In the Siluroid genera Clarias and Heterobranchus the accessory organ takes the form of branched, arborescent and highly vascular structures, developed as outgrowths from the dorsal extremities of one or two branchial arches, and enclosed within a posterior and dorsal expansion of the proper branchial cavity (Fig. 171).

Another example of these interesting structures occurs in Chanos salmoneus and a few other Clupeidae^[320] in the shape of a coiled gill-like organ ("gill-helix"), which is supported by the dorsal segment of the fourth branchial arch, and enclosed in a similarly curved caecal extension of the branchial cavity. Each gill derives its blood from the fourth afferent branchial artery, the corresponding efferent vessel joining the fourth efferent branchial artery. A similar spirally-coiled "gill-helix" is found also in Heterotis ehrenbergii,^[321] amongst the Osteoglossidae, and in several species of Characinidae.^[322]

In other Teleosts the accessory breathing organ assumes the condition of paired lung-like outgrowths of the branchial cavity. Thus, in one of the Symbranchidae, the Indian "Cuchia Eel" (Amphipnous cuchia),^[323] there is a pair of small bladder-like sacs, with membranous and vascular walls, each of which opens into the branchial cavity above the first gill-cleft, and is supplied with blood by the afferent branchial artery of the gill-less first branchial arch. An extreme modification in the same direction is presented by the Indian Siluroid Saccobranchus.^[324] In this Fish a long caecal diverticulum of the branchial cavity extends backwards on each side from the dorsal region of the first branchial cleft to the tail, and in its course is situated internally to the lateral trunk musculature, and close to the vertebral column (Fig. 172). The walls of the caeca are vascular, but no special respiratory structures are developed within their cavities, which, during life, only contain air. In S. singio the right caecum is supplied with blood by an extension backwards of the dorsal portion of the first afferent branchial artery of that side; the left, on the contrary, being supplied by the corresponding portion of the fourth afferent artery of the same side. In S. fossilis^[325] both air-sacs are supplied by the fourth afferent branchial artery. The efferent vessels join the fourth efferent branchial artery, right or left as the case may be.

With perhaps one or two exceptions, the accessory respiratory organs of Fishes seem to exist for the purpose of enabling their possessors to breathe in air. This is certainly the case with the labyrinthiform organs of Anabas and its allies, and also in such Fishes as Amphipnous, Saccobranchus, and the Ophiocephalidae, and probably in others. Nearly all these Fishes are tropical in geographical distribution, more or less amphibious in their habits, and usually possess a remarkable capacity for sustaining life out of water, under conditions which are promptly fatal to ordinary Fishes. Thus, Anabas scandens may be kept alive for days in earthen pots without water, and when free is able to travel short distances on land, especially in the early morning when the dew is on the ground, while Amphipnous frequents marshes, lurking in holes in the grass and about the sides of ponds. In fact, even when in the water, access to air, which is probably swallowed and passed over their accessory breathing organs, is indispensable to their existence. Experiments conclusively prove that if the Fish is artificially prevented from obtaining air in this way asphyxiation speedily ensues.^[326]

In addition to breathing air through the agency of special organs evolved for the purpose, there are many freshwater Fishes which, like those just mentioned, periodically rise to the surface and swallow air in order to saturate the water which bathes the gills with oxygen.^[327]

CHAPTER XI

THE AIR-BLADDER

In the Crossopterygii, Chondrostei, and Holostei, in the Dipnoi, and in the great majority of Teleosts, there is situated on the dorsal side of the coelom, between the alimentary canal below and the kidneys and vertebral column above, a more or less elongated sac with membranous walls, an internal epithelial lining and gaseous contents—the air-bladder (Figs. 154 and 173). Usually developed in the embryo as a caecal outgrowth from the dorsal surface of the oesophagus, the air-bladder grows anteriorly and posteriorly, and may either retain throughout life its primitive connexion with the alimentary canal by means of a longer or shorter tubular canal, the ductus pneumaticus, or become completely separated therefrom in the adult by the atrophy of the duct. Its walls sometimes, but rarely, contain muscle-fibres, as in Lepidosteus, Amia, and the Dipnoi, and are always more or less vascular, while laterally and ventrally the organ is invested externally by the peritoneum (Fig. 173). In addition to the muscle-fibres distributed in its walls, the bladder is often provided with powerful extrinsic muscles, more especially in those Fishes in which it is used as an organ for sound-production. In the different groups of Fishes in which it is present the air-bladder frequently undergoes remarkable structural modifications and becomes adapted for various distinct functions.

