which doubtless lodges the anterior end of a gristly mandibular' core, b u t this was n o t investigated
in th e preparation of my material. On the ventral surface of th e dorsal limb of the bone is inserted
th e strong ligament th a t has its origin on the internal surfaces of the second and th ird infraorbital
bones. The bone is traversed by the mandibular latiero-sensory canal and lodges b u t two sense organs
of the line.
The articular is a s tout bone with a pointed anterior end which fits into th e V between the
two limbs of the dentary, lying on the dorsal surface of th e v entral limb of th a t bone. On th e ventral
half of its internal surface is a deep V-shaped groove which receives th e hind end of th e ventral limb
of th e dentary. Near th e hind end of this groove a canal traverses th e bone from its ventral to its
dorsal surface, b u t what it transmits was not determined. The bone is without coronoid process,
and nearly its entire hind end is occupied by the articular facet for th e quadrate. The bone is t r a versed
by the mandibular latero-sensory canal and lodges one organ of th a t line.
The angular is a small bone th a t fits against the ventral surface of the articular, there forming
a low, longitudinal and rounded ridge. At about th e middle of its length there is a low transverse
ridge, the posterior surface of which gives insertion to a ligament th a t has its origin on a rod-like
bone th a t must represent th e interopercular. From th e h ind end of this little bone a ligament extends
posteriorly and has its origin on the external surface of th e dorsal ossification of th e ceratohyal, near
its dorso-posterior end; the bone thus being intercalated in th e mandibulo-hyoidean ligament, as a
p a rt of th a t ligament, b u t seeming nevertheless to represent th e interopercular.
The OPERCULAR is, as Gill has said, a flexible subtriangular bone, th e external surface of
which is covered with scales.
The SUBOPERCULAR is said by Gill to be „almost membranous* mostly concealed“ , and
to lie internal to the ventral end of th e opercular. I find it as a relatively s to u t and curved bone,
th e dorsal end of which lies internal to th e ventral end of the opercular, .while the ventral end
curves forward and lies against the internal surface of the preopercular; this la tte r end of th e bone
being directed toward b u t widely separated from the hind end of th e interopercular, In Figure 71,
which alone shows these bones, the line separating th e subopercular and opercular has been omitted.
The INTEROPERCULAR is apparently represented, as Gill has s tated and as ju s t above
described by me, b y a rod-like bone intercalated in th e mandibulo-hyoidean ligament.
4. L A T E R O - S E N S O R Y C A N A L S .
The main infraorbital canal begins a t th e dorsal edge of the lachrymal, directly antero-ventral
to the anterior end of th e large nasal opening between th e nasal and ectethmoid bones, there lying
directly opposite and close to the anterior opening of th e supraorbital canal. I t runs a t first ventro-
laterally, th en tu rn s sharply backward a t a right angle, and so continues in a nearly horizontal
position until it has traversed the lachrymal and th e anterior two-thirds of the second infraorbital
bone. There it turns sharply upward and forward,, a t an acute angle, and curving backward and
upward borders th e hind edge of th e orbit, traversing th e second, third, and fourth infraorbital bones.
At the anterior end of the horizontal p a rt of its course, it sends a short branching tubule forward
in th e lachrymal, and a t the hind: end of this p a rt of its course, it sends a long branching tubule
directly backward nearly to th e postero-ventral corner of th e second infraorbital bone,. Numerous
other tubules arise from the canal in this p a rt of its course, some of them being simple tubes while
others branch repeatedly, forming complicated dendritic systems.' The tubules are scattered along
the canal instead of being grouped together, and it is impossible from th e tubules alone to determine
th e number of primary tubes. Furthermore, here, as in other parts of the latero-sensory system
of this fish, th e large dendritic systems arise from th e canals in the bodies of the related bones,
frequently near th e middle point of th e bone, and almost never in the sutural line between two
bones; this being distinctly a characteristic of th e canals in th e chondrostean ganoids, here found
in a teleostean fish. A full knowledge of the development of the system would accordingly be of
considerable interest.
The main infraorbital canal, having left th e dorsal one of the infraorbital bones, enters and
traverses th e postfrontal, a t th e dorsal end of which bone it turns sharply backward and enters the
pterotic, anastomosing a t the bend with what appears to be the terminal tube of the supraorbital
canal. The main infraorbital th e n traverses in succession th e pterotic and suprascapular, lying for
a short distance, as it passes from one of these bones to the other, in a groove on th e lateral edge
of the lateral extrascapular. I t th en traverses a short tubular bone th a t lies along the lateral edge
of the suprascapular and th a t apparently represents the latero-sensory component of the supraclavicular,
as already set forth. As the canal traverses the pterotic i t anastomoses with the dorsal
end of the preopercular canal, and as it traverses th e lateral edge of the lateral extrascapular, it
gives off th e supratemporal commissure.
In the main infraorbital canal there are three sense organs lodged in the lachrymal, two in the
second infraorbital bone, three in th e third infraorbital and one in the fourth infraorbital, all innervated
by th e ramus buccalis. In th e postfrontal there is one organ innervated by the ramus oticus,
and in the'pterotic three organs ¿ two innervated by th e oticus and one by th e supratemporalis lateralis
vagi. In th e section of canal th a t lies in the groove in the lateral extrascapular there is apparently
no organ, b u t the organ usually found here may be represented in a p a rt of the one organ found in this
bone. This organ is unusually large, begins in the main canal, and from there extends postero-
mesially in the supratemporal commissure, thus certainly belonging, in p a rt a t least, to th a t commissure.
I t is innervated by a single branch of the supratemporalis lateralis vagi. In the suprascapular
there is one organ, innervated by th e supratemporalis lateralis vagi.
The supratemporal canal begins a t th e lateral edge of th e lateral extrascapular, opposite the
sutural line between the pterotic and suprascapular, and running postero-mesially and then mesially
traverses the lateral extrascapular and then the mesial extrascapular, anastomosing, a t the mesial
edge of the la tte r bone, with its fellow of the opposite side. Each of these two bones lodges a single
sense organ innervated by a branch of th e supratemporalis lateralis vagi.
The supraorbital canal begins a t the lateral edge of the single median nasal bone, and from
there runs upward and mesially, and then upward and backward in th e nasal, curving around the
anterior and then the mesial border of the large nasal opening of the skull. I t then enters the anterior
edge of the frontal and runs almost directly backward to the middle point of th a t bone. There it
tu rn s sharply laterally, and then curving laterally and backward reaches th e postero-lateral edge of
th e frontal, where i t anastomoses, by what is apparently its terminal tube, with th e main infraorbital
canal. In the full length of the canal there are, as in all the other fishes of the group, six sense
organs, one in the nasal and five in the frontal, all innervated by the ophthalmicus lateralis; b u t in
Dactylopterus the relations of the organs in the frontal to th e frontal commissure are not as in those
o th e r fishes. Here, three organs lie anterior to the commissure and two posterior to it, the commissure