WALLICH, ON TRICERATIUM. 245 
towards the same extremity of the frustule on all four sides. 
Valve slightly convex on its surface; cornua well defined, and 
projecting in both front and end views ; spines elongate, not 
marginally disposed; valve deeply constricted between the 
bases of the cornua and its free margin, which is everted ; 
connecting membrane projecting boldly. Length -0094; 
breadth ‘0063 ; diameter of each side of valve ‘0059 to ‘0070. 
Cellulation 9 to 11 in ‘001. Taking the characters given 
in the ‘ Synopsis’ as our guide, it would be more easy to recon- 
cile this form with Amp/hitetras than with Triceratium, inas- 
much as “the cubical outline,’ to quote the text, “ distin- 
guishes it from all other forms.” - But in Amphitetras, the 
frustules cohere into a zigzag filament ; the connecting mem- 
brane is imperfectly annulate and indefinite; the cellules are 
circular and inconspicuous at the angles of the valve. 
Whereas in the St. Helena species, the frustules never form 
a filament: the connecting membrane is definite and consists 
of four distinct plates; the cells on both the valves and con- 
necting bands are similarly marked with a conspicuous and 
regular hexagonal cellulation ; and lastly there exist the well- 
developed cornua and spines not seen in Amphitetras. 
Again, viewed as a Triceratium, the chief distinguishing 
type of that genus falls to the ground; for whilst the species 
under notice occurs abundantly in the locality named, in no 
instance has a three-sided frustule presented itself. There 
must be some limit to type, and therefore when the character 
fails, as it is here shown to do, upon which the individuality 
of a genus in a great measure rests, the alternative remains 
of either cancelling that character, or of separating the form 
in which so constant an anomaly exists. In this instance it 
must be borne in mind that the four-sided form is therefore 
the typical one, and yet that analogies of structure clearly 
indicate its position amongst the Triceratia. 
The nearest approach to its characters, I find in the 
‘Smithsonian Contributions, entitled ‘Notes of New Species 
and Localities of Microscopic Organisms,’ by Professor 
Bailey, of New York. A plate is there given of a frustule 
of Triceratium setigerum which, on a cursory examination, 
might be considered identical with the St. Helena Diatom. 
But, in the first place, it is to be inferred that the connecting 
band in that species offers no peculiarity, inasmuch as no 
allusion is made to such ; and, in the next, the characters given 
indicate its distinctness, the “bases,” as they are termed, 
‘being triangular, bearing three large obtuse projections or 
horns, at the base of each of which is placed a setiform 
process.” 
