MARINE DIATOMS OF THE PHILIPPINE ISLANDS 77 



ENDICTYA OCEANICA Ehrenberg 



(Ehrenberg, Mikrogeologie, pi. 35A, sec. 18, figs. 6-7; Schmidt, Atlas, pi. 65, 

 figs. 12-13.) 



Genus EPITHEMIA Kiitzing 



EPITHEMIA ZEBRA (Ehrenberg) Kiitzing 



(Smith, Brit. Diat., pi. 1, fig. 4; Van Heurck, Synopsis, pi. 31, figs. 9, 11-14; 

 Schmidt, Atlas, pi. 252, figs. 1-21.) 



The natural habitat of this diatom is probably fresh water; but it 

 often occurs in marine gatherings, not as detritus, but as a member 

 of the flora. It also appears in fresh-water fossil deposits. It is rare 

 in the material examined for this paper. 



Genus EUODIA Bailey 



EUODIA JANISCHII Gurnow 



See Leudugeria janiscMi. 



Genus FRAGILARIA Lyngbye 



FRAGILARIA ANGUSTATA Cleve and Grove 



(Le Diat., 1891, pi. 8, figs. 19-20.) 

 Scarce at the Philippine Islands. 



FRAGILARIA DUBIA Gronow 



(Grun, Oester. Diat., pi. 7, fig. 28; Van Heurck, Synopsis, pi. 36, fig. 18.) 



As the specific name suggests, it is doubtful if this diatom belongs 

 to the genus Fragilaria. 



Genus GLYPHODESMIS Greville 



No adequately sharp line of demarcation between this genus and 

 Plagiogramma can be drawn. The ideal Gly phodesmis may be said to 

 be G. tvilliamsonii (Gregory) Grunow, and the ideal Plagiogramma, 

 P. validum Greville, and between these two there is a perfectly sat- 

 isfactory difference. But this is not maintained between other species 

 of the two genera. The most pronounced mark of demarcation is 

 that Plagiogramma has an evident hyaline area across the middle of 

 the valve, often with a ridgelike border, and bearing at its center 

 a large disk or elevation, while Glyphodesmis has no hyaline central 

 area; its beading also is heavy, closely set and liable to be quadrate 

 rather than round, as in Plagiogramma. But in such forms as P. tessel- 

 atum Greville the beading is subquadrate and the hyaline central area 

 is reduced to the dimensions of a large oval median disk or pseudo 

 nodule. It might be well to put this genus subgenerically under 

 Plagiogramma, the older of the two. 



