TIOMPOIIAL ARCUKS OF THE REPTJLIA. 19 



allied Ctqytorhinus. But whether or not the reptiles have been 

 derived from a Labyrinthodont or even a SeymouriaAike ancestor 

 need not at present concern ns, as we have a temporal roof in 

 Pantylas primitive enough to have been the origin of all the 

 later types of temporal regions seen in higher reptiles. Those 

 who believe that in many later reptiles, such as IcJuhyosaitrus, 

 rieurosauriis, and others, we still have a well-developed supia- 

 temporal element, must necessarily derive them from an ancestral 

 type such as we have in Fantylv.s, where the supi'atemporal is 

 still preserved. Personall}^, I am of opinion that all the so-called 

 supratemporal bones in later reptiles are really tabulars, that the 

 supratemporal was very early lost never to reappear, and that a 

 temporal region such as we have in Captorkimis gives us a better 

 idea of the ancestral reptilian temporal roof than any other 

 well-known animal. 



It is, of course, a point which might be debated at considei^able 

 length whether it is the tabular which is lost and the supia- 

 temporal retained in the primitive Diapsids or the supratemporal 

 lost. The evidence is not as convincing as one would desire; 

 but, as I am showing elsewhere, there seems good reason to believe 

 tha,t the upper bone in the temporal region of the lizard is the 

 tabular, and in the most lizard-like of the Cotylosams such as 

 Procoloj^hon it is certainly the tabular that is retained and the 

 supratemporal lost, so we are pi-obably justified in concluding that 

 except in a few primitive Ootylosaurs such as Pantylus and 

 Diadectes a supratemporal is never found in reptilian skulls. 



Gaptorhinus, though it has all the ancestral elements of the 

 temporal region and onl}^ those that are met with in most later 

 reptiles, is considerably specialised. The squamosal is probably 

 considerably larger than in the more genei^alised ancestor, and 

 the tabular is certainly more reduced than it must have been. 

 Still the type as seen in Captorhinus is our best starting-point for 

 tracing the evolution of the temporal region. 



In all the Maiiimal-like reptiles there seems to me no reason- 

 able doubt that the temporal condition has arisen by an opening- 

 forming between the postorbital and the squamosal, exactly as 

 figured bj'^ Versluys in his " tSynapsider Typus I." *. This leaves, 

 when the opening is well formed, a temporal fossa bounded above 

 by the postorbital and • squamosal, and below by the jugal and 

 squamosal. In Dtametrodon and Gorgonops and Gcdepus we find 

 this type of fossa, and there is no doubt that, even where as in 

 Cynodonts and some others, the parietal forms part of the temporal 

 border, the condition is a secondary one. 



The Mammal-like reptiles with this single lower temporal fossa 

 form a very natural grovip, to which the name Synapsida has been 

 given by Osborn. 



There is another group of reptiles which, having also a single 



* Versluys, I. " Uebev den Phylogeiiie der Schlafenvuben unci Joclibogen bei den 

 Reptilien," Sitzungsber. der Heidelberger Akad. der Wisseiischaften. Heidelberg, 

 1919. 



