370 AMERICAN TESTUDINATA. Part II. 



violate every natural relationshiji. At the same time, a renewed examination 

 would afford ample opportunity, even to the most skeptical, to satisfy himself that 

 the characters upon which these genera are founded have thus far, for the most 

 part, escaped notice, and constitute a real addition to our knowledge, whatever 

 be the view taken of the genera themselves. 



As to the families adopted in this revision, they bear to one another exactly 

 the same relations as all natural families have to one another in any natural 

 order of the animal kingdom. They are consequently more readily distinguished 

 by their habitus, as all natural families should be, that is to say, by their form, 

 than are the artificial groups thus far called families among Testudinata by any 

 special characters assigned to them. Why, according to present classifications, 

 Chelydra and Cistudo, for instance, should belong together to the same family 

 with our Terrapins, is not any more obvious than why the latter genus should not 

 be referred to another grouj), the Testudinina, for instance ; for there certainly are 

 as striking differences, and even differences of a higher order, between Chelydra 

 and Cistudo, or Chelydra and the common Terrapins, than between Vultures and 

 Falcons. The same may be said of Ozotheca and Cynosternum taken together 

 I when compared with either of them. And I cannot suppose that any naturalist 

 will contend that different classes of the same great tyjDe of the animal kingdom 

 should be classified upon different principles, however great the difference in the 

 nature of the characters may be. 



From what I have said in the opening of this section, it might be inferred 

 that I consider the North American species of Testudinata as too well known to 

 require much further attention and study. I am fiir from entertaining any such 

 opinion. On the contrary, I consider, in general, an accurate knowledge of species 

 as of such difficult attainment, that I do not yet venture upon sketching descrip- 

 tions of our Turtles, as I understand that specific descriptions should be, even 

 though I have already spent years in their investigation. What I offer in the 

 following pages I wish to be considered merely as contributions towards a fuller 

 illustration of this subject. It will still require long and patient studies before 

 our Turtles are known as they ought to be, in order to draw a complete pic- 

 ture of the habits, growth, and variations of every species.^ 



As to the synonymy of the species,^ it is not my intention to swell this vol- 



^ It is one thing to draw up perfect descriptions known at the time of the publication of his work is 



of species, and another and a very different tiling to very learnedly discussed by J. D. SciicErrr, in his 



write mere diagnoses, or simply to point out the j)ecu- Historia Testudinum, Erlanga^, 1792, 1 vol. 4to. For 



liarities by which closely allied species may be distin- the North American species consult Dr. Holbrook's 



guished. Comp. Part I., Chap. 2, Sect. G, p. I(i3. North American Herpetology, or Dumeril and Bi- 



^ The older synonymy of all the Testudinata bron's Erpetologie generale. 



