SIWALIK AND NARBADA BUNODONT SUINA. 
5—39 
differ from tlie tjqDical s|)ecimen,s by the relative position of the molar series : thus m. 3 
is placed entirely behind the posterior border of the zygomatic root, and extends 
backwards as far as the line of the jiosterior border of the palate. The line of the 
cheek-teeth is straighter ; the true molars are of a more elongated form, their length 
exceeding their breadth ; ptn. 1 is placed closer to pm. 2 | while pm. 4 is of nearly the 
same size as pm. 3 instead of being considerably smaller. The following table 
exhibits the difference in the dimensions of the true molars of these specimens, and 
those of the two typical skulls figured in the “ F.A.S.”; vis. : — 
Type form. 
Aberrant form. 

Width 
Length 
Width 
Length 
Width 
of 
m. 
1 
1-35 
1-8 
1-78 
? ) 
? 5 
n 
1-75 
1-5 
1-62 
J J 
n 
2 
1-85 
1-65 
2-1 
2-14 
>) 
>> 
2'11 
2-08 
1-75 
1-85 
) > 
n 
3 
1-8 
1-9 
2-1 
J f 
1 > 
J 1 
2 - 0.5 
1-95 
There are several other crania in the British Museum {e.g. Nos. M. 491 ; 16,381 ; 
16,378) exhibiting precisely similar characters ; and it appears from these that the 
cingulum on the inner side of the true molars is less developed than in the type 
cranium. These crania are on the whole rather smaller than those of the type form. 
Intermediate crania . — Were it not for the two specimens to be now mentioned, it 
would appear that the crania last described belonged to a distinct species from H. 
sivalensis ; but the former exhibit such transitional characters that it appears 
impossible to accept this view. The two crania in question are in the British 
Museum (Nos. 16,379, 16,382): in the first m. 3 , although entirely behind the 
zygomatic root, is in advance of the posterior border of the palate, and pm. 4 is of 
the small size characteristic of the tyjDical skulls : while in the second m. 3 is partly 
behind the posterior border of the zygomatic root, and considerably in advance of 
the posterior border of the palate : in both the true molars are of the elongated type. 
It thus seems that the wide-toothed Siwalik hippopotamus is so closely connected 
in certain resj)ects by these two skulls with the narrow-toothed form, that, in the 
absence of other well-marked differences, it appears im})ossible to consider them as 
more than varieties, or races : it may be convenient to designate the former, or 
typical race as var. latidens, and the latter as var. angustidens. In provisionally 
regarding these two forms as races of one si^ecies it must, however, be borne in 
mind that no sufficient distinction can be drawn between any of the different species 
of the genus from the pattern of their molars, and that thus a character of much 
A^alue among other genera as a means of specific distinction is lost ; and it may, 
therefore, still be possible that there may be more than one si^ecies necessarily 
included in IT. sivalensis. The large number of species by which the other genera 
of large Ungulata were represented in the Siwaliks renders this supposition the more » 
probable. 
In the crania represented in plate LX. of the “ F.A.S.” the orbit is more 
prominent, and the post-orbital portion of the skull relatively shorter : these 
