22 ROPER, ON BRITISH MARINE DIATOMACES. 
valve so hyaline, when mounted in balsam, that it has probably 
hitherto escaped notice, from being considered a detached 
ring or connecting membrane of C. radiatus or eccentricus. 
I was at first inclined to refer this species to Podosira, but 
the slight convexity of the valve, and the absence of the appa- 
rent perforation at the apex characteristic of that genus, are, 
T consider, sufficient to preclude its bemg so classed. The 
frustules being always separate and never in filaments, distin- 
guish it from Melosira, and the want of any process or 
pseudo-nodule separate it from Eupodiscus It differs from 
the finest-marked specimens of Coscinodiscus eccentricus 1m 
the absence of the eccentric limes and spinous margin, and 
from all other species of that genus in not having distinct 
cellular markings. 
Coscinodiscus (?) ovalis,n.s.— Valves oval, with finely dotted 
striz radiating from the centre to the circumference ; of a dull 
slate colour when dry, and light brown in balsam. Length 
00158" to :0023" ; breadth -00128” to :00149” (fig. 4). 
Marine. Caldy, Pembrokeshire, and dredged off Tenby, 
Rev. J. Guillemard. 
The valves of this species occur abundantly in the Caldy 
gathering, and in the washings of Vesicularia dredged in five 
fathoms off Tenby. Professor Smith informs me that he does 
not see any satisfactory evidence for referring this species to 
the Diatomacee ; and although I differ from so high an au- 
thority with great reluctance, I still record it, though with 
some doubts as to its generic position, in the hopes that the 
attention of observers in other localities may be directed to 
it, in order to clear up the doubtful poimts in its structure. 
That it belongs to the Diatomacez I think admits of little 
doubt ; the frustules are siliceous, composed of two valves 
very slightly convex, and occur abundantly in gatherings, 
almost confined to various species of marme Diatoms. The 
radiating strize on the surface of both valves are delicate, and 
require a magnifying power of 400 diameters to make them 
out satisfactorily; but the arrangement of the dots or 
cellules is very similar to that of many other species of the 
class. 
The general outline of the valves agrees with that of some 
species of Cocconeis, but the absence of a median line and 
central nodule separate them from that genus. It is very 
probable that it may be entitled to rank as a distinct genus ; 
but as I have not had any opportunity of examining it in a 
living state, I place it provisionally in Coscinodiscus, to 
which, in general structure, it appears most closely allied. 