In the Cyclostomata there is no trace of an air-bladder, and, unless represented in certain Sharks (e.g. Mustelus, Galeus, and Acanthias),^[328] by a small caecum embedded in the dorsal wall of the oesophagus and communicating with its cavity, it is also absent in all Elasmobranchs. In the Crossopterygii (e.g. Polypterus),^[329] the air-bladder is double, but while the right sac is long and somewhat tubular, the left is much smaller and oval in shape (Fig. 174). Near their anterior extremities the two sacs fuse into a single unpaired chamber, beyond which they again project in the form of two short caeca. The median chamber opens into the oesophagus on the ventral side by an orifice (gl) bounded by prominent lips and furnished with a muscular sphincter. The organ is devoid of internal sacculations. In the Chondrostei (e.g. Acipenser) the air-bladder is oval in shape, with a smooth, non-sacculated, inner surface, and a lining of ciliated epithelium, and it communicates with the oesophagus by means of a relatively wide, dorsally placed, funnel-like orifice.

In the Lepidosteidae the single air-bladder extends the whole length of the abdominal cavity, and, as in Polypterus, communicates with the exterior through a larynx-like vestibule provided with a glottis,^[330] which, however, opens dorsally into the oesophagus (Fig. 175). A strong fibrous band runs along the median line of the inner surface of its dorsal wall, from which extends ventrally on each side a series of transverse fibro-muscular ridges, forming the boundaries of a double row of regularly arranged alveoli (Fig. 176). The bottom of each alveolus is still further sacculated by finer branches of the principal fibrous bands.^[331] In the Amiidae the bladder is very large, and, except that a short median cleft divides it in front into two short caeca, it is unpaired. Internally, its walls are much sacculated, but the alveoli are smaller and arranged less regularly than in Lepidosteus. The aperture of communication with the oesophagus is dorsally situated.

It may be mentioned that in all the preceding Teleostomi the ductus pneumaticus is remarkably short, the connexion between the air-bladder and the oesophagus being almost direct by means of a larger or smaller orifice, which, except in Acipenser, is more anteriorly placed than in most other Teleostomi; and further that, unlike many Teleosts, there are no special "retia mirabilia," "red bodies," or "red glands."

In the Dipnoi the structural resemblance of the air-bladder to a true lung, which to some extent is indicated in Polypterus, Amia, and Lepidosteus, becomes still more marked.

In Neoceratodus^[332] the organ is not unlike that of Lepidosteus, and takes the form of a spacious unpaired sac, extending from one end of the abdominal cavity to the other. On its inner surface two fibrous bands, one of which is dorsal and the other ventral, traverse the whole length of the bladder, and project slightly into its cavity. Between these median ridges extend a number of transverse septa, forming the boundaries of a series of pairs of bilaterally symmetrical oval alveoli, the walls of which are still further sacculated by a network of finer ridges (Fig. 177). The short ductus pneumaticus seems to be an anterior continuation of the right half of the bladder, and opens into the oesophagus by a small glottis, situated on the ventral side, a little to the right of the median line.

The more complicated and much more lung-like air-bladder of Protopterus (Fig. 178)^[333] is essentially double, consisting of an anterior unpaired portion, and of two sac-like prolongations which extend backwards the whole length of the coelom, gradually tapering towards the cloaca. Anteriorly, the unpaired portion of the organ is continued into a vestibule or pneumatic duct, which, after passing ventrally on the right side of the oesophagus, opens into the latter by a ventrally-situated, slit-like glottis, immediately behind the last pair of gill-clefts. The margins of the glottis are provided with radially-arranged dilator muscles, and in connexion with its anterior border there is an epiglottis-like fibro-cartilaginous plate.^[334] The central cavity of each lung (Figs. 178 and 179) communicates with a series of larger or smaller alveoli in the lung-wall, and each of the latter opens in succession into smaller tubular cavities, and then into still smaller terminal caecal sacculi. Hence, much more than in Neoceratodus, the lungs approximate in structure to those of the higher terrestrial Vertebrata. Non-striated muscle cells, pigment cells, and blood capillaries are abundantly present in the connective tissue external to the lining epithelium of the lung-cavities.

