346 Mr. J. Ralfs on the Diatomacese. 



There are specimens in tlie British Museum collection varying 

 from three-quarters of an inch to a whole inch in length. This 

 species or variety comes very near the S. gigas, Leach, ' Edw. 

 Crust.^ iii. p. 205, from which it principally differs in the more 

 elongated and narrower outer plate, and in the grooved elevation 

 at the base of the more arched last joint of the abdomen. 



11. SeroUs paradoxa (Fabr. spec), Leach. 



Mr. Wright informed me, on his return from the Falkland 

 Islands, that this singular flat crustaceous animal is very com- 

 mon in the Carnash, Berkley Sound, Falkland Islands; it is 

 found in shallow places with a bottom of light sand and mud, 

 among which it bmTows, and in which, when disturbed, it buries 

 itself very rapidly. He assured me he had seen specimens at least 

 six inches long. This species was first found by Sir Joseph Banks 

 on one of Cook^s voyages, off the coast of Tierra del Fuego. We 

 have one specimen in the Museum labelled as coming from Se- 

 negal, the others were given by Mr. Wright from the locality 

 above mentioned. 



XLIV. — On the British Diatomacese. By John Ralfs, Esq., 

 M.B.C.S., Penzance*. 



[Continued from p. 276.] 

 [With a Plate.] 



Meloseira, Ag. (Gallionella, Eh.) 

 Filaments cylindrical, siliceous, jointed, fragile ; one or two lines 



passing round each frustule near its centre. 



This genus in its cylindrical filaments differs from the other 

 Cymbellece, and thus connects them with the ConfervecB ; but it 

 agrees with them in being generally of a brow^n or yellowish co- 

 lour when recent, and especially in its siliceous filaments and in 

 the presence of striae ; characters which sufficiently point out the 

 propriety of its present situation amongst the Diatomacece. 



The filaments have no proper margins marked by distinct cha- 

 racters as in the other genera of the Cymhelleoe ; and the strise 

 when present on* the junction- surfaces are not transverse but 

 radiated. 



The cylindrical form of the filaments, and also the other differ- 

 ences just mentioned, compel me, in describing this genus, to use 

 the terms length and breadth in the sense in which they are ap- 

 plied to a Conferva, and consequently in one the reverse of that 

 which they have in respect of the other genera of this family. 



This genus may be divided into two sections, which differ in 

 several respects. 



* Read before the Botanical Society of Edinburgh. 



