10 RHODOMELACE^. v. 



detected in all the plants of this order, and affords a mark by which they may 

 generally be recognised when not in fruit. The growing ends of the branches and 

 ramuli are clothed with articulated, mostly dichotomous, raonosiphonous hairs or 

 ramelU, and sometimes the whole surface of the frond is thickly covered witli them. 

 In most cases they are deciduous, falling away soon after the point from which they 

 spring becomes fully developed, but sometimes (as in Dasya) they are persistent, 

 either remaining unchanged, or, (in the section Stichocarpus) eventually becoming 

 polysiphonous and changing into ramuli. 



Considerable uniformity prevails throughout the Order in the structure of the 

 organs of fructification. The concepticle in all is of the form technically called 

 ceramidium, that is, an ovate or urceolate hollow case, pierced by a terminal pore, 

 and containing a tuft of spore-threads radiating from a basal, more or less elevated 

 placenta. These conceptacles are formed at the ends or sides of the branches or 

 ramuli, and are always transformations of shortened ramuli, or of the apices of 

 branches. The spores are pear-shaped, tapering greatly to the base, and are formed 

 singly from the terminal cells of the spore-threads. The tefraspores in all cases, so 

 far as I am aware, are tripartite or unequally triradiate — three of the sporules 

 only being visible at either face. They are never equally dispersed through the 

 frond, and only in a few cases lodged in its unaltered ramuli. Most usually they 

 are produced in little ovate or lanceolate, pod-like receptacles called stichidia, de- 

 veloped either from the midrib, the disc, or the margin of the leafy species ; or 

 formed Avithin the axils of the ramuli ; on the ends of the branches ; or even on 

 the permanent hairs (as in Dasya) that clothe the surface. The distinctive generic 

 characters often depend on the occurrence, the form, or the position of the stichidia. 

 Antheridia have been found in several Polysiphonia', Rhodomela', Dasya^, ^x. They 

 occur in the form of minute pod-like bodies, resembling the stichidia in shape, but 

 much smaller, filled with exceedingly minute granular cells, and are usually of a 

 pale yellow colour. They are mostly developments of the dichotomous hairs that 

 crown the growing ends of the branches. 



The order Ehodomelacea^, though including within its limits many brilliant Algfe, 

 contains a large proportion of darker and more soberly coloured species. It takes 

 its name from one of the genera {Rhodomela) which is noted for changing from 7'ed 

 to black during the process of drying, and very many others have similar proper- 

 ties. In OdontJmUa the growing frond is blood red, while the dried is dark vinous 

 purple. Many rosy or purplish leafy kinds change to dull brown. Brown again, 

 of various depths and shades, is the proper colour of several Polysiphonice ; while 

 others of that genus, though equally brown when dry, are a pure crimson or pur- 

 ple when growing. In so large a genus all hues of red, as may be supposed, are 

 found, and are more or less lost or preserved in drying. The genus Bostrychia is 

 remarkable for the depth and dulness of its purple colour, the only varieties being 

 tinges of green, or of black or very dark brown. Dasya among the American 

 genera shines out in greatest lustre, its soft fronds (as in D. ekgans), clothed with 

 a rich fleece of rosy purple ; a colour which is preserved or even heightened in the 

 drying specimen. 



None of the North America genera of this order are peculiar to the coasts of 



