OBSERVATIONS IN THE SCOTIA SEA 319 



the autumnal secondary increase appears to show this irregularity (cp. Harvey et al. 

 1935, p. 439). It would appear to be far more dependent upon prevailing weather con- 

 ditions than the main increase. 



Before leaving the South Georgia area it may be mentioned that in the exceptional 

 spring of 1930-1, when pack-ice actually extended some way to the north-east of the 

 island, Phaeocystis was found in moderate quantity in the ice. It has not been observed 

 there on other occasions, but may be expected in small quantities whenever the pack 

 gets so unusually far north. The Chlorophycean Halosphaera viridis was recorded by 

 Hardy in enormous numbers, but from three stations only and from subsequent work it 

 would seem to be so local that it can hardly be considered a regular constituent of the 

 phytoplankton. 



THE SCOTIA SEA 



Eighty-nine estimations of pigment content are available from this area ; they were 

 obtained in different seasons, but being fairly well distributed over the whole of the 

 productive period appear to give a good idea of the probable seasonal cycle. The 

 relevant figures are given in Table 9, and are also plotted in Fig. 14. It must again be 

 noted that the graph has had to be constructed on a smaller scale than that used for the 

 oceanic Northern Region, but larger than that used for the South Georgia area. 



5000 



4000 — 



3Q0O 



2000 



1000 — 



JuIlj August September October NovemberDecember January Februarq March April MaL| June 



Fig. 14. Scotia Sea. Seasonal variation in plant pigments per m.^, means of available observations at mean 

 dates. Numbers of observations in brackets. Note necessarily smaller scale than that used for oceanic 

 regions. 



8-2 



