162 



IOWA GEOLOGICAL SURVEY 



the Dipnoan type, and that veritable homologies exist between 

 their cranial roofing plates and those of the living Neoceratodus. 

 This position is maintained in the following discussion of the 

 group, as it seems to be most nearly in accord with the prepon- 

 derance of known facts. 



Fig. 23. 



Fig. 23. Mylostomavariabile Newt). Cleveland shale; Sheffield, Ohio. Complete tritoral 

 dentition arranged in natural position, but the elements of both pairs of palato-pterygoid 

 plates belonging to different individuals . The presence of a third pair of vomerine elements 

 in advance of the two here shown has not yet been established by positive evidence, x 1-1. 



Concerning the systematic arrangement of Arthrodires, it 

 need only be said that, owing to faulty preservation, we are 

 still too imperfectly acquainted with the details of structural 

 organization in different families to permit of more than a 

 provisional scheme for illustrating successive stages of ad- 

 vancement. Analogy with other, and especially higher, groups 

 leads one to expect the dentition to furnish not only reliable 

 clues as to relationship, but also a convenient and serviceable 

 basis of minor classification. Experience proves that the ex- 

 pectation is only partially justified. As between the two well- 

 marked types of dental structure, which find close parallels in 

 the existing Neoceratodus and Protopterus respectively, there 

 need be little hesitation in recognizing the former, or triturating 

 type, as the more primitive, it being one of the most constant, 

 persistent and distinctive features of the Dipnoan stem. Nothing 

 is more natural than to regard the trenchant or sectorial type 



