386 DESMIUE.E. 



2. MiCRASTERIAS KADIATA HaSS. 



Plate XC. Fig. 2. 



Char. Fi-ond orbicular, smooth. Segments hemispherical 

 each Jive-lohed. Lobes approximate, deeply bifid. Divi- 

 sions divergent, linear, and slightly bidentate. 



Mici'asterias melitensis Ealfs, in Annals, vol. xiv. p. 260, 

 pi. vi. f. 2. 



Hab. Amongst Diatoma fenestratum in a small pool a 

 little below the outlet of Llyn Genernon, near Dolgelly, 

 very rare : Mr. Ralfs. 



Tills species is described by INIr. Ealfs under the name of 

 Micrasterias melitensis, and as the Euastrum crux melitensis 

 of Ehrenberg, which species it seems to me that it as- 

 suredly is not. In the true Micrasterias crux melitejisis the 

 frond is oval, rather than circular, and each segment is three- 

 lobed instead of five-lobed, and the lobes themselves are not 

 approximate but divergent; see PI. xc. fig. 7. The plant 

 wliich Ehrenberg figures as the young of his species seems 

 to be referrible to the following genus. 



63. HOLOCYSTIS Hass. 



Char. Fronds much compj-essed, formed of two deeply divided 

 segments not incised. Segments not divergent. Ij£)bes 

 angular, not radiant. Angles often spinous. 



Denvation. From oKos, entire, and Kvaris, a cell ; in allu- 

 sion to the fact of the segments being entire at their 

 extremities, and not emarginate as in 3Iicrasferias. 



This genus bears considerable resemblance to 3Iicrasterias 

 Ag., from which it differs mainly and essentially in the absence 

 of the radiant disposition of the lobes, which is the chief 

 character of that genus. To it should be referred perhaps 

 the Euastntm crux melitensis, pi. xii. fig. 3 c. of Ehrenberg ; 

 Arthrodesmus conoergens Ehr. approaches likewise ratlier 

 closely to the genus Holocystis. 



