56 DISCOVERY REPORTS 



frustules were large, and auxospore formation was observed at comparatively few 

 stations, while microspore formation was quite common. It is reasonable to suppose that 

 the first maximum of production for this species had already passed, and that the large 

 form was about to be succeeded by a wave of smaller frustules derived from microspores, 

 such as had been found in the Bellingshausen Sea at mid-season. These in their turn 

 would in normal circumstances doubtless have given way to the spineless vegetative 

 form towards the end of the season. This season, however, was if anything unusually 

 cold, and from work done by the R.R.S. 'WiUiam Scoresby' later in the year it seems 

 that a succession of waves of the normal spiny form of Corethron invaded South Georgian 

 waters, their chief source of origin being the eastern Weddell Sea. The only stations 

 worked during this season, at which the spineless chain form was markedly dominant 

 were away to the north-west, in the vicinity and to the north of the Antarctic con- 

 vergence. While odd individuals have been taken in the far south in late summer, it 

 seems that the abundant occurrence of this form is limited by the water temperature, 

 which reached 3-0° C. or over at all the stations at which it has been abundantly 

 recorded during this investigation, except for the southernmost portion of the dense 

 patch recorded to the south-east of the island during the previous season. 



FragUaria aiitarctico (Fig. 24) was another form reaching its greatest abundance in 

 the western Weddell Sea water round the southern end of the island. It also occurred 

 in moderate numbers in the eastern Weddell Sea water, and in the small hauls in the 

 Bellingshausen Sea water to the north-west, where it formed an important part of the 

 comparatively scanty phytoplankton. The minimum for this form was apparently in the 

 oldest surface water in the extreme north of the area investigated. 



This distribution is closely paralleled by F. antorctica f. botivet (Fig. 25), except 

 that it was much rarer in the eastern Weddell Sea water than the type species. Very 

 large catches of this form were obtained to the south and west, while it formed quite 

 a large proportion of the small hauls obtained in the Bellingshausen Sea water to the 

 north-west. 



Next in order of total abundance on this survey comes Nifzschia seriata, which evi- 

 dently reaches its maximum for the area somewhat later in the year than the other 

 species ; on the survey dealt with by Hardy it totalled over 200 millions, while on this 

 occasion only about 88 millions were recorded. It will be seen from Fig. 26 that though 

 widely distributed in all three types of surface water, this species showed a maximum 

 of abundance to the south-west in western Weddell Sea surface water. But from the 

 description of the catches on the Bird Island line (p. 44) it will have been seen that it 

 reached its highest proportions in the catches towards the outer end of that line, in 

 water of Bellingshausen Sea origin. At the anomalous St. 493 on the Larsen line, 

 A^. seriata reached its maximum in numbers for the whole survey, and was one of 

 the dominant forms. As it was comparatively poorly represented in the catches from 

 the old eastern Weddell Sea water to the south and to the west this fact requires some 

 explanation. From the Station List (1932, pp. 87-9) it will be seen that this line was 

 worked some two days later than the Bird Island line on which Bellingshausen Sea 