The air-bladder of Lepidosiren closely resembles that of Protopterus, and, as in the latter Dipnoid, the glottis seems to be furnished with an epiglottis.^[335]

In all the Dipnoi the air-bladder is highly vascular, but nevertheless presents no trace of "red bodies" or "red glands."

The most striking features in the remarkably polymorphic air-bladder of Teleosts relate to (a) its presence or absence; (b) differences in shape and relative size; (c) the development of caecal outgrowths; (d) the subdivision of its cavity by the formation of internal septa; (e) the retention or suppression of the ductus pneumaticus, and the occasional development of secondary ducts communicating directly with the exterior; (f) the presence of "red glands" or "red bodies"; (g) its connexion with the auditory organ; (h) its adaptation as an organ for sound-production.

(a) The air-bladder is by no means universally present in Teleosts. It is absent in several entire families,^[336] such as, for example, the Flat Fishes or Pleuronectidae, the Scopelidae, and the "Lump-suckers" (Cyclopteridae). In a few families, as in the Mackerels (Scombridae), the "Blennies" (Blenniidae) and the Polynemidae, the organ is present in most genera, but absent in a few, or even present or absent in different species of the same genus. Thus, of the three British species of Mackerel, viz. the Spanish Mackerel (Scomber colias), S. pneumatophorus, and the common Mackerel (S. scombrus), an air-bladder is present in the first two, but absent in the third.^[337]

(b) As might be anticipated, the shape of the air-bladder is extremely different in various Teleosts, and usually conforms to the shape of the body, while differences in relative size are of frequent occurrence, even in closely related species. Sometimes the organ is more or less tubular, fusiform, ovoid, or heart-shaped; occasionally it is shaped like a "dumb-bell," consisting of two lateral sacs connected by a median tubular portion, as in the Siluroids Clarias and Callichthys; or it may be horse-shoe-shaped, as in the Silurid Ailia.^[338] Not unfrequently a transverse constriction divides the air-bladder into two intercommunicating sacs, as in most of the Carp family (Cyprinidae), or three such sacs may be formed by two constrictions (e.g. Ophidium). In the "Electric Eels" (Gymnotidae) there are two sacs, connected by a slender canal, from which the ductus pneumaticus takes its origin.^[339]

The air-bladder is either more or less free in the abdominal cavity, or firmly attached to the vertebral centra and their rib-bearing processes by fibrous extensions passing between the two structures. Not rarely the organ extends from the abdominal cavity into the tail, sometimes penetrating for a short distance into the expanded haemal canal of the anterior caudal vertebrae, or extending unsymmetrically along either the right or left side of the tail. More frequently, perhaps, where the air-bladder is prolonged into the tail, it assumes the form of two bilaterally arranged and symmetrical caeca, which extend backwards for a variable distance internal to the caudal muscles and in contact with the adjacent skeletal elements, as in Notopteridae, and in some Sparidae, Carangidae, and Scombridae. The extension of the air-bladder into the tail is often associated with a short, laterally-compressed trunk, which, if the bladder is to attain its normal degree of development, necessitates its prolongation into the caudal region.

(c) A characteristic feature in the air-bladder of many Teleosts belonging to widely different families is the development of a more or less complex system of simple, or variously branched, caecal outgrowths, which, like the internal septa, are specially characteristic of those Fishes in which the bladder is used as a vocal organ without, however, being peculiar to them.

In some of the Gadidae, as in the Cod (Gadus morrhua), the air-bladder divides anteriorly into a pair of caecal prolongations which extend forwards to the head, and are often curiously coiled. Somewhat similar caeca are also present in species of Berycidae, Sparidae, Siluridae, Clupeidae, and Notopteridae. Caecal prolongations may also be developed from the hinder end of the bladder, and, as already mentioned, extend into the tail; or even from both ends in the same species (e.g. Notopterus).^[340] In the Silurid, Rita crucigera,^[341] a long tubular caecum is developed from each side of the heart-shaped bladder, and thence is prolonged backwards to the anus. In certain species of Doras of the same family (e.g. D. maculatus),^[342] an elegant series of variously sized branched caeca fringe each of the lateral margins of the bladder. It is, however, in the Physoclist family of the Sciaenidae^[343] that the branching of the air-bladder attains its greatest development in extent and variety.

