Decembeb 5, 1913] 



SCIENCE 



825 



transformed into veritable museums of tera- 

 tological specimens. 



Eaymond J. Pool 

 The IjNrvERsiTY op Nebraska, 

 October 10, 1913 



AN ANCESTRAL LIZARD FROM THE PERMIAN OF 

 TEXAS 



There has been no more vexed problem in 

 vertebrate paleontology than the origin of the 

 scaled reptiles. The theory generally ac- 

 cepted has been that the lizards arose from 

 the double-arched or rhynchocephalian type by 

 the loss of a primitive lower arch, a theory of 

 which I have been skeptical for many years 

 past. I have urged in various publications for 

 the past ten years that the lizard phylum is a 

 very ancient one, predicting that it would 

 eventually be discovered in the Permian, a 

 prediction that I am now able to verify. Three 

 years ago I described briefly a peculiar reptile 

 from the Lower Permian of Texas under the 

 name Armoscelis. It has only been recently 

 that the stress of other material has permitted 

 the full preparation of the several more or less 

 complete skeletons upon which the genus was 

 based, a study of which has disclosed more de- 

 cisively than in any other A m erican Permian 

 reptile the structure of both skull and skeleton. 

 Armoscelis was an extraordinarily slender, long 

 legged, cursorial and arboreal reptile of about 

 eighteen inches in length. The skull is re- 

 markably lizard-like in appearance and struc- 

 ture, with a typical upper temporal vacuity 

 bounded precisely as in the mosasaurs. The 

 sides of the skull below the arch, instead of 

 being open, as in the lizards, are covered over 

 by a broad expansion of the squamosal bone, 

 which is rather loosely united to the quad- 

 rate. The quadrate is supported, as in lizards, 

 by the tabulare and opisthotic; it is rather 

 free and is broadly visible from behind. The 

 lacrimal bone is small, as in lizards, a char- 

 acter hitherto unknown among ancient rep- 

 tiles; and the palate has rows of teeth on all 

 the different bones. The neck has seven or 

 eight more or less elongated vertebrae, the dor- 

 sal region twenty. The sacrum is almost 

 indistinguishable from that of lizards. The 



pectoral and pelvic girdles differ chiefly in 

 their old-fashioned characters. The tail was 

 slender and long. The feet have an elongated 

 calcaneum and a reduced astragalus, unlike 

 those of the known contemporary reptiles. 

 Finally the attachment of the ribs, one of the 

 most peculiar characters of the Squamata, is 

 by a dilated head, articulating with both arch 

 and centrum. 



To convert Arceoscelis into a modern lizard 

 would require the reduction of the squamosal 

 bone from below to a slender bone articulating 

 with the postorbital ; the closer fusion of the 

 postorbital with the postfrontal; the greater 

 freedom of the quadrate; the loss of the pos- 

 terior coracoid bone and a modernizing- of the 

 girdles, every one of which characters we are 

 quite sure must have been present in the an- 

 cestors of the Squamata. 



Arceoscelis can not be placed in any known 

 order of reptiles, unless it be admitted to the 

 Squamata. But, I do not think that the dif- 

 ferences from the Squamata will justify its 

 ordinal separation, if we are to classify or- 

 ganisms phylogenetically. I would rather 

 modify the definition of the order Squamata 

 to include the genus as the representative, 

 doubtless with Kadaliosaurus also, of a dis- 

 tinct suborder, the Arceascelidia. Several 

 years ago I recognized in another Permian 

 vertebrate a primitive salamander, bearing 

 about the same relations to the modern 

 TJrodela that Arceoscelis does to the modern 

 lizards. The urodelan character of Lysorophus 

 has now been generally accepted, and I be- 

 lieve that after I have published the full de- 

 tails of the structure of Arceoscelis I shall 

 find concurrence in its phylogenetic associa- 

 tion with the Squamata. 



I regret much to add that Dr. Broom's inex- 

 perience with the American Permian verte- 

 brates has led him into several errors in his 

 recent discussion of the afiinities of Arceo- 

 scelis, based upon the meager details which 

 have been published. Had he heeded Dr. 

 Case's warning I do not think he would have 

 so readily assumed that the skull and skeletal 

 bones which he described as Ophiodeirus really 

 belong together. They probably do not, for 



