WALLICH, ON TRICERATIUM. 253 
Cocconema and Gomphonema, which are, in like manner 
distinguished by these remarkable appendages. 
The only genera with which Hydrosera can at all be con- 
founded, are—Terpsinoé of Ehrenberg, Anaulus of the same 
author, and Tetragramma of Professor Bailey ; the last being 
in reality, however, nothing more than a variety of Terpsinoé 
musica, and therefore not demanding further notice. 
In Terpsinoé the frustules are described as “ tabular and 
obsoletely stipitate,’ a character which might apply to 
H. compressa, but which at once fails in H. triquetra. The 
filaments, however, assume a “ zig zag”’ form, and the cellular 
structure is “ very minutely punctate,’ with no appearance of 
reticulation. 
I admit that it was a question resting chiefly on how far 
H. triquetra can be safely separated from [7. compressa, that 
induced one to remove the latter species from Terpsinoé, to 
which it bears a strong resemblance in its “tabular” or 
rather compressed form, but from which it differs materially 
in the presence of the lateral appendages, the spinous angles, 
and the direct nature of its filament. 
In Lithodesmium the valves are described as triangular, but 
they are distinguished from those of the present genus by 
their “extreme smoothness,” transparency, and their not 
being cellulate. Two sides only beimg symmetrical and 
“ undulated,” whilst the third “is doubly excised or notched.” 
Lastly, in Anaulus “the frustules never form a filament, but 
are single, and neither furnished with tubular processes, 
nodules or apertures.* The separation of Hydrosera is how- 
ever completed, I submit, by the presence of the very re- 
markable appendages I have described, and which afford a 
character so very distinct from what is to be seen in any 
other alluded genera. 
* Vide ‘Micrographic Dictionary,’ and Kiitzing, “Species Algar.” 
