14 Dr. K. H. Traqiiair on. the Oenera 



sphenoideum " (parasphenoid) to have been inserted. Finally, 

 though he compares the teeth to those of Lepidosireyi, he does 

 not seem to consider Holodus as a Ctenododipterine, saying that 

 " the remarkable structure of the head, so strikingly different 

 from every thing hitherto known from the older formations, 

 points to a new family, whose dermal coverings, as in the 

 Ctenodipterini, consisted of a bony substance outwardly pro- 

 tected by a kosmine layer." 



I have already mentioned that it is clear that Paloedcvphus 

 insignis, v. Ben. and De Kon., and Holodus Kiprijanoxoi^ 

 Pander, are closely allied, and represent corresponding parts 

 of the head in the animals to which they respectively be- 

 longed ; the strongly Ctenododipterine aspect of the dental 

 plates in both cannot fail to strike every observer. But while 

 there is nothing to prevent PafefZa/)//?**' devonieyisis from being 

 the palatal tooth-plate of a fish allied to Dijpterus or Ctenodus, 

 the aspect of P. insignis^ as of i/. Kiinijanowi^ is certainly in 

 many points very unlike that of the snout or " upper jaw " of 

 a fish of that family. So Dr. Giinther, in considering what 

 other genera of fossil fishes should accompany Dijpterus to its 

 new position among the Dipnoi, remarks, " At first I thought 

 that HoIoduSy Pander, was another Dipnoous genus ; but I 

 changed this opinion after having compared it with Palceda- 

 fhus of Van Beneden and De Koninck. These two genera 

 are evidently closely allied ; and the position of their nostrils 

 (so far as we can judge from the fragmentary remains) appears 

 to have been different from that of the Dipnoi : these openings 

 were more lateral and outside of the mouth. It seems also 

 that there would not have been room for a pair of vomerine 

 teeth, at least not in Palcedajplms''^'^ . 



Prof. J. S. Newberry, however, maintains the Ctenodo- 

 dipterine nature of Palcedaplius devoniensis, but at the same 

 time considers it to be both generically and ordinally distinct 

 from P. insignis. For the former he proposes the generic 

 xidiuiQHeUodus^ to include also a new species H. Lesley t^ Newb., 

 from the Devonian of North America, retaining the genus 

 Palcedaphus for P. iiisignis, with whose original describers he 

 agrees in considering it to be probably the head of a large 

 Plagiostome f- 



That two forms which in the aspect of their dental plates 

 bear so evident a resemblance to each other and to the Cteno- 

 dodipterini should be so widely separated as Prof. Newberry 

 supposes, does seem a priori a little improbable. A step to- 

 wards the solution of the problem may, however, be attained 

 by simply asking ourselves the following questions : — Are we 



* Op. cit. p. 557. 



t raliTeontology of Ohio, vol. ii. (Columbus, 1875), pp. 62, 03. 



