24G AMERICAN TESTUDINATA. Part II. 



they retain undivided fingers, such as the embryos have, even exaggerating this 

 feature, in the adult, into an elongated paddle for the anterior limbs. Chelonii con- 

 stitute, then, the lowest sub-order in the order of Testudinata ; and it will presently 

 be seen that its characters are not derived from the form of its representatives. 

 Those Avho are sufficiently conversant with the subject will be aware that when 

 characters derived from the form have been added to the other characters in order 

 to distinguish the Chelonii, they have answered but indifferently ; indeed, the form 

 of Sphargis and that of Chelonia differ much more than that of Emjdoidae compared 

 with Testudinina. The scaly Chelonii, the ChelonioidjB proper, have their shield 

 more or less heart-shaped, and the posterior angle is not prolonged into a projecting 

 point extending far over the tail, as is the case among the naked Chelonii, the 

 Sphargididse. For this and other reasons which it would be superfluous to mention 

 here, as my object is not now to characterize every group of Testudinata minutely, 

 I hold that Chelonioidoe proper and Sphargidida?, which differ by their form, are two 

 distinct families in the sub-order of Chelonii, and that this sub-order exhibits struc- 

 tural features of inferiority when contrasted with the other Testudinata. Gray and 

 Bell, in their early publications, had, in my opinion, correctly distinguished SphargidiB 

 and Chelonida3^ as families, even though they afterwards gave up that distinction 

 and placed them incorrectly upon one level with Trionyx, Emys, and Testudo. In 

 this respect, Fitzinger presented this matter in a more correct light when, like 

 Oppel, he contrasted the united Chelonii with the other groups of the order; but 

 I believe he was mistaken in urging the reunion of the families of Sphargidaj and 

 Chelonida3. If the view which I have presented of the case is correct, tlie marine 

 Turtles would constitute a sub-order, for which a variety of names had been 

 proposed : that of Pterodactyli l^y Fr. Meyer, that of Thalassites by Dumeril and 

 Bibron, that of Oiacopodes by Wagler, that of Eretmochelones by Ritgen, that of 

 Pinnata by Merrem, and that of Chelonii by Oppel, all of which are perfectly 

 synonymous. That of Oppel, which is the oldest, having been proposed in 1811, 

 should have made all the others superfluous, and ought now to be retained. This 

 sub-order includes two families, the Chelonioidaj and the Sphargididte, as these differ 

 in form. Their characteristics are fully illustrated in the next chapter. 



The scarcity of Trionyx in European museums seems to have prevented so accu- 

 rate a study of that group as of the others. It is, at least, surprising that some 

 of the ablest herpetologists have failed to perceive how greatly they differ from 

 the other fresh-water Turtles. Wagler unhesitatingly unites them with the Emyds, 

 while quite recently Major LeConte has united them with Chelydra.^ Yet, as 



^ When I quote the systematic names of original ^ LeConte, (Major,) Catalogue of the North 



writers, I follow their spelling ; in other cases, I American Testudinata, in Proc. Ac. Nat. Sc, Thila. 

 adopt that which seems to me correct. vii., 18o4. 



