Greville^ on New Diatoms. 
33 
tive feature. The minute^, circular, brilliant dots are arranged 
in radiating lines, but at considerable distances both between 
the lines and from each other ; and occasionally the lines are 
interrupted, as if one or more dots had dropped out. The 
whole substance is somewhat hyaline. The margin is stri- 
ated, or rather, the terminations of the strise of the frustule, 
as seen in the front view, appear more like puncta. Dis- 
tance between the angles *0045'^ 
Triceratium araneosum, n. sp., Grev. — Minute ; valve with 
somewhat convex sides and rounded angles ; central portion 
filled with a fine, cobweb -like, irregular network of unequal 
cellules, which are connected with the margin by a few short 
lines. (Fig. 17.) 
Hab. Barbadoes deposit, Cambridge estate ; in slides com- 
municated by C. Johnson, Esq. 
The only species with which the present diatom can be 
contrasted is T. labyrintlmum, which is double the size ; and 
although it possesses a network of cellules not reaching to 
the margin, the cellules are much larger, of a totally different 
shape, and filled with puncta. 
These two curious diatoms resemble each other only in 
outline, and in the circumstance that the cellules form an in- 
ternal cluster, which is connected with the margin by a few 
radiating lin^s. Distance between the angles 'OOIG'". 
Entogonia. 
Entogonia elegans, n. sp., Grev. — Valve with convex sides 
and somewhat produced, obtuse angles ; the border-compart- 
ments filled with very numerous, minute cellules ; central 
triangle with regular radiating costse. (Fig. 19.) 
Hab. Barbadoes deposit, Cambridge estate; C. Johnson, 
Esq. 
The convex sides and produced angles give an outline to 
the valve distinct from that of every other species. In some 
respects it approaches E. marginata, but the lines of the 
inner triangle are not moniliform, and the cellules of the 
border- compartments are much more numerous and more 
closely disposed. In the latter character it differs from E. 
pulcherrima, as well as in the very conspicuous pseudo- 
nodular, circular, blank space. The regularly radiating 
costse, independently of other characters, separates it from 
E. amabile, in which the costae are more or less interrupted. 
In the specimen I have drawn, a number of spines may be 
seen, mostly near the margin, but they are of no diagnostic 
value. Spines present themselves occasionally in various 
diatoms, as, for example, in Eupodiscus Jonesianus, to which 
