some species these interruptions are more numerous, being found
along the entire line of suture, and are often connected with
minute canals, hollowed out between the siliceous epiderm and
internal cell-membrane, and apparently formed by waved flexures
of the epidermal envelope. These are noticed under the genus
Epithemia, p. 11. They are very conspicuous in Epithemia
lonyicornis, Supp. Plate XXX. fig. 247, and form distinctive characters
in the genera Siirirella and Campylodiscus. I have employed
the term “ Foramina,” in reference to these perforations
in the siliceous epiderm, and the epithet “ Canaliculi,” to designate
the grooves into which they occasionally open.
The connecting membrane, being of later formation, is usually
less firmly siliceous than the epidermal valves, and rarely presents
much complexity of structure. Not so with the valves
themselves; we have here the greatest variety of form and
diversity of structural arrangement.
These variations in form and structure will be best understood
by an inspection of the figures, and will be noticed more particularly
under each genus. I shall confine my remarks at present
to a few general characters.
In all, we find the valves presenting a surface of a greater or
less degree of convexity. This follows as a necessary consequence,
from their enclosing, when applied to each other, the
space occupied by the internal cell. This convexity is often
that of a regularly curved surface, forming a segment of a circle
of a greater or less radius. The segment is occasionally that of
a perfect hemisphere, and the valves, when applied to each
other, form a miniature globe; but more frequently the curvature
is slight, and extends over a part of the valve only, which
then appears as a portion of a hollow tube, closed at one end,
and which may be either cylindrical or compressed, triangular
or quadrilateral.
The intimate structure of the valve presents an amazing
variety, in every genus, and almost in every species, offering
distinct features, which serve as the best and most facile means
of distinction and identification. Striai or lines frequently mo-
niliform, dots arranged in a radiate or concentric manner, and
minute divisions presenting perfectly hexagonal outlines, are
among the most frequent appearances. Great diversity of opinion
has prevailed respecting the nature of these markings, and
various functions have been assigned to them, as they have been
regarded as apertm’es, protuberances, or depressions. I am,
however, disposed to regard them all as modifications in the
arrangement of the silex of the valve, arising from the mode of
development peculiar in each case to the membrane with whicli
the silex is combined. It is a well-established fact, that all increase
in vegetable and animal tissues takes place by the production
and addition of cells; and the law seems to prevail even
in the formation of the envelopes of the minute organisms witli
which we are now concerned. Certainly no one can look at the
structure exhibited by the siliceous valves of Triceratium Pavm,
PL V. fig. 44, and those of Isthmia enervis or I. nervosa, to be
given in the second volume of the present work, without feeling
assured that he has before him the representation of a cellular
membrane, precisely homologous with the epidermal tissues of
many vegetable organs. An object-glass of sufficient power
and definition shows the same structure in the valves of many
species of Pleurosigma; and the generalization is at once forced
upon the observer, that the valvular markings in every case arise
from modifications of cellular tissue,—a conclusion which a
fuither and closer examination does much to confirm. In no case
have I been able to satisfy myself of the existence of perforations
m the general surface of the valves; and the existence of depressions
or elevations (except in a few cases, when such appearances
arise from the wavy outline of the surface,) seems to me to be
equally problematical; but no difficulty presents itself to the
supposition, that the moniliform striae of Epithemia, Navicda,
and others, the circular markings of Coscinodiscus eccentricus,
1 late III. fig. 38, and the irregular star-like structure of Eupo-
dtscus Argus, Plate IV. fig. 39, are all modifications of cellulai-
tissue; and even in the costae of Pinnularia and the unresolvable
Eupodiscus sculptus, Plate IV. fig. 42, and others, it is
not difficult to conceive that we have confluent cells whose union
gives rise to the appearance of lines or bands.
With the appearances thus arising from the cellular structure
0 le tissue with which the silex is combined, we must not coiib