CrEBViLLEj Oil Ntw Diatoms. 
27 
auy thing about the production of sporangial frustules in this 
group. Were we to assume that the frustules which have 
given rise to these remarks are indeed sporangial, it would 
not account for the deviation from the usual characters 1 have 
pointed out. It would certainly be a remarkable circum- 
stance, that, if sporangial, none should appear but in the least 
developed form of the species. Although I have examined a 
considerable number of individuals like the smaller of the 
two valves I have figured, I have not seen any of a size inter- 
mediate between them and the larger one, which is so enor- 
mous that it might well give rise to some suspicion of error 
in measurement. 
Triceeatium. 
Triceratium conveooumj n. sp,, Grev. — V alve with very con- 
vex sides, narrow, obtuse, angles, and small pseudo-nodules ; 
granules faint and minute, in radiating lines, 16 in '001". 
(PI. Ill, fig. 6.) 
Hab, Ceylon, Dr. Macrae. 
I am quite aware that the Triceratium I have just defined 
Avill, at a hasty glance, be regarded as occupying a rather 
ambiguous position. I myself thought that it was T. orbicu- 
latum of Shadbolt; but the precise statement of that ob- 
server, that his species exhibited a structure similar to that 
of Coscinodiscus radiatus,'' only less regularly hexagonal, and 
that the reticulations were " largest at the centre, and 
diminished in size gradually towards the margin of the 
valve,^^ led me to hesitate ; for in my present species the 
puncta are so minute that they could never be termed reti- 
culations.''^ And, more than that, they gradually increase in 
size from the centre to the margin, exactly opposite to the 
rule in T. 07'biculatim. Then, again, the lines of puncta in 
the present species, minute though they be, are distinctly 
radiating ; whereas in T, orbiculatum they are represented as 
following no order. And Mr. TufFen West, who engraved 
the figure from nature Mic. Trans.,^ vol. ii, pi. i, fig. 6), 
could never have committed such an error had they been de- 
cidedly radiating. However closely, therefore, the two diatoms 
resemble each other in general aspect, they cannot be united. 
While upon this subject I maybe allowed to express my con- 
viction that T. orbiculatum of Shadbolt and the diatom pub- 
lished under the same name by Mr. Brightwell are really 
distinct. The difference in the shape of the valves is con- 
siderable. In T. orbiculatum the puncta, as we have seen, 
do not radiate* The pseudo-nodules, judging from the 
