96 GENERAL HISTOKY OF THE INFUSOP.IA, 



the argument. The rectangular form of the Diatomeae is doubtless a conse- 

 quence mainly of the silicious composition : yet it is far from univei^al among 

 them ; for some species are rather orbicular, others sections of cylinders, others 

 capsular, and others again not unhke square sacs with bulging sides and rounded 

 corners. Even where a rectangular outline exists, it is most frequently only in 

 one view ; and the most that can be said is, that the Hues of junction are in 

 many instances acute. On the other hand, examples of a rectangular outline 

 are to be found among the Desmidieae and their allies : the junction-surfaces 

 of Hyalotheca and Didymoprium are at right angles to the sides of the frond ; 

 the end view of Staurastrmn tumiduin is as angular as the front \iew of a 

 Triceratium ; and the front view of Euastrum euneatum presents decidedly 

 rectangular truncate extremities. So too in the genus Pediastrum, formerly 

 enumerated among the Desmidieae, although now detached as a subfamily 

 and placed between them and the Palmelleae, examples of an angular outline 

 occur, as in the Pediastrum Tetras and other species. As to the production 

 of spines, sufficiently numerous examples exist among the Diatomeae to prove 

 it no distinctive peculiarity of the Desmidieae ; and although warty expan- 

 sions or elevations of the surface precisely like those of some Desmidieae, may 

 not be noticed in Diatomeae, yet certain exaggerated inflations of the surface 

 are seen in some Diatomeae, e. g. in Biddidpliia pidcheUa and B. regina. 

 The two next distinctions indicated by Mr. Ralfs are of more consequence, 

 but nevertheless cannot be admitted as demonstrative of an entire difference 

 in nature. They are thus stated : — '' Their internal matter is usually brown 

 when recent ; and although some species are greenish, or become green after 

 they have been gathered, none are of a truly herbaceous character. Their 

 vesicles bear some resemblance to those in the Desmidieae ; but they are 

 of a yellower colour, and no starch has been detected in them." The last 

 section of this statement must be held as still sid) judice ; the chemistry of 

 the endochrome is too imperfect to afford a safe argument, and the chemical 

 relations of starch and isomeric compounds too little understood. The con- 

 cluding distinction, " that the Diatomeae do not conjugate," the researches of 

 Mr. Thwaites have negatived. 



To employ the summary of the affinities of the Diatomeae presented by 

 Prof. Smith {Synops. vol. ii. p. xxi) : — '' The Diatomaceae, with specialities of 

 their owti, have also intimate alliances with the other orders of the Proto- 

 phyta, resembling the Zygnemaceae and Desmidiaceae in the reproductive pro- 

 cess, — the Nostochaceae in the tendency sho^Ti by several genera to surround 

 their frustules with frondose masses of mucus, within which linear series of 

 cells are subsequently developed, — the Oscillatorieae in their movements, — the 

 Palmellaceae and all the orders I have named, in the self-dividing act by 

 which the indi\iduals of the species are multipHed, or the aggregate of spe- 

 cific life maintained and increased." 



Deteemination of Species and Genera ; Varieties ; Classification. — 

 The question has been very much discussed of late, what characters of the 

 frustules and of their contents are to be employed in the construction of 

 species ? Ehrenberg generally proceeded on the principle of notifying every 

 departure from any one form, assumed to be specific, as representing another 

 species ; but this loose plan has been found productive of error and of ex- 

 cessive multiplication of species, inasmuch as shape, or outline, or markings of 

 the surface are not nearly so permanent and distinctive as formerly imagined. 



Although in some species the size and figure seem pretty constant, yet in 

 many they are subject to endless variations. Prof. Gregory cites as examples 

 of changeableness of form the three species, Eunotia gihha, Pinnularia 

 divergeyis, and Himantidmm bkJens ; and he would comprehend several pre- 



I 



