348 Mr. J. Ralfs on the Diatomacese. 



MeloseircE they are attenuated ; the cells too are not divided as 

 in Meloseira by lines, either single or double. 



" The above I think are some of the principal differences be- 

 tween the Conferva which has received ]\Ir. Thompson's name 

 and the Meloseira!, and these are in my o]jinion too considerable 

 to admit of its holding a place with these. Would it not be more 

 correct to place it with the species of the genus Lyngbya V 



* Filaments moniliform ; frustules united in pairs. 



1 . M. nummuloides, Ag. Frustules spherical, having a line near each 

 end, but less strongly marked than the central one. Ag. Syst. 

 p. 8 ; Consp. Diatom, p. 65 ; Kutz. in Linnsea 1833, p. 70. fig. 72; 

 Grev. in Hook. Br. Fl. vol. ii. p. 401 ; Harv. Br. Alg. p. 195. 

 M. discigera, Ag. Syst. p. 8 ; Berk. Br. Alg. p. 3i. t. 11. fig. 2 ! 

 M. Borreri, Wyatt, Alg. Damn. no. 232 ! Conf. nummuloides. 

 Dill. p. 45. t. B. Conf. moniliformis, Jurg. Dec. 1. no. 7 ! Gal- 

 lionella nummuloides. Eh. Die Infus. p. 167. t. 21. fig. 1 ; Pritch. 

 Infus. p. 199. G. moniliformis, Bailey, American Bacil. part 2. 

 p.4. pl.ll.fig.3. 



In the sea or in brackish water. Margate, Rev. M. J. Berkeley ; 

 river Dart, Mrs. Wyatt ; salt marshes near Larne, co. Antrim, and 

 near Dublin, Mr. D. Moore. Penzance ; Bangor, N. Wales. 



It is brov.n when recent and becomes gi-een in drying. Fila- 

 ments free, slender, moniliform, constricted between each pair of 

 the frustules. Fi-ustules globular, united in pairs, and each di- 

 vided into hemispheres by a central line, and crossed by a fainter 

 line at each end. 



Plate IX. fig. 1. Meloseira nummuloides ; a, frustules deprived of colour- 

 ing matter; h, different stages, showing the mode of growth ; c. dotted lines, 

 marking new hemispheres; d, new liemispheres just formed. 



2. M. Borreri, Grev. Frustules rather longer than broad, cylindrical, 

 rounded at the ends, with a central, strongly marked line. Grev. 

 in Hook. Br. Fl. vol. ii. p. 401 ; Harv. Br. Alg. p. 196. Coyif. num- 

 muloides, Eng. Bot. t. 2287 ! (bad). 



On marine algae. Shoreham Harbour, Mr. Borrer ; Chichester 

 Creek, Mr. Jenner. Penmaen Pool near Dolgelley. 



Tliis species much resembles the preceding, but is I believe 

 distinct. The filaments are stouter ; the frustides rather longer 

 than broad, cyhndrical, with their ends less convex, and marked 

 only with a central line. Sometimes the ends at their junction 

 are compressed, and then the frustules appear angular. In this 

 character it seems to approach the M. moniliformis', the ends 

 however of the frustules in that species are figured by Ehrenberg 

 as truncated cones — more convex therefore than those of M. nwn- 

 muloides, and consequently still more so than those of this plant. 



Plate IX. fig. 2. Meloseira Borreri. 



3. M. globifera, Harv. Frustules nearly globular, with numerous 



