AlR-in.ADDEIi OF FISHES. 273 



"Maigre" and other si^ecies of Scicena, with most of the Corvince, 

 have very numerous lateral pneumatic ca^.ca, which are more or less 

 ramified.* In some species of Cheilonemus and Gadus blind pro- 

 cesses are continued from both the sides and ends of the air-bladder 

 (see the anterior ones in Gadus callarias, Jig. 69. a, p). In Gadus 

 Navavaga the lateral productions expand, and line corresponding 

 expansions or excavations of the abdominal parapophyses, thus foi-e- 

 shadowing the pneumatic bones of birds. 



The proper walls of the air-bladder of ordinary osseous fishes 

 consist of a shining silvery fibrous tunic, the fibres being ai*ranged 

 for the most part transversely or circularly, and in two layers 

 (yfig. 58, q r) : they are contractile and elastic ; but the walls of the 

 anterior compartment of the air-bladder of Cyprinoids (ib. p) are 

 much more elastic than those of the posterior one. The air-bladder 

 is lined by a delicate mucous membrane, with a ' plaster-epithelium ; ' 

 it is more or less covered by the peritoneum. Its cavity is commonly 

 simple ; in the Sheat-fish it is divided by a vertical longitudinal 

 septum along three-fourths of its posterior part, f The lateral com- 

 partments are subdivided by transverse septa in many other Siluroids 

 (e. g. genus Bagrus) : the large air-bladder of some species of Ery- 

 tlirinus (e. g. E. salvus, E. tcBniatus) is partially subdivided into 

 smaller cells. The cellular subdivision is such in the air-bladder of 

 the Amia, that Cuvier compared it to the lung of a reptile:}: ; and 

 the transition from the air- or swim-bladder to the lung is com- 

 pleted in the Protopterus or Lepidosiren annectens, in which the 

 cellular subdivision and multiplication of the vascular surface are com- 

 bined with a complete bilateral partition of the bladder into two 

 elongated sacs, with a supply of venous blood from a true pulmonary 

 artery, and with the communication of the ductus pneumaticus, as in 

 the Polypterus, with the ventral surface of the oesophagus. 



At the first introduction into the Animal Kingdom of a true lung, 

 or air-breathing organ communicating with the pharynx or a3So- 

 phagus, much variety of form and structure, much inconstancy even 

 as to existence, might be expected, especially in that class in which 

 the normal function of the new organ could be so seldom in any 

 degree exercised, and in which, tlierefore, different accessory or 

 subordinate ofiices predominate in such rudimental representative of 

 the pulmonary organ. There is no swim-bladder, for example, in the 

 orders Dermopteri, HoloccphaU, and Elugiostomi ; it is present in one 

 of tlie families {Gadidce) of the thoracic suborder of Atiacanthinif 



* See xxxix. i. p. 94., after Cuvier et Valenciennes, xxiii. pi. 1:58, 139. 

 t cxiv. ii. p. 33. pi. G. fig. 4. \ xxiv. ii. p. 377. 



VOL. II. T 



