_Variation._--The species vary in shape nearly, but not quite as much
as in Balanus, for we very rarely here see cylindrical varieties; but
in the same species, we have extremely depressed, steeply conical, and
convex forms; the orifice varies in relative size; the state of the
surface, whether ribbed or smooth, whether well preserved or corroded,
the upfilled parietal tubes being thus exposed, varies more than in
Balanus. The colour also varies; specimens of _T. porosa_ (Pl. 10, fig.
1 _a_ to 1 _f_) being dark purple, or even inky black, white, pale
pinkish-purple, and green: as far as I have seen, the pinkish varieties
of _T. porosa_ are confined to the eastern half of the world. Even the
ridges and crests on the under side of the scutum vary in some degree;
and the tergum in _T. porosa_ varies considerably, and in some of the
other species, slightly, in general shape: I believe that the tergum
becomes narrow and elongated, when the shell is steeply conical, with
the orifice of small diameter. After having spent several days in
disarticulating and closely examining the many specimens of _T. porosa_
in my possession, I concluded that its varieties formed at least four
species, these being in external appearance extremely distinct; but
ultimately the many intermediate forms compelled me to unite all into
a single species. Again, I may instance the conical, ribbed variety
of _T. purpurascens_ (Pl. 11, fig. 1 _b_), with the outer lamina of
the shell preserved and with the radii widely developed, as having not
the smallest resemblance to the other common depressed variety (fig. 1
_a_), having a granulated surface, produced by the exposed tips of the
upfilled parietal tubes, and without a trace of the radii or even of
the sutures; I may add that, according to the characters used by some
authors, these two varieties would be classed in two separate genera.
To distinguish the species of this genus, the chief reliance must be
placed (as in the case of most other sessile cirripedes) on the general
outline of the opercular valves, and on the ridges and crests on their
under sides: the internal structure, however, of the radii, and in
a lesser degree of the parietes, must not be overlooked. I have not
found the parts of the mouth, or the differences in the cirri, of much
service; and it will be shown under _T. porosa_, that the relative
numbers of the segments in the several cirri, and even their shape, is
extraordinarily variable.
_Affinities._--All the species are pretty closely allied, and there is
no ground for making any sectional division of the eight species, more
especially not on the ground whether or no the radii are developed.
_Tetraclita purpurascens_, taking into account all the characters,
including the animal's body, is perhaps the most distinct species in
the genus, but it is closely allied to _T. costata_. In the well-marked
character of the parietes being formed of a single row of large tubes,
_T. rosea_ differs from all the other species. The genus is closely
allied to Balanus; I can point out no difference in the animal's
body, nor any constant difference in the opercular valves. The ridges
and crests on the under sides of the scuta are more prominent in
Tetraclita; and all the species, excepting two, have crests (though
sometimes very slight) for the attachment of the rostral depressor
muscle, and these occur only in two species of Balanus: crests, also,
for the attachment of the lateral depressor muscle are common in
Tetraclita, but rare in Balanus. The main character, however, of the
genus, as compared with Balanus, is derived from the existence of
only four compartments, and in a lesser degree from the several rows
of parietal pores; but, as just stated, _T. rosea_ has only a single
row, and, on the other hand, in _Balanus crenatus_, there is a slight
tendency exhibited, in the dividing of the longitudinal septa near
the outer lamina, to form a second row of parietal pores; and in _B.
cariosus_, moreover, we actually have two or three rows of irregular
and very short pores. The thin, yet solid calcareous basis which
occurs in all the species of Tetraclita, excepting _T. purpurascens_,
resembles the basis in _Balanus flosculus_ and _imperator_, but I
suspect that the structure of the cement-ducts is different in the two
genera.