2i2 DISCOVERY REPORTS 



Hendey's classification of the class Bacillariophyceae (diatoms) aims at providing a system wherein 

 the great structural differences between the most abundant plankton forms are more clearly empha- 

 sized than they were in earlier systems; among which that of Schiitt (1896) with minor modifications 

 by subsequent workers, is still the most widely accepted arrangement. 



Schiitt's scheme divided the whole group into two suborders Centricae and Pennatae, based upon 

 cell-symmetry around a supposedly central point in the one group, and a more or less iso-bilateral 

 symmetry in the other. Most marine plankton forms belong to the Centricae; and the majority of 

 the attached marine forms, freshwater and soil diatoms to the Pennatae. To this extent the scheme 

 shows a reflection of the earlier ideal, ultimately found impracticable, of classifying diatoms principally 

 upon the mode of life, as Smith (1853-6) had tried to do. The main morphological basis of Schiitt's 

 scheme soon shows up the incompatibility : although most of his Pennatae are not to be found as normal 

 constituents of the marine phytoplankton, some lesser subdivisions of the group are well represented 

 therein; so that the agreement between systematic classification based on morphology and primary 

 habitat differences, although better than it is in many other classes of organisms, is by no means 

 complete. 



Hendey (1937, 1951) has pointed out that the major subdivisions of the Centricae (variously ranked 

 as sections, families or tribes in later modifications of Schiitt's scheme) show structural differences so 

 marked that their inclusion within a single suborder seems questionable. Some of them (Biddul- 

 phineae, Soleniineae) even show a type of symmetry that it is very difficult to relate to a supposedly 

 central point. Moreover, even the Pennatae include groups of very diverse structure, although they 

 do not show such extreme divergence from the main characters of the suborder as do some of the 

 Centricae. 



Hendey, therefore, proposed that the class Bacillariophyceae, regarded as including only the one 

 order Bacillariales, should be divided into ten suborders of equal systematic status. The first five of 

 these include the forms assigned to various minor rankings in the more recent modifications of Schiitt's 

 Centricae (e.g. Hustedt, 1927-37) while the last five were included in Schiitt's Pennatae. Hendey 

 (1937) has himself pointed out that with regard to rankings below the level of suborders, his arrange- 

 ment is not materially different from the earlier ones, and he has been at pains to make due acknow- 

 ledgement to Schiitt and other workers in this field. The formal terminations of the family and sub- 

 family names that he has adopted merely serve to render his classification of the Bacillariophyceae 

 consistent with generally accepted usage among other classes of algae. 



Table 1 1 shows Hendey's classification, excluding the genera not represented in the material that 

 he was working on when he first formulated it. That material consisted of a large selection of plankton 

 samples from most of the areas traversed by ships of the Discovery Investigations in the Southern 

 Hemisphere, prior to 1935. 



In Table 1 1 the genera and corresponding supra-generic rankings observed in our Benguela current 

 material have been printed in ordinary type, leaving the groups that have as yet been recorded only 

 from other parts of southern seas distinguished by italics. 



Supra-generic rankings, either at the level of suborder or family, were found to lend themselves to 

 the definition of four of the five arbitrary groupings used later in this report, for descriptive purposes 

 only. They are preceded by numbers in brackets, and printed in bold type in Table 1 1 . 



The fifth group is a lumping of such members of the last five of Hendey's suborders as occurred in 

 the Benguela current samples. In this report they have been termed ' Pennatae ' for the sake of brevity 

 not from any desire to revert to the older classification. 



Whether one thinks of them as the Pennatae of Schiitt, or the last five suborders of Hendey's 

 system, those familiar with marine phytoplankton will know that nearly all the planktonic members 



