98 DISCOVERY REPORTS 



It will be seen that these stations lend themselves to consideration in pairs, both from 

 their positions and the nature of the phytoplankton observed. At St. WS 538, worked 

 at the edge of the pack-ice some 60 miles east of Candlemas Island, Corethron valdiviae 

 was the dominant species, with lesser numbers of Chaetoceros criophilum, Ch. neglectus 

 and Rhizosolenia styliformis, the phytoplankton agreeing closely with that already de- 

 scribed from St. WS 536, between Zavodovski and Visokoi Islands. At St. WS 559, 

 worked some 15 miles to the south of this position twelve days later, Corethron valdiviae 

 constituted almost the whole of the catch, and thus the phytoplankton was evidently of 

 the same nature as that found by the ' Discovery II ' slightly farther south a fortnight 

 later. Evidently a zone of phytoplankton characterized by the dominance of this species 

 invested the more northerly islands of the South Sandwich group during this second 

 season. 



The next two stations, Sts. WS 539 and WS 540, were worked some distance to the 

 south-east from St. WS 538, and at the edge of the pack-ice, just before the ' William 

 Scoresby' succeeded in working round it to the east. This proximity to the ice is re- 

 flected in the low salinities and surface temperatures recorded. The phytoplankton 

 hauls were richer as regards numbers than the two described above, and again charac- 

 terized by the pronounced dominance of a single large species, in this case Chaetoceros 

 criophilum. From Table 17 it will be seen that at both these stations the phytoplankton 

 was almost of identically the same constitution, as would be expected from their close 

 proximity, but the haul obtained at the first was nearly twice as big as at the second 

 station, yet another instance of the great variability of the plankton in nature, as the 

 difference is too large to be due to experimental error alone. Besides the dominant form, 

 Rhizosolenia styliformis and Corethron valdiviae were the only species occurring in 

 large numbers at these two stations. 



Having rounded the pack-ice to the east, at St. WS 541 a haul was obtained which 

 showed yet another great change in the constitution of the phytoplankton, though the 

 number of individual diatom frustules recorded was nearly the same as at the preceding 

 station. Here Rhizosolenia styliformis was strongly dominant, although Chaetoceros 

 criophilum and Corethron valdiviae persisted in moderate numbers, while at the last 

 station of this set of tables, worked some 50 miles to the southward, the change was 

 complete, Rhizosolenia styliformis forming the bulk of an extremely rich catch in which 

 the only other form of note was Thalassiosira antarctica in the small gelatinous colonial 

 form. The occurrence of two of the rarer forms at this station is also of interest, namely 

 Rhizosolenia bidens and Thalassiothrix antarctica. The former because although large 

 and easily recognized it has been found to be extremely sporadic in its occurrence, and 

 the latter because it was not found farther to the southward during this piece of work. 

 Each of these species formed roughly 2-1 per cent of the total phytoplankton at 

 St. WS 542. 



The remainder of the stations worked during this cruise followed at varying intervals 

 between 59 and 69° S, the general direction of the course steered being slightly to the 

 east of south. The time interval being naturally much smaller, the few stations worked 



