This demonstration, which dispels all doubt, establishes the physiological principle, that the presence of epithelium is compatible with the respiratory office of the part Which it clothes. This law prevails throughout the class of fishes ; it has also been reduced to actual fact by the author throughout the whole sub-kingdom of the invertebrata. But it must not be forgotteri that its office on the breath ing organs is almost exclusively mechanical. In no known example among vertebrated animals does the epithelial investment of a respiratory surface develop itself into any of the forms of a secreting organ. No " follicles" are, at any time or under any circumstances, discovered on these localities, in this class of animals, but in the inver tebrata follicular glandules are constant on the surfaces of the respiratory organs. The constituent Leales are therefore function ally- passive. Nuclei and a granular proto plasm would find no purpose to subserve if they were present in a highly developed form. Thus is exemplified the law of " demand" and " supply :" disuse entails attenuation on all living structures. Either the gases by which the epidermis of respiratory localities is traversed suppress the glandular office of the scales, or these latter, from the first, re ceive a special organisation. The scales of the bronchial epithelium contain nothing but a pellucid fluid. This fluid condenses, fluidi fies, the respiratory gases in transitu. This is the office of the refined covering under study. The " principle" that the epithelium of the breathing organs is required by the physical conditions of its office to be reduced to the state of the utmost thinness receives new proofs from the study of the internal branchire.
From all that is known it is probable that in minute structure the branchim of the per enni-branchiates conform to the plan of the temporary organs just described. The ge neral arrangement of the primary bronchial vessels and the structure of the heart are identical.
The internal temporary Branchiw of tIze Amphibia.— The process by which these or gans are withdrawn into the interior of the bronchial chamber is not simply that of short ening. It is the labour of a new organisation. The internal gills of the tadpole differ in type of structure from the external. The ultimate vessels of the latter are differently' looped.
In both they are simple loops, but a distinc tion is obvious ; so evident as to render it impossible that the traosition from the ex posed to the concealed state of the branchim can consist in a bodily retractation. On the external organs, preparatorily to their disap pearance, the vibratile cilia first cease, the epidermis then increases in density, the meshes between the blood-capillaries enlarge, and the vessels become obliterated. These declining changes are not limited to the extreme distal ends of the branchial lobules. They occur simultaneously on every part of the surface. Temporary arches (c) of delicate cartilage now arise within the branchial chamber. It is from the convexities of these arches after the manner of the pisciform t3pe that the new vessels (a) of the internal temporary gills proceed. They are appended under the character of de licate flocculi. Enlarged, they appear as minute digitations. Each carries a looped vessel, and is loosely invested with a de licate membrane (d). This membrane belong,s to the mucous, not to the epidermal, class and yet it differs in a striking manner from that which lines the rest of the branchial chamber. Nowhere is it ciliated. That co vering the branchial vessels is remarkably thinner than the parietal portion. The fonner, however, is true epithelium. Its constituent scales are distinctly traceable by their out lines, though they are as structureless as a basement membrane. It is not often that it happens that the epithelium of a breathing organ overlies, as in this instance, perfectly homogeneous parts. Nothing but the proper coats of the vessel lie underneath. They are literally structureless and hyaline. The cells of the superficial epithelium, therefore, admit of indisputable definition. It is not " base ment membrane," but epithelium, though attenuated, that here invests the respiratory vessels. By this demonstration a principle is established. Epithelium is not supplanted by any other structure on the organs dedicated to respiration. No other instance, however, is known within the limits of the vertebrate kingdom in which this epithelium is ciliated, than that afforded in the case of the tem porary external gills of the Amphibia. On those of fishes these motar organisms do not exist. Wherefore this distinction ? Why should they exist on the external and not on the internal gills ? It is not a law of the mucous membrane that they should not exist, for they occur in other tracts of this sarue structure. These are questions of ultimate
design which it is not given to science to answer.
