PHYTOPLANKTON OF BRANSFIELD STRAIT 121 



US face to face with another equally important concept due to Gran — that of the re- 

 generation of an association on the spot. It can be shown that the form in question has 

 a strong autumnal maximum in the BeUingshausen Sea (p. 133), at which season it 

 penetrates into the Bransfield Strait in small numbers as shown by the February survey. 

 On this spring survey moderate numbers coming up from the south-west were again 

 found at the western end of the strait, but there was a very definite gap between them 

 and the maximum observed to the north-east. It is known that the general north- 

 easterly drift of the Antarctic surface water is at a minimum during the winter months, 

 so that it would seem that the most rational explanation of the observed distribution 

 would be as follows : sufficient individuals remain from the autumnal invasion to give 

 rise within Bransfield Strait to a marked spring increase, that drifts away to the north- 

 east as the strength of the current again increases. A similar increase probably takes 

 place somewhat later in the northern part of the Bellingshausen Sea, and this was pro- 

 bably represented here by the re-invasion apparent at the western end of the strait. 



The distribution of Biddulphia striata is shown in Fig. 60. It will be seen that its 

 maximum was reached at the eastern end of the strait, and that while it was present in 

 small numbers to the west and north-west it was almost entirely absent from the middle 

 line. As this and certain other species were plentiful at the station worked within the 

 eddy of western Weddell Sea water, it might be thought that their presence in large 

 numbers to the northern end of the eastern line was due to mixing of the two types of 

 surface water, which would be promoted by large masses of pack-ice farther east, 

 always a probability at this time of the year. However, the salinities given in Table 

 XXXVI are alone sufficient to show that on this survey the Weddell Sea influence most 

 certainly did not extend north beyond the middle of the eastern end of the strait, so that 

 the most reasonable explanation of the large numbers of B. striata at Sts. WS 477 and 

 WS 479 appears to be regeneration on the spot, as in the case of Rhizosolenia gracillima. 

 This is not, as might at first appear, in contradiction to the distinctions between the 

 phytoplankton of the various types of surface water drawn throughout this paper. From 

 the first (pp. 20, 21) it has been pointed out that all the more important phytoplankton 

 species appear to be common to all types of Antarctic surface water. It is in the relative 

 proportions of the dominant forms that the distinction lies, and it will be seen that 

 while Biddulphia striata was twice as numerous at St. WS 477 as at St. WS 481, its 

 proportion of the total catch was 2-2 per cent at the latter as against i-i per cent at 

 the former. 



The distribution of Thalassiosira antarctica, shown in Fig. 61, closely parallels that of 

 Biddulphia striata. It reached its maximum estimated numbers to the north-east and 

 was generally present in fairly large numbers at the eastern end of the strait, reaching 

 by far its greatest relative abundance (16-2 per cent) at St. WS 481 in the eddy of 

 Weddell Sea water. Yet another species of some importance, having a distribution of 

 this type, was Eucampia antarctica, as will be seen from Fig. 62. Though by no means 

 so widely distributed as the two species described above, it showed an even more pro- 

 nounced maximum at St. WS 481. This species was entirely absent from the middle 



