. Monthly averages of phosphate content in the surface water at Grytviken, South Georgia in 1925-6. 



The two surveys of November 1930 and 1933 (Table IV) give an indication of the 

 annual variation of phosphate in the surface layer. If we attempt to compare these 

 November results by reducing the number of stations made in 1930 so that the 

 positions of the stations in the 2 years are more nearly comparable, we find that twenty- 

 three stations in 1930 give an average of 92 mg. for the 0-100 m. layer compared with 

 an average of 109 for twelve stations in 1933. Possible factors which might be responsible 

 for the difference of 17 mg. in these 2 years are difference of average date of the stations 

 involved, meteorological differences, amount of phytoplankton and date of main out- 

 burst in each year, and hydrological differences in the surface layers. The average date 

 the twenty-three stations taken in 1930 which correspond with the positions of the 

 twelve stations of 1933, was 17 November as opposed to 30 November in 1933. In the 

 absence of exact information as to the dates of the main outburst in the 2 years, the fact 

 that the average date of the survey in 1930 was 13 days earlier than in 1933 would tend 

 to argue that the phosphate content of the surface layer should have been greater in 1930 

 than in 1933 because the phosphate content decreases towards the summer. Thus the 

 factor due to the difference of date of the two surveys should have given rise to more 

 phosphate being removed in 1933 by phytoplankton activities, whereas the surface layer 

 was found to contain more phosphate in 1933 than in 1930. In the opinion of Dr Hart 



