288 DISCOVERY REPORTS 



concentrations mentioned above, however, the genus is unimportant numerically. 

 Auxospore formation is more often to be seen among the solenoids than in any other 

 group, and good examples of this phenomenon in Rh. alata are particularly common. 



Dactyliosolen antarcticus Castracane (Hendey, 1937, pp. 323-4) = D. antarcticus 



Castracane + Z). laevis Karsten + D. flextiostis Mangin in Hart (1934) and Hardy, 



in Hardy and Gunther (1935). 



The forms described as separate species are treated by Hendey as phases of the 



'type', an opinion which I had come to as a result of the work in the field during the 



third commission. In some one or more of these phases, D. antarcticus is to be found 



throughout the Antarctic zone. It is most abundant in the South Georgia (neritic) 



area at the time of the main increase, but is more important, relative to the total 



phytoplankton present, in oceanic areas in autumn and winter. It should therefore 



probably be regarded as an oceanic species. The less strongly silicified laevis phase has 



a more southerly distribution than the type, which is the reverse of what one would 



expect from the silica content of the water. 



Corethron criophilum Castracane (Hendey, 1937, pp. 325-9, shows how all previously 

 recorded species appear to be but phases of the type) = C. valdiviae Karsten, 1905 ; 

 Hardy in Hardy and Gunther, 1935; Hart, 1934. 



The most important solenoid diatom of Antarctic surface waters, to be found, mainly 

 in the hystrix, type and inerme phases described by Hendey, throughout the whole of the 

 Antarctic zone at all seasons in varying numbers. It is most important in neritic areas, 

 where it sometimes forms almost the whole of the phytoplankton (Hart, 1934, pp. 40, 

 135), but from the wide distribution of most phases it must be regarded as an essentially 

 oceanic species. Living examples have been seen in pack-ice. Like some other members 

 of Group II this species is locally more abundant as one proceeds southwards, in the 

 open ocean. 



There is no doubt that Hendey is correct in applying Castracane's name to the 

 species, but it happens that the taxonomic type phase (that first described) does not 

 correspond to the phases most frequently encountered in nature. For this reason I find 

 some parts of Hendey 's descriptions, relating to the other phases, somewhat misleading. 

 In my experience the 'average' Corethron of the Antarctic zone is intermediate, as 

 regards size and strength of frustule, between Hendey 's hystrix and type phases. 

 Auxospores developed from the type phase always approximate more to the hystrix 

 phase in these respects, and I find the convexity of the valves too variable within each 

 phase to help in drawing even these elastic distinctions. Karsten's ' species ' C. valdiviae 

 is certainly nearer the ' average ' Corethron of Antarctic surface waters than the small 

 fragile C. criophilum Castracane that constitutes the type. C. valdiviae becomes part 

 of the hystrix phase in Hendey's system. 



Hendey describes the inerme phase, which I had previously referred to as the ' spine- 

 less chains' of C. valdiviae, as having 'robust cells, usually strongly siliceous'. This is 

 true enough in comparison with the type, but the minute, fragile, extremely weakly 



