Chap. II. THE SPITARGIDID^. 323 



the fourth pair meet at their hind ends \nider the pelvis. So we have an 

 irregular ellipse of true bone, narrowed backward. This ring does not touch the 

 ribs. The ribs are broad and Hat, firmly supported and kept in position by 

 the corium resting upon them. The first pair is free from the second, and so 

 is the tenth from the ninth. The ninth extends back by the side of the pelvis, 

 and thus strengthens the narrow end of the carapace. The specimen examined 

 has only some of the Jieck vertebraj preserved, among which is the last ; this 

 has verj' little motion at the joint Avith the first dorsal vertebra. There are 

 no scales on any part of the skin ; at least, there are none on the skin of the 

 only genus thus far known to belong to this family. 



The skeleton is light; the shield narrow and small in proportion to the size of 

 the animal, and so placed with reference to the limbs as to be as little cum- 

 bersome as possible ; the surrounding thick layer of corium is filled with fat ; the 

 body is rounded, and the wings and paddles are large and free. These characters 

 seem to indicate that the animal travels far and fast. This assumption \\ould cer- 

 tainly l)e justified, if it can be shown, as I shall attcmj)! to do,^ that the speci- 

 mens of Sphargis coriacea, observed in Europe, had travelled across the Atlantic 

 from the coasts f)f North America. 



The head is high, short, and very broad at the hind end. As in the other 

 members of the sub-cu-der of Chelouii, the parietal, pustfrontal, jugal, temporal, and 

 mastoid bones are so extended as to form one continuous bony roof over the 

 whole head back of the eyes, protecting the temporal muscles, and projecting 

 somewhat back over the first neck vertebra?. In Sphai-gididaa the parietal bones 

 are almost exclusively devoted to the formation of that arch, as they enlarge the 

 cavity of the brain-box only by a depression in their thickness, and a sulcus formed 

 by t\vo low ridges, and do not reach doAvn to the floor of the skull, the upper 

 occipital l)one extending entirely across the brain-box under them. The temporal 

 bones are small, and reach out\vard, so as to add rather breadth than length to 

 the bony arch, thus making more room for the temporal muscles. The lower 

 edge of the temporal and jugal Iiones, at their meeting, is deeply concave, thus 

 allowing a broader attachment of the )nuscles for the lower jaw, and leaving 

 them here somewhat exposed. The tloor of the skull is carrieil far forward, con- 

 siderably be}ond the end of the roof The prefrontals do not extend l)eyond 

 the frontals, but the front edges of both make the front end of the top of the 

 skull ; the roof formed by thou does not extend more than half way over the 

 nasal cavity. The os quadratum descends low flown, and carries the artifulation 

 of the jaws far below the general level of the floor of the skull. The outer 



» See, Ir'Iow, Cli.ii). 3. 



