ANIMAL NATURE OP DIATOMEiE. 397 



allied to the Synedne, that there remains no character to 

 distinguish them except that they are free, whilst the 

 latter are parasitic and affixed. 



I confess that I cannot comprehend by what motive 

 Kiitzing divides the numerous species that follow into 

 two sections, (oblongse, ellipticse et ovatse.) Were there 

 a difference, how small soever, but constant, this might 

 have a systematic value as rendering the distinction of 

 species more easy ; but we have only to compare together 

 the two species craticula and bifrons, which figure in 

 the two sections, to convince ourselves that the distinction 

 proposed is not based on a constant character. And it 

 causes real surprise to find enumerated in the second of 

 these sections a species {^S. angustd), which it is true has 

 not the median contraction, but, from similitude of form, 

 would by any one be supposed to belong to the section 

 preceding, which we say is allied to the Synedrce, and in 

 which section we have the S. regula and S. multifasciata 

 both equally wanting in that character. Comprising, 

 then, in a single group, both these sections, we assert 

 that here exists that important character of the secondary 

 surfaces exceeding the primary, which, as before stated, 

 forms the contrast with the DenticulcB. But even among 

 the latter there is not wanting an instance of the opposite 

 condition {J), undulata.) The new Surirella Jenneri, of 

 Hassall, as well as Denticula constricta, has the primary 

 surfaces perfectly equal in dimension and form to the 

 secondary. After all I believe that affinities and dis- 

 tinctions are rather to be sought in internal organisation 

 than in external form, the latter having no organogra- 

 phical value except as an indication of the former. With 

 regard to structure, though we are still far from having 

 sufficient data whereon to establish any principle of 

 classification, yet we find in Surirella, more, perhaps, than 

 in any other genus of Diatomese, a multiplicity of organs 

 and a complex organisation. The 8. striatula, of which 

 Turpin and Corda have given figures of little accuracy 

 and in a great degree imaginary, supplied to Ehrenberg 



