Full text of "Indian Tertiary and post Tertiary Vertebrata"

See other formats

'ow-' '4

''4'/

i'^

^ im-

sa;

h ' ■ ■•f !' ' .

:■> ".""O''

■d

.« >; .,

' ty '

v>\'^ .' '■

'IV.UZ

Jii;

:.y

.'(yy,

T. -tAr.

MEMOIRS

OF THE

GEOLOGICAL SURVEY OF INDIA.

BEING

FIGURES AND DESCRIPTIONS OF THE ORGANIC REMAINS PROCURED DURING
THE PROGRESS OF THE GEOLOGICAL SURVEY OF INDIA.

PUBLISHED BY OKDEP. OF HIS EXCELLENCY THE GOVERNOR GENERAL OF INDIA IN COUNCIL.

Ser. X.

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

Pt. i, 1874 —RHINOCEROS DECCANENSIS.-By R. B. FOOTE, F.G.S., Geological Sarveg
of India. *

Pt. ii, 1876.— MOLAR TEETH AND OTHER REMAINS OP MAMMALIA;

Pt. iii, 1878.-CRANIA OF RUMINANTS;

Pt. iv, 1880.— SUPPLEMENT TO CRANIA OF RUMINANTS ;

Pt. V, 1880.— SIW ALIK AND NARBADA PROBOSCIDIA.— By R. LYDEKKER, B.A.,

Geological Survey of India.

CALCUTTA:

OFFICE OF SUPERINTENDENT OF GOVERNMENT PRINTING;
GEOLOGICAL SURVEY OFFICE, AND BY ALL BOOKSELLERS;
LONDON: TRUBNER & CO.

MDCCCLXXX.

FEINTED AX THE OFFICE OF SCPEEINTENHBIfT OF GOVBBNMEKT PKINIING, HASTINGS STBEBT, CALCUTTA.

TABLE OF CONTENTS,

Page.

Peeface . . . . . . . . - . • • . vii

List of Plates . . . ........ xxi

List of Abbreviations used, and works and memoirs quoted ...... xxiii

‘ Corrigenda and Addenda xxix

Part I. Ehinoceros deccanensis, by E. B. Foote, F. G. S. . . . . • • 1

„ II. Molar teeth and other remains of Mammalia, by E. Lydekker, B.A. ... 19

Introductory remarks ........... 19

Rhinoceros, remarks on .......... 21

„ 'palceindicus 22

„ sivalensis 26

„ platyrhinus '29

„ iiamadicus (= indicus) ........ 32

„ iravadicus .......... 36

„ 'planidens Acerotlierium ■perimense) ....... 41

„ sp. var. . . * . . . . . . . • .43

Acerotherium ' . . . . ... . . . . .51

„ perimense .......... 51

Vishnutherium iravadicum ... . . . . . . . .55

60
61
62
62
64

65
67

„ pentapotamice . . . . . . . .70

Dinotheriiim . . . . . , . . . . .72

„ pentapotamice . . ... . . . . . .

Sunitherium , . . . . . . . . . . .76

„ schlagintweitii . . . . . . - . .76

List of fossil Indian Saina . . . . . . . . .78

Tetraconodon . . . . . . . ... .78

„ magnum .......... 79

Manis . . . . . . . . . .82

„ sindiemis . . . . . . . . , .82

Amphicyo-n palceindicus .......... 84

Camelopardalis

„ sivalensis

Rramatherium

„ perimense

Camelus sivalensis
Dorcatherium

„ majus .

„ minus .

Cervus

„ latidens

„ triplidens

simj)licidens .
Listriodon

TABLE OE CONTENTS.

Part TIL Crania of Euminants, from the Indian Tertiakies, by E. Lydekker
Introductory remarks ........

List of Indian Tertiary Euminants ......

Jios . . ......

„ namadicus .......

„ planifrons . . . , . .

„ acutifrom ..... ...

„ flatyrhinus ......

Bison ..........

„ sivalensis ... .....

