Mr. J. Ralfs on the Diatomaceae, 271 



are smaller than those of the sides. On each side of this central 

 portion are seen the lateral sui'faces, with their reticulations and 

 striae arranged transversely. As these lateral sm-faces are tui'gid 

 and compressed, they appear at lu'st sight to form part of the 

 fi-ont of the frustule. This may be observed in some slight degree 

 in Diatoma vulgare, but not to the same extent as in this genus. 



That it is in fact a portion of each lateral sm-face which is thus 

 seen in front will be found on a careful exammation of the exterior ; 

 but it is stiU more evident on a \iew of the interior of a fractm'ed 

 frustule, when the junctions of the sides with the central portion, 

 marked externally by mere lines of sepai'ation, form internally 

 projecting plates or rims. 



The mode of groA\1;h in this genus is very cm'ious. In Diatoma, 

 as well as in most of the Diatomacece, the plant increases by a 

 di\ision of the fi'ustules ; but in this genus, and also in Biddulphia 

 and Amphitetras, two new frustules are formed within the old 

 one, and, as they enlarge, rupture it, when it falls off. In these 

 the front portion is at tu'st very narrow and merely a broad line, 

 but it increases greatly in breadth until the new frustules are 

 fully fonned. 



1. I. enervis. Eh. Frustules generally much broader than long, the 

 sides with large reticulations, not striated, or else with the striae 

 between each row of cells. Isthmia enervis. Eh. Die Infus. p. 209. 

 t. 16. f. 6 ; Pritch. Infus. p. 223. f. 183. Isthmia obliquaia, Hook. 

 Br. Fl. p. 405 ; Han-. Br. Alg. p. 201. Isthmia obliquata, (3 te- 

 nuior, Ag. Consp. Diatom, p. 55. Biddulphia obliquata, Gray, Arr. 

 p. 294. Conf. obliquata, E. Bot. t. 1869. 



/3. suhcylindrica. 



Ballywater, co. Down, Ireland, Mr. Thompson ; Torquay and Ilfra- 

 combe, Devonshire ; Bracelet Bay near Swansea. 

 ft. Torquay, Mrs. Griffiths. 



Ehi'enberg considers the 'Eng. Bot.' figure to belong to the 

 next species, but I have no doubt that it is intended for this 

 plant, of which it is a very good representation. The highly mag- 

 nified figm-es in ^Eng. Bot.' undoubtedly represent it; those that 

 are less magnified are too broad for the frastides of his /. obli- 

 quata, whilst under a low power this species appears striated like 

 them. 



The trivial name obliquata, which was originally apphed to this 

 species, should I think have been retained ; but, on account of the 

 celebrity of Ehrenberg's work, I have not thought it expedient to 

 change his name. 



Under this species, as I am informed by ]\Ir. Dahymple, 

 Ehrenberg gives no synonyms and only one habitat, — Gothen- 



