profundi and also the encephalic branch of th e jugular vein. The antero-ventral edge of th e trigemino-
facialis chamber is either notched, or perforated by a foramen which transmits th e palatinus facialis
from the chamber into the myodome, this nerve here, as in Trigla, issuing from th e cranial cavity
through the facialis foramen an d th en running forward along the floor of th e trigemino-facialis chamber.
The edge of th e o rbital portion of the bone, dorsal to its mesial process, is notched to transmit the
oculomotorius, and dorsal to this notch, near the ventral edge of th e alisphenoid, there is. a second
b u t shallower notch for the nervus trochlearis.
The MYODOME has prootic and basioccipital portions, and, excepting in th a t th e basisphenoid
bone and th e prepituitary portions of the mesial processes of th e prootics are replaced by
membrane, th e canal is th e exact equivalent of th e canals of Trigla and Scorpaena. There being
no basisphenoid bone, th e myodome, in th e prepared skull, opens into th e hind end of th e orbit by
a wide median opening, bounded, on either side, b y the ascending process of th e parasphenoid. In
th e middle line of th e floor of th e opening there is the median tooth-like process of the parasphenoid.
The hypophysial fenestra extends nearly th e full length of the myodome, b u t is much narrower in
th e basioccipital region th a n in the prootic. The fenestra is completely closed by th e underlying
parasphenoid, th e myodome n o t opening, posteriorly, on th e ventral surface of th e skull. The roof
of th e basioccipital portion of th e canal is formed by a th in plate of bone, which separates this p a rt
of th e myodome from th e overlying cavum sinus imparis. The cavum sinus imparis extends posteriorly
slightly further th a n the myodome, the pointed ends of b o th canals being directed toward
th e point of th e conical vertebra-like depression on th e hind end of the basioccipital.
The BASIOCCIPITAL is normal, b u t presents, in median-vertical section, a marked feature.
The shallow conical vertebra-like depression in the hind end. of th e bone, in such sections, is lined
by a superficial layer of dense bone differing markedly in appearance from the deeper portions of
th e bone. A thinner layer of similar bone lines th e deeper conical depression, in th e anterior end
of th e bone, th a t forms th e hind end of th e myodome. The ends of these two cones approach each
other, and the dense bone lining them is continued, in th e middle line of th e bone, from one cone
to th e other. In Scomber I described (’03, p. 102) a similar b u t much less pronounced line, which,
in th a t fish, connected th e bottom of the cavum sinus imparis with th e vertebra-like depression in
th e hind end of th e basioccipital, and I said th a t this seemed to indicate th a t th e cavum sinus imparis
might be the remnant of the anterior conical depression of a vertebral body. In Peristedion it is not
th e cavum sinus imparis, b u t th e hind end of th e myodome, th a t has th e appearance of being such
a depression on the anterior surface of a vertebral element; and if it be such a depression, it would
offer a rational explanation of th e basioccipital extension of th e myodome. Tha t this extension
of th e myodome is due simply to the fact th a t th e rectus externus, deriving great advantage from
a slight additional posterior shifting of its point of origin, has extensively excavated th e basioccipital,
has never appealed to me. And if a simple posterior extension of its point of origin is of such considerable
advantage to th e rectus externus, why should it n o t also be of some advantage to the rectus
internus, which muscle, in Scomber and in all of the mail-cheeked fishes th a t I have examined,
never acquires this posterior extension? But, if there were a pre-existing depression in the anterior
end of the basioccipital, its occupation and subsequent enlargement by one only of the two muscles
would seem most natural. Similar reasoning, applied to th e prootic, would account for the
origin of the prootic p a rt of the myodome, as will be further discussed in the section devoted to the
myodome of fishes.
The cavum sinus imparis extends, in Peristedion, b u t slightly beyond the middle of the length
of the basioccipital, th e bone posterior to it having a relatively broad, concave dorsal surface, the
posterior portion of which forms the ventral boundary of the foramen magnum and the floor of the
cranial cavity immediately anterior to th a t foramen. Lateral to the cavum sinus imparis the bone
lodges a small portion of th e short subcircular saccular groove.
The basioccipital is, as usual, bounded dorsally, on either side, by th e exoccipital, and anteriorly
by the prootic. Ventrally i t is overlapped externally by the parasphenoid.
The EXOCCIPITAL is bounded by th e basioccipital, prootic, pterotic, opisthotic, epiotic
and supraoccipital, and it is perforated by separate foramina for th e glossopharyngeus, vagus and
occipital nerves. The vagus and occipital foramina have positions similar to those in Trigla, the
vagus foramen, in all my specimens, being divided into two parts by a transverse bar of bone. The
glossopharyngeus foramen lies directly anterior to th e vagus foramen, a t one half or two thirds the
distance to th e anterior edge of the bone. Immediately dorsal to the vagus and glossopharyngeus
foramina there is a slight horizontal ridge along the outer surface of the bone; and dorsal to this
ridge, nearly th e entire lateral surface of the brain case is occupied b y a large subtemporal depression
which, as in th e other fishes of th e group, gives origin to the adductor hyomandibularis and adductor
operculi muscles, and probably also to the fourth and fifth levators of the branchial arches; b u t the
origins of th e levator muscles of the branchial arches were n o t investigated. The fossa on th e prootic,
so well developed in Scorpaena, is apparently represented, in Peristedion, by a slight groove along
th e anterior edge of th e subtemporal depression.
On th e internal surface of th e exoccipital there is a mesial process, b u t i t has almost completely
coalesced with th e lateral wall of the bone, th u s here giving to the bone a thick and distinctly double
ventral edge which suturates with th e basioccipital. A t th e anterior end of this thick ventral edge,
th e mesial process separates slightly from the side wall of the bone and so bounds a small dorso^
posterior portion of th e saccular groove.
The OPISTHOTIC forms p a rt of th e th in ventro-laterally projecting portion of the posterolateral
edge of the skull. I t lies in a nearly transverse position, filling a large and somewhat square
interval between th e ventral edge of th e posterior process of the pterotic and a right-angled incisure
in th a t portion of th e exoccipital th a t forms p a rt of th e postero-lateral edge of the skull. Because
of its nearly transverse position, the bone forms p a rt of the flat posterior surface of the skull. A
process on its postero-mesial surface gives articulation to th e opisthotic process of the suprascapular,
suturating with it. The antero-mesial edge of the bone expands, Y-shaped, and overlaps externally
th e adjoining edges of the pterotic, exoccipital, and epiotic, covering also an interval of cartilage
between those bones. The interval of cartilage forms p a rt of th e wall of th a t recess of th e cranial
cavity th a t lodges the hind end of th e sinus posterior utrieuli and th e related ampulla posterior,
and if the cartilage were to be suppressed th e opisthotic would form p a rt of th e bounding wall of
the recess.
The EPIOTIC is normal.
The PTEROTIC is bounded, in its deeper, primary portion, by the sphenotic, prootic,
exoccipital and epiotic, th e opisthotic overlapping, externally, th e ventral edge of th a t p a rt of the
pterotic th a t encloses the posterior portion of the external semicircular canal. The dermal portion
of the bone is bounded by th e frontal, postfrontal, parieto-extrascapular, lateral extrascapular and