256 AMERICAN TESTUDINATA. Part II. 



with its thin, soft, and wet matrix, the stratum Malpighii. Then, immediately under 

 this, we find a bony plate. Now this bony plate consists of two elements, very differ- 

 ent in their anatomical and physiological character ; namely, first, of parts of the true 

 skeleton, the vertebra?, the ribs, and the bones of the sternum ; secondly, of ossifica- 

 tions of the skin, or rather of the outer walls of the body, which overlie the true 

 skeleton and fill out its framework, thus making one continuous bony shield of the 

 vertebrse and ribs, and another of the sternal bones. These ossifications of the skin, 

 commonly called the dermal skeleton, are divided into many fields, like a pavement, 

 by sutures, the direction and extension of which are entirely independent of the 

 underlying framework of the true skeleton. These fields are larger where they over- 

 lie the bones of the true skeleton; they become smaller and thus relatively more 

 numerous where they reach beyond it, namely, in the margin of the upper shield. 

 As already stated, these marginal bony plates are mere ossifications of the skin 

 extending beyond the ribs. The relative direction and extension, as well as the 

 number of all these fields of the ossified skin, are very similar in the different 

 families of Testudinata. 



This composition of the shield, from the elements described above, is common 

 to all the land Turtles, to the Emydoidae, to the Cinosternoidae, to the Chelydroidte, 

 and to the South American, Eastern, and Australian Pleuroderas, the Chelyoidae 

 and Hydraspididaj. Thus far, we know only three groups which present any differ- 

 ences in these respects, the Chelonioida3, the Sphargididoe, and the Trionychidje. 

 Though we find that in the Chelonioida3 all the elements named above take part 

 in building up their shield, still their dermal skeleton is very much reduced, while 

 in land Turtles it makes up by far the largest part of the bony shield and actually 

 grows into the true bony skeleton at the expense of the latter, in such a manner 

 that parts of this disappear and are replaced by the ossification of the skin. In 

 the Chelonioidae, on the contrary, the dermal skeleton fills only imperfectly the 

 spaces between the ribs, but then it forms a regular row of marginal plates, 

 and again scantily fills the spaces between the sternal bones. In Trionychidae, 

 we observe the same partial development of the dermal skeleton, as it fills only 

 to some extent the intercostal spaces and the spaces between the stei-nal bones, and 

 forms but a few marginal plates, which may even be entirely wanting, as is the 

 case in the Southeast African Cycloderma, recently discovered by Dr. Peters, and 

 in our own Trionyx ferox and muticus. Finally, in Sphargis the dermal skele- 

 ton is developed in a very different way, namely, as one continuous shield above, 

 and another beneath, nowhere resting immediately upon the true skeleton, there 

 remaining between the dermal and the bony skeleton a thick layer of corium, 

 which never ossifies. This structure constitutes the most striking contrast when 

 compared with Testudo, where the dermal shield actually grows into the true bony 



