166 GREVILLE, ON DIATOMACE. 
In Californian guano. 
This very interesting species is nearly related to C. turris, 
but differs in the larger, more truly cylindrical and truncate 
frustules, and in the considerably smaller areolation; the 
areolz in C. turris being 7, in C. turgida 11 in 001”. In 
both they are beautifully hexagonal. At the time my 
drawing was made I had only met with isolated frustules, 
but I have since seen them in union. The colour (in balsam- 
mounted slides) is very pale yellowish-brown; and the 
substance is so transparent, that it 1s easily overlooked. 
C.? ferox, n. sp., Grev.—Frustules oblong; valves sub- 
globose, campanulate, thin and hyaline at the suture; 
connecting processes spine-like; areolation large, hyaline, 
walls of the areole passing at the angles into short spinous 
processes, so as to give a hispid character to the frustules ; 
length :0020” to :0025"; breadth -0012” to -0015”"; areole 5 
im, OU (hice: loro.) 
In Californian guano. 
Single valves of this species are common, generally pre- 
senting themselves in a vertical position, in the form of a dise, 
when the connecting processes are lost sight of, but the 
campanulate or expanding margin is well seen. Occasionally 
entire frustules are observed ; but I have only twice had an 
opportunity of examining them in connexion. The species 
is so very different in habit from the other Cresswellie that 
it may eventually prove to belong to a distinct genus. The 
comparatively very large and strong areolation, suggestive of 
Polycystinee rather than Diatomacee,—the spine-hke con- 
necting processes, and the expanded margin of the valves 
forming a sharp keel at the line of suture, indicate other 
affinities. There seems also to be a striking difference in 
the substance.  C. turris and turgida have a more flexible 
appearance, the latter especially; while the present form 
is hyaline, and rigid, as if it would be readily fractured 
under pressure. The most remarkable feature about it, 
however, is the hispid character of the frustule when seen in 
profile, and which at first sight is exceedingly difficult to 
account for, as there are no traces of genuine spines. A 
minute examination has brought me to the conclusion that 
the effect is owing to the walls of the hexagonal areolz being 
produced at the angles into spine-like processes. 
After a careful study of the very brief characters given of 
the Pyzidicule by Ehrenberg and Kiitzing, Iam quite unable 
to identify the species now described with any one of them. 
