CHELONIA. 47 



admits, with respect to the portions of Trionjjx found abundantly in the gypsum of 

 the environs of Paris, associated with the Palseotheres, Anoplotheres, and other 

 extinct animals of the Eocene epoch, that he could find nothing in those fragments 

 to authorize him to fix their specific characters.* The compilers of the labours of 

 practical palaeontologists have, as usual, been affected by no such scruples, and have not 

 hesitated to assume a knowledge, which Cuvier did not feel himself entitled to claim, 

 viz. that of the fact of the specific distinction of the Trioni/x of the Montmai'tre 

 quarries : but I do not find that they have added anything to its history except the 

 name of Tri. Parisiensis. It is probable, from the analogy of our own Eocene 

 deposits, that more than one species of Trionyx may have left its remains in the 

 Parisian localities of the corresponding geological formation. 



The fossil remains of Trionyx from the tertiary deposits of the Gironde.f Lot-et- 

 Garonne,| Montpellier, and Avary, were not sufiiciently characteristic to permit 

 the great anatomist and founder of Palaeontology to infer more than the existence of 

 the particular genus of fresh-water Chelonia in question in those formations. The 

 only specimen of fossil Trionyx in which Cuvier recognised characters distinguishing 

 it from the known existing species, is that which M. Bourdet first described § under 

 the name of Trionyx Maunoir, from the Eocene quarries at Aix. Cu\'ier has given 

 reduced views of a large proportion of its carapace and half its plastron in the 

 ' Ossemens Fossiles,' torn, v, pt. ii, PI. XV, figs. 1 and 2. This description and the 

 figure of the carapace serve to elucidate by comparison the characters of the more 

 perfect specimen of Trionyx here described from the Eocene of Hordwell. 



In the first place, the contour and proportions of the entire carapace of the 

 Tri. Henrici differ from those of the Tri. Maunoir. The carapace of the Tri. 

 Henrici (PI. 6), which is formed, as usual, by the neural plates (si, «2, &c.) and eight 

 pairs of costal plates (^Vi — s), measures 10 inches 8 lines in length, and 11 inches 

 2 lines in breadth, in a straight line across the third (j'^/s) costal plate, where 

 it is widest. In Tri. Maunoir, the neural plates (plaques vertebrales) rise a httle 

 above the plane of the carapace, as in the Tri. ferox {Tri. carinatus, Geoffr.||) : in the 

 Tri. Henrici there is no trace of this carinate structure ; the neural plates are flat, and 

 on a level with the broad costal plates articulated with them ; in which characters it 

 resembles the Tri. yangeticus, Cuv., and Tri. javanicus, Cuv. 



The first costal plate {jjh) is broader than it is long in Tri. Maunoir ; in Tri. 

 Henrici its breadth is little more than half its length, and decreases as it recedes from 



* "Mais je n'ai rien trouve dans ses debris qui m'autorisat a en fixer les caractferes specifiques." 

 (Ossemens Fossiles, torn, v, pt. ii, p. 223.) 



t Cuvier, torn, cit., pp. 225, 227. 



X The skull of the Trionyx from this locality showed a slightly different profile from that of any of the 

 existing species. 



§ Bulletin de la Societe Philomathique, 1821. 



II Anuales du Museum, torn, xiv, pi. 4, 1809. 



