SCOPE OF THE METHODS EMPLOYED 277 



variations in quantity. Its disadvantages are obviously those inseparable from the use 

 of any form of tow-net — loss of nannoplankton forms and small solitary diatoms 

 through the mesh, and a certain degree of clogging where the phytoplankton is very 

 dense. Thus the values obtained will always be minimal. There is considerable evidence 

 that nannoplankton forms and dinoflagellates are never present in such numbers as to 

 be important producers (as compared with the diatoms) in the Antarctic zone. The 

 Meteor results and my own centrifuge counts may be useless for comparative purposes 

 as Nielsen maintains, but would certainly have shown up the presence of a large pro- 

 portion of nannoplankton forms if it was in any sense a general occurrence. Moreover, 

 the colonial habit is strongly developed in most of the small Antarctic diatoms, though 

 this is not always readily apparent in preserved samples. Even the difficulty due to 

 clogging rarely arises, for the design and dimensions of the Harvey net are such that 

 the proportion of filtering surface to effective aperture is more than three times as great 

 as in an ordinary tow-net (cf. Hardy, 1939, p. 47). During the three commissions some 

 800 observations within the Antarctic zone have been obtained by this method. When 

 these are grouped regionally and in time sequence as in this paper, the general picture 

 they present agrees so well with the changes in the physical and chemical factors of the 

 environment, studied by entirely independent methods, that it seems certain that they 

 must be roughly comparable to the true value of the standing crop. I should be the 

 first to admit that in warmer seas where nannoplankton forms may predominate, and 

 dinoflagellates are important, the method would be inadequate. 



The advantages of Harvey's method for our particular purpose are more readily 

 appreciated if one considers the weak points of other methods available. If one had 

 obtained sedimentation counts from some eight hundred stations (none too many 

 considering the size of the area concerned) the time spent in the actual collecting at 

 sea, which extended, in conjunction with our other work, over more than five years, 

 would have been considerably increased. All the counting would have had to be done 

 in a shore laboratory and, owing to the uneven distribution of phytoplankton with 

 depth, at least six counts from each station would have been needed to give a true 

 picture. Each count takes from two to three hours according to Nielsen (1933), so 

 that the working up of such a volume of material would occupy the whole time of an 

 experienced worker for at least a further five years. From this practical consideration 

 alone it is evident that such refined methods can only be employed to advantage after 

 the general conditions have been made known in broad outline, so that the detaded 

 work can be limited to manageable series of observations where conditions are probably 

 typical of larger areas. A minor drawback of the sedimentation method (Nielsen, 1933, 

 1935. P- 5)' that certain small naked forms must always be lost or become un- 

 recognizable when working with preserved material, need not concern us; but the 

 difficulties he experienced when Chaetoceros spp. were numerous would prove a serious 

 handicap in polar waters. 



While census-taking will always remain an essential part of the study of the phyto- 

 plankton, it is subject to some general objections inseparable from all purely numerical 



