124 INTRODUCTION TO CRYPTOGAMIC BOTA^ r. 



matters, is quite convincing. Some of Ehrenberg's arguments 

 we do not think of sufficient importance to bring forward here. 

 As regards practical use, none at present is known ; though 

 doubtless the species exercise their proper part in the general 

 economy of the world, helping to purify the water, and to 

 afford food to moUusca and other animals. In some species of 

 Closteriwm, a distinct circulation has been observed. 

 3. DiATOMACE^, Ag. (Naviculacea, Ehrb.)* 

 Propagation and division of cells as in Desmicliacece. Walls 

 of cells containing a large quantity of silex, often beautifully 

 sculptured. Endochrome golden brown. 



99. These curious productions are often imited with the fore- 

 going into a single group, but they seem so clearly distinguished 

 by the flinty shell, which is almost always curiously striated, 

 and the absence of pure green colom-ing matter, that they seem 

 to me to form a group of at least as much importance as the 

 others adopted in this place. As in Desmidiacew, there are 

 solitary species, and others grouped into lines or membranes ; 

 and in the few which have been observed to produce new plants 

 by means of spores, the new productions do not exhibit at first 

 the normal character of the species. The division of the fronds 

 or cells, where it can be observed, takes place precisely in the 

 same way as in Desmidiacece, and no genus perhaps exhibits 

 this more distinctly than Isthmia (Fig. 7, b). The genera ai-e 

 more numerous than amongst the Desviidiacece, and are fre- 

 quently most acutely characterised ; and not only from their 

 peculiar striation, or singular form, but perhaps also from the 

 perfection in which dried specimens exhibit the minutest cha- 

 racters, they are favourite microscopical objects. Both in the 

 solitary and associated species there is frequently a distinct 

 pellucid peduncle, which in those which have a flabelliform frond 

 is dilated above, and in some cases forked, or repeatedly dicho- 

 tomous. In such cases every new frustule remains attached, the 

 base dilating as need reqxiires ; but in the filiform species, the 

 attachment is confined to a single corner, and probably the 



* The work of Mr. Smith ou Diatomacete may be quoted here with 

 the same praise as that of Mr. Ealfs on Desmidiacece. Like that, it is 

 perfectly original, and an important contribution to science. 



