THE MOST PRIMITIVE KNOWN REPTILE. 



295 



processes, the fenestra ovalis retaining its position quite at the 

 lower surface of the skull. 



The palate of Limnoscelis has advanced far beyond that of 

 Seymouria in the wide interpterygoid vacuity and in the 

 pronounced pterygo-transverse flanges. 



The lower jaw seems to have lost the post-splenial and pro- 

 bably the anterior coronoids. 



The vertebrae in retaining a quite long spine are probably more 

 primitive than those of Seymouria, for all the Temnospondyls 

 have high neural spines. 



The single-headed ribs are, however, an advanced feature. 



In the shoulder-girdle Limnoscelis is advanced in having a 

 posterior coracoid element, but retains in its cleithrum a primitive 

 bone lost by Seymouria. 



The pelvis of Limnoscelis much resembles that of Seymouria in 

 the general form of the ventral surface, which is very reminiscent 

 of Rachitomi io the short pubes. The posteriorly produced ilium 

 is also a primitive feature retained by both reptiles. 



The hind leg of Limnoscelis resembles that of Seymouria in 

 having a separate intermedium, but the conditions are different 

 in the non-ossification of the tibiale in the former animal. 



Limnoscelis is thus in most ways more advanced than Sey- 

 mouria, although it is still a very primitive reptile. 



Diadectidce. — The Diadectids, known from New Mexico, Texas, 

 and probably also Europe (Stephmospondylus), are obviously 

 much more advanced reptiles than Seymouria. They have a 

 remarkably specialised palate and occiput — which region I have 

 recently discussed in detail. The brain-case is theoretically 

 derivable from that of Seymouria, but is so extremely modified 

 that no useful end will be served by a further discussion. 



Pariasaurido3. — With the Middle and Upper Permian group of 

 the Pariasaurs, we pass to reptiles which have advanced much from 

 the primitive Seymouria structure in all features of skull and 

 skeleton. Some of these differences I have already pointed out, 

 others are at once obvious from any comparison. One important 

 feature which deserves emphasis is the high position of the 

 fenestra ovalis, which lies on the front of the paroccipital process 

 far removed from the basioccipital and the tuber basisphenoiclalis. 

 The opening is small and the stapes is a slender rod, without the 

 much expanded base-plate it must have had in Seymouria. 



Procolophonidce. — Procolophon and its allies are the most 

 advanced Cotylosaurs known. They differ extremely from Sey- 

 mouria in nearly all features. The differences which at present 

 most interest me are those in the brain-case and otic region. In 

 essentials these regions resemble the condition in Sphenodon, 

 and differ from that in Seymouria in ways I have already pointed 

 out, perhaps most markedly in the slender paroccipital process 

 with a grooved lower surface in which the stapes lies, and in the 

 high position of the fenestra ovalis. 



The Cotylosaurs so far considered all differ markedly from 



21* 



