196 ]\Iessrs. Hancock & A'tthey, on Diptems . ojid. Ctenodus, 



entire absence, then, in Ceratodus of any such bony support 

 would seem to indicate that none had ever existed. 



So long as Ctenodus and Dipterus were represented merely 

 by the dental plates, they were placed ^\\i\\ the Placoids ; and 

 no one would have been justitied in placing them elsewhere. 

 Therefore, until some further information is obtained respect- 

 ing Ceratodus^ it would seem best to allow it to remain as at 

 present located, along with the Selachians. The form of its 

 dental plates, too, is sufficiently characteristic to justify its 

 separation generically from this interesting Australian animal 

 as well as from Dij)teriis and Ctenodus. Certainly in the 

 dental organs all three approximate to the Selachians ; but 

 the Ganoid characters so predominate that we apprehend no 

 naturalist would hesitate to place them in that order, though 

 they may be considered to a certain extent " synthetic " forms, 

 as suggested by M. Agassiz*. 



In the present state of knowledge respecting Ceratodus^ it 

 is, then, evidently hazardous to place Mr. Krefft's fish in that 

 genus ; but its affinity with Dipterus and Ctenodus^ more 

 especially witli the latter, is clear enough. All three are 

 covered with large cycloidal scales ; the fins are arranged 

 much in the same manner ; the skeletons are nearly in the 

 like state of partial ossification ; the dental plates are much 

 alike, there being four ridged plates — two palatal, two man- 

 dibular. And when the so-called Ceratodus Forsteri has been 

 fully examined, there can be little doubt that this affinity will 

 become only the more evident. 



Nevertheless the relationship is perhaps closest with Cte- 

 nodus. Like most of the members of that genus, the Aus- 

 tralian fish is large, measuring from three to six feet in 

 length. And it is only necessary to look to the mandibles 

 (PI. XIII. figs. 1 & 2) of the two forms to be convinced how 

 close this relationship is. In Ctenodus the ramus (fig. 2) is a 

 stoutish bone, flattened vertically, with the upper margin 

 turned over towards the external surface, to give support to 

 the large dental plate ; it is therefore channelled on the outer 

 surface, and somewhat convex on the inner. • The posterior 

 extremity projects backwards beyond the dental plate a little 

 more than half the length of the latter ; and is for the greater 

 part occupied by the glenoid surface, which extends from the 

 upper margin, and is a deep, Avide, circular notch, inclining 

 backwards and downwards. In front the symphysial surface 

 is straight, extending the whole depth of the ramus, and is 

 gi'ooved transversely. The dental plate is about two-thirds 

 the entire length of the ramus, and is placed nearer the sym- 

 , _ * 'Nature,' No. fi], vol. iii. p. IGO fl«70). 



