DIET ADOPTATIONS IN VARANUS NILOTICUS. 71 



it projects a good deal above and beyond the anterior end 

 of the supraoccipital while in V. salvator the suture between 

 the parietal and the supraoccipital is perfectly visible from 

 above. In V. niloticus the muscles have encroached upon 

 the parietal in its posterior upper part more and more so 

 that the muscles (the innermost portion of temporalis) from 

 both sides even seem to meet with their areas of insertion 

 in the middle line along a space of about 12 mm. The plain 

 upper surface of the parietal has thus the shape of a short 

 and broad transversal triangle, the apex of which is continued 

 in a ridge, with sloping sides, measuring 12 mm. in length, 

 to the projecting posterior edge of the parietal. But in 

 V. salvator the muscles do not require so much space for 

 their attachment to the parietal so that the plain upper 

 surface of this bone is continued all the way to its own, not 

 at all projecting posterior margin and the muscnlar areas of 

 either side are thus separated by a median plain surface of 

 a minimum breadth of more than 9 mm. In the quite young 

 V. niloticus the muscles have as yet encroached very little 

 on the parietal so that its upper surface has almost the same 

 width behind as in front. Through this comparison the 

 phylogenetic development of the parietal of V. niloticus be- 

 comes quite clear. 



The wings formed by the posterior parietal processes 

 and the squmosum (sitpratemporale) have in V. salvator a 

 width at the middle of 9 mm., in V. niloticus of almost 14 

 mm. Quadratum has, in the latter, a length of 33 mm. in 

 the former only 24 mm.,* the basal transverse thickness 

 of the same bone is resp. 16 and 11 mm., and the antero- 

 posterior diameter of the upper end is resp. 23 and 14 mm. 

 This may be enough to indicate the difference, in stoutness, 

 of the posterior region of the skull of these two lizards. 



The only portion which is stronger in V. salvator is the 

 basilar occipital expansion formed by basioccipitale and occi- 

 pitalia lateralia which is not quite 1 mm. broader than in 

 V. niloticus, and at the same time has a more uneven surface 

 with crests and deep grooves for the insertion of muscles 

 and tendons. This stånds no doubt in connection with the 

 habits of this lizard of preying on more vigorons and lively 



* These measurements agree elosely with those expressing the length of 

 cohtmella? cranii of corresponding skulls. 



