208 DISCOVERY REPORTS 



workers were gradually building up concepts capable of wider application than the older ' plankton 

 types ' of Cleve. 1 



Gran's system itself involves the use of some of Haeckel's terminology. ' Holoplanktonic ', ' mero- 

 planktonic' and 'tychopelagic', defining the degree to which organisms are dependent on the sea 

 bottom at some period of their history, are widely accepted. ' Oceanic ' and ' neritic ', however, can 

 soon lead to anachronisms in dealing with the phytoplankton, if applied in a rigidly restricted sense 

 (cf. Hart, 1942, p. 283). Many undoubtedly holoplanktonic 'oceanic' diatoms, to be found in the 

 open sea at all seasons, may attain considerable abundance in coastal ('neritic') sea-areas. Con- 

 versely many ' meroplanktonic ' (and therefore truly ' neritic ') species seem able to go on reproducing 

 vegetatively for so long that they may often form a considerable proportion of the phytoplankton in 

 some ' oceanic ' areas far beyond the edge of the shelf. 



To avoid the confusion that could arise through the necessity of listing some species as both oceanic 

 and neritic, we suggest the use of the word ' panthalassic ' to describe these ubiquitous forms. The 

 word was used by Johnstone, Scott and Chadwick (1924) and is given in Henderson' 's Dictionary of 

 Scientific Terms (1953) as: 'living both in coastal and offshore waters; neritic and oceanic', but we 

 have not yet discovered who first used it in this sense. 



To distinguish species of very wide distribution, whether neritic, oceanic or panthalassic, we have 

 here used the adjective ' cosmopolitan ' in its general sense, when their known distribution shows a 

 higher degree of tolerance towards temperature and salinity differences than is necessarily shown by all 

 the panthalassic species. There are many regions where occupation of both inshore and offshore 

 waters does not involve very wide tolerance of these two best-known parameters of the surface-waters. 



Use of the expressions 'dominant' or 'predominant', practically unavoidable in discussing dif- 

 ferences between phytoplankton populations, provides another example of the difficulties involved 

 when words so useful in their generally accepted sense acquire an arbitrary, specialized connotation, 

 through the attempt to define them objectively. Numerical preponderance at some arbitrarily chosen 

 level can be misleading (from the point of view of potential productivity) of orgmisms varying so 

 widely in size and shape, and invites the further criticism that all known counting methods involve 

 errors. Yet, if we are to attempt anything less subjective than the addition of such descriptions as 

 ' abundant ', ' common ', or ' rare ' to the organisms identified, some form of counting — and acceptence 

 of arbitrary levels in drawing deductions from the counts — is unavoidable. Conversely, it should not 

 be forgotten that admittedly subjective observations by the earlier naturalists lead to the first recogni- 

 tion of any recurring pattern of plankton distribution. 



The whole subject is one to raise echoes of the resounding controversies, as to methods of quanti- 

 tative plankton investigations, that followed publication of Haeckel's famous Plankton Studien, and 

 the replies of Hensen and others, of which the best brief account that we have come across is given by 

 Johnstone (1908). 



Throughout the subsequent development of plankton study the early methods have been periodically 

 subjected to destructive criticism, too often by those who have never learnt to count plankton, and 

 constructive reviews like that of Gran (1932) have been few. Considerations which should help to 

 place the matter more fairly in perspective are: 



First, the truism that the ideal comprehensive quantitative method may indeed be approached, but 



1 Cleve's ' Plankton Types ' constitute the first clear epitome of the close relation between plankton population and con- 

 ditions of the milieu, that may be evident whenever neighbouring water masses are exceptionally well defined. When further 

 work made it clear that water-movements and seasonal changes frequently produced more complicated inter-relationships, 

 it was against the background of Cleve's concept that the new evidence was marshalled. Cleve's 'types' are still distinguish- 

 able in essence wherever abrupt differences in environment recur with the consistency usual around southern Scandinavia, 

 but few sea-areas provide such clear-cut examples. 



