water fishes; but the proportion of oxygen varies at different times, and is thought to act as a reserve against a time of need. This would be indirectly a respiratory function.
Certain fishes, the pike and eel for example, have the air-bladder communicating with the stomach, while the carp and many others have it in communication with the œsophagus. In the gar-pike, a cellular air-bladder communicates with the mouth by a trachea and glottis; and the mud-fishes have the same, not only resembling thus in structure the lungs of reptiles, but performing the same function. Thus we find in the single class of fishes a progression from a closed sac of mechanical function to a double lung like that of reptiles, and the point of communication rises from the stomach to the mouth.
Fig. 6.—Anabas scandem; head, with k the gill-cavity laid open, and l the contiguous cavity containing the foliated labyrinthine structure. b, Tadpole; c, young Polypterns from the Nile; d, embryo of the shark. All these have external gills, br. (After Semper.)
All amphibians possess lungs in the adult state, but with varying degree of usefulness. Those having permanent gills may use the lungs very little, as the Proteus and Menobranchus, some of them perhaps not at all; while others, as the Siren, for instance, use them mainly. Other amphibians without gills may also quite dispense with the lungs,