Buhalus . . . . . . . .

,, platyceros

„ jpalceindicus

Perihos

„ occipitalis (= Semihos occipitalis)
Hemibos ....

„ triquetriceros (= S. acuticornis)

Amphihos ....

„ acuticornis {= Hemibos sp.)

Antilope ....

„ sivalensis

„ patulicornis .

„ porrecticornis

Hydaspitherium

„ megacephalum

Homology of horns of Sivatheridce and Camelopardalis
Capra ....

sivalensis

perimensis

IV. Supplement TO Crania of Euminants, by

Introduction ....

Bos planifrons and B. acutifrons
Bubalus platyceros
Hemibos ....

„ occipitalis

„ acuticornis

„ antilopinus

Leptohos falconeri
,, frazeri

Bramatherium perimense
Hydaspitherium grande and leptognathus
Bucapra daviesii
List of Indian Tertiary Euminants

V. SiwALiK AND Narbada Proboscidia, by I
Dentition and affinites of the Hinotheridm
Binotherium pentapotamice

„ indicum

„ sindiense

Dentition of the Elephantidce
Mastodon

„ falconeri
„ pandionis

R. Lydekker

Lydekker

Page.

100

112

119

122

127

132

141

145

150

154

157

158

159
1591
166
169

169

170

171

172

172
17a

173

174 ,,
174 i
176
178

178

179
179

179

180
180
182
182
183
192
196
198
202
202
213

TABLE OF OONffENTS.

\i‘ I

yfpiU

Mastodon latidens
„ perimensis
„ sivalensis
Elephas
Stegodon

„ cliftii

- „ hombifrons

„ insignis

„ ganesa

Loxodon

„ planifrons

Euelephas

„ hysudricus

„ namadicus .

Table of dimensions of molars of Indian Mastodons .

„ „ „ Stegodons

,, Recent and fossil Proboscidia ....

Table showing the number of ridges in the molars of the Proboscidia
Conclusion ........

References to non-Indian species of fossil Proboscidia

Appendix . . .....

Page.

227

239

248

250

262

268

273

275

278

287

280

282

283

284

286

289

292

V '

INDIAN TERTIARY AND POST-TERTIARY VERTERRATA.

VoL. I.

PREFACE.

In completing this first volume of the series of the “ Palseontologia Indica,” entitled
“ Indian Tertiary and Post-Tertiary Vertehrata,” I avail myself of the opportunity
afforded of making certain corrections and additions, which subsequent investigations
have rendered necessary, in regard to some of the descriptions. I may mention
that at tlie time of publication of the second fasciculus, there were contained in the
Indian Museum only very fragmentary remains of many species, now represented
by a much larger and more complete series : by the help of these more ample mate-
rials, I am now enabled in several cases to correct certain errors into which I had
previously fallen. I may further add that had I had any idea, at the time of publi-
cation of that fasciculus, that the Indian Museum was at all likely to obtain such
a magnificent collection of Siwalik vertebrate fossils as now enrich its cases, I
should not have described remains of different orders in the heterogeneous manner
in which they are there placed, hut should have devoted a fasciculus to each order
or sub-order, as has been subsequently done.

I cannot hut regret that the execution of many of the plates in the second and
tliird fascicuh is so poor ; it was, however, the best that could be done at the time.
A change in our arrangements has produced better results in the last fasciculus.
Three of the worst of the earlier plates have been reissued with the fourth and
fifth fasciculi.

In the references given below, and also in the index to the volume, the number
of the pages refers to the continuous paging of the volume, and not to the separate
paging of the five component parts.^ The plates, with the exception of the first
three, have the same heading, “ Tertiary Mammalia,’’ as forming one continuous
series of illustrations of mammalian remains. The measurements given in this
volume are all in inches and tenths.

^ In the first fasciculus, there is of course only one system of paging ; in the second, the volume numbering is at
the bottom of each page ; and in the succeeding parts, on the outer side of the part numbering.

Vlll

PEEPACE.

