468 ANIMAL NATURE OF DIATOMEiE. 



ance of interrupted and capitate vittse. The appearance 

 of four transverse series is produced by the form of the 

 frustules {bacilli). The transverse striae are continuous 

 even across the vittse and canals ; and it is only by atten- 

 tion to the slight projection of these, that, in withdrawing 

 the object by slow degrees from the microscope to bring 

 different parts into focus, we can see first the striae of the 

 intermediate spaces and not those of the canals, then the 

 latter and not the former, and vice versa if the object be 

 brought near again. They then appear always in inter- 

 rupted longitudinal series, but sometimes in the inter- 

 mediate spaces, and sometimes on the vittse ; and we can 

 never see them on both at once. Near the margins, the 

 difference of surface diminishes, and therefore the striae 

 appear continuous. 



In the figure which Ehrenberg gives of B. arciiatum, 

 (pi. XX, fig. 8,) the secondary surface is represented 

 simply elliptico-acute, and the vittse appear entirely con- 

 fined to the extremity. I suspect this figure to be taken 

 from the external aspect rather than from direct observa- 

 tion ; and from analogy with our species, I should be 

 induced to believe that the primary surfaces remain 

 parallel to each other quite to the attenuation of the 

 extremities, and, therefore, that the series of the apparent 

 vittae becomes double rather than quadruple. 



The four genera, [Striatella, Tessella, Hyalosira, 

 BJiabdomena,) which constitute the family of Striatelleae, 

 are certainly connected by a great affinity. Still it is dif- 

 ficult to determine their mutual relations, or the connexion 

 of the entire family with others. Ktitzing places the 

 Striatellese in comparison with the genera Fragillaria, 

 Diatoma, and AcJinanthes, as they do not specially difier 

 from the two former except in the abundance of vittse, to 

 which, I believe, the two usual longitudinal canals are 

 correspondent. From AcJmanthes they differ as well in 

 form as in the absence of a central aperture. We will 

 compare them with the next family when we have treated 

 of it, contenting ourselves at present with intimating that 



