400 S. W. WILLISTON 



crotaphous reptiles, provided these bones are not both identical 

 with those I call the tabulare and squamosal in the lacertilians 

 and Araeoscelis. If they are identical, then the two phyla may have 

 originated together. 



The problem remains. What shall be done with Araeoscelis? 

 To throw it into a common receptacle with Dimetrodon, Edapho- 

 saurus, and Casea may be an easy way to dispose of the genus, but 

 it is hardly scientific. Some classification, even the most conserva- 

 tive, it must have; and while I am vigorously opposed to the still 

 more reprehensible practice of giving every form that is hard 

 to understand a new group name, it is evident that to place 

 Araeoscelis with the Pelycosauria is to increase the confusion. 



I have urged that the resemblances of Araeoscelis to the 

 Squamata would justify its inclusion in that order as a suborder, 

 under the name AraeosceHdia, co-ordinate with the Lacertilia and 

 the Ophidia. And I beHeve that will be its final disposition under 

 some subordinal designation. But it seems to me that the rela- 

 tions with Protorosaurus and Kadaliosaurus are too definite, too 

 pronounced, to warrant their dissociation. I would therefore 

 propose to unite these three genera, together with, provisionally, 

 Haptodus and Callihrachion, under the order Protorosauria of 

 Seeley, and place the order immediately before the Squamata in 

 any serial classification of reptiles. 



RESTORATION 



The restoration of the skeleton shown in Fig. 7 is based upon 

 the material described in the foregoing pages. I have omitted the 

 tail because, while there are a sufficient number of caudal vertebrae 

 preserved to show conclusively that it was long and slender, there 

 are not enough preserved in connection to determine its length. 

 I have little doubt that the number of vertebrae exceeded sixty. 

 The precise lengths of the digits are also in a measure conjectural, 

 since in no specimen is a digit preserved complete. Sufficient 

 phalanges, however, have been recovered to enable one to recon- 

 struct the fingers and toes with little chance of error, following the 

 arrangement determined in other contemporary reptiles. The 

 first digit of the hand is purely conjectural. It must have been 



