RHODYMENIACE^E. 121 



palraately cleft; it is rarely pinnatifid, and when so cleft the 

 laciniae are alternate. The lower part of the frond is fre- 

 quently narrow and cylindrical, or contracted into a stipe, 

 and the outline in such species, is often a segment of 

 a circle, all the tips being of equal length. The substance is 

 seldom delicately membranous : it is more commonly rather 

 thick, composed of several strata of cells. The internal cel- 

 lular structure is rather lax, the cells being of large size, fre- 

 quently empty, though sometimes filled with granular matter, 

 polygonal, about as long as broad, and either lying close to- 

 gether or separated by wide air-cells or passages. Towards 

 the circumference the substance becomes gradually more 

 dense ; the cells smaller and more filled with colouring mat- 

 ter ; and the cells of the outermost layers are always of very 

 minute size. In some genera (as Gracilarid) the outer 

 strata of cells which form the periphery, are arranged in 

 lines, or filaments, perpendicular to the surface ; and these 

 genera indicate a passage into Cryptonemiacew, a group 

 which touches the present order (as at present constituted) 

 at many points. In the leafy species the coccidia are either 

 confined to the margin or scattered over the surface. They 

 are always prominent, very convex, and contain a mass of 

 spores, various in character and in degree of perfection. 

 Sometimes the spores are exceedingly minute and numerous, 

 the whole substance of the nucleus breaking up into a pow- 

 der ; sometimes (as in Rhod. ciliata) they are formed from 

 the terminal cells of radiating filaments. In the filiform spe- 

 cies the coccidia are either lateral, or they are, more rarely, 

 lodged in the centre of the branches, forming nodose swell- 

 ings at intervals. The tetraspores are never collected into 

 well-defined sori, and are most usually dispersed over the 

 smaller branches of the frond. When sori exist they are 

 spreading and cloud-like, without exact limit. In several 

 species the tetraspores are transversely parted or zoned ; a 

 character which may, perhaps, be advantageously employed 

 in defining genera, but which is not taken up in the present 

 work. Zoned tetraspores exist in Rhod. ciliata, R. jubata, 

 and R. bi/ida ; in Hypnea ; and in several others. 



The Rhodymeniacea are widely dispersed ; all our genera 

 having representatives in very distant countries with very 

 various climates. Rhodymenia is an ill-defined genus, as it 

 stands at present, and will probably be eventually broken up 

 into several. Its species are most numerous in temperate la- 

 titudes, between the parallels 40 and 50 at either side the 



