﻿INTRODUCTION. XXI 



Group of the United States are in favour of the supposition that 

 they are truly autostylic. As originally pointed out by Newberry x , 

 the dentition of Dinichthys is most nearly paralleled by the existing 

 Dipnoan Protqpterus. The recently discovered triturating plates of 

 Mylostoma would have been assigned to the Dipnoi or Chimaeroidei, 

 if they had not fitted certain associated mandibular bones identical 

 in shape with those of the Dinichthys-tyipe : on one page, indeed, 

 Newberry terms the fish a " Placoderm " 2 , while on another it is a 

 " Dipterine Ganoid " 3 . The bones of the cranial shield, while 

 apparently homologous throughout the group, cannot be described 

 by the terms that are applicable to all Teleostomi, except perhaps 

 the modern Acipenseroids ; but these bones are symmetrically 

 disposed with respect to the median longitudinal line, and are thus 

 worthy of a nomenclature. In short, the evidence in favour of the 

 autostylic character of the Coccostean fishes has now accumulated 

 to such an extent, that we venture to regard them as an order of 

 Dipnoi, bearing the same relation to the Sirenoidei that the Acan- 

 thodians seem to hold with respect to the primitive Elasmobranchs 

 (Tchthyotomi), or the Actinopterygians with respect to the primitive 

 Teleostomes (Crossopterygii). For this order the name Arthrodira 

 is suggested, in allusion to the ginglymoid articulation by which the 

 cranial shield is hinged upon the anterior border of the armour of 

 the abdominal region in the typical and best known genera. 



TELEOSTOMI. 



It is generally admitted that the Crossopterygian Teleostomi are 

 closely related to the Dipnoi, and the Devonian representatives of 

 this order tend in some degree to lessen the hiatus between the 

 two great subclasses. Since, however, all the early Crossopterygii 

 hitherto discovered conform to the normal Teleostome type in the 

 arrangement of the bones of the cranial shield, it seems probable 

 that the two groups had already diverged before the development 

 of membrane-bones commenced. 



The most interesting feature of the Crossopterygii consists in the 

 mode of specialization of their fins ; and this, as pointed out by 

 Cope, affords a satisfactory basis for the definition of the suborders. 

 In all the known Palaeozoic and Mesozoic members of the order the 



1 J. S. Newberry, Eep. Geol. Surv. Ohio, vol. ii. pt. ii. (1875), p. 6. 



2 J. S. Newberry, ' Pakeoz. Fishes N. America ' (1889), p. 122. 



3 Ibid. p. 161. 



