146 DISCOVERY REPORTS 



these stations naturally included the first and the last worked in the Bellingshausen Sea 

 during that season, and accordingly the conditions will be seen to have varied con- 

 siderably. At the first station (St. 556) worked to the north of Anvers Island in December, 

 a moderate phytoplankton very poor in species was encountered. Corethron valdiviae 

 was strongly dominant and Chaetoceros neglectus the only other species of much im- 

 portance. It will be noted that an association of this type was encountered at several 

 stations in the more ice-bound portions of the Bellingshausen Sea farther to the south- 

 west, and also at the western end of the Bransfield Strait early in the season. It would 

 seem that it is the first association to develop when the ice begins to clear away, except 

 perhaps in the extreme west, where, as will be seen from Table LI I, a phytoplankton 

 of quite a different type was encountered only a week later. 



The next station in east to west order was St. 604, worked near the mouth of Neumayr 

 Channel towards the end of January. Here a catch of small proportions but rich in 

 minute species was obtained, of the same type that we have learnt to associate with the 

 channels of the Palmer Archipelago and the littoral waters of the islands farther south. 

 Chaetoceros neglectus was dominant, and the other more numerous species were 

 Fragilaria antarctica, Thalassiosira antarctica, Corethron valdiviae, Nitzschia seriata, 

 Rhizosolem'a tnincata and Nitzschia closteritim, in that order of importance. 



A phytoplankton of rather similar type but much poorer in quantity was found at 

 St. 557 worked well out to sea three weeks earlier. Here the most numerous species 

 were Fragilaria antarctica, Chaetoceros neglectus, Thalassiosira antarctica, and a small 

 branching colonial form of Nitzschia closely resembling the northern A^. frigida, which 

 we have never observed far from land or extensive pack-ice. The striking difference 

 between the phytoplankton observed at this station, and that from its near neighbour 

 St. 556, appears to be explicable on the grounds of the deflection of the inshore current 

 along the islands to the southward by the extensive masses of pack-ice surrounding 

 them, though we have no direct evidence of such deflection. It is, however, certain that 

 an essentially neritic association was present at St. 557, of a type clearly distinguishable 

 from the associations sometimes met with along the ice-edge which have a certain 

 general resemblance to neritic floras. 



At St. 603, the next to be considered from Table XLIX, it will be seen that a rich 

 phytoplankton of the Corethron — Rhizosolefiia gracillima type was encountered, with a 

 higher proportion of Fragilaria antarctica than had been observed at the stations worked 

 later in the previous year, and Corethron greatly in excess of Rhizosolem'a. However, 

 the phytoplankton observed at this station and at St. 596 made it fairly evident that 

 surface water bearing the Corethro?t — Rhizosolenia gracillima association was drifting 

 away from about the middle of the Adelaide Island line to the north-east, and as we have 

 already seen (p. 133) a similar phytoplankton was observed as far north as the southern 

 half of Drake Passage some three weeks later. 



At St. 558 on the outward voyage, worked near the pack-ice to the north of the Biscoe 

 Islands, Corethron valdiviae was very strongly dominant in a fairly rich haul in which 

 the only other species of much importance were Fragilaria antarctica and Thalassiosira 



