THE BLACK OR SHIP RAT 



591 



of a parietal, measured along a crest, is considerably less than, instead 

 of about equal to, the greatest distance between the crests. The 

 parietals are vaulted noticeably above the general dorsal profile, a 

 feature emphasised by a slight flattening of the frontal and inter- 

 parietal regions. The large inter-parietal extends relatively far back- 

 wards, and its boldly convex posterior border produces a characteristic 

 backward deflection (from the plane of the mastoid processes) of the 

 central part of the lambdoidal crest ; as a consequence the occiput is 

 vertical, the condyles are completely hidden in the dorsal view, and 

 the occipito-nasal length amounts to from 105 to 106-5, instead of 102 

 to 104-5 per cent, of the condylo-basal length. The rostrum is relatively 

 slender; its width amounting to from 17-6 to 20-3, instead of 19-5 

 to 22-3 per cent, of the condylo-basal length ; the blade-like outgrowths 

 of the premaxillae beneath the nasals are rather large ; the zygomatic 

 arches as a whole are more slender ; and the palatal length and the 

 diastemata (in E. r. rattus, alexandrinus and frugivorus) are relatively a 

 little shorter than in the Brown Rat. The auditory bullae are relatively 

 a little larger and more inflated than in norvegicus, and they com- 

 press to some extent the anterior part of the basi-occipital, the width 

 of the latter bone in front being rather less instead of slightly more 

 than half its median length. 



As in all other mammals, growth produces many changes in the 

 proportions of the skull, the facial region being the chief seat of post- 

 natal growth. The following measurements from Hossack illustrate 

 the process in the present species : — 



Apart from its smaller size the mandible does not diff'er importantly 

 from that of norvegicus. 



The cheek-teeth (Plate XXVI 1 1., Fig. 8) have the outer cusps slightly 

 less reduced, and in the upper teeth, the median tubercles slightly less 

 developed than in norvegicus. In wi}- cusp i is distinct and usually 

 nearly as large as ;tr ; as a rule there is no trace at all of an anterior 

 cingulum. In m^ cusp 5 is still evident although partially fused with 

 z ; in one case Forsyth Major saw this cusp placed so far forwards that 

 It blocked up the entrance to the transverse sulcus. In w^ a minute 

 trace of cusp i may be present occasionally, but 5 is completely lost. 

 In m^ cusp 6 is apparently a constant feature, but n is always more 



