426 
ACTINOPTEEYGII. 
bony splints. The maxilla (inx) is a narrow, elongated element, 
much expanded behind the eye ; the premaxilla ( pmx) is compara¬ 
tively small and insignificant. Surrounding the eye is a narrow riug 
of four thin bones (circumorbitals) bounded behind by others of the 
“ suborbital ” series ( 5 . 0 ); and the space between the latter, the 
cranial roof, maxilla, and opercular apparatus is covered by a single 
bent bone, interpreted by Traquair as preoperculum (p.op). In 
the mandible the articular portion of the meckelian cartilage is os¬ 
sified, and the rest is ensheathed outside by a verj T large dentary 
(cl) and a small angular (cig), while its inner face is equally covered 
by an extensive laminar splenial. Toth the splenial and dentary, as 
a rule, bear teeth. The hyomandibular element of the suspensorium 
is well ossified superficially and is thus usually preserved, but no 
symplectic has been noticed; the former is much elongated, and 
somewhat bent at about the position of the lower border of the 
operculum. The pterygo-quadrate seems to be ossified at least at the 
quadrate articulation ; and there is evidently a large, elongated 
membrane-bone ensheathing its inner or oral aspect. The oper¬ 
culum (op) is suspended from the hyomandibular, and is usually 
narrow, bounded below by a large suboperculum (i.op.j, often with 
indications of a feebly developed interoperculum. Beneath the sub¬ 
operculum, the opercular fold is strengthened by a series of lamelli- 
form branchiostegal rays (&/•), which meets the corresponding series 
of the opposite side in front, and terminates in an anterior azygous 
element at the symphysial angle of the mandible. The branchial 
arches are sometimes seen to be ossified. 
In the axial skeleton of the trunk, the notochord must have been 
persistent, and there is as yet no definite evidence of ossifications in 
its sheath. The neural arches and spines throughout the trunk, and 
the haemal arches, with their spines in the caudal region, are super¬ 
ficially ossified, and are thus observed when there is no obscuring squa- 
mation ; but there are no traces of ribs in any genus, the abdominal 
haemal arches being merely small short pieces of cartilage. In the 
only genus in which they have been well displayed ( Coccolepis ), the 
neural spines are not fused with the supporting arches in the abdom¬ 
inal region, but both these and the haemal spines are firmly fixed to 
their arches in the tail; at the base of the caudal fin the haemals are 
much enlarged for the direct support of the dermal rays, while the 
neurals become gradually aborted, and there is a series of distinct 
supporting ossicles beneath the fulcra of the upper caudal lobe 1 . 
1 The characters of the axial skeleton are to some extent shown in Ann. Mag. 
Nat. Hist. [6] toI. v. pi. xxi. figs. 2 , 4; but more satisfactory information will 
appear in Mem. Geol. Surv. N.iS. Wales, Pakeont. no. 9, pis. i., ii. 
