448 PROF. W. H. FLOWER ON THE CRANIAL AND [May 16, 



In the former, besides the greater general size and greater length 

 and height of the cranium as compared with breadth, the ascending 

 ramus of the mandible is considerably higher in proportion (a cha- 

 racter much relied on by Blyth),sothat the whole skull, when mounted 

 upon the lower jaw, is more elevated. The occipital surface also is 

 markedly higher and narrower than in R. sondaicus. But in ad- 

 dition to these well-known and obvious characters, there are certain 

 features in the conformation of the base of the skull which are 

 eminently characteristic of the two species, and which when once 

 recognized cannot be mistaken, and are superior for diagnostic 

 purposes to those derived from the general form and proportions, 

 or from parts of the skull the form and dimensions of which are in- 

 fluenced by muscular development, the size of the horns, &c, and 

 consequently very liable to individual variation. 



In R. unicornis (fig. 3) the mesopterypoid fossa is always narrower 

 than in R. sondaicus (fig. 4) ; and the same condition extends back- 

 wards throughout the basi-sphenoid and basi-occipital bones, not 

 onlv relatively to the size of the skull, but absolutely, the point of 

 junction between these two bones being, in large skulls of R. unicor- 

 nis, actually narrower from side to side than in much smaller speci- 

 mens of R. sondaicus, though generally making a more salient pro- 

 jection downwards. Furthermore, the free ends of the pterygoid 

 processes (pt) are compressed and deeply grooved in R. unicornis, 

 whereas in R. sondaicus they are more flattened and laterally ex- 

 panded. The hinder margin of the palate is more regularly concave 

 in the former, and has a projection in the middle line in the latter. 

 But the most absolutely diagnostic structural difference is seen in the 

 hinder end of the vomer (vo), which in R. unicornis is thickened 

 and firmly united by its sides to the base of the pterygoid processes, 

 while in R. sondaicus it is thin, lamelliform, pointed, and free, so 

 that in museum specimens it is very often injured or destroyed. 



The upper molar teeth of R. unicornis and R. sondaicus are 

 remarkably unlike for species otherwise so nearly related*; but the 

 same kind of difference exists between the two best-distinguished 

 species of the African forms, R. simus and R. bicornis ; so that the 

 characters of the teeth alone, which have been so much relied on in 

 the case of the extinct species, are not, when taken by themselves, 

 good tests of affinity. 



In R. unicornis, in the first and generally in the second molar, the 

 crochet (or posterior combing-plate) (c' J ) curves forwards and usually 

 unites with the crista (anterior combing-plate) (c 1 ) developed from the 

 lamina, so as to cut off an " accessory valley " (a) from the extremity 

 of the median sinus f. The premolars and milk-molars present a 



* Professor Owen says truly in his ' Odontography,' p. 594 (1845) : — " Even 

 in existing species so nearly allied as the unicorn Rhinoceroses of India and 

 Java, each might be determined by a single detached molar tootb." But bis 

 views must have been subsequently modified ; for in the descriptive catalogue of 

 t lie Museum of the College (1858), the skulls of both Bpeciea are described 

 under the common name of E. mdicus. 



I- For an explanation of these terms see Busk, P. Z. S. 1869, p. 4.10. "Notice 



