OSTEOLOGY OF AMERICAN PERMIAN VERTEBRATES 395 



it was on this assumption and the equally unjustified assumption 

 of a superior temporal vacuity in Paleohatteria, that these two 

 genera and others even more indeterminate have been almost 

 unreservedly classed either among the Rhynchocephaha or as a 

 group of the Diaptosauria, of which the true Rhynchocephaha are 

 another member. As I have elsewhere said, Sphenodon and the 

 Rhynchocephaha have been, in the past, a cloak which has covered 

 a multitude of taxonomic and phylogenetic sins. We have tried 

 to trace all reptiles back to a primitive double-arched condition, 

 on the assumption that Sphenodon is a very primitive reptile. It 

 now seems more than probable that a single perforation of the 

 temporal roof was the more primitive condition among reptiles, 

 and that the Squamata are, in this respect, more primitive than the 

 Rhynchocephaha. It has been assumed, indeed, that the Squamata 

 themselves were derived from a primitive double-arched condition 

 by the loss of the lower arcade, an assumption that in itself would 

 seem to be improbable. Is it reasonable to suppose that the 

 quadrate lost its fixity and became streptostylic suddenly, which 

 would necessarily have been the case by the final loss of the 

 quadratojugal arch ? It would be far more reasonable to suppose 

 that the gradual loss of a large squamosal, such as is found in 

 Araeoscelis, permitted the gradual acquisition of strep tostyly. 



Seeley also figures a small antorbital vacuity in Protorosaurus, 

 but with doubt; there is certainly none such in Araeoscelis. The 

 lacrimal, according to Seeley, is small in Protorosaurus. So far, 

 then, as I can see, there is nothing known in the skull of Protoro- 

 saurus that would prevent the immediate association of the genus 

 with Araeoscelis. 



In the skeletal structure Protorosaurus shows some remarkable 

 resemblances to Araeoscelis: in the elongated cervical vertebrae, 

 in the attachment of cervical and dorsal ribs, and in the general 

 hollowness of the bones of the skeleton. These characters, in 

 connection with the probable, or at least possible, resemblances 

 of the skull, are, I believe, more than coincidences; they are 

 genetic, not homoplastic. 



Unfortunately httle is known of the structure of the girdles in 

 Protorosaurus. The vertebrae show distinct differences in the 



