The FRONTAL is an eight-sided bone, each side being straight or slightly concave. One
of these sides forms the middle portion of th e dorsal margin of the orbit, another suturates with the
frontal of th e opposite side, th e others suturating with th e ectethmoid, nasal, supraoccipital, parietal,
pterotic and postfrontal. The bone has no ventral flange. I t rests upon th e expanded dorsal edges
of th e body and brace-like internal process of th e alisphenoid, on a small portion of the sphenotic,
and on a small projecting shelf from the deeper layers of th e anterior edge of the supraoccipital. I t
is traversed by the supraorbital latero-sensory canal and lodges five organs of th e line, as will be
la te r fully described.
The POSTFRONTAL is a small dermal bone th a t lies upon and is inseparably fused, in all
of my ad u lt specimens, with a lateral portion of th e dorsal surface of the sphenotic. I t suturates
with th e frontal and dermo-pterotic, and is traversed b y th e main infraorbital latero-sensory canal,
lodging one organ of th a t canal, innervated b y th e oticus lateralis.
The PARIETAL is a sub-oval bone, bounded anteriorly by th e frontal, laterally by the
pterotic, posteriorly b y th e lateral and mesial extrascapulars, and mesially by th e supraoccipital.
I t rests upon th e dorsal surface of th e epiotic, upon the dorsal edges of those flanges of th e proofcic
and supraoccipital th a t form the antero-mesial wall of th e labyrinth recess, and also upon a small
shelf projecting mesially from what are apparently the deeper layers of th e dermal portion of the
pterotic.
The LATERAL EXTRASCAPULAR is a small oval bone traversed by th e lateral portion of
th e supratemporal commissure of th e latero-sensory canals, and lodging one organ of th a t canal.
I t is not traversed by th e main infraorbital canal, b u t a groove on its lateral edge lodges a short
section of th a t canal, apparently without related organ, as will be further explained when describing
the canals. I t is bounded antero-mesially by the parietal, antero-laterally by the pterotic, postero-
laterally by the suprascapular, and postero-mesially by the mesial extrascapular. I t has no bounding
relations either to th e subtemporal fossa or to th e large fossa on th e posterior surface of th e skull,
its ventral surface being entirely covered by the epiotic, pterotic and suprascapular, on which bones
i t rests.
The MESIAL EXTRASCAPULAR is a large subrectangular bone with stra ig h t and nearly
parallel lateral and mesial edges. I t is traversed by th e mesial section of the supratemporal-commissure
and lodges one organ of th a t commissure. I t suturates, in the middle, line, with its fellow
of th e opposite side. Anteriorly, it is bounded by th e supraoccipital and parietal, and laterally by
th e lateral extrascapular and suprascapular. Its mesial third, approximately, lies upon the broad
flat dorsal surface of the spina occipitalis, its lateral th ird resting upon a shelf-like portion of the
mesial edge of th e suprascapular. Between those two bones it forms p a rt of the roof of the large
fossa on th e corresponding side of the posterior surface of th e skull.
The SUPRASCAPULAR is a large bone w ith a prolonged and pointed hind end. I t is bounded
anteriorly by th e pterotic, antero-mesially by the lateral extrascapular, and mesially, along the
anterior half only of its length, by th e mesial extrascapular. The lateral half of the bone is bent
downward a t an angle to th e mesial portion, as already stated, and along this angle there is a stout
ridge which begins a t th e anterior qua rte r of th e bone and extends backward to its pointed hind end.
The bone is traversed by the main infraorbital latero-sensory canal and lodges one organ of th a t line,
innervated by a branch of the supratemporalis lateralis vagi. This organ, in 5 cm specimens, is much
larger th a n the other organs of the line, and may perhaps represent the two organs usually found
one in this bone and the other in the supraclavicular in the other fishes of the group.
On the ventral surface of th e bone there is a stout V-shaped flange, projecting ventro-antero-
mesially. The line of origin of th e anterior limb of the V begins a t th e rounded antero-mesial corner
of the bone, and from there runs postero-laterally until it reaches the angle between the mesial and
lateral portions of the bone. There it tu rn s postero-mesially, nearly a t a right angle, and so continues
until it reaches th e mesial edge of th e bone a t about its middle point. From this right-angled line
of origin, th e V-shaped flange projects ventro-antero-mesially, its two limbs and the overlying
body of th e bone enclosing a sub-pyramidal space which forms th e lateral, recess-like corner of the
large fossa on the hind end of the corresponding half of the skull. The angle of ihe V-shaped flange
is thickened, and its ventral end suturates with the exoccipital, the dorsal portion of the mesial edge
of th e anterior, limb of th e flange suturating with the postero-lateral end of th a t flange on the internal
surface of the epiotic th a t represents th e posterior surface of th a t bone. At the dorso-mesial corner
of the anterior surface of the anterior limb of the flange on the suprascapular, there is a small recess
which lodges b u t is not in synchondrosis with th e hind end of the band of cartilage th a t lies between
th e pterotic and exoccipital. The mesial edge of the suprascapular here closely approaches the lateral
edge of the dorsal plate of the epiotic, b u t apparently does not touch th a t bone. The V-shaped flange
of the suprascapular must accordingly certainly contain the opisthotic process of the bone, and it
probably represents both th a t process and th e epiotic process, joined by a web of bone which entirely
closes the space usually occupied by the posterior opening of th e temporal fossa. However this may
be, a posterior opening of a temporal fossa is wholly wanting in this fish, and if any portion of the
fossa exists it must open on th e lateral surface of the skull and hence be represented in the posterior
portion of the large subtemporal fossa. In certain specimens of Cottus octodecimospinosus, I have
already shown th a t the posterior opening of the temporal fossa may be entirely closed by the invading
growth of its bounding bones, th e fossa then opening wholly on th e lateral surface of the skull. In
such a fish, if th e subtemporal fossa were to be greatly deepened, as it is in Dactylopterus; it would
inevitably absorb and incorporate in itself a posterior portion of the adjoining temporal fossa. But
in th a t case the epiotic should form p a rt of the bounding wall of the fossa, and I can not find th a t
i t does so in Dactylopterus; the epiotic here apparently being everywhere covered by the hind end
of the pterotih-^xoccipital band of cartilage which lies in the bottom (roof) of the fossa. The lateral
edge of the epiotic, immediately anterior to the pedicle of the suprascapular, comes ^close to the
lateral edge of the pterotic-exoccipital cartilage and may there perhaps be exposed, thus forming
p a rt of the bounding wall of th e subternporal fossa. Furthermore, the subtemporal fossa, as defined
b y Sagemehl, is said to be an excavation of the cranial wall within the arch of the external semicircular
canal, but, in Dactylopterus, the fossa has a large portion which lies posterior to th a t canal.
This posterior extension may therefore represent the posterior portion of the temporal fossa of the
fish, here incorporated in the subtemporal fossa by the unusual development of the latter..
On the ventral surface of the suprascapular, close against or even cutting into the lateral edge
of the base of the posterior limb of the V-shaped pedicle, there is a deep circular pit, so deep th a t it
shows, on the dorsal surface of the prepared skull, as a circular translucent spot in the bone. This
p it gives origin to strong fibrous tissues which have their insertion on the dorsal end of the clavicle,
th a t end of th e clavicle n o t apparently entering the pit, in prepared specimens, b u t the p it being
q uite certainly developed in articular relation to it. At the anterior margin of the p it there is a slight