EASTMAN: MYLOSTOMID DENTITION. 227 



The Mylostomid is properly regarded as the more primitive type of 

 dentition on account of its obvious agreement with the Ceratodont, a 

 fact previously noted by Newberry. And indeed, as he rightly observed, 

 their resemblance is such as " will strike any one who examines them, 

 and no closer analogy suggests itself in the whole range of ichthyic den- 

 tition." The combined evidence of dental, cranial, and most of the 

 skeletal characters (the only marked exception being that of dermal 

 armoring) furnishes wellnigh irresistible proof of the Dipnoan affinities 

 of Arthrodires. A close parallel exists between the dentition of Mylo- 

 stoma and Ceratodonts on the one hand, and Dinichthys and Protop- 

 terus on the other. The coincidence ceases to be remarkable when it is 

 understood that other facts as well point to a common origin for 

 Ceratodonts and Arthrodires. Very interesting also is Semon's ob- 

 servation that in the young of Neoceratodus the upper dental plates 

 are at first divided into two pairs, as in Mylostomids, these afterwards 

 combining into a single pair of palato-pterygoids. 1 



Adopting the view that the Arthrodiran type of dentition is strictly 

 homologous with the Dipnoan, it is desirable to employ uniform desig- 

 nations for the dental parts. The tooth usually called " premaxillary " 

 in the former group is therefore to be identified with the vomerine of 

 typical Dipnoans, and the one or two succeeding pairs of " rnaxillaries " 

 as the case may be, with the palato-pterygoid dental plates. The latter 

 may be supposed to have been supported by the palato-pterygoid carti- 

 lage in Mylostoma precisely in the same manner as in Ceratodonts and 

 Ctenodipterines. Their homologues in Dinichthys were no doubt situated 

 well within the interior of the headshield, as we have a right to expect, 

 at least, from analogy with Mylostoma. Some device is evidently 

 required, owing to their large size, to take up the strains due to impact 

 against the powerful lower dentition, and this we find actually provided 

 for by the massive ridges developed on the under side of the headshield, 

 whose position, direction, and inferred function suggest comparison with 

 similar ridges in Neoceratodus. 



Conclusions. — The consequences depending upon Dean's reconstruc- 

 tion of the Mylostomid dentition are as follows : The complete upper 

 dentition of Mylostoma consists of two pairs of tritoral plates only, one 

 of which is the functional, but not the morphological equivalent of the 



1 Consult in this connection the recently published conclusions of Prof. J. Gra- 

 ham Kerr on the genetic affinities of lower Gnathostomata, in his paper entitled 

 Embryology of certain of the lower Fishes, and its bearing upon Vertebrate Mor- 

 phology. Proc. Roy. Phys. Soc. Edinb. 1906, 16, pp. 191-215. 



