PHYTOPLANKTON OF THE WEDDELL SEA AREA 89 



At St. 359 a rich phytoplankton was encountered, with the spineless chain form of 

 Corethron valdiviae strongly dominant. Other forms estimated to run into millions in 

 the net haul were: Fragilaria antarctica, Thalassiothrix antarctica and Nitzschia seriata. 

 St. 342, the next in the table, was the outermost station on the Cooper Island line of 

 the South Georgia survey. Here Corethron valdiviae (spineless chains) and Thalas- 

 siothrix antarctica were co-dominant in a moderately rich phytoplankton. The com- 

 paratively high temperature of the upper layers at this station, together with the low 

 salinity and phosphate content, indicate that old Bellingshausen Sea water from the 

 westward may, by mixing with the old water of eastern Weddell Sea origin, have been 

 responsible for the localization of the moderately rich phytoplankton to the east of 

 South Georgia. 



A similar phytoplankton was encountered at St. 360, with Corethron dominating over 

 Thalassiothrix, and very few other organisms present ; but at the next station to the east- 

 ward, St. 361, the phytoplankton became extremely poor, and it was evident that the 

 limit of the s^mdtss-Corethron and Thalassiothrix belt had been reached. At the 

 succeeding station it was even poorer, though the falling temperature indicated that 

 more typically Weddell Sea conditions were being approached. Here the normal spiny 

 form of Corethron valdiviae was encountered for the first time among the hauls taken 

 on board the ' Discovery H ', though in very small numbers. 



During the running survey of the South Sandwich Islands it was only found possible 

 to make phytoplankton collections from three full stations. The material obtained, how- 

 ever, was very interesting, as it showed a phytoplankton which may possibly be taken 

 as typical of old surface water of the eastern Weddell Sea type in late summer. It may 

 be that during this mild open season the remnants of a formerly rich phytoplankton 

 hung round the immediate neighbourhood of the islands, like the numerous small bergs 

 that were encountered, as the catches obtained in the open sea to the westward on the long 

 line previously described (Table XIX) were very much poorer. This South Sandwich 

 phytoplankton bore a close general resemblance to that normally found in the tongue of 

 eastern Weddell Sea water to the east of South Georgia in the latter half of the season, 

 in strong contrast to that actually present in the latter region on this occasion. The 

 analyses of the net hauls made at the South Sandwich Islands are contained in Table 

 XXV, where they are arranged in north to south order. 



It will be seen that at St. 365 between Visokoi Island and the Candlemas Islands small 

 Chaetocerids, notably Chaetoceros dichaeta, were numerically dominant, but the most 

 remarkable feature of the catch was the presence of Coscinodisciis boiivet in large numbers. 

 This form was also taken in the larger closing nets down to the unusual depth of 750- 

 500 m. Similar findings of this large heavily silicified form at great depths have already 

 been described by Hardy (in press), who noted that as the depth increased the number 

 of " living " individuals, as estimated by taking the percentage of the frustules containing 

 chloroplasts, decreased. This observation led to the formulation of the important theory 

 that the diatom population of Antarctic surface waters might be kept up by the return 

 of resting spores in the upper layers of the warm deep water. The formation of minute 



