whilst within the sack of the parent, vary in length from .007
to .009 in Lepas, to .023 of an inch in Scalpellum: my chief examination
of these larvae has been confined to those of _Scalpellum vulgare_; but I
saw them in all the other genera. The larva is somewhat depressed, but
nearly globular; the carapace anteriorly is truncated, with lateral
horns; the sternal surface is flat and broad, and formed of thinner
membrane than the dorsal. The horns just alluded to are long in Lepas
and short in Scalpellum; their ends are either rounded and excessively
transparent, or, as in Ibla, furnished with an abrupt, minute, sharp
point: within these horns, I distinctly saw a long filiformed organ,
bearing excessively fine hairs in lines, so exactly like the long
plumose spines on the prehensile antennae of the larvae in the last stage;
that I have not the least doubt, that these horns are the cases in which
antennae are in process of formation. Posteriorly to them, on the
sternal surface, near each other, there are two other minute, doubly
curved, pointed horns, about .004 in length, directed posteriorly; and
within these I again saw a most delicate articulated filiformed organ on
a thicker pedicel: in an excellent drawing, by Mr. C. S. Bate, of the
larva of a Chthamalus (_Balanus punctatus_ of British authors), after
having kept alive and moulted once, these organs are distinctly shown as
articulated antennae (without a case), directed forwards: hence, before
the first moult in Scalpellum, we have two pair of antennae in process of
formation. Anteriorly to the bases of these smaller antennae is seated
the heart-shaped eye, (as I believe it to be,) .001 in diameter, with
apparently a single lens, surrounded, except at the apex, by
dark-reddish pigment-cells. In some cases, as in some species of Lepas,
the larvae, when first excluded from the egg, have not an eye, or a very
imperfect one.

There are three pairs of limbs, seated close together in a longitudinal
line, but some way apart in a transverse direction: the first pair
always consists of a single spinose ramus, it is not articulated in
Scalpellum, but is multi-articulate in some genera; it is directed
forwards. The other two pair have each two rami, supported on a common
haunch or pedicel: in both pair, the longer ramus is multi-articulate,
and the shorter ramus is without articulations, or with only traces of
them: the longer spines borne on these limbs (at least, in Scalpel