Mr. J. Ralfs on the Diatomacese. 349 



striae, which are most evident on the siliceous covering. Harv. Br. 



Alg. p. 196. Rosaria gloUfera^ Carm. in Hook. Br. Fl. vol. ii. 



p. 372. 



On marine algse. Torquay, Mrs. Griffiths ; Hastings, Mr. Jen- 

 ner ; Ardmore, Miss Ball (according to a specimen from Mr. W. 

 Thompson). Ilfracombe. 



Filaments affixed by a short stipes ; frustules nearly globular, 

 united in pairs, each with a strongly marked central line. Nu- 

 merous close strise pass round the filaments, and are most con- 

 spicuous on the siliceous remains of the old frustule, by which 

 the new ones are connected in pairs. 



The specimens which I have seen are apparently defective, as 

 in all only one pair of frustules is attached to the stipes. 



Ehrenberg represents the frustules of his Gallionella lineata as 

 longer than broad, cylindrical, with a quadrangular appearance, 

 whereas in this species they are nearly globular ; otherwise I 

 might have supposed he intended to describe this species. 



Some specimens I have received from Mr. Jenner have their 

 frustules generally longer than broad, but in other respects agree 

 with the above description. 



I have received from M. Montague a Meloseira {M. hormoideaj 

 Montag.) gathered in Cayenne, which appears to me to be iden- 

 tical with this species, and Mr. Berkeley is of the same opinion. 

 As however the filaments are composed of several frustules, it is 

 in a more perfect state than any British specimen which I have 

 examined. 



Plate IX. fig. 3. Meloseira globifera. 



** Filaments not moniliform ; frustules cylindrical. 



4. M. arenaria, Moore MS. Filaments stout ; frustules broader than 

 long, with a single central line ; junction-surfaces closely united, 

 striated. Gallionella varians, Eh. Die Infus. p. 167. (in part) t. 21. 

 fig. 2; Pritch. Infus. p. 199. fig. 131^ (junction-surface). 

 In fresh water. On a wet bank near Lame, co. Antrim, Mr. D, 

 Moore ; inner extremity of the Giant's Causeway, Mr. W. Thomp- 

 son ; very sparingly in a freshwater stream near Shoreham, Kent, 

 Mr. Jenner. 



Brownish when recent, pale green when dried. 



Filaments much stouter than any other species in this genus, 

 distinct to the naked eye, when rubbed between the fingers feel- 

 ing rough like grains of sand, whence its specific name. 



Frustules closely united, broader than long, discoid, appearing 

 slightly granulated with a broadish smooth central line ; their 

 junction appears like a dentated suture ; the junction- surfaces are 

 striated in a radiating manner ; the striae are most evident at the 

 margin, and sometimes very faintly marked in the centre. 



