i 3 o REPTILES 



beins p mere hollow shells, with the cavities filled with air derived 

 from the windpipe. A small perforation in each of these pneu- 

 matic bones marks the point of entry of the air- tube. The occur- 

 rence of this pneumatic condition in the bones of both birds and 

 pterodactyles is perhaps the most remarkable instance of the inde- 

 pendent development of an important structural adaptive feature 

 to be met with in nature, implying as it does the correlated 

 modification of several totally distinct parts of the organism. 

 It may be added that the breast-bone, or sternum, of pterodac- 

 tyles is furnished in the middle line with a prominent vertical 

 ridge, or keel, for the attachment of the powerful muscles 

 essential to flight, thereby presenting another adaptive resem- 

 blance to birds. 



All tortoises and turtles, with the exception of the leathery 

 turtle and its extinct relatives, present the peculiarity that the 

 dermal armour is welded with the ribs so as to form one solid 

 whole. This being the case, it will be obvious that the shoulder 

 and pelvic girdles cannot occupy their normal position between 

 the ribs and the skin (that is to say the dermal armour). The 

 difficulty, as already mentioned, has been got over by trans- 

 ferring the bones of the shoulder-girdle and the pelvis to a 

 position inside the ribs. 



Many profound modifications of the skeleton have been 

 brought into existence in connection with the various modes of 

 motion characteristic of different groups of reptiles. 



The most striking and conspicuous of these occur in the 

 limbs of the marine forms, among which it is interesting to notice 

 how the same end has been attained by different means. One 

 remarkable modification in the structure of the vertebrae 

 seems to be connected with motion. In most groups of reptiles, 

 as in the higher vertebrates generally, the vertebras are 

 movably connected with one another by means of two pairs of 

 flattened articular facets or surfaces, technically known as prezy- 

 gapophyses and postzygapophyses. Such a mode of articula- 

 tion serves perfectly well for creatures with bodies of moderate 

 length and flexibility ; but it is conceivable that it would be in- 

 secure and liable to dislocation in reptiles with the elongated 

 and flexible bodies of snakes. It is, therefore, not surprising to 

 find that in those reptiles the vertebrae are provided with two 

 additional pairs of articulations, namely the projecting zygo- 



