68 Transactions of the Royal Microscopical Society. 
the valves attain a great relative depth, we find not only that the 
connecting zones are more largely developed, but that the valves are 
furnished with a constricted rim, or groove, to which the margin 
of the connecting zone belonging to each valve is attached, as if to 
afford a more fixed point of resistance from which it may extend 
itself. ji 
In Amphitetras and certain forms of Triceratium and Biddul- 
phia, I went on to say, the undeviating presence of marginal rows 
of puncta on that part of the connecting zone in close proximity 
to the suture between it and the valve, proves that the growth of 
each half of the connecting zone takes place only at the margin 
farthest from the valve to which it belongs. For, were it not so, 
the rows of puncta would recede farther and farther from the 
“sutures,” as the zones increase in depth, a result which never 
follows. Growth takes place in both zones simultaneously, the 
actual extent of the overlapping being dependent on the rate at 
which the new valves, and endochrome within, happen to be de- 
veloped. In the newly detached frustules (notably in such genera 
as Amphitetras and Biddulphia), one valve, namely, the newly 
formed one, may still be seen enveloped by the half of the connect- 
ing zone of the parent frustule within which it was generated ; 
this half remaining adherent sometimes for a considerable period. 
I believe that a similar structure obtains in nearly all genera, 
although more easily traceable in some than in others, owing to 
their greater size and bolder outlines and markings. It may be 
thus seen in Himantidium, Odontidium, Denticula, Eunotia, 
Grammatophora, Amphitetras, Biddulphia, Isthmia, Hydrosira, 
Ooscinodiscus, &c. I added, in passing, that the beautifully 
executed figures in ‘The Synopsis of British Diatomacex’ show 
distinctly the average appearances of all these genera, with excep- 
tion of Hydrosira (which was found by me in Bengal after that 
work was published) ; but that the author had evidently failed to 
recognize either the true structure or important uses of this por- 
tion of these organisms.* 
Again, in my observations “On the Structure of the Diatom 
Valve ” (1860), | dwelt on the share taken in the secretion of the 
siliceous valve by the “ primordial utricle,” though admitting that 
our knowledge on the subject was singularly meagre. We knew, 
for instance, that the frustules of the Diatomacez, like the fronds 
of the Desmidiacew, are encased sometimes in a gelatinous in- 
vestiture ; and that, in some genera, this investiture is largely 
* It is extremely difficult, sometimes impossible, to demonstrate the presence 
of the two halves of the connecting zone in the more delicate and minute navicu- 
loid and discoid forms; but I have always succeeded in assuring myself that they 
are invariably present,—in other words, that there is no such thing as a ‘ hoop,” 
in the sense of its being single. 
a 
