226 bulletin: museum of comparative zoology. 



tritoral plates in Mylostoma might together be equivalent to the single 

 " maxillary " element, as it is commonly called in Dinichthys, and that 

 a third pair of upper dental plates representing the " premaxillaries " 

 were either actually or potentially present in Mylostoma. At all events 

 greater weight was placed upon assumed numerical correspondence of 

 dental plates in the two genera than upon morphological equivalence, 

 for it is impossible to recognize the least similitude in form between 

 the usual Arthrodiran " premaxillary " and either of the tabular crush- 

 ing plates of Mylostoma. Denying that Arthrodires belong to Pisces 

 proper, and that they have gill-arch jaws, he holds that their dental 

 apparatus is non-homologous with that of all other fishes. 



According to the present writer's interpretation, the two pairs of 

 Mylostomid palatal plates are together equivalent to the single pair 

 of " maxillary " elements in the Coccosteid type of dentition, this 

 latter being regarded merely as a modification of the Mylostomid, 

 adapted for cutting instead of crushing. That the upper dentition of 

 Mylostoma consists in all of three pairs of plates, the foremost of which 

 is the precise morphological equivalent of the so-called " premaxillary " 

 pair in Dinichthys, is to be inferred not only fx*om analogy with 

 Dinomylostoma, in which this number has been definitely proved to 

 obtain/ but from the remarkable constancy of form displayed by the 

 more forwardly placed element in all Arthrodiran genera where it is 

 known to occur. Unacquainted though we be with actual specimens, 

 the existence of vomerine teeth in Mylostoma, real or potential, is an 

 assured fact. 



It follows from the point of view just stated that the Mylostomid and 

 Dinichthyid types of dentition are reducible to a common plan, and this 

 plan is further seen to be identical with that found in Dipnoans. The 

 one element which by virtue of its function retains a constant form 

 among Coccosteans generally, and in at least one genus of Mylostomids, 

 is that commonly known as the " premaxillary," in reality the vomerine 

 tooth ; and the palatal plates (or more properly the palato-pterygoids) 

 of the more primitive Mylostomid type are seen to have become fused, 

 turned upright, and sharpened into a cutting edge along their functional 

 margin in the more specialized Dinichthyid type. 



1 Tlie recently published suggestion on the part of Dr. Hussakof (Mem. Amer. 

 Mus. Nat. Hist. 1906, 9, p. 119) that the type of Dinomjjlostoma beecheri includes 

 portions of more than one individual, all embedded in a single block of shale, 

 is now abandoned by that writer as the result of further study of the original 

 material. This statement is made here with Dr. Hussakof's consent. 



