Chap. II. THE TESTUDININ A. 361 



head is witlKlrawn far back, but the .skhi does not i-oll ofl' from the neck so ftar 

 as to Ibid together before it, as in the Einydoida^. The humerus is carried round 

 l)efore the body, the knees brought together before the head, and the forearm and 

 hand turned back iipon the humerus. See PI. 3. The knees meet before the 

 humerus reaches directly across the body, and they are somewhat raised above tlie 

 shoulders, which is made necessary by the rise of the jilastron forward, so that 

 tlie humerus reaches somewhat outward and upward, and not exactly across the 

 body. The blade formed of the forearm and hand is nearly as broad as the 

 opening about this end of the body, and when the knees are brought together 

 the opening is almost entirely closed, and the surface of the forearm and hand 

 exposed before it. The femur is carried to the side of the pelvis, reaching upward 

 as well as forward, so that the knee is raised high uj) within the carapace ; the 

 foreleg is turned down and back upon the femur, and the foot and hip thus brought 

 together occujiy the whole open space by the side of the plastron, so that the bot- 

 tom of the foot and the liind surface of the hip only are exposed. The short, 

 stubbed tail is bent directly forward (when longest a little curved) between the 

 hips, so as to cover most of the surftxce behind the pelvis. All the parts exposed 

 when the limbs are thus withdrawn are covered with thick, hard scales. 



The food of this fomily is exclusively vegetable. They seem to pi-efer the 

 succulent stems of plants and fleshy fruits to leaves or grass. I have often seen 

 our Gopher gnawing the stumps of cabbage and the apples falling from the trees, 

 in my garden, as the squirrels do, holding them between their feet. This vegetable 

 diet seems to affect essentially the structure of the digestive ajiparatus, for in our 

 Gopher (the only genus examined) the large intestine is longer than all the rest 

 of the alimentary canal, including the stomach and oesophagus, whereas in no one 

 of the many genera which have been examined of the families of Eraydoida^, Cinos- 

 ternoidi\>, and Chelydroida^, does the proportion reach as high as one to five. The 

 lungs are very much largt-r in the Testudinina than in any other family of the sub- 

 order, which is undoubtedly due to the exclusively terrestrial habits of the animal. 



These two peculiarities of structure, the great length of the large intestine, and 

 the large size of the lungs, directly traceable to the haljits of life, go far towards 

 giving the middle region of the body its peculiar size and form. A connection 

 will readily be seen also between the ]iri>poi'tions of the terminal regions, which 

 are high and short, and the maimer of walking and of withdrawing the limbs, inas- 

 much as the legs move in a plane so nearl}^ perpendicular, and the knee and 

 elbow joints are raised Avlien retracted so liigh up ^\itllin the carapace. Again, the 

 equilibrium throughout the body is clearly connected with the stead}-, straightfor- 

 ward motion in walking. Thus this family exhibits, more closely than any other, 

 the direct relation which exists between the form and structure. 



46 



