8 INTKODUCTION. 



In order to do tVn, we will return to the more striking ot 

 the examples here given. There can certainly be no doubt of 

 this — that the lungs of mammals have not been developed by 

 modification from those of land snails ; we know on the contrary, 

 or will assume, th:at the lungs of all th§ Vertebrata are identical, 

 and to be regarded as modifications of the air-bladders of the 

 bony fishes (fig. 3), although these organs do not serve, or at 

 any rate do not mainly serve, for respiration. On the contrary, 

 fish breathe by their gills. But the lungs of mammals differ 

 remarkably in structure from those of birds, and yet more from 



Flo. 4.— », the bono o( n. Cat, showing the marrow-tube; b, that of a Bird with 

 cavities containing air instead of marrow; c, the slvull of a Buccros sawn through. 

 Air-cavities traverse every part of the bone. 



those of the lower reptiles or the amphibia. In these last they 

 are usually simply capacious sacs opening into the mouth by a 

 very short passage (the trachea) ; in mammals they exhibit a 

 spongy structure, and often a highly complicated arrangement 

 of extremely long air-tubes ; in birds also the lungs have a 

 spongy structure, and connected with them there are always 

 , numerous air-cavities which lie pai'tly in the cavity of the body 

 and partly, in the form of canals, deep in the bones of the skull 

 (fig. 4) and of the vertebral column, or penetrate to the end 

 of the extremities, forming what are known as pneumatic 

 bones. Now these differences in the structure of the luniS's of 



