i86 DISCOVERY REPORTS 



southern limits are less well defined, but may very probably be found to coincide with 

 the northern limits of the pack in winter (cf. Schott, 1926, Taf. vi). A glance at 

 Schimper's notes on the numerous bottom samples obtained by the ' Valdivia ', quoted 

 by Karsten (1905), reveals the fact that in almost every sample Fragilaria antarctica 

 predominated, and the only other important species were Thalassiothrix antarctica and 

 discoid forms, mainly Coscinodiscus spp. With this our own preliminary observations 

 in the South Atlantic are in entire agreement. Now it is precisely these forms that are 

 found unchanged in crustacean stomachs and in the guano of sea-birds, and it seems 

 natural to suppose that they have much more strongly silicified skeletons than the 

 more typical plankton forms. Direct evidence of this is forthcoming for Coscinodiscus: 

 Wailes (1929, p. 25) says: "...half-a-dozen nearly pure gatherings of plankton 

 diatoms,. . .dried for one hour at 110° C, proved on analysis to contain an average of 

 40 per cent of silica, these were mostly small species of Chaetoceros [the predominant 



planktonic genus in the south] Gatherings composed mostly of large Coscinodiscus 



gave. . .around 55 per cent; if dried until there was no further loss in weight, the per- 

 centage of silica is about 75 per cent of the dry weight ". From our numerous plankton 

 gatherings in the south we know that apart from local swarms of Coscinodiscus bouvet, 

 Coscinodiscus spp. rarely form more than 1-2 per cent of the phytoplankton, and that 

 such dominant forms as Chaetoceros, Corethron and Rhizosolenia styliformis are but rarely 

 represented in the bottom samples. It follows that opinions, based on examination of 

 bottom samples and guano, that Coscinodiscus is dominant in certain areas, stand in need 

 of revision. 



Among the other chemical constituents of sea water which have recently come under 

 review as possibly limiting phytoplankton production, iron may be mentioned (Gran, 

 1932, p. 353). Little is known of this subject at present, but it has been suggested that 

 it may help to explain the observed richness of the neritic plankton in north temperate 

 waters, and the fact that this plankton frequently begins its main increase some time 

 before that of the oceanic plankton offshore, the land being regarded as a source of iron 

 in organic combination akin to the humic substance in soil. The preliminary results of 

 both Gran {loc. cit.) and Harvey (personal communication) indicate that the most 

 minute traces of iron and of soil washings may exercise a pronounced beneficial effect 

 upon the growth of diatom cultures. It is thus conceivable, though it can hardly be 

 thought probable, that the large number of morainic bergs observed in the Weddell Sea 

 as compared with the Bellingshausen Sea, may have some bearing upon the fact that 

 the phytoplankton of the former is richer. 



PHYSICAL FACTORS 

 It is when we come to consider the physical features of the environment that the 

 factors exerting the strongest influence upon phytoplankton production in the far south 

 appear. These factors are closely interrelated, and this makes it difiicult to assess their 

 individual importance. The strongest influences appear to be due to weather, currents, 

 light, ice conditions and temperature. 



