ANIMAL NATURE OF DIATOMEjE. 471 



we observe such a frustule {bacillus) on one of its 

 secondary surfaces, the large central aperture eludes the 

 eye ; still it may be seen, but only across the transparent 

 surface itself. 



In this species we may frequently see two similar 

 diaphragms. In the T. Jlocculosa, as many as nine may 

 be met with (Hassall). They are not all equally easy to 

 be seen in specimens that have been dried and softened. 

 A green substance seems to be coagulated within them ; 

 and where this is wanting, the transparency prevents our 

 distinctly seeing either the canals or the diaphragms. 



It is a frequent condition of the T. Jlocculosa that this 

 coloured substance is found in one lateral half, and not 

 in the other, of the same canal ; hence the alternating 

 character of the so-called vittse. So there does not exist 

 a canal continuous to open extremities, which traverses 

 the frustules {bacilli), as authors describe and delineate, 

 but many apertures {fenestrce), arranged in a series, in 

 which bubbles of air remain imprisoned, unable to make 

 their escape without rupture of the shield. 



In \hQT.fenestrata, every individual has two diaphragms 

 when the development is complete. When deduplication 

 takes place, two individuals are produced, each having 

 one diaphragm. 



The second diaphragm afterwards appears, and, at a 

 later period, another deduplication. It is still a ques- 

 tion whether both the new individuals equally divide; 

 whether a new diaphragm appears in both or in one 

 only ; whether on the inner or outer side ; and whether 

 among the four individuals resulting from the second 

 deduplication, those are equally prolific which retain 

 the primitive diaphragms, or those which have new ones 

 only, — or both together? In the T. Jlocculosa it is rare 

 to see one diaphragm only ; and, again, it often happens 

 that two contiguous frustules {bacilli) possess different 

 numbers of them. Therefore, besides the inquiries indi- 

 cated above, another might be instituted on this species, 

 — whether or not the deduplication is always subsequent 



