198 THE AIR-BLADDER OF FISHES 



essential function, and is on the road towards extinction. On no 

 otlier theory can we explain its great diversity in nearly-related 

 species. 



If, then, we may look upon the air-bladder as an organ which 

 has partly or wholly lost its original function, the question 

 follows, What was that function ? There are certain good reasons 

 for believing that the h-eathing of air was the original purpose of 

 this orgaji. In mature Teleosteans this is occasionally indicated 

 by the existence of a pneumatic duct connecting with the oeso- 

 phagus. This duct is usually of no functional use, and varies 

 from partial to complete disappearance. But all fishes with an 

 air-bladder possess a duct in the early stage of embryological develop- 

 ment. In the mature stage it is lost by all Teleosteans except the 

 Physostomes. The extraordinary development of retia mirabilia 

 in the inner tunic of many air-bladders, though now used only for 

 the secretion of gases, may be survivals of ancient pulmonary 

 capillaries, which have changed their character with their function. 

 And the apparatus for compressing and dilating the bladder may 

 have been originally developed purely for respiratory purposes, as 

 this embryological persistence of the pneumatic duct would lead 

 one to suppose. 



From the evidence afforded by embryology, we proceed to 

 that given by paleontology. The ancestors of the modern Teleos- 

 teans, the Ganoids, all possess an air-bladder, which remains a 

 fully-developed air-duct in the mature stage. In the sub-order, 

 Dipnoi, the air-bladder is functionally developed as a lung, and as 

 counterparts of these fishes existed in the Devonian age, it is pro- 

 bable that they breathed air then as they do now. 



In the higher animals the duct connects with the ventral side 

 of the oesophagus, whereas in most of the Ganoids the air-duct 

 opens into the dorsal side of the oesophagus. But Wilder shows 

 there is a series of forms, mostly Ganoids, leading from Amia and 

 Lepidosteus, with the air-duct entering the throat or the dorsal 

 side, to Lepidosirea, in which it enters on the ventral side, as in 

 the higher vertebrates. In all the fishes just named, the air- 

 bladder functions as a lung ; in Polypteros it has lateral divisions ; 

 in Lepidosteus, the American Gar-Pike, it is cellular and lung- 

 like. The fish keeps near the surface, and may be seen to emit 