In Otolithus (Fig. 180) two short tubular canals are given off from the antero-lateral angles of the bladder, each subsequently dividing into two elongated, tapering sacs, of which one is directed forwards and the other backwards. In Corvina lobata (Fig. 181) the lateral margins of the bladder are everywhere fringed with a series of tufts of caeca, each tuft being connected by a short common canal with the cavity of the organ. In the "Drum" (Pogonias chromis) (Fig. 182) each side of the anterior third of the air-bladder has a series of digitately branched caecal appendages, the most posterior of which on each side are connected by a tubular canal, also bearing branched caeca, with the corresponding postero-lateral extremity of the bladder.

Collichthys^[344] has a still more remarkable arrangement. In this Sciaenoid (Fig. 183) twenty-five tubular branches are given off from each side of the bladder, all of which soon subdivide into a dorsal and a ventral division. These still further divide, and their branches either end blindly or are prolonged into a series of arches to the mid-dorsal or mid-ventral line as the case may be, where they become continuous with the corresponding branches of the opposite side. The series of dorsal branches, enveloped in their peritoneal investment, extend between the body of the air-bladder and the roof of the body-cavity, while the corresponding ventral branches, similarly invested, surround that part of the coelom which contains the stomach, intestine, and liver.

(d) In addition to the subdivision of the cavity of the air-bladder by the externally obvious, transverse, or longitudinal constrictions already described, or by the growth of simple or branched prolongations, the organ is often chambered or sacculated by the development of internal septa or partitions.

In many of the Gurnards (Trigla)^[345] the cavity of the bladder is divided into two intercommunicating compartments by a transversely-disposed and centrally-perforated diaphragm. The large air-bladder of some species of Erythrinus^[346] is subdivided internally into numerous alveoli or sacculi. In Notopterus a longitudinal septum divides the cavity of the abdominal portion of the bladder into two lateral chambers, which, however, freely intercommunicate anteriorly. In the great majority of the Siluridae^[347] the cavity of the organ is divided by a characteristic T-shaped arrangement of a primary transverse and a longitudinal septum into three communicating chambers, of which one is anterior and transversely disposed, and two are posterior and longitudinally arranged (Fig. 222). The posterior compartments in many genera are still further divided by the growth of secondary transverse septa, extending outwards from the median longitudinal septum, without, however, reaching the external lateral walls of the chambers. In a few genera, as in certain species of Pangasius,^[348] additional fibrous bands and ridges passing between the primary and secondary septa give to the cavities of the lateral compartments the appearance of being occupied by a coarse spongy network.

(e) In its relations to the oesophagus and to the air-bladder the ductus pneumaticus exhibits striking modifications in different Teleosts. With very rare exceptions, an open ductus is wanting in the Heteromi, Catosteomi, Acanthopterygii, Opisthomi, Pediculati, Jugulares, and the Plectognathi, for which reason the term "Physoclisti" has often been used as a collective name for these Fishes. On the other hand, a permanently open ductus is generally present in the Malacopterygii, Ostariophysi, Apodes, and the Haplomi, which, in consequence, have been designated "Physostomi." It must be emphasised, however, that all Teleosts are Physostomous in the embryonic condition, and whether they eventually become Physoclistous or remain Physostomous depends entirely on the abortion or retention of the primitive communication between the air-bladder and the alimentary canal. When present in Teleosts, the ductus pneumaticus, with a few exceptions (e.g. Notopterus), where it is both short and relatively wide, is almost invariably much longer and narrower than in the other orders of Teleostomi and in the Dipnoi, sometimes passing directly from the air-bladder to the oesophagus, but not infrequently describing a sigmoid curve, as in some Cyprinidae, or an even more tortuous course. The opening into the alimentary canal is, with perhaps a single exception, dorsal, but may vary from the commencement of the oesophagus to the hinder end of the stomach. In Erythrinus the oesophageal aperture is lateral. In two instances the air-bladder has acquired secondary openings to the exterior, and of these one occurs among the Physostomi and the other in the Physoclisti. In the Herring (Clupea harengus),^[349] in addition to the proper ductus, which is connected with the distal end of the caecal stomach, a tubular canal leaves the hinder extremity of the bladder and opens externally on the left side of the genital aperture; consequently, in this Fish the air-bladder has a secondary and direct connexion with the exterior in addition to the primary and indirect communication by means of its proper duct. The Horse-Mackerel (Caranx trachurus)^[350] is even more peculiar. This Teleost has no true pneumatic duct, but instead a special duct which passes from the bladder to open into the right branchial cavity by a very minute aperture. In neither case is anything known of the mode of origin or morphological nature of the secondarily acquired duct.

