Chap. II. THE EMYDOID^. • 3-33 



nocted with the nasal rogioii by a long, narrow sulcus, for the passage of the 

 olfactory nerve. The palatines rise continually from the suture with the pter}-- 

 goids to the prefrontals, but at their front ends they are considerably lower down 

 from the top of tlic slvull than in the Cinosternoidae. The prefrontals meet from 

 the foramen olfactorium down to the vomer; they retreat below the foramen. The 

 upper maxillaries and the intermaxillaries do not, as in the Cinosternoidse, retreat 

 in such a manner as to carry the mouth far inward under the head, but are 

 more nearly perpendicular, thus leaving the mouth larger ; the jugals come down 

 between the maxillaries and the temporals, except that sometimes a very narrow pro- 

 cess from the former projects back under the jugals, and meets another from the 

 temporals. The jaws vary widely, but never terminate in the long, strong, sharp 

 points which exist in the Cinosternoidas. 



The shield is not completely ossified till late in life, and the bony plates are 

 very constant and regular in their arrangement. The carapace consists of the usual 

 eight costal plates on each side, of eight vertebral plates attached to the fixed ver- 

 tebrae, and of two more ^^lates not so attached, wdiich continue this row backward 

 to the marginal rim ; in the rim there are eleven pairs of plates and one odd one 

 at each end, making in all, twenty-four marginal plates. The number of plates in 

 the vertebral row varies a little, but the row itself is always continuous from the 

 odd marginal plate at the front end to the one at the hind end. The plastron 

 consists of nine 2:)lates, four pairs and one odd one. The first pair lies across the 

 front end, before the shoulder apparatus, and under the extended neck ; it is the 

 shortest and smallest. The second and third pairs, as in the other families, reach 

 clear across the body, and unite with the carapace on either side ; these two pairs 

 are much longer in the body of the plastron than in the bridge which extends 

 from thence to the carapace ; they make more than two thirds of the whole plastron. 

 The bridge sends off from each end a long process, which is fixed into the cara- 

 pace above ; when the plastron is hinged, these processes are very small, or entirely 

 wanting. The hinge, when it exists, is always between the two middle pairs, and 

 never, as in the Cinosternoidfc, between them and the adjoining pairs.' "When there 

 is a hinge, the edges of the carapace and plastron are united by a narrow, flexible, 

 unossified dermal ligament. The odd plate is just back of the suture which unites 

 the first pair to one another, and betw^een the fore part of the edges of the next 

 pair ; it sends back a slender, pointed process for some distance over the suture of 

 the second pair. The fourth pair lies under the pelvic region ; it is larger than 

 the first pair, but smaller than the second or third. 



Large epidermal scales cover the outside of the whole shield, the form and 



' Compare the note of p. 348. 



45 



