152—4 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA. 
The ridges are relatively small, and are divided by a median longitudinal cleft, 
Avhich is, however, less marked in the third ridge. The valleys are partially blocked 
by accessory columns (a, a), and there are traces of cement in all ; the presence of 
the latter being a character of the molars of M. iierimends^ although in many 
specimens the cement has fallen off. 
If the present specimen be compared with the penultimate upper milk-molar’ of 
M. latidens represented in plate XXX.VII., fig. 4 of the first volume, it will be seen 
that it differs by its narrower form, its much higher and more completely divided 
ridges, and less open valleys. The tooth is quite different from the molars of M. 
sivalensis ; and as it agrees precisely in general characters with the upper true molar 
of M. perimensis figured in plate XVI., fig. 2, it may be referred with great 
probability to that species. The length of this tooth is 2'3 ; the width at the first 
ridge P2, at the third P4 ; and the height of the middle ridge 0'95 inch. 
In plate XVII., fig. 1, there is. represented a fragment of the right maxilla of a 
young tetralophodont mastodon, containing the second and third milk-molars, in a 
much-worn and somewhat battered condition. Allowing for its different condition 
of wear the third milk-molar in this jaw agrees well with the last specimen ; and this 
circumstance, taken with tlie place of origin of the present specimen, leads to the 
conclusion that it very probably belongs to the present species. From its worn and 
battered condition it is impossible to say whether the third milk-molar differed 
essentially from the homologous tooth of M. latidens represented in plate XXXVII., 
fig. 4 of the first volume ; but it seems probable that the ridges when unworn were 
higher, and resembled those of the specimen represented in plate XVII., figs. 3, 3a. 
The second premolar is too much worn to permit of any comparisons. The length 
of the third milk-molar is DS, and its greatest width D38 inches; these measure- 
ments agreeing well with the dimensions of the last specimen. 
Penidtimate lower premolar . — The broken tooth represented in plate XVII., fig. 4, 
is contained in the fragment of the right ramus of the mandible, in that imperfect 
and battered condition which is frequently so provoking to the palaeontologist. The 
figured tooth has had the anterior portion of its crown hannnered off, and is but 
partly protruded : it is succeeded posteriorly in the jaw by a larger and fully 
protruded tooth, of which unfortunately the crown is now entirely wanting ; the 
latter tooth being again succeeded by a fragment of a tooth still in alveolo. The 
broken tooth appears to have been the last milk-molar, and the figured tooth will 
accordingly be the j^enultimate premolar, as it evidently belongs to the second series. 
The latter has a smooth enamel, showing that the specimen does not belong to M. 
pandionis ; and it is accordingly thought probable that it may belong to the common 
Perim species ill. perimensis ; which from the evidence of the fine specimen figured in 
plate XL. of the first volume is known to have developed preniolars. The specimen 
under consideration shows the alveolus of a small mandibular tusk. 
1 Tormcd 2 in vol. I. 
