FISHES. The following remarks are condensed from Professor Owen's Anatomy of the Verte brates (London, 1866-68). In the Cyclostomi, which, if we except the lancelet, constitute the lowest order of fishes, and include' the hag and lamprey, the gills are sacciform, with external openings, and six or seven in number on each side. Each gill-sac receives its proper artery either from the branchial artery or one of its branches. "The leading condition of the gills in other fishes may be understood," says Owen, "by supposing each compressed sac of a myx ine to be split through its plane, and each half to be glued by its outer smooth side to an intermediate septum, which would then support the opposite halves of two distinct sacs, and ex pose their vascular mucous membrane to view. If the septum be attached by its entire margin, the condition of the gill in the Plagiostomi (sharks, dogfish, rays, skates, etc.) is effected. If the septum be liberated at the outer part of its circumference, and the vascular surfaces are produced into pectinated lamelligerous processes, tufts, or filaments proceeding from the free arch, the gill of an ordinary osseous fish is formed. Such a gill is the homologue, not of a single gill sac, but of the contiguous halves of two distinct gill-sacs, in the myxines: Already in the lam preys the first stage of this bipartition may be seen, and the next stage in the sharks and rays; consequently, in these fishes a different artery goes to the anterior branchial surface of each sac or fissure from that which supplies the pos terior branchial surface of the same fissure; while one branchial artery is appropriated to each supporting septum or arch between the fis sures, as it is to the liberated septum or bran ehial arch in the ordinary osseous fishes." The lampreys, myxinoids, sharks, and rays are termed fishes with 'fixed gills,' because in them each supporting septum of the anterior and posterior branchial mucous surfaces is at tached to the pharyngeal and dermal integu ment by its entire outer margin, and the streams of water flow out by the same number of fissures in the skin as those by which they enter from the pharynx. In the osseous and in the ganoid fishes there are the outer border of the supporting branchial arch being unattached to the skin, and playing freely backward and forward, with its gill-surfaces, in a common gill-cavity, which has a single outlet, usually in the form of a vertical fissure. In the myxinoids (see illustration under HAG) six or seven bran chial sacs open on each side, and their outlets are produced into short tubes, which open into a longitudinal canal, directed backward, and dis charging its contents by an orifice near the middle line of the ventral surface. Between the two out lets is a third larger one, which communicates by a short duct with the end of the esophagus, and admits the water, which passes from that tube by the lateral orifices leading into the branchial sacs. These sacs, which are developed from the oesophagus, and which may be regarded as the simplest form of piscine gill, have a highly vascular, but not a ciliated, mucous membrane, which is arranged in radiating primary and sec ondary folds, so as to increase the surface. In
the lampreys there is a further separation of the respiratory from the digestive tract, for each internal blind duct communicates with a median canal beneath and distinct from the oesophagus.
In all the higher fishes the inlets to the bran chial interspaces lie on each side of the gullet, and are equal in number with the interspaees; while, except in the elasmobranchs, there is only one outlet on each side. These outlets vary ex tremely in size, being relatively largest in the herring and mack erel families, and smallest in the eels and lophioid fishes, as the angler (q.v.). The length of time that different fishes can exist out of water depends on the modifications for retaining water in the branchial chambers. As a general rule, the chamber is largest where the outlet is smallest, as in the eels, hlennies, and lophioids, and these are the fishes that survive the longest out of water, except in such cases as the climbing fish (q.v.), in which the bran ehial apparatus possesses complex lahyrinthie appendages. The main object of the gills of fishes being to expose the venous blood, in very thin-walled vessels, to streams of water, the branehial arteries rapidly subdivide into capil laries, which constitute a network in one layer, supported by an elastic plate, and covered by a tessellated but non-ciliated epithelium. This cov ering and the capillary wall are so thin as to admit free interchange to take place between the blood, loaded with carbonic acid, on the one hand, and the aerated water on the other. The extent of respiratory surface is increased in various ways, of which by far the most common is "by the production of the capillary-supporting plates from each side of long, compressed, slen der, pointed processes, extending, like the teeth of a comb, but in a double row, from the convex side of each branchial arch." The number of vascular plates or lamellae attached to each branchial process has been estimated at 135 in the carp, 700 in the eel, 1000 in the cod, 1400 in the salmon, and MOO in the sturgeon.
We now pass on to the considera tion of these organs in amphibians. In the lower or perennibranchiate members of thin order, the gills exist permanently, but in the great majority they are mere temporary organs. In the newt three pairs of external gills are developed, at first as simple filaments, each with a capillary loop, but speedily expanding and giving off loop lets. The gill is covered with ciliated epitheliuM; which loses the cilia before the absorption of the organ, and this takes place after a few days of larval existence. In the larval frog the gills, which are on a simpler plan, diminish about the fourth day, and disappear on the seventh. The parts of the branchial framework which support the deciduous gills never get beyond the carti laginous stage. They thus readily shrink, and become more internal as the head increases in size. As the gills of the perennibranchiate am phibians in all essential points resemble those already described, it is unnecessary to notice them. See AMPIIIBIA DIPNOI.