128 



FRESH-WATER BIOLOGY 



20(23,24,25) Central nodule small, rounded, or slightly elongated. ... 21 



21 (22) No lateral longitudinal areas of transverse septa. . . Nmicula Bory. 



A form which grows in gelatinous tubes is regarded by some 

 authors as a different genus Schizonema but others regard it as 

 a true Navicuki. 



Fig. 77. Navicula rhync/tocepftala Kiitzing. X 557- (Original.) 



22 (21) Two lateral longitudinal areas of transverse septa. Mostly imbedded 



in a gelatinous pseudothallus. . . . Mastogloia Thwaites. 



In shape, Mastogloia resembles Xavicula, but is distinguished from it by 

 the gelatinous envelope and the presence of lateral, transverse, siliceous septa 

 or plates which divide the lateral regions of the cells into- small compart- 

 ments. There are transverse striations on the valves. It is not a very 

 common genus in .\merica. 



Fig. 78. Maslogloia smiihii Thwaites. X about 300. (After Smith.) 



23 (20, 24, 25) Central nodule broad, extending to near the margin of the valves. 



Stauroneis Ehrenberg. 



Stauroneis occurs frequently in all 

 bodies of water and is a constituent of 

 the diatomaceous flora which forms large 

 siliceous deposits at the bottom of lakes. 



Fig. 79. Stauroneis anceps Ehrenberg. 

 X 600. COriginal.) 



24 (20, 23, 25) Central nodule elongated to a short rod. Borne on gelatinous 



stalks Brehissonia Grun. 



Fig. 80. Brebissonia 

 sp. X 580. (Original.) 



25 (20, 23, 24) Central and end nodules elongated, enclosed with the raphe 

 between two longitudinal, parallel, siKccous ribs. Frus- 

 tules sometimes borne in gelatinous tubes. 



Vanheurckia Brebisson. 



Fig. 81. Vanheurckia rhomboides Ehrenberg. X 370. 

 (After West.) 



26 (13) Valves asymmetrical with reference to the raphe or to a longitudinal 



axis; raphe arched, or nearer one margin than the other. 



Family Cymbellaceae . . 27 



27 (28) Valves greatly convex; girdle side elliptical or oval. 



Amphora Ehrenberg. 



A Van Heurck regards A mphora as one of the most difficult 

 genera of diatoms and notes that over 200 species have 

 been placed in this genus. It is believed that it origi- 

 nated from CymbeUa. 



^ Fig. 82. Amphora ovalis Kiitzing. a. Valve side; b. girdle 

 side. X 600. (Original.) 



28 (27) Valves flat or only slightly convex 29 



