Introduction,- 
FISHES. Structure. 105 
Ctenoid denotes an arrangement of processes like 
the teeth of a comb, and designates scales with a row 
of acute points on their margins, or on the concentric 
lines of their discs. These teeth or serratures are 
almost always confined to the uncovered posterior part 
of a scale, and do not extend to the anterior part of 
the disc which remains in the follicle, and which is 
most commonly radiately furrowed, as in the cycloid 
scales. There is a general association between ctenoid 
scales and spinous processes on the bones of the head, 
shoulder, and gill-cover, as well as with firm and pun- 
gent or bony rays in the fins, these pieces of structure, 
variously combined, being more or less constantly pre- 
sent in the order of fishes named Acanthopteres. It 
is to be observed, however, that with a few exceptions, 
the skeleton of the Acanthopteres is formed of what 
Kolliker calls a “ homogeneous or tubular osteoid 
substance ” whieh does not contain bone-corpuscles, 
and is not, therefore, of so truly bony a nature as the 
skeleton of the cycloid fishes. The cycloid and ctenoid 
scales belong to the “ True fishes ” of Agassiz, and are 
not now considered by him as of equal value for the 
purpose of classification with the ganoid and placoid 
kinds, which serve to characterize two groups of verte- 
brals to which he assigns the higher value of classes. 
The ganoid scale has a hard shining enamelled 
surface, with a bony base containing radiated corpuscles, 
or, as Kolliker expresses it, the exo-skeleton of the 
ganoids is composed of real bone, partly associated 
with a homogenous osteoid substance (ganoin) and 
with dentinal tubes. 
Placoid scales want the enamel of the ganoids. 
They are sometimes in form of large plates, sometimes 
in shape of small points — forming, when prepared with 
the skin for use in the arts, the. substance called 
“ shagreen at other times, they have the shapes of 
prickles, thorns, or spines. They occur in existing 
fishes in Agassiz’s fourth class of vertebrals, the 
Selachians. None of these fishes, according to Kolliker, 
possess real bone-cells in their hard parts, which are 
formed merely of cartilage-cells in a matrix indurated 
by the deposition of earthy matter. The spines of the 
Selachians are constructed of dentine. 
All the hard or horny productions of the skin are 
considered, when spoken of generally, to be scales ; the 
designations of spines, thorns, prickles, bristles, hairs, 
plates, shields, scutes, &c., being merely descriptive of 
form. In some fishes the cranial or pelvic bones, or 
other parts of the endo-skeleton, appear on the surface ; 
and when that is the case, their discs are very often 
modified so as to assume the appearance and function 
of a dermal production, by various kinds of sculpturing, 
or the existence of muciparous cells, &c. The cranial 
scutes of sturgeons, though evidently of dermal origin 
externally, have internal surfaces of smooth compact 
substance like an endo-skeleton. 
Fins are of two kinds, differing from each other in 
their relations to the endo- and exo- skeletons. One 
kind is always developed in the mesial plane of the fish, 
aud is formed of a duplicature of the skin either solely 
or with the addition of rays, which may be solid and 
pungent when they are named spines ; or binate, 
flexible and jointed, and simple or branching, when 
Von. 11 70 
they are termed soft rays. In the Dermopteres the 
consistence of the mucoid rays is often so small that 
they can scarcely be traced. These vertical fins appear 
in the embryo fish as a cutaneous seam edging the 
whole upper and under outlines of the body. Parts of 
this seam are arrested in their growth, and disappear, 
while other parts continue to increase ; and thus several 
fins are formed. Those on the dorsal aspect are named 
dorsals, and those on the ventral one anals — the mar- 
ginal seam, with one or two exceptions only, uniformly 
vanishing before the vent, behind which opening there- 
fore the anal or anals are always situated. The ter- 
minal fin is named the caudal. In some fishes, as in 
the common eel, the continuity of the fin membrane 
remains during life along the back, round the tip of the 
tail, and along the ventral surface forwards to the vent 
— (Plate 1, figs. 4 and 6). The caudal fin may be 
pointed, crescentic, or forked. When the tapering 
extremity of the vertebral column is prolonged through 
the upper lobe of the caudal, and the lower lobe is 
developed in its greatest breadth further forward, the 
caudal acquires a lengthened unsymmetrical form called 
“ heterocercal” — (see Plate 17 e) — and is characteristic 
of the fossil fishes detected in the ancient strata. 
The form of caudal which prevails in existing fishes 
is named “ homocercal,” the lobes being alike nearly in 
size and shape, and having an even, crescentic, forked, 
or a tapering and acute termination — (see Plate 17 c). 
This symmetry is produced by a peculiar arrangement 
of the irregular flat bones called “ interneurals,” to 
which the rays of the caudal are attached ; and in the 
salmon, and many other fishes which have homocercal 
caudals, the tip of the vertebral column may be seen 
inclining upwards among the terminal interneurals 
without affecting the external equality of the caudal 
lobes. In the mackerel tribe — (Plate 10, No. 51) — the 
end of the back-bone is more direct, and the upper and 
under lobes of the caudal are more distinctly separated. 
Homocercal fishes have not been discovered in the 
magnesian limestone, or in any formation beneath it. 
They make their first appearance in the oolitic strata. 
The lateral fins are not, like the vertical ones, mere 
dermal productions, but are appendages to the endo- 
skeleton, and are strictly homologous with the limbs of 
a quadruped. The pectorals representing the fore-feet 
are attached to the inverted bony or cartilaginous arch, 
composed ehiefly of the eoracoids, and suspended in 
osseous fishes through the scapula and suprascapula to 
the skull ; but having a position under the foremost 
spinal vertebrae in sharks, and being firmly united to 
these vertebrae in the skates. 
Attached to two triangular pieces called pubic bones, 
the ventrals represent the hind feet of a quadruped, yet 
are less constant in their position than the pectorals, 
being sometimes far behind the coracoids and connected 
with them only by intervening soft parts, when they 
are named abdominals; at other times, they are situated 
under the face, and receive the designation oi jugulars; 
and when under or nearly under the pectorals, they 
were termed by Linmeus thoracic ventrals. Cuvier 
changed the latter appellation to suh-brachuin, and 
extended it to all that have pubic bones joined to the 
coracoids, even though the fins might be somewhat 
