ANIMAL NATURE OF DIATOME^E. 387 



Taking also into account the two doubtful species (C.minu- 

 tula, C.Rotuld), which Kiitzing justly suspects may be frus- 

 tules of Melosira, we have thus nine species in this genus. 



11. PYXIDICULA. Individua singularia velbinatim con- 

 juncta, non concatenata, libera vel versilia ; latus prima- 

 rium obsoletum (nullum] later a secondaries convexa. (Lorica 

 bivalvis, valvis convexis, annulo interstiali destituto.) 



It is really surprising that although Agardh, Kiitzing, 

 Ehrenberg, and Brebisson, have described and figured the 

 Cymbella or Frustulia operculata as the type of this 

 genus, and corresponding perfectly with the foregoing 

 generic character Kiitzing himself, without acquainting 

 us with the mistake of others, or his own, should now 

 form the type of the preceding genus from this same 

 species, preserving the name at first proposed, but 

 altering the sense ; and at the same time should describe, 

 as a new species of Pyxidicula, the P. operculata of Bailey, 

 which seems either identical with that of Ehrenberg, or 

 very similar to it, and to that formerly so described and 

 figured by the authors named above. The fact is, that 

 the specimens of Brebisson and Lenormand, in my pos- 

 session, belong to Cyclotella, and not to Pyxidicula. The 

 other species (P. major) described and figured by Bailey 

 and Kiitzing, seems rightly to belong to the tribe of 

 areolated Diatomeae. Of the species included in flints (P. 

 globata, P.prisca,) we can say nothing organographically, 

 so uncertain is their nature. The P. adriatica is a being 

 equally singular, of which we know nothing in an organo- 

 graphical and physical point of view. The same may be said 

 of Ehrenberg's two genera (Goniothecium, Rhizosolenia), 

 which Kiitzing places doubtfully at the end of Pyxidiculte. 



12. PODODISCUS. Individua singularia vel concate- 

 nata, stipitata ; latus primarium obsoletum (nutturn) latera 

 secondaria convexa. Stipes lateralis. (Lorica bivalvis, 

 stipitata valvis convexis, subhemisphericis.) 



The only species (P. jamaicensis) is too incompletely 

 known for anything positive to be inferred. 



