2 32 DISCOVERY REPORTS 



55. Cells with internal septa in valvar plane, septa often perforated S 6 



55. Cells without septa "■ 59 



56. Valves broadly clavate, girdle cuneate Licmophora (337) 



156. Valves not clavate ■•• "■ "" 57 



57. Cells concavo-convex in girdle view, septa 6-20 Entopyla (339) 



57. Cells slightly biconvex or plane in girdle view 5 8 



58. Cells in zigzag chains; septa 4-12, straight or nearly so Rhabdonema (339) 



58. Septa wavy, usually 2 at each end of the cell Grammatophora (338) 



59. Cells with bilateral symmetry about the transapical axis 6° 



59. Cells asymmetrical about the transapical axis ° 2 



60. Cells in ribbon-like chains; valves usually linear Fragilaria (330) 



60. Cells solitary 6l 



61. Cells with small terminal spines Thalassionema (336) 



61. Cells without terminal spines ••• ••■ S?y»«/ra (334) 



62. Cells in stellate colonies, valves clavate Asterionella (333) 



62 Cells solitary, spirally twisted, often in dense masses ; one end of the cell inflated only a little 



more than the other Tfudassiothrix (335) 



63. Cells in ribbon-like chains, valve surface bearing 7-12 transverse bars which alternate 

 with a double row of puncta Fragilariopsis (331) 



63. Cells in zigzag chains, valve surface bearing usually 2-6 transverse bars, spaces between the 



bars sometimes hyaline or with fine subradial punctation Anaulus (288) 



SYSTEMATIC ACCOUNT 



The following arrangement includes all the species of diatoms that have been ob- 

 served in the phytoplankton material examined. No attempt has been made to give an 

 exhaustive list of references or to cite all the accepted synonyms. The references given 

 include the original description of the species, while the others have been chosen be- 

 cause they provide excellent descriptions and illustrations or throw interesting light 

 upon the history of the name now in use. 



General observations regarding abundance or rarity, peculiarities of structure, ap- 

 pearance, or occurrence are included. Descriptions of the species are provided together 

 with notes upon distribution and habit. Dimensions are expressed in microns, and are 

 not intended as defining the limits of size of a species from any locality, but refer only 

 to the specimens actually measured in carrying out this work. In order to give a more 

 clear idea of the specimens examined, dimensions, particularly of diameter, are fre- 

 quently expressed as two extremes, and that which was most frequent. For example: 

 Diameter of valve 20-60/x, mostly 52^, indicating that although specimens occurred as 

 small as 20/x and as large as 60/x the majority of the specimens were actually 52/x in 



diameter. 



Notes upon taxonomy and nomenclature are included in so far as they are of im- 

 portance to the work in hand. Geographical distribution is indicated in general terms 

 while the occurrence of a species is indicated by the station numbers, at which each 

 species was observed. The station numbers afford reference to the Station List Dis- 



