I30 FRESH-WATER BIOLOGY 



'7 (36) Girdle side geniculate. Valves straight, linear, or fusiform; frus- 

 tules either free or stalked. . . Family Achnanthaceae. 



Only one genus Achnanthes Bory. 



Cells so curved that the two valves are not alike, the one concave with raphe, middle and 

 nd nodules; the other convex, without a middle nodule, but with a pseudo-raphe. Girdle view 

 symmetrical with reference to a transverse axis. Cells single or in bands, 

 ^^ mostly on gelatinous stalks., 



<^N\\A The cells may be solitary, though they usually form long, sessile chains or 



N^^\3J bands attached to the surface of green algae. The genus includes both marine 

 I I and fresh-water forms. 



' Fig. 89. Achnanthes exUis Kutzing. X 600. (Original.) 



38 (11, 34) No middle nodule present on either valve, except in Ceratoneis, 

 or at most consisting of a slight, ring-like elevation. . . 39 



39(40,41,62) Valves asymmetrical with reference to a longitudinal axis, in 

 that on one margin there is a longitudinal row of bead-like 

 thickenings (keel points) while on the other margin they are 



lacking Family Nitzschiaceae. 



Only one genus Nitzschia Hassall. 



Valves linear, sometimes curved, keeled, with canal raphe. Cells rhomboidal in cross sec- 

 tion. 



"T"^T'"TTr,imtTT>^ Fig. 90. Nitzschia linearis Smith. X 57S' 

 ■ — -^ fOrieinal.) 



(Original.) 



40 (39, 62) Valves with median, sigmoid keel, compressed, strongly arched, 



bearing raphe. . . . Family Amphiproraceae. 



Only one genus Amphiprora Ehrenberg. 



Valves fusiform, with central and two end nodules on raphe. 

 Girdle side sharply constricted at center. 



Fig. 91. Amphiprora sp. X 400. (Original.) 



41 (39, 62) Valves symmetrical with reference to a longitudinal axis . . 42 



42 (47) Valves each with two wing-Eke keels, strongly costate, with pseudo- 



raphe but no nodules. . . Family Surirellaceae . . 43 

 Cells mostly large, ovate, or elliptical. 



43 (44) Cells bent in saddle shape Campylodiscus Ehrenberg. 



Though the shape of the cells seems more or less tri- 

 angular, they are in reality circular, and their seeming 

 angularity is due to the curvature of the frustules. It 

 is a very large genus, some 92 species having been 

 recorded: the species are mostly marine, though a number 

 are found in fresh water. Their large size makes them 

 among the most conspicuous of the diatoms. 



Fig. 92. Campylodiscus cribrosus W Smith. X dbout 300. 

 (After Smith.) 



44 (43) Cells not bent or spirally twisted 45 



