2 GENEEAL HISTORY OF THE INFUSOEIA. 



beauty. This will be best illustrated by reference to the Plates I. and II. ; for 

 description alone would fail to convey even a tolerably accurate conception. 

 In Micrasterlas (I. 18, 20, 21) the frustule has a general circular outhne, but 

 is bipartite and variously cut. In Euastrum (I. 23, 24, 26 ; II. 10) it is 

 bipartite, and each valve deeply sinuated. In many species of Cosmarium 

 (1. 1, 2 ; II. 33) the constriction is much shallower, the valves hemispherical, 

 and their margin entire. In Staurastrum (I. 31-34 ; II. 3, 7) each segment 

 is more or less irregularly produced at the extremities into horn-like pro- 

 cesses. In Penium, Docidium, and Clostermm (II. 1, 2, 9, 14) the frond is 

 elongated and wand-Hke, without constriction, or with only a very faint one, 

 and in many species is, moreover, curved or crescentic. Not a few genera 

 present numerous fronds united together ; the outline of the compound being 

 will consequently vary, both according to the figure of each individual frond, 

 and especially to the mode in which the several fronds are united. Thus in 

 HyaTotluca, Desmidium, and other genera (II. 35, 37, 39), the quadrate 

 fronds are united side by side in single series, so as to foim a chain or 

 filament, in other words are concatenated. 



The lateral view or cross-section of the fronds furnishes valuable characters, 

 and is largely made use of by Mr. Ralfs with that object, especially to distin- 

 guish between the several filamentaiy species. His figures show that the 

 fronds may be more or less compressed, and consequently offer on a transverse 

 section (end view) an oval and more or less acuminate form (I. 25 ; II. 23, 

 29), further modified by the elevations and depressions which the surfaces 

 possess (I. 25 ; II. 23). In other cases the section is circular, e. g. in ffya- 

 lotheca and Didymoprium (II. 32, 38), whilst in others, again, three or four 

 sides exist which are commonly concave, as m Desmidmm (II. 40). 



The end view exhibits the arrangement of the mass of chlorophyll, which 

 in some instances would appear to be pecuhar and determinate of species. 



The appearance of the Desmidiese is much modified by the sinuosities, 

 eminences, depressions, and processes, as well of the sm-face as of the margin 

 of the fronds, and also by the depth and width of the central constriction. 

 The surface may be dotted over irregularly, or more often regularly : the dots 

 themselves are in most cases elevated points, and in fewer instances depressions. 

 An irregular distribution of minute dots produces a granular-looking surface 

 (I. 24 ; II. 23, 30). Where the spots are larger their elevated character 

 becomes e\ident on the margin, to which they give a finely-toothed or dentate 

 appearance, e. g. in CosmciHum (I. 1, 2, 3). In some elongated fonns, such 

 as Tetmemorus and Penium (II. 15), the puncta are disposed in lines parallel 

 to the length : in Docidium, however, the disposition, so far as regular, is 

 transverse. In several examples the surface is marked by elevated hues or 

 by furrows (II. 6). Such markings seem peculiar to the elongated genera, 

 particularly to Closterium. 



Many apparent lines are resolvable by higher magnifying powers into rows 

 of puncta. Where the lines are fine, they are said to produce a striation of 

 the surface, as in Closterium attenuatum and C. acerosum ; where they are 

 more distinct they are termed costa?, and the surface they cover is costate or 

 ribbed, as in Closterium costatum and C. angiistatum. In general, in order 

 to discover the striation of the surface, the fronds must be viewed when 

 empty ; sometimes indeed the Hues can be made out at the extremities which 

 are unoccupied by chlorophyll. 



The strige and costae of Closterium and Penium referred to are disposed 

 longitudinally, but frequently they are intersected at one or more points by a 

 transverse line. In these spindle-shaped genera, where no constriction is 

 found, one such transverse line, usually central, is constant, and indicates 