(f) The air-bladder differs greatly in its degree of vascularity in various Teleosts, as well as in the extent to which its capillary blood-vessels accumulate at special points on the inner surface to form the so-called "red bodies" or "red glands." In some Teleosts the distribution of capillaries is uniform or nearly so; in others, as in the Carp (Cyprinus carpio) the vessels are arranged in fan-like, radiating tufts over almost the whole extent of the inner surface; in others again, as in the Pike (Esox lucius) the tufts are larger and more definitely localised. A more extreme modification occurs in some of the Physostomi, in which a remarkable concentration of capillaries takes place at one or more points on the inner surface of the bladder, which project into the cavity of the organ in the form of variously shaped blood-red masses. These "red bodies" are essentially retia mirabilia, consisting of masses of interlacing, tightly-packed capillaries with their afferent arteries and efferent veins. The flattened lining epithelium of the bladder is continued over these bodies without undergoing any special modification. In the common Eel (Anguilla vulgaris) there are several of these bodies, of which the largest are near the entrance of the pneumatic duct.

In the Physoclisti the "red bodies" seem to be replaced by true glands,^[351] which nevertheless in appearance closely resemble the former. Some of the Gadidae, such as the Cod (Gadus morrhua), the Haddock (G. aeglefinus), and the Hake (Merluccius vulgaris), have a single large "red gland" projecting into the interior of the bladder from its dorsal or ventral wall (Fig. 184, A). The John Dory (Zeus faber) has five such glands, worm-like and curved in shape, with their concavities facing a central point between them (Fig. 184, B). In these Fishes a "rete mirabile" of blood-vessels forms the vascular basis of the glands. The ordinary pavement epithelium of the bladder becomes replaced by faintly granular, columnar, and evidently glandular cells as it passes over the retia mirabilia, and at the same time becomes invaginated into the mass of capillaries in the form of a number of simple caecal glands (Fig. 185). So far as is at present known, the "red glands" are only found in those Teleosts in which the air-bladder has no ductus pneumaticus, whereas in those Fishes which retain the ductus throughout life there are either no special retia mirabilia, or, as in the Eel, only the so-called "red bodies."^[352]

(g) and (h) The structural modifications involved in the connexion of the air-bladder with the auditory organ, and its adaptation for sound-production, as well as its use in respiration, are considered elsewhere.^[353]

The Gases of the Air-Bladder.—The gaseous contents of the air-bladder consist of oxygen and nitrogen, but the relative proportions of the two gases differ in different Fishes, and even in the same Fish, under different conditions. Normally the proportion of oxygen is considerably less in freshwater than in marine Fishes, and amongst the latter the proportion of oxygen is often enormously greater, amounting in some cases to 87 per cent., in deep-sea species as compared with their shallow water congeners. A trace of carbon dioxide is also usually present. The gases are derived from the blood as the latter circulates through the capillaries in the walls of the bladder, and it is highly probable that the "red glands" take an important part in the process; at all events, experimental research has shown that the "secretion" or diffusion of gas into the air-bladder, as well as the absorption of gas from the bladder into the blood, take place most rapidly in those Fishes in which "red glands" or "red bodies" are present.^[354]

The Functions of the Air-Bladder.—Probably no single organ in any group of Vertebrata is associated with the performance of a greater variety of functions than the air-bladder of Fishes. Originally evolved, it may be, as a glandular caecum in certain Sharks, the air-bladder in the Dipnoi, and some of the more generalised Teleostomi (e.g. Amia and Lepidosteus), and perhaps also in a few of the more specialised members of the latter group (e.g. certain Teleosts), is to a greater or less extent an accessory respiratory organ. In not a few Teleosts it is an organ for sound-production, and in others again it is sometimes regarded as having an important relation to the sense of hearing. But omitting such subordinate functions which, as it were, have been grafted on to the air-bladder, there can be no doubt that in the great majority of Fishes its primary use is to act as a hydrostatic organ or "float." From this point of view experimental investigations^[355] seem to justify the following conclusions:—

