Wallich, on the Diatom-valve. 
139 
striation of Navicular and the canaliculate structure of 
Surirella, associated on distinct portions of the same valve. 
In a specimen in my possession, from Bengal, and a similar 
one from the Channel Islands, the central third of the valve 
is distinctly striated, whereas the two outer thirds exhibit 
the canaliculate character, in respect of which this form has 
been referred to Surirella. In vol. ii, p. 97, of the ^ Journal 
of the Microscopical Society,' Professor Gregory has described 
a similar form as occurring in the ^' Mull deposit and 
alludes to the so-called " canaliculi being ''^bars." Under 
my view of the structure in Pinnularia, to which genus the 
diatom in question bears much closer resemblance than to 
Surirella, the spaces between the bars are the analogues of 
the " costse,'' whilst the bars constitute the intra-costal 
spaces. I would here beg leave to state that, in retaining 
the term costse," as ordinarily applied, which is much more 
applicable to the intra-costal spaces than to the parts which 
have hitherto received it, I am guided solely by the desire of 
avoiding inconvenience invariably attendant on changes of 
the kind. 
In Pleurosigma, on the other hand, the intra-linear spaces 
constitute the strongest portions of the valvular plate. In 
P. formosum there exists good evidence to prove that the 
intra-linear spaces are occi)ipied by elevated rhomboidal 
papillae, which present facetted surfaces ; whereas in P. 
Balticum, instead of rhomboidal elevations, we have four- 
sided flattened pyramids, presenting, as in the former case, 
four sets of lines, of which those bounding the spaces, and 
not those crossing them, are the predominant ones. 
Again, taking into consideration the secondary internal 
valvular plate, the existence of which is constantly seen in 
some species, it is not improbable that such a structure may 
occur throughout the whole family, although incapable of 
separation in most examples. From the modified impress of 
the markings of the external plate found upon that beneath 
it — as, for example, in Cocconcis — it is clear that, to a 
certain extent at all events, the markings of the external 
surface of the primary plate are traceable on its internal 
aspect. Much additional evidence must, however, be forth- 
coming, before this question can be satisfactorily settled. 
A good deal of misconception has arisen from the sup- 
posed analogy between the markings in Triceratiumj in which 
hexagonal structure really occurs, and those in Pleurosigma, 
in which the appearance of hexagonal cellulation is only 
observable under deceptive instrumental adjustments. In 
the one case, the hexagonal spaces constitute the thinnest 
