Apeil 21, 1911] 



SCIENCE 



631 



■ PERMIAN REPTILES 



The relationships of the South African and 

 American Permian faunas have long been of 

 profound interest, from both the geological 

 and evolutionary standpoints. Students have 

 now generally come to an agreement in the 

 union of all the Permian and Triassic stego- 

 crotaphic reptiles (exclusive of the Chelonia) 

 under the single order Cotylosauria, but, until 

 recently, the zygocrotaphic terrestrial reptiles 

 have been distinguished under five or six or- 

 dinal names, though Broom is vfilling to unite 

 the African forms under the ordinal or 

 superordinal term Therapsida. 



Within the past few years, beginning with 

 the important collections made by Professor 

 Case, the University of Chicago has acquired 

 a very rich representation of the American 

 Permian reptiles and amphibians, some six or 

 seven of the thirty odd genera acquired repre- 

 sented by practically complete skeletons. 

 Among the latest of these collections, discov- 

 ered the past year by Mr. Paul Miller in the 

 vicinity of Seymour, Texas, is a very nearly 

 complete articulated skeleton of the remark- 

 able genus described a few years ago as Sey- 

 mouria by Broili, from two incomplete skulls. 

 Almost nothing of the remainder of the skele- 

 ton has hitherto been known. Within the 

 past year I described and figured a consid- 

 erable part of a skeleton of a very small rep- 

 tile, based upon two specimens of nearly iden- 

 tical size, under the name Desmospondylus, 

 suggesting at the time the possible identity 

 with either Seymouria or Pantylus. The 

 name proves to be a synonym of Seymouria. 

 Although the two specimens described are 

 scarcely a third of the size of the adult, and 

 both of the same size, they doubtless are 

 juvenile, or embryonic. I also suspect that' 

 Conodectes Cope is the same genus, or at 

 least is a closely allied genus; but the name 

 is unworthy of priority, since the type was 

 never really described or figured. 



Seymouria presents such extraordinary 

 characters, that, if we raise the Diadectidae, 

 Pariotichidae and Pantylidse to subordinal 

 rank, as would be justified from the charac- 



ters used to distinguish the South African 

 groups, we must also elevate the Seymouriidse 

 to the same rank. I am, however, opposed to 

 the erection of so many ordinal names; they 

 are in large part merely confessions of igno- 

 rance. The family Limnoscelidae, for in- 

 stance, recently described by me from the 

 Permian of New Mexico, shows certain inter- 

 mediate characters between the Diadectidae 

 and Pareiasauridse. For the present, it seems 

 to me that the following classification will 

 sufiice : Order Cotylosauria, families Diadec- 

 tidffi (Nothodontidfe Marsh), Limnoscelidas, 

 Pariotichidae, Seymouriidae, Pantylidae, Parei- 

 asauridae and Procolophonidae, the last two ex- 

 clusively African and European, the others 

 exclusively American. 



And I would go still further ; possibly some 

 will think too far. The possession of a very 

 perfect skull of Edaphosaurus, hitherto knovm 

 from imperfect material only, convinces me 

 that Broom is right in his acceptance of the 

 views previously held by Cope, but which for 

 some years have fallen into desuetude, that 

 the African and the American therocrotaphic 

 reptiles (that is, those with a lower temporal 

 vacuity only) are likewise related in the same 

 way and perhaps in the same degree. Broom 

 would still retain their ordinal distinction, but 

 I am disposed to go further and reunite them 

 under the name originally applied by Cope to 

 the Pelycosauria and Anomodontia — Thero- 

 mera or Theromorpha. The working out of 

 the genus Casea, recently described by me, has 

 disclosed many aberrant characters, separating 

 the genus more widely from the Pelycosauria 

 than are any two groups of Broom's Therap- 

 sida. But, I repeat, I am not willing to make 

 so many new orders; it serves no useful pur- 

 pose ; and both Edaphosaurus and Casea would 

 require ordinal distinction if we accept the 

 groups of Therapsida as orders. I therefore 

 propose the following classification of the 

 therocrotaphic reptiles (excluding the Therio- 

 dontia) : order Theromera; suborders Pelyco- 

 sauria, Poliosauridae, Edaphosauridae, Caseidae, 

 Araeoscelidffi(?), Therocephalia, Anomodontia, 

 Dinocephalia and Dromasauria, the first five 

 American, the others African. I leave the 



