8 BULLETIN 34, UNITED STATES NATIONAL MUSEUM. 



There are always a distinct so-called ceratohyal ; one or more axial ele - 

 ments or basibranchials, and lateral pieces, or ceratobranchials. There 

 is no glossohyal. Further than this nothing can be said of the class, 

 as the orders differ among themselves in the details. 



In the carpus and tarsus there is always an os centrale, except in the 

 tarsus of the Salientia. In the extinct Carboniferous genera Eryops^ 

 and Archegosaurus,^ there are two centralia in each foot {pes of Eryoj)s 

 unknown). The intermedium is either present or fused with one of the 

 adjacent bones of the first row. There is a series of carpalia and of 

 tarsalia, giving attachment to the metapodials, whose number and con- 

 dition differ in the different families. The number of digits is generally 

 not more than four in front and five behind. There is very often a rudi- 

 meutal thumb on the anterior foot in the Salientia, and in the extinct 

 rhachitomous genus Eryops there are five well-developed digits on the 

 manus. (Plate 45, fig. I.) In the Salientia there is often a rudimental 

 sixth toe internal to the hallux. (Plate 67-8.) 



The shoulder-girdle is not connected with the skull in the Batrachia, 

 excepting in the genus Hemisus. There is a large supra- scapula. The 

 osseous coracoid is of various proportions, and it has various cartilagi- 

 nous extensions, as epicoracoid and procoracoid. These are ossified in 

 some of the extinct forms. There is much variety in the pieces which 

 occupy the middle line of the scapular arch. The orders may be ar- 

 ranged as follows on this basis : 



An episternum and no sternum : Ganocepliala, Ehachitomi, Emiolomeri, Stegocephali. 

 An omosternum and sternum ; no episternum, Salientia. No median sternal elements : 

 Trachystoniata, Froteida, Urodela (except Trematodera) . 



The pelvis is always furnished with an ilium, but the pubis is want- 

 ing or represented by rudiments, except in the extinct forms, where it 

 is present. The ischium is primitively an undivided cartilaginous plate. 

 No obturator foramen. There are some characters which are common 

 fo all or nearly all Batrachia, but which may be found on further knowl- 

 edge of the extinct forms not to have been always present. One of 

 these is the continuity or fixed articulation of the quadrate cartilage 

 or bone with the skull. The proximal part of this bone is intercalated 

 between the squamosal and exoccipital, and the pterygoid when present, 

 so as to present only its distal extremity free. In the Salientia it is an 

 insignificant element, being generally cartilaginous. 



The vomeropalatine bones are always double, except where wanting, 

 which is only the case in the Trachystomata. They are nearly always 

 dentigerous. 



The orbitosphenoid bone is always well developed. 



In the existing orders the atlas is undivided. I have ijut forth the 

 hypothesis^ that the vertebral bodies in the existing and most of the 



> Cope, American Naturalist, 1888, p. 436. 



^Baur, Carpus u. Tarsus der Vertebrateu, 1887, Batrachia, pp. 6-12. 

 ^On the Ihtercentrum of the Terrestrial Vertebrata, Transac. Amer Philosoph. 

 See, 1886, p. 243. 



