74 Prof. Owen on the Reptilian Fossils of South Africa. 



racters can I detect any trace in the specimen before me. I am disposed therefore 

 to conclude, that the resemblance to the Chelonian type of cranium, so strikingly 

 exemplified in the edentulous lower jaw and intermaxillary border, may have been 

 extended to the composition of the bony palate. 



Returning to the upper surface of the portion of skull under consideration 

 (PI. V. fig. 2.\ we find the region between the upper border of the ascending process 

 of the intermaxillary and the middle of the inter-orbital space formed by a broad 

 rhomboidal tract, including the nasal, 1 .5, pre-frontal, 14, and part of the mid-frontal, 

 12, bones : the median suture dividing the latter is discernible. The rhomboidal tract 

 has a more unequal surface than the corresponding narrower part of the skull of 

 the Dicynodon lacerticeps : the prominences developed from the pre-frontals at the 

 anterior part of the orbits are more marked, and they are divided by depressions 

 from an intermediate prominence at the middle of the tract. The inter-orbital tract 

 is traversed by two longitudinal shallow channels, divided by a median gentle lon- 

 gitudinal rising, and bounded laterally by the raised supra-orbital ridges. 



The back part of the cranium has been broken away from the present skull 

 through the parietal foramen, and just behind the commencement of the smooth 

 depressions that mark the origins of the strong temporal muscles {t, t). The 

 thickness of the parietal bone at this part is nearly half an inch ; such strength 

 of the cranial bones is both a Chelonian and Crocodilian character. The orbits, 

 0, have a smaller vertical diameter than in the Dicynodon lacerticeps ; their lower 

 boundary is thick and rounded off; a gradually widening and nearly flat floor 

 of the orbit extends obhquely inwards, downwards and backwards ; the lower 

 boundary of the orbit extends straight, outwards and backwards, for two inches ; 

 forming, at the fractured end, the beginning of the zygomatic arch, but showing 

 no tendency to curve upwards here, to form the posterior boundary of the orbit, 

 nor any mark of union with such a boundary: if this boundary was completed, as 

 is probable from the analogy of the preceding specimen, posterior to the fractured 

 part of the zygomatic arch, the antero-posterior diameter of the orbit must have 

 been both absolutely and relatively greater than in the Dicynodon lacerticeps. The 

 Crocodilia and Ophidia vera have no bony floor to the orbit : most Lacertia and 

 Chelonia have that structure, but it is much closer to the palate than in the Dicy- 

 nodon. The expansion of the malo-maxillary bone both vertically and transversely, 

 to form the socket for the tusk, gives a peculiar form and development to the 

 under part of the orbit of the bidental Reptiles of the African sandstones, which 

 will be readily appreciated by comparison of the present specimen or the figures 

 (PI, IV. & v.), with the skull of any existing Reptile ; it will be seen, however, that it 

 is merely an adaptive modification, in subordination to the support of the tusks, 

 which gives so mammalian a character to this part of the skull of the Dicynodon. 



