DEPTH DISTRIBUTION: PREVALENCE OF THE COLONIAL HABIT 323 



These figures strongly suggest that by far the greater part of the phytoplankton pro- 

 duction in the Antarctic zone takes place in the upper 10 m. of the surface layer. This 

 is in striking agreement with what would be expected from the classic oxygen con- 

 sumption experiments of Marshall and Orr (1928) and others, when we remember 

 that both higher latitude and increased scattering due to rough weather will both tend 

 to reduce penetration of light to a greater extent than in north temperate regions. The 

 importance of the loss of light due to scattering and reflexion at the surface where 

 rough weather prevails was first clearly recognized by Atkins (1926, p. 456) who is 

 responsible for the development of so many of our concepts concerning the growth of 

 phytoplankton in relation to its environment. We may say that in comparison with the 

 conditions studied experimentally in north temperate seas, the euphotic layer is centred 

 higher in the water column. The optimum depth, in the Antarctic zone, would appear 

 to be around 5 m. as a general rule. The effects of systrophe in lessening production 

 above the optimum are evidently less than in north temperate waters, while it is 

 probable that for most species the lower limit of the productive layer, or compensation 

 point, will not be below 35m., even at the height of the southern summer. Summing 

 up, we may say that the figures provide some concrete evidence that the suggestions 

 put forward in earlier work (Hart, 1934, pp. 189-91) regarding the effects of light and 

 interrelated factors upon the depth distribution of the Antarctic phytoplankton are, in 

 the main, correct. 



THE COLONIAL HABIT IN RELATION TO ENVIRONMENT 

 It will have been evident from the notes on the individual categories of phyto- 

 plankton organisms dealt with that many of the most important forms show a pro- 

 nounced development of the colonial habit, which is most marked at the height of the 

 main increase. It would seem that the hardening of protoplasmic connexions following 

 fixation in formalin renders the colonies brittle, so that they disintegrate easily, for 

 very much longer chains or larger colonies may be seen in fresh material than in pre- 

 served samples. 



The phenomenon would appear to be bound up with the rapidity of binary fission 

 when conditions are at their optimum. Many of the ' ribbon-forming' species, notably 

 Fragilariopsis antarctica and Eiicampia balaiistium moelleria phase, show it in an extreme 

 degree that involves marked torsion of the chains. The shorter chains common at other 

 seasons are straight, or curved in one plane only. Besides the typical ' ribbon-forming ' 

 species, many of the larger Group II diatoms show a similar increase in length of 

 chains, but with few exceptions ; this is more marked after the peak of the main increase, 

 when they reach their maximum relative importance. 



Rhizosolenia alata graciUima phase often forms very long chains far south at the time 

 of the main increase, but these are composed of few extremely elongated frustules. 

 Farther north, at the corresponding period, Rhizosolenia antarctica and Rh. chunii also 

 form very long chains, but with these species larger numbers of frustules are conjomed. 

 The large pennate diatom Thalassiothrix antarctica is usually found in rafts of from four 