The function of the air-bladder is to render the Fish, bulk for bulk, of the same weight as the water in which it lives. In this condition of equilibrium, or plane of least effort, the Fish floats in the water, and therefore it is able to swim with a minimum of muscular effort. It is obvious, however, that as a Fish rises or sinks it becomes exposed to an increase or a diminution of hydrostatic pressure, which will necessarily bring about the expansion or contraction of the volume of gas in the air-bladder, and, therefore, by decreasing or increasing the specific gravity of the animal, will tend to remove the Fish from its plane of least effort. To counteract this, and to restore the Fish to a plane of equilibrium at the new level, gas is either absorbed from the air-bladder, or more gas is secreted into the bladder, as the case may be. According to Moreau, by this process of automatic adjustment a Fish will always find, sooner or later, a plane of least effort, whatever may be its depth in the water; and further, this process takes place much more readily in those Fishes which possess "red glands" or "red bodies", and with extreme slowness in those in which these organs are absent. Nevertheless, it seems doubtful if this process of adjustment can be of much use to a Fish in ordinary vertical movements, inasmuch as gaseous secretion and absorption are comparatively slow processes, the length of which in different Fishes, and under different conditions, varies from a few hours to several days. On the whole it seems more probable that adjustment to the varying pressures of different depths by such means is far more likely to be useful during such slow and gradual changes of level as are encountered in the course of diurnal, seasonal, or other periodic migrations than during the rapid changes of level which may take place in ordinary vertical locomotion.^[356] In the generality of Fishes, and more especially in the Physoclisti, it may be concluded that the possession of an air-bladder restricts freedom of movement in the vertical direction, and confines ordinary locomotion within more or less well-defined vertical limits above or below the plane of least effort for the time being. As illustrating this point, and as a proof of the danger incurred by a too rapid rise in the water, the following remarks with reference to the "Kilch," a small Salmonoid (Coregonus) inhabiting the Lake of Constance, and a favourite article of food, may be quoted:^[357]—The Fish "are caught in nets, and brought to the surface of the water; they come up invariably with the belly much distended, the air in the swimming-bladder, being relieved from the pressure of the column of water, has expanded greatly and occasioned this unnatural distension, which renders the Fish quite incapable of swimming. Under these conditions the Fish is naturally unable to live for any length of time. But the fishermen of the lake have a very simple remedy; they prick into the air-bladder with a fine needle; the air escapes with some force, the distension subsides, and the fishes are enabled to live under totally changed conditions as to pressure, even in quite shallow water and at the surface, swimming quite as freely as their companions, the natives of the surface water. Hence the Kilch is confined to a certain depth, because it is not capable of accommodating the tension of its swimming-bladder to the change of pressure in the column of superincumbent water."

It is not improbable that the Physostomi, or at any rate most of them, are somewhat more advantageously placed in this respect. From the general absence of "red glands" in this group, it may be inferred that whatever capacity for gaseous secretion or absorption they possess must be exercised with exceptional slowness, and, therefore, as a means of pressure-adjustment may be neglected. On the other hand, they seem to possess the compensating advantage of being able to substitute for absorption the mechanical liberation of gas through the ductus pneumaticus. It would seem, therefore, that the Physostomi have a distinct advantage over the Physoclisti in that during ascent in the water they can more readily adapt themselves to the diminished pressure of a higher level by ejecting the needful amount of gas than by relying upon the process of gaseous absorption.^[358] This conclusion is in harmony with the results of experiment and with much that is known of the habits of these Fishes and their greater freedom of locomotion in the vertical direction.

These briefly summarised conclusions as to the hydrostatic function of the air-bladder must, however, be accepted only in a general sense. There are many structural anomalies in the air-bladder of Fishes which are very difficult to explain, or to correlate with any variations in the habits or in the locomotor activities of its possessor.

In this connexion it may be mentioned that the presence or absence of an air-bladder in different Fishes seems to some extent to be governed by two causes. First, whenever the requirements of a Fish necessitate exceptional freedom of locomotion in all directions the restrictions imposed by the presence of an air-bladder are removed by its partial or complete suppression; a result produced, secondly, by the assumption of a bottom feeding or ground habit on the part of the Fish. Fishes like the Flat Fishes or Pleuronectidae, when not in motion by the exercise of their fins, habitually rest on the sea-bottom, and, as an air-bladder is useless under such conditions, it has, in consequence, undergone complete atrophy. Not a few Siluridae, and some Cyprinidae, inhabit the comparatively shallow waters of rapidly flowing mountain torrents, and are often provided with suckers for attachment to stones or rocks. To such Fishes as these a hydrostatic organ is obviously useless, and it has hence become greatly reduced in size, and in other respects approaches the condition of a vestigial organ.