STRUCTURE OF FISHES. 7 



distinct names : first there is the basi-hyal in the middle, and from this a bone termed the glosso-lival 

 usually extends into the substance of the tongue. At the sides rise up the long liorn-like bones 

 termed cerato-hyal, above this is the epi-hyal, and yet higher still the stylo-hyal. All these tracts are 

 not universally met with. In the Conger Eel, for instance, stylo-hyal is a ligament, and the basi-hyal 

 is blended with the cerato-hyal. In other fishes, like the genus Mursenophis, the glosso-hyal is wanting. 

 From the hinder margin of the cerato-hyal and epi-hyal a number of slender, long, curved bones are 

 prolonged backward and outward. These are termed the branchiostegal rays. They support the 

 membrane which forms the external cover to the chamber which contains the gills. Their number is 

 very variable, but most usually seven, as in the Cod. In the Herrings of the genus Elops there are 

 more than thirty rays in each gill-cover. In the Carp they are flat and broad, and reduced to three 

 in number. In the Angler they are enormously long. Behind the hyoid arch, and more or less 

 connected with it, are the branchial arches. Originally there were six of these arches, one behind 

 the other, with clefts between them, but only five ai-e commonly developed. The first four support 

 the gills, the fifth, margined with teeth, guards the entrance to the gullet. The lower ends of these 

 arches are united to a chain of little bones prolonged backward from the basi-hyal element of the 

 hyoid. This part, usually termed basi-branchial, most frequently consists of three bones. Each 

 branchial arch rises from this outward and upward. It consists of three or four separate pieces of 

 bone, though the fifth arch commonly consists of one bone only. Sometimes these arches become 

 complicated in fishes which live long out of water, such as the Climbing Perch, by developing at their 

 upper margins large bony folds, in the recesses of which water is contained, so that it may trickle from 

 them over the gills. Occasionally the branchial arches remain cartilaginous, and all six pairs retain 

 the cartilaginous condition in Lepidosiren, but the second and third arches do not support gills, though 

 they are found on the last arch. The scapular arch of the fish is often attached to the side of the skull, 

 or occasionally to the basi-occipital, though in the cartilaginous fishes it is usually removed farther 

 back. It consists of several bones, which have received different names from the several anatomists 

 who have described them. In Sir R. Owen's system, the uppermost piece is the supra-scapula, 

 which sometimes consists of two short columnar bones attached to the auditory region of the skull. 

 The next piece is termed the scapula, and these two bones are always blended together in the Siluroid 

 fishes. The lower bone Sir R. Owen terms the coracoid. They are sometimes blended together at 

 their lower margins, but more frequently these bones are joined by ligament, though in the Siluroids 

 they unite by a toothed suture. The bones which Owen names scapula and supra-scapula Huxley) 

 with good reason, calls clavicle and supra-clavicle. The scapula and coracoid in all animals form the 

 arch which gives attachment to the base of the limb. These bones support and defend the heart in 

 all fishes, and give attachment to the diaphragm which separates the cardiac cavity from the abdominal 

 cavity. They also furnish a margin against which the operculum shuts, enclosing the cavity which 

 contains the gills. 



The vertebrae in fishes present many curious modifications. Thus, in the Sturgeon the first five 

 or six neural arches are blended together so as to form a sheath of cartilage which encloses the 

 spinal cord and the front part of the notochord, the tapering end of which is prolonged into the 

 base of the skull. The ribs are attached only to about the first twelve of the trunk vertebrae. 

 They join the vertebrae by simple heads, and often consist of two or three jointed pieces. The 

 same kind of union of the earlier dorsal vertebrae into a continuous cartilaginous sheath around 

 the notochord is formed by the first ten vertebrae of the Chimseroids. In some of the Sharks the 

 ribs become very numerous, extending in Acanthias to forty pairs. Among the Skates of tfie 

 genus Rhinobatie, Sir R. Owen finds but a single arch over the bodies of two vertebrae, and in 

 the Chimsera the slender rings which represent the bodies of the vertebrae in the cartilage covering the 

 notochord are more numerous than the neural arches which extend over them. In the Blue Sharks 

 the vertebrae are most perfectly ossified, having only four notches for the neural arch and transverse 

 processes. In most bony fishes the vertebrae are conically cupped at both ends ; often the body of the 

 vertebra remains distinct from the neural arch. In most animals the front of a vertebra is easily recog- 

 nised by the processes called zygapophyses, which yoke the bones together in front and behind, the 

 articular surfaces in front always being directed upward or inward, but in the Perch the reverse condi- 

 tion is met with. The posterior zygapophysis here looks upward, and receives upon its surface the 



