134 DISCOVERY REPORTS 



from the Weddell Sea had proceeded down the south shore of Bransfield Strait and 

 penetrated the channels of the Palmer Archipelago. The low surface salinities and rather 

 high temperatures at these stations render this explanation improbable, however, and 

 hydrological results obtained within the Bransfield Strait itself all point to the prob- 

 ability that the Weddell Sea influence does not extend much beyond Astrolabe Island 

 in a south-westerly direction. The only parallel we have encountered to this rich and 

 varied phytoplankton association, with so many typically spring forms persisting in large 

 numbers late in the summer in water that must have ultimately originated in the 

 Bellingshausen Sea, was at St. WS 501 worked just to the south of Anvers Islands near 

 the southern entrance to Neumayr Channel at midsummer of the previous year. It 

 seems possible that a rich local association may persist among the channels of the 

 Palmer Archipelago throughout the major part of the season. Mangin has described a 

 somewhat similar association from material collected in Scotia Bay at the South 

 Orkneys, where Chaetoceros socialis was dominant throughout the season (see Mangin, 

 1922, pp. 94, 95), though this was in Weddell Sea water. If a localized phytoplankton 

 rich in small forms does in fact persist within the channels, further investigations there 

 may prove very interesting, as I am informed by Mr F. C. Fraser that the abundance 

 of those species included here under the general term " small forms " shows some degree 

 of correlation with the abundance and the size of the developmental stages of Euphaiisia 

 superba. 



CONCLUSIONS 



If the observations described above represent an approach to the normal conditions 

 in the strait, we may conclude that the course of the phytoplankton development pro- 

 ceeds somewhat as follows. 



Early in the year, when the western end of the strait is still blocked up by pack-ice, 

 a moderate development takes place in the old Bellingshausen Sea surface water within 

 the strait. The dominant species of this spring increase is Corethron valdiviae, but the 

 other more important forms, Thalassiosira antarctica, Rhizosolenia gracillima, Biddulphia 

 striata and Fragilaria antarctica are all more abundant in proportion than at any 

 other time of the year in surface water of this nature. The influence of the Weddell 

 Sea current round Joinville Island does not reach its maximum until December, but 

 is already apparent in the greatly increased proportions of Thalassiosira antarctica, 

 Eiicampia antarctica, etc., to be found in that locality. The maximum for the year in the 

 old Bellingshausen Sea water occupying the greater part of the strait is probably reached 

 about the last week of November or first of December. 



As the season advances the pack at the western end tends to break up, the influx of 

 Bellingshausen Sea water into the western end of the strait is increased, the phyto- 

 plankton formed by the spring increase drifts away to the north-east in the direction of 

 Elephant and Clarence Islands, and is not immediately replaced. About the second week 

 in December the phytoplankton development in the eddy of Weddell Sea water reaches 

 its height. This is characterized by the pronounced dominance of small Chaetocerids, 



