GREVILLE, ON DIATOMACES. 165 
Kiitzing seems to have been under this impression, when he 
reunited the whole under the old name (‘Sp. Alg.,’ 1849). 
But this proceeding had only the effect of perpetuating for a 
time the union of a number of objects whose history and 
habits continued to be enveloped in mystery, and many of 
which had manifestly no real affinity with each other. Were 
we now constrained to select one of Ehrenberg’s sub-genera 
as aresting-place for Cresswellia turris, and another species to 
be immediately described, we should have to take Stephano- 
pyxis, of which the author gives the following character : 
“ Pyxidicule generis bivalves turgid aut subglobosz 
forme, que valvularum teste structura cellulosa msignes 
sunt et denticulorum, aculeorum aut membrane coronam in 
media quavis valvula gerunt in hoc Pywidicule subgenere 
colliguntur.” 
Now this character was prepared with special reference to 
certain existing forms; and if we examine these forms, we 
shall perceive how indefinite the character becomes. Sz. 
appendiculata (‘ Mikrogeol.,’ tab. xviii, fig. 4) is distinguished 
by the valves being furnished with only a single short blunt 
horn “extra medium posito.” St. cristata (l. c., tab. xvii, 
fig. 6) appears to have a sort of limbus surrounding the 
whole frustule,—nothing in the shape of a “corona” of 
any kind. St. aculeata is said to be “undique parvis 
aculeis hispida, nec cellulosa.”” With regard to the only 
remaining species, St. diadema, of which no figure has ever 
been published, it would appear to come generically nearer 
to our Cresswellia turris, but we are not aware that it has 
been seen in a perfect state or whether the frustules are 
united in chains. Kiitzing, in his generic character, says 
positively of all the Pywidicule that they are “ non-con- 
catenata,”’ thereby excluding Cresswellia. Upon the whole, 
therefore, seeing how ill the species of Stephanopyxis agree 
generically among themselves, and that we do not even know 
which of them was regarded as the type, we think that by 
establishing the genus Cresswellia we have adopted the only 
course open to us. It was impossible for us to place our 
new Diatom in Stephanopyxis without altering the character 
to such an extent, that while the new species was admitted, 
all those for which it was actually constituted would be 
either excluded, or at best doubtfully retained. Such a 
proceeding would have been, of course, inadmissible. 
C. turgida, n. sp., Grev.—Frustules cylindrical oblong, 
obtusely truncate; connecting processes dilated ; length about 
0032”; breadth :0021”; areolee 11 in ‘001”. (Fig. 14.) 
