GttEviLLE^ on New Diuloin^. 
space and the margin^ and again bent back near the margin 
at each end; costse rather distant, 4 in -001". Diameter 
•0036". (PI. Ill, fig. 2.) 
Hah. Ceylon, Dr. Macrae. 
This species belongs to the second section of Ealfs^ first divi- 
sion ( ^ Pritch. Infus./ 1861), and is clearly different from any 
one there described. In the remarkable nndulation of the 
surface of the disc, it is nearly allied to Gregory^s C. angu- 
laris. By the first bending, a line or ridge is produced on 
each side, Avhich extends in a curve from one end of the 
median space to the other. The second bending of the disc 
becomes most evident towards the ends, where the edge is 
bent back as in C. angularis. The costse are given off from 
the median line at right angles, throughout its whole length, 
and the latter is remarkably clearly defined. In all the 
specimens I have seen there is a tendency in a number of 
the central costse to become forked at the margin, but as the 
apices again unite, it would be more correct to say that they 
formed little loops. The whole frustule has a very diapha- 
nous and brilliant appearance, with the markings sharply 
defined. 
DiCTYOPYxis {Elir.), Grev. 
IVustules two-valved, cellulose, cohering by simple, more 
or less convex, junction-surfaces, into a filament. Valves 
cup-like, destitute of any silicious connecting band. 
In consequence of the discovery by the Rev. R. Cresswell 
of the diatom which noAv bears the name of Cresivellia turris, 
Professor Walker- Arnott and myself were enabled to define 
with some precision a little group of Pyccidiculcs, which 
Ehrenberg had partially indicated, but, from want of perfect 
materials, had left in great obscurity. Of his genus Stephano- 
pyxis, no two of the four species referred to it appear 
certainly to agree generically with each other ; and as his 
character was drawn up in ignorance of the perfect state of 
any of them, and as, besides, it was not known which of 
them was to be regarded as the type, we adopted what 
seemed, under the circumstances, to be the most satisfactory 
course, and constituted the genus Cresivellia upon good and 
perfect specimens. Three other species have been subse- 
quently added by myself, and my friend, Mr. Ralfs (who 
sustains Ehrenberg's name), mentions under Stephanopyxis 
diadema ( ^ Prit. Infus. ,^ 4th edit., p. 826) that he has seen 
two frustules connected by the coronal spines. It con- 
sequently, Avithout doubt, belongs to the same group. 
