54 BRITISH DIATOMACE^E. 



3. Podosira? maculata, W. Sm. Dry valve purplish, with alter- 

 nate bands of shading, distinctly punctate. 



Marine : not unfrequent in deep water. Stomach of Pecten, very abun- 

 dant, Coast of Sussex, March 1853 ; Stomach of Pecten, Kinsale Bay, April 

 1855, &c, W. Sm. Wisbeach, Dec. 1853, Mr. S. Smith. Frith of Clyde, 

 Feb. 1 854, Mr. Hennedy. Pembroke Harbour and Thames, Mr. Roper. 

 Newbie Deposit, Prof. Harkness. (Cherbourg, M. De Brebisson.) 



Having never been able to detect this species in a fresh state, I cannot refer 

 it with confidence to the present genus ; its loose valves are, however, so 

 similar to those of P. Montagnei, that I have felt constrained, in the absence 

 of fuller evidence, to place it near that species. Cyclotella Scotica, Kiitz. 

 Bacill. plate i. fig. 2 & 3, is also exceedingly similar, and may prove to be the 

 present species. 



Plate XLIX. 328. 



Genus 49. MELOSIRA, Ag. 



Filament cylindrical, of numerous frustules, attached or free ; frus- 

 tules spherical or cylindrical, often geminate ; valves hemispherical 

 or subcylindrical, more or less convex at the junction-surfaces. 



In an admirable paper on the conjugation of the Diatomaceae, 

 inserted in the e Annals of Natural History' for March 1848, Mr. 

 Thwaites suggests the propriety of separating from the genus Melosira 

 all those species " whose frustules are not convex at their extremities, 

 and which form by their close contact an uninterrupted cylindrical 

 filament." To the genus thus formed he gave the name of Orthosira : 

 this suggestion I have adopted in the present work. 



Mr. Thwaites also proposes to construct a genus, " Aulacosira" 

 for the reception of such species " as are characterized by the absence 

 in the frustule of an evident central line (suture) indicating the place 

 of future fissiparous division, but each frustule having two somewhat 

 distinct sulci or fossulae passing round it." This genus I cannot 

 adopt, as I believe the characters have no real existence, and owe 

 their apparent presence in the species Mr. Thwaites adopted as his 

 type, viz. Melosira crenulata, Kiitz., only to accident, or observation 

 under imperfect illumination. A careful study of the specimens from 

 Aberdeen, upon which Mr. Thwaites himself founded his remarks, 

 and gatherings from various other localities, has failed to satisfy me 



