1 1 04 CLASS REPTILIA. 



the areas which they overlie on the carapace are conspicuously 

 fluted. The last genus, Tropidemys, while having the roof- like 

 posterior neurals of Pelobatochelys, is distinguished by the almost 

 regularly hexagonal form of these bones, owing to the lengthening 

 of the anterior and the shortening of the posterior lateral surfaces. 

 One species from Hanover referred to this genus shows the unique 

 feature of a median row of numerous small intervertebral shields 

 dividing the normal vertebrals into two lateral series. This genus 

 occurs in the Lithographic Limestone, in the English Kimeridge 

 Clay, and the Lower Greensand of Switzerland (T. valanginiensis). 

 The detached neurals are readily distinguished not only by their 

 shape, but by their excessive thickness. 



Family Chelydrid^. — This family is confined at the present 

 day to North, and part of South America, where it is represented 

 by the Alligator-terrapins (CAe/ydra), and the larger Snapper {Macro- 

 clemmys). The skull is more or less triangular, and very broad 

 behind ; the temporal fossa is partially roofed over, but the squa- 

 mosal is widely separated from the parietal ; the bones of the palate 

 do not develop plates to floor the narial passage ; and the tympanic 

 ring of existing forms, like that of the Testudinidcz, is in great part 

 closed behind. The nuchal bone of the carapace develops a long 

 rib-like (costiform) process on either side, which underlies the mar- 

 ginals ; and the complete union of the marginals with the costals 

 does not take place for a long period ; while in the plastron 

 vacuities frequently persist till late in life, and the plastron itself 

 may be relatively small, and unites with the carapace by gomphosis, 

 the hyo- and hypoplastrals not giving off buttresses to underlie the 

 carapace. In existing forms the caudal vertebrae are opisthoccelous ; 

 and there is no bony union between the ischium and pubis of the 

 same side to enclose an obturator foramen. The head cannot be 

 completely retracted within the shell. The humerus is not flattened, 

 and has its radial and ulnar processes large, and directed towards 

 the ventral aspect. 



If all the genera provisionally included in this family are rightly 

 referred, it will be the oldest representative of the order which still 

 exists. The genera may be divided into three subfamilies, accord- 

 ing to the presence or absence of sculpture or epidermal shields on 

 the shell. It is, however, by no means certain that all these forms 

 should be included in this family ; but until we know the skull and 

 caudal vertebrae of all the genera it is impossible to decide this 

 point, and it may eventually prove that there is a more or less com- 

 plete passage to the Dermatemydidce from this family, since there 

 already are some indications of a connection between them. Thus 

 the American Cretaceous genus Toxochelys is stated by Dr Baur to 

 be a true Chelydroid, but to have the proccelous caudal vertebrae 



