MEMBRANE BONES OF SKULL 579 



development has taken place. As the brain-box expanded to con- 

 tain the enlarging brain, the squamosal and parietal spread on the 

 surface of the cranium and became reduced externally, so that 

 the parietal has come to lie completely, and the scjuamosal largely, 

 inside the jaw muscles. At the same time the postorbital and 

 postfrontal are lost, so that the orbit and the lateral temporal 

 fossa become confluent ; the latter, in fact, is no longer recognis- 

 able as an opening, for it is simply an empty area above the zygo- 

 matic arch, which is all that remains of the plate of bones originally 

 covering the jaw muscles. In monotremes there is an interesting 

 relic of the post-temporal fossa in the form of a narow canal 

 between the squamosal and the periotic (Fig. 452). In artiodactyls 

 and primates the lateral temporal 



fossa is secondarily separated off Postsquamosal arcade 



from the orbit by downgrowth ^^s-v;-. J • — ^^ / 



of the frontal which meets the /p|p^" \ yJ 



zygoma. The interparietal present Jr } >/x 



in some mammals, such as the ^j x 1 ^ -'/./ J 



dog, represents fused dermosupra- ^^^^^ "^^W /^^ 



occipitals. ) .^sS^^^^>^'5 



In the third main reptihan type, — - — x^-/ 



the diapsid (Fig. 449 D), there are Fig. 452.— SkuU of Ornithorhynchus. 



- 7^ - , , f posterior view, showinj^ the post- 



two lateral temporal tOSSSe, one squamosal arcade, probably repre- 



above and one below the post- senting the post-temporal fossa of 



orbital. The lower is in the same ^^^ ^ ^^' 

 position as that of the synapsids, while the upper pierces the bony 

 arcade of the cranium. The dinosaurs and pterodactyls had skulls 

 of this type, and it is found at the present day in Sphenodon and 

 in the Crocodiha (Fig. 453). The birds are derived from the diapsid 

 condition by a similar development to that seen in mammals. The 

 frontals, parietals and squamosals have spread over the cranium 

 inside the jaw muscles, and the loss of the postorbital makes 

 the orbit and the two lateral temporal fossa? all confluent. 



The fourth and last type, the parapsid or anomapsid. has only 

 a single lateral temporal vacuity, but this time it is above the 

 postorbital and so corresponds to the upper vacuity of diapsids 

 (Fig. 449 C). It is now generally agreed that the parapsids do not 

 make a natural group, the single upper temporal vacuity havmg 

 arisen more than once in widelv divergent forms. The Squamata 

 have a skull which is parapsid in form (Fig. 454). ^^"t they are 

 almost certainly derived from diapsid ancestors. Ihe extinct 



