CuAP. I. SPECIAL CLASSIFICATION OF TESTUDINATA. 243 



lies, with two sub-fainilics of Canino, correspond exactly to the lour families of 

 Dunieril and Bibron, the difference lying only in the separation, as families, of 

 the Chersites and Elodites by Dum^ril and Bibron, while they constitute two sub- 

 families of the Testudinidae of Canino. Again, the Chersites, the united Potamides 

 and Elodites of Dumeril and Bibron and their Thalassites represent the divisions 

 of Ritgen and Wagler. I do not mean by this to say, that the separation of 

 the Potamides and Elodites is not natural, but only to allude to the fact that 

 Dumeril and Bibron's Thalassites correspond exactly to Ritgen's Eretmochelones and 

 to Wagler's Oiacopodes, while their Chersites ansAver to Ritgen's Podochelones and 

 to Wagler's Tjdopodes, the Potamides and Elodites of the French herpetologists 

 corresponding together to the Phyllopodochelones and Steganopodes of the two 

 German writers. 



The agreement, and the discrepancies between these different systems, then, 

 consist in this, that Oppel and Merrem and with them Bell, admit two higher 

 subdivisions in the order of Testudinata, those with oar-like feet and those with 

 distinct fni'>;ers, while Ritgen and Wajrler admit three, distinguishing between those 

 the visible fingers of which are webbed, and those in which they are entirely 

 separated, while Dumeril and Bibron introduce a farther distinction between those 

 with webbed feet and a scaly body and those with a naked carapace, the Emj^ds 

 proper and the Trionyx. Canino maintains this distinction between the naked 

 and scaly fresh-water Turtles, but as he unites all the scaly ones together, whether 

 their fingers are webbed or not, his division includes the Chersites of Dumeril 

 and Bibron as well as their Elodites. The sub-families which Dumt^ril and Bibron 

 introduce among the Elodites are founded upon the mode of motion of the neck, 

 which exhibits differences already noticed by Wagler in 18-30. Bell, Gray, and 

 Fitzinger, Avho have a still larger number of groups which they call fiimilies, have 

 founded them upon the same features which have led Dumeril and Bibron to 

 subdivide the Elodites. I do not here speak of the classifications of Fleming^ 

 and Latreille,^ which are too artificial to deserve special notice. 



Beyond these divisions, all authors mention only genera and sub-genera. Now, 

 it must be obvious, from the agreement of all these writers in some points of 

 their subdivisions of the Testudinata, that this order is not so homogeneous as to 

 exclude higher divisions than genera in its classification. The point on which all 

 agree is, the separation of the Turtles with oar-like, natatory organs of locomo- 



^ Flkmino, (J.,) The Philosophy of Zoology, ' Latreille, (P. A.,) Families naturoUos du 



London, 1822, 2 vols., 8vo., divides the Chei.osea, regno animal, Pari.s 182."), 1 vol., 8vo., divides the 



as he calls the Testudinata, into those with a movable Chkloxians into those whieh can retract their legs, 



and those with an innnovalile sternum. Cryptopodes, and those which cannot, Gyninopodes. 



