44 NATURAL HISTORY OF VERTEBRATES. 



which project widely outwards from the branchial arches, and do not come into 

 relation to tlie clefts separating them. 



The Air-Bladder. 

 Various accessory respiratory organs occur in fishes which are exposed to peculiar 

 physical conditions, such as the drying up of the rivers or ponds in which they live. 

 These exceptional structures will be described under the genera in question; but 

 almost all fisli-like forms, with the exception of the sharks, possess a bladder formed 

 from the anterior part of the alimentary canal, which, as it contains air, and has blood- 

 vessels in its walls, is often of respiratory importance. This air-bladder is generally 

 considered to be homologous with the lungs of higher vertebrates, having come first 

 partially, and then entirely, into use instead of the gills. We shall consider this 

 homology after ha\ing discussed the form and function of the bladder as it ordinarily 

 occurs. 



In the Ganoids and more primitive Teleosts (Physostomi) the air-bladder communi- 

 cates with the intestine, generally in front of the stomach, and either by a slit, as in the 

 Ganoids, or by a duct. On the other hand it is completely closed from the intestine in 

 the less primitive Teleosts (Physoclysti). Whether provided with an air-duct or not, it 

 lies between the intestine and the vertebral column, and when it opens into the intes- 

 tine does so, with but few exceptions, on the dorsal wall. Confoi'mably with its devel- 

 opment it presents the same coats as the intestine, viz., a connective tissue coat, a mus- 

 cular coat, a submucous coat containing vessels, and a mucous coat. The last rarely 

 presents any glands, and may often be formed of a single layer of flat epithelium ; so, 

 also, the submucous coat may contain very few vessels, and thus be of no importance 

 for respiratory purposes. But in certain instances the vessels are more numerous, 

 and may be extremely abundant, as in the bow-fin and gar-pike {Amia and Lepidos- 

 teus). Indeed, in the former genus, almost all the blood sent through the last pair of 

 gill-arches is diverted into the walls of the air-bladder ; the same is the case in the 

 African Ganoid Poly2)terus, where the air-bladder opens on the ventral wall of the 

 oesophagus, and we have only to conceive the series of gill-filaments on the gill-arches 

 to be absent, and the blood to pass through the corresponding aortic arches without 

 being aerated, to have the air-bladder supplied with blood in exactly the same way as 

 are the lungs in the air-breathing vertebrates. 



In many forms destitute of an air-duct, the cajjillary vessels are not distributed 

 regularly over the whole surface of the bladder, but form a dense raeshwork at one or 

 more spots, known as retia mirahilia. These are thought to serve for the absorption 

 or secretion of air, which thus comes from the blood, in the absence of a direct com- 

 munication with the outside. Even when a duct is present, it is frequently so narrow 

 and tortuous that it is hard to conceive air being taken in through it, although, occa- 

 sionally, bubbles are, no doubt, ejected by this channel, so that we have to conclude 

 that also, in such cases, the blood vessels are the source of the air contained in the 

 bladder. But where a wide opening exists, as in the Ganoids, it is obvious that the 

 respir.atory importance of the bladder may be much greater. 



The muscular coat is often undeveloped, but in certain cases it may contain strijied 

 fibres, continuous with those of the oesophagus, or it may be functionally replaced by 

 muscles arising from the vertebral column and distributing themselves over the blad- 

 der. In many instances, however, the air contained in the bladder does not ap]iear to 

 be controlled by voluntary muscles. Such is apparently the case where the outer 



