CLASSIFICATION OF DIATOMACE-ffi. 289 



from one side to the other ; and in Gomphonema (Fig. 139), in 

 which a foramen fulfilling the nutritive office is found at the larger 

 extremity only, the movement (which is only seen when the frustule 

 is separated from its stipes) is a hardly perceptible advance in 

 intermitted jerks in the direction of the narrow end." 



222. The principles upon which this interesting group should be 

 classified, cannot be properly determined, until the history of the 

 Generative process — of which nothing whatever is yet known in a 

 large proportion of Diatoms, and very little in any of them, — shall 

 have been thoroughly followed-out. The observations of Focke* 

 render it highly probable that many of the forms at present con- 

 sidered as distinct from each other, would prove to be but different 

 states of the same, if their vshole history were ascertained. On the 

 other hand, it is by no means impossible that some which appear 

 to be nearly related in the structure of their frustules and in their 

 mode of growth, may prove to have quite different modes of repro- 

 duction. At present, therefore, any classification must be merely 

 provisional ; and in the notice now to be taken of some of the most 

 interesting forms of the Diatomacece, the method of Prof. Kutzing, 

 which is based upon the characters of the individual frustules, is 

 followed in preference to that of Prof. W. Smith, which was founded 

 on the degree of connection remaining between the several frustules 

 after self-division, f — In each Family the frustules may exist under 

 four conditions ; (a) free, the self-division being entire, so that the 

 frustules separate as soon as the process has been completed ; (b) 

 stipitate, the frustules being implanted upon a common stem 

 (Fig. 126), which keeps them in mutual connection after they have 

 themselves undergone a complete self-division ; (e) united in a fila- 

 ment, which will be continuous (Fig. 131) if the cohesion extend 

 to the entire surfaces of the sides of the frustules, but may be a 

 mere zig-zag chain (Fig. 127) if the cohesion be limited to their 

 angles ; (d) aggregated into a frond (Fig. 141), which consists of 

 numerous frustules more or less regularly enclosed in a gelatinous 

 investment. It is not in every Family, however, that these four 

 conditions are at present known to exist ; but they have been 

 noticed in so many, that they may be fairly presumed to be capable 



* According to this observer ("Ann. of Nat. Hist.," 2nd Ser., Vol. xv., 

 1855, p. 237) Navicvla bifrons forms, by the spontaneous fission of its in- 

 ternal substance, spherical bodies which, like genimules, give rise to 

 Surirella microcora. These by conjugation produce N. splendida, which 

 gives rise to N. bifrons by the same process. He is only able to speak 

 positively, however, as to the production of N. bifrons from JV. splendida; 

 that of Surirella microcora from N. bifrons, and that of N. splendida 

 from Surirella microcora, being matters of inference from the pheno- 

 mena witnessed by him. 



> t The method of Kutzing is the one followed, with some modifica- 

 tion, by Mr. Ralf s in his revision of the group for ' ' Pritchard's His- 

 tory of Infusoria," 4th Edition; and to his systematic arrangement the 

 Author would refer such as desire more detailed information than the 

 necessary limits of the present treatise permit him to give. 



U 



