Chap. I. THE SKELETON. 269 



lower end rests in a socket, in the foremost part of tlie sternnin ; but in Turtles 

 the whole shoulder apparatus being drawn inwards and backwards, this bone had 

 to be removed from the sternum, and lies free in the muscles. 



The humerus is short, crooked, and turned inwards in such a way that it 

 moves inwards in one plane with the scapula and coracoid. The forearm is articu- 

 lated \ipon the large lower epiphysis of the humerus, but its position is peculiar 

 to the Turtles, its transverse diameter standing vertically. This is effected by an 

 overlying of the fibula upon the radius. In the structure of the hand, we find 

 again, in the same manner as in the forearm, the transverse diameter standing ver- 

 tically, the ulnar side above, the radial side below. This singular conformation of the 

 shoulder, the arm, the forearm, and the hand, makes it possible for the fore leg to be 

 drawn back under the iipper shield by the bending of all the joints in the plane 

 of the scapula. This motion is more or less extensive in different families, accord- 

 ing to the degree of expansion of the carapace. 



The conformation of the hand varies much in different families, according to its 

 function as a jiaddle, as a fin, or as a jiillar.^ 



The pelvis is much easier to understand than the shoulder. It is formed, on 

 each side, by three pennanently distinct bones, meeting in the cond3'loid cavity. 

 Two pairs of these bones are flat and more or less horizontal, and rest upon the 

 sternum, to which they are more or less closely attached. The larger pair, the 

 ossa pubis, leans forwards, the smaller pair, the ossa ischii, backwards. The 

 bones of each pair unite respectively with one another in the middle, in a 

 median line, while the two bones of the same side, meeting laterally, form the 

 lower part of the cavity for the femur. The upper part of this cavity is formed 

 by the third pair of the pelvic bones, the ossa ilii ; these are smaller cylindrical 

 bones, much enlarged at both ends, running upwards and backwards, and meeting 

 with the long transverse processes of the sacrum. 



The bones of the hind leg agree generally with those of the fore leg, tliough 

 the femur is straighter than the humerus. There are, however, great differences 

 in different flimilies, in respect to the relative size of the two jjairs of the legs. 

 These diflferences are .so strongly marked between the marine Testudinata on one 

 side, and the fluviatile and terrestrial types of the order on the other side, that 

 they cannot be considered as family characters, but rather point out a natural 

 subdivision of the whole group, alread}- hinted at above,^ and to which I shall 

 asrain call attention hereafter. 



•d" 



' Sec the Family Cliaracters, below, Cliap. 2. ' See, above, Sect. 2, p. 241-249. 



