60 PALEONTOLOGY. 



Considering the more typical representatives of the family, the 

 earliest to appear is Chitonodus, which began during the Lower Bur- 

 lington epoch and is last met with in the St. Louis. It is clearly 

 the forerunner of Paicilodus and Cochliodus, which appeared at later 

 epochs, and which are modifications based upon the same plan, 

 and which might readily be supposed to possess features that 

 entitle them to at least sub-family rank in contradistinction to the 

 before-mentioned genera. While Cochliodus was apparently of com- 

 paratively brief duration, since its remains have thus far only been 

 met with in the St. Louis and Chester formations, Paicilodus, the 

 earlier of the two to make its appearance, dates from the Warsaw, 

 the last representative being derived from the upper Coal Measures. 

 Throughout this vast geological range its species manifest the least 

 tendency toward differentiation, on the contrary maintaining the 

 distinctive characteristics of the genus with extraordinary fidelity to 

 the type. The same may be said of Deltodus, whose inception is 

 contemporary with Chitonodus, with which its relations are, perhaps, 

 the most intimate. This is especially pronounced in the case of the 

 homologous terminal forms of the mandible of either genus, Delto- 

 dus maintaining a simple convex anterior lobe in lieu of the 

 median coronal prominence of Chitonodus, which is a mere modifi- 

 cation of the former, and which reached its extreme differentiation 

 in Cochliodus; but in the supposed terminal form of the upper jaw 

 we are again presented with a marked departure from the Coch- 

 liodus type, the significance of which is sought in the previously 

 mentioned genus Sandalodus, while a less intimate relationship with 

 Deltoptychius and Orthopleurodus is noted, but sufficient to arrest 

 attention with suggestions of their common origin. Deltodopsis 

 evidently was a later modification of Deltodus, and, although it is a 

 well defined group, it is not clear whether it should be deemed of 

 greater rank than subgeneric. 



The foregoing observations are rather- what the comparatively 

 meagre facts seem to suggest, than actual demonstrations of a com- 

 plete system of derivation and generic relations. It is therefore with 

 no small degree of trepidation we venture to exhibit these supposed 

 relationships in diagramatic form ; but it will at least subserve the 

 purposes of illustrating in a vivid manner the stratigraphical range 

 of the various generic groups alluded to. 



