134 PROCEEDINGS OF THE ACADEMY OP [1885. 



batrachian. But from that time forward the dominion of the Iish 

 on the land must have steadily decreased. The fin could not 

 compete with the leg and foot as an organ of land motion, and 

 the Dipnoid fishes were probably driven back to the water. As 

 a result of this change of condition a retrogressive evolution 

 took place in the air-breathing organ. Some fishes continued to 

 use it occasionally as a lung, of which we have instances in the 

 modern Dipnoi. Yet with the Ganoids, as a rule, it probably 

 never attained a lung-like development, and was used only for 

 temporary breathing purposes. This is its condition in most of 

 the few existing Ganoids. But with their successors, the Teleos- 

 teans, it lias lost all air-breathing capabilities, and has passed 

 through every stage of degeneration, from a condition closely 

 resembling that of the Ganoids to complete extinction. And in 

 this process of degeneration it has been, in certain cases, adapted 

 to minor uses, some of the most probable of which have been 

 above enumerated, while there ma}' be others as yet unknown 

 to us. 



A consideration of the gaseous contents of the air-bladder may 

 lead to a conception of one such possible use. It is somewhat 

 remarkable that it contains nearly pure nitrogen in fresh-water 

 forms, while in the deep-swimming sea fish oxygen forms its main 

 contents, often to a very large percentage. There must be some 

 sufficient cause of this difference of contents. It is not due to 

 any difference in the gases contained in water at various depths, 

 for the percentage of nitrogen is closely the same at all depths, 

 while oxygen diminishes in quantity from the surface downward. 

 Thus, if its contents depended on the relative quantity of gases 

 present, nitrogen should predominate below as well as above. It 

 is probable, however, thai the presence of oxygen in the bladder 

 of deep-sea fishes is really due to the smaller quantity of oxygen 

 there present in the water. The bladder may serve as a comple- 

 mentary aerating apparatus, as suggested by Semper, a reservoir 

 of oxygen for the use of the Iish during sleep, or when, from any 

 cause, no1 actively breathing, or in poorly aerated water. Such a 

 fund ion would be of little or no importance to surface fish, which 

 can readily obtain water rich in oxygen. Ami these fish, for this 

 reason, may secrete only the useless nil rogen into the air-bladder. 

 But for deep-water fishes this function may be highly necessary. 

 When actively breathing they probablj obtain little more ox3*gen 