Air-bladder of fishes.— This organ repre sents the prototypal form of " the lung" in the animal kingdom. It is present in nearly all osseous fishes. It is always tensely filled with gas. In that of marine fishes, oxygen predominates ; in that of fresh water, nitrogen. Humboldt found the gas in the air-bladder of the electric gymnotus to consist of 96 parts of nitrogen, and 4 of oxygen. Biot found 87 parts of oxygen, nitrogen, and carbonic acid in the deep Mediterranean fishes. No hydrogen has ever been detected in this organ. It occupies the roof of the abdomen, between the kidneys and chylopoxtic viscera, and sometimes (gymnotus ophiocephalus, coius), beneath the caudal vertebrm to nearly the end of the tail. In some species of diodon, tetradon, dactylopterus, penwlodus, and poia notus, it is bifurcated. In arias gagora, poly pterus and lepidosiren,it is divided lengthwise into two bladders. In the cyprinidw and characinidx it is divided transversly into two communicating compartments. Many other varieties of form occur. (Vide art. PtseEs.) The proper walls of the air-bladder consist of a shining silvery fibrous tunic, the fibres being arranged for the most part transversely and circularly, and in two layers. They are con tractile and elastic. This coat yields the finest gelatine. Its fibres belong to the white variety : they " swell" under the action of acetic acid. A stratum of vessels is inter posed between the mucous membrane and the fibrous layer. The meshes formed by these vessels are considerably larger and more oblong than those of the pulmonary capil laries. In the latter instance the meshes exceed the vessels in diameter. The arteries of this organ are derived sometimes from the abdominal aorta, sometimes from the cmliac artery, sometimes from the last branchial vein ; and in thelepidosiren they are continued from the aortic termination of the two non-rami fied branchial arteries, and therefore convey venous blood to the cellular, lung-like, double air-bladder (Owen). The veins of the air bladder return, in some fishes, to the portal vein ; in some to the hepatic vein; in some to the great cardinal vein ; and in the lepido siren, they penetrate by a common trunk the great portal vein formed by the confluence of the visceral and vertebral veins of the trunk. In the protopherus and ganoid fishes the vessels of this organ forrn no retia nzirabilia and vaso-ganglions, but rather a diffused capillary network, more close and rich in the anterior than the posterior part. In the osseous fishe.s, several varieties of the vascular system of this organ occur. That of the carp forms tufts of capillaries throughout the whole interior of the organ, a variety of which tufts occurs in the pike. The perch and cod exhibit a vaso-ganglion, a body pe culiar to the air-bladder of fishes. In the cod-fish, a large artery, a branch of the cceliac, and a still larger vein, which empties itself into the mesenteric, perforate together the fibrous tunic of the bladder. Before they reach the inner surface, they divide into some branches which then radiate and sub divide upon the mucous membrane. The arterioles frequently anastomose with each other. Both are inextricably interwoven, and form the basis of the so-called " air-gland," which is essentially a larger "bipolar rete mirabile" (Miiller), or vaso-ganglion. In the cod the ultimate vessels of this gland have a loop-like arrangement, their free surface (a a), being covered over with another layer of vessels and epithelium This organ, however, is further composed of a number of pecu liarly arranged, elongated corpuscles, which descend in two rows from each vascular branch, and are bound together by a loose cellular tissue : the corpuscles are beset vvith fine villiform processes. Thus it should be noticed that the veins as well as the arteries concur to form the vaso-ganglions. The vaso-ganglions of the eel and conger are placed at the sides of the opening of the air-duct, are " bipolar," and consist of arteries and veins ; their efferent trunks do not ramify in the immediate margin of the vaso-ganglion from which they issue, as in the vaso-gan glion of the cod, turbot, acerine, and perch, but run for some distance before they again ramify to form tbe common capillary system of the lining membrane of the air-bladder. In the parasitic and suctorial derrnopteri and pleuronectidm and ray-tribe the air-bladder does not exist.