76 
GrevillEj on New Diatoms, 
this point. This diatom is so exceedingly peculiar that_, not- 
withstanding the deficiency of the front view, and, conse- 
quently, our ignorance of its precise place in the system_, I 
do not hesitate to make it known. 
Peponia Barbadensis, n. sp., Grev. (Fig. 25.) 
Hab. Barbadoes deposit, from Cambridge estate (not un- 
frequent) ; C. Johnson, Esq. 
Valve yellowish, more or less circular, the transverse dia- 
meter being often the longest, suddenly contracted at the 
opposite ends into a triangular, subacute apiculus. Whole 
surface filled up with rounded cellules, largest in the middle. 
In size the valves vary greatly, from "0020" to upwards of 
•0030" in diameter. Between the body of the disc and the 
apiculus there is the appearance of a septum. The apiculus 
itself is generally, but not always, destitute of cellules. 
The individual figured is the finest I have seen. 
Thaumatonema, nov. gen., Grev. 
Frustules united into a filament; valves disciform, each 
producing a central, forked process, which is articulated at 
the apices with that of the opposite valve. 
From the front- view position of the frustule it is difficult 
to ascertain precisely the nature of the structure of the dis- 
coid portion of the valve, but it appears to consist of radiating 
lines of minute puncta. The genus, I presume, must be 
placed among the Melosirece. 
Thaumatonema Barbadense, n. sp,, Grev. (Fig. 26.) 
Hab. Barbadoes deposit, from Cambridge estate, in a slide 
communicated by C. Johnson, Esq. 
An extraordinary diatom, unlke anything previously de- 
scribed. If the valves of Melosira Westii, instead of being 
joined by their narrow, truncated ends, were each provided 
with a central process widely forked, the apices of the forks 
somewhat dilated and articulated with the forks of the oppo- 
site valve, a tolerably accurate idea would . be conveyed of 
this singular production. Length of two valves, with pro- 
cesses in situ, -0038". Breadth of valve, and also of the space 
between the articulations of the forks, -0017". Length of 
process, including the fork, the one being about equal to the 
other, -0013". 
