680 



PEOE. T. H. HUXLEY OB SEE V AXIOMS 



each premaxilla sends up a broad and strong process, which, united 

 with its fellow, ascends in front of the nasal septum to the nasal 

 bones, and thus the two nasal apertures become widely separated by 

 bone. Laterally, on the other hand, the premaxillae are united with 

 the maxillae by a short process, which has no ascending branch. 



In Hyperodapedon, as in Sphenodon, the median end of the oral 

 face of each premaxilla is elongated downwards into a recurved 

 process (Plate XXVI. figs. 2, 3, 5, r.p) ; and the two processes, 

 closely applied together by their inner surfaces, form the beak, or 

 rostrum, in which the upper jaw terminates. In Sphenodon, 

 these rostral processes are broad and flattened from before back- 

 wards. Their broad truncated ends descend but little beyond 

 the dentary edge of the mandible, when the mouth is shut, and 

 present two tooth-like eminences. Moreover, their inner sides, 

 closely applied through the greater part of their length, diverge 

 so as to leave a median notch below. In Hyperodapedon, the 

 rostral processes are thick and semicircular in section, the inner, 

 mutually applied, sides being flat and the outer faces convex. Each 

 tapers to its extremity, and the apices of the two are conjoined to 

 form the sharp conical extremity of the beak. The rostrum, as a 

 whole, is much more curved than in Sphenodon, and descends far 

 lower in front of the mandibular symphysis. There is no dentary 

 substance upon any part of the rostrum of Hyperodapedon ; and the 

 aspect of its external surface leads me to believe that it was coated 

 with a horny sheath. 



In Sphenodon, the anterior end of the mandible lies altogether 

 behind the broad premaxillary rostrum, when the mouth is shut. In 

 Hyperodapedon, on the contrary, the rostrum is received between 

 the rostral processes of the mandible. 



The lateral view of the skull (Plate XXYI. fig. 2) shows that 

 the orbits looked more upwards in Hyperodapedon than in Sphe- 

 nodon ; the lateral temporal fossae (l.t) are relatively much smaller 

 and have a very peculiar ear-like contour, while the bony boun- 

 daries of these parts are much stronger. In Hyperodapedon, I 

 can find nothing but bone in the region which should be occupied 

 by the posttemporal fossae. The contour of the alveolar edge of 

 the maxilla in Sphenodon is slightly convex downwards and 

 slightly concave inwards. In Hyperodapedon, the contour is very 

 strongly convex both downwards and inwards, and this feature, 

 combined with the great length and recurvature of the pre- 

 maxillary rostrum, confers a very peculiar aspect on the skull 

 (see figs. 1 & 3, p. 682). 



In the under view of the skull of Sphenodon (fig. 6, p. 683), the 

 maxillary and palatal regions present little difference from those of ordi- 

 nary Lizards. The palatal area slowly widens from before backwards, 

 and its maxiilar} T boundaries are slightly concave inwards. Punning 

 parallel with the posterior half of each of these is the dentigerous 

 ridge of the palatal bone, bearing a single series of some seven or 

 eight teeth, which, like those of the maxilla, are acutely conical. 

 Between these and the posterior maxillary teeth is the deep groove 



