CERATODUS. 25 



The known difference between A and B relates, then, to the presence or absence of 

 gular plates — a point sufficient to denote a family, but hardly of ordinal value. I am 

 inclined to hold that, on the whole, A and B are not ordinally separable, but that they 

 represent two families or suborders of Ganoids, which may continue to be called 

 Sirenoidei (Dipnoi) and Ctenododipterini. According to present knowledge we may 

 arrange such a series as this, extending between the Plagiostomi and the Teleostei : — 



Plagiostomi. 



Chimcsra. 



C Lepidosiren. Ceratodus. "^ 



Dipterus. Ctenodus. I 



Ganoidei.-J ~ .. |>Ganoidei. 



Orossopterygn. 



Lepidosteidae. Acipenseridae. Amiadae. J 

 Teleostei. 



This series is interrupted by many gaps, of which that between Chimara and Lepi- 

 dosiren is perhaps the most considerable. Although the dentition is capable of being 

 described in nearly the same terms in both genera, it differs greatly in details and 

 proportions ; and the absence of scales, the entirely cartilaginous skull, the absence of a 

 swim-bladder, and the claspers of the male, separate Chimara very decidedly from the 

 existing Dipnoi. 



The dental plates of Ceratodus are the sole means of discriminating the fossil species, 

 and were till lately the only source of information respecting the genus itself. There is 

 a palatal pair attached to the pterygo-palatine bones, and a similar pair in the 

 mandible, receiving support from the splenial, and directly or indirectly from the angular 

 bone. In the recent species the connection with the angular bone is by connective- 

 tissue only, and the bony attachments of all the dental plates, though extensive, are so 

 slight that they are readily detached without fracture. This was apparently the ease 

 with many of the Triassic species. Sometimes, however, we find a fossil tooth which 

 retains a part of its bony base. Such specimens are figured in PI. III. We owe to 

 these rare examples the means of certainly distinguishing the palatal from the 

 mandibular plates. In figs. 1 — 4 the splenial element can be identified, while on the 

 outer and upper side of the palatal teeth (fig. 5) a boss of bone, which passes up towards 

 the frontal region, is to be distinguished. 



The only constant difference between the palatal and mandibular dental plates 

 seems to lie in the number of denticles. In nearly all the recent and fossil examples 

 which can be certainly identified as belonging to either category there is one denticle 

 more in the upper than in the lower. 



The tips of the denticles are external, and the larger denticles are in front, so that 

 any tooth can be readily assigned its true place in the mouth. 



Do similarly situated teeth succeed one another vertically ? or were the young 



