94—28 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA. 
plate LXXXII, figure 1 of the “ Eauua Antiqua Sivalensis,” under the name of 
E. sivalensis, appears to agree with the last mentioned specimen. 
Einally, taking the Narbada specimen as the type of the teeth of E. ncmiadicus, 
it seems highly probable that the specimen represented in plate XY, figure 4, belongs 
to the same species, and this being so, it is difficult to separate the specimen repre- 
sented in plate XIY, figure 3, which certainly does not belong to E. sivalensis , and 
if not referred to E. namadieus must belong to a new’ species. 
Tipper milk-molars. — If the permanent molars described above have left any 
doubt as to there being two species of fossil Indian true horses, this doubt is at once 
dispelled if the upper milk-molars of a horse from the Narbada represented in figure 2 
of plate XY are compared with the corresponding teeth from the Siwaliks represented 
in figure 1 of the same plate. These teeth are implanted in a fragment of a right maxil- 
la, collected by Mr. Hacket in the pleistocene deposits of the Narbada valley. At the 
right side of the figure is seen the small first milk-molar (mm. 1), immediately 
behind which a fragment of the second milk-molar (mm. 2), and the germ of the 
displacing second premolar can be seen ; the two next teeth (mm. 3, mm. 4) are the 
third and fourth milk-molars, as is proved by their state of wear, and by the germs 
of replacing premolars below them ; while the last tooth (m. 1) is the first true 
molar. 
By comparing these teeth with the upper milk-molars represented in figure 1 of 
the same plate, and referred to E. sivalensis f the essential difference between the two 
w r ill be at once apparent. The present milk-molars differ indeed from those of any 
species of true horse with which I am acquainted by the nearly square form of their 
grinding surfaces, and thereby approach the milk-molars of some of the hippotheres, 
like II. antilopinum. 
As no other species of horse, besides E. namadieus , is known from the Narbada 
these teeth are provisionally referred to that species : taken with the Siwalik milk- 
molars they unquestionably prove the existence of two species of fossil Indian 
horses. 
Comparisons. — Assuming that all the specimens described above belong to 
E. namadieus, we may predicate ' of that species that the upper molars are distin- 
guished by the relatively great length of the grinding surfaces of the anterior 
‘ pillars/ a character especially well-marked in the premolars, but that there is a 
certain amount of variability in this character. The species is further distinguished 
from all living horses by the square form of the crowns of the upper milk-molars ; 
and is also marked by the very general retention of the first milk-molar in the per- 
manent series, though, judging from the specimens figured in the “ Fauna Antiqua 
Sivalensis,” this character is not always noticeable ; when present the persistent 
first milk-molar is always of much larger size than in existing horses. 
In Equus caballus the grinding surfaces of the anterior ‘pillars’ of the 
upper true molars are about equal in size to those of E. namadieus. In the 
premolars they are, however, smaller. The enamel islands in the existing horse 