The first fasciculus of this volume bears a different serial title from the suc-
ceeding fasciculi ; this discrepancy is owing to the circumstance, that at the time
of the publication of the first fasciculus, there was no idea entertained, that the Indian
Museum was likely to acquire the large collection of Siwalik fossils which are now
exhibited in its cases.

In the introductory remarks to the second fasciculus, it was stated that the fourth
fasciculus would be devoted to the description of the remains of Carnivora, and that
a classified synopsis of the fossil Mammalia of South-Eastern Asia would he ap-
pended. Since the publication of the second fasciculus the collection of fossil
Proboscidia in the Indian Museum has increased to such an unexpected extent, that
it has afforded ample materials for a large fasciculus of itself : the collection of fossil
Carnivora, on the other hand, is still very imperfect, and its description has accord-
ingly been postponed : the publication of the synopsis has likewise been deferred.

No strict systematic arrangement has been adopted in the second, thu’d, and
fourth fasciculi, many of the specimens having been obtained while the work was still
in progress, and described out of their proper serial succession : in the fourth fasci-
culus the same arrangement of the ruminants has been adopted as in the third, for
convenience of reference. Similarly in the lists of ruminants given on pages 92 and
ISO, the same grouping has been adopted : the reader will understand, therefore,
that the arrangement in those lists is in no wis'e a systematic one . in a systematic
list the goats and sheep would of course be placed with the other Camcornia. In
the last fasciculus, where the materials were all at hand at the commencement of
tlie work, a strictly systematic arrangement has been adopted.

Ehinoceros deccanensis. — Professor Elower^ classes this species under the
generic, or suh-generic, division Atelodus, in which are included the living
B. Ucornis and B. simus of Africa, and the fossil B. pachygnathus, B. etruscm,
B. leptorhmus, B. hemitcechus, and B. tichorhinus of Europe. The lower jaw
figured in Plate LXXIV, fig. 6, of the “ Eauna Antiqua Sivalensis,” under the
name of B. sivalensis,^ seems also to indicate an animal belonging to the same
group.

Narbada Ehinoceros. — A recent re-examination of the two upper molars of
a rhinoceros from the pleistocene rocks of the Narbada valley, figured on Plate IV,
figs. 5 and 6, of this volume, and described on page 32 under the name of BJiino-
ceros namadicus^ has convinced me that these teeth are specifically indistinguish-
able from those of the living B. indicus, the very slight differences which I pointed
out as existing between the recent and fossil last molars not being more than indi-
vidual varieties. If, therefore, the similarity in the teeth can he relied on, we

^ Proc. Zool. Soc., 1876, p. 457.

2 Althougli in the text I have adopted Falconer’s determination of the lower jaws of the Siwalik Ehinocerotes, I
am quite unacquainted with the grounds on which such determinations were made.

® The name R. namadicus of Falconer occurs' in the introduction to the “ Fauna Antiqua Sivalensis” (Pal. Mem.,
Vol. I, p. 21), and was, I believe, as stated in the sequel, applied to limb-bones from the Narbada.

PREPACE.

have evidence of the existence of R. indicus in the pleistocene of India, ^ contem-
poraneously with the extinct mammals.

I cannot discover the history of the two molars under discussion : when I joined
the Geological Survey, they, in company with other fossils, were in the Museum of
the Survey and were labelled “ Nerhudda valley.” Both teeth are in a highly
mineralised condition, and the earlier one (fig. 6), is embedded in a block of hard
brown clayey sandstone, like many other Narbada specimens ; I have therefore no
doubt as to their origin.

An ultimate upper molar of R. indicus has been obtained by Mr. Eoote from
the alluvium of Madras, showing, in conjunction with the Narbada specimens, that
the former geographical range of the species must have been very extensive.^

A right humerus of a fossil rhinoceros from the Narbada, in the collection of
the Indian Museum, differs considerably in form from the corresponding bone of
R. indicus, and would, therefore, seem to indicate tlie former existence of a second
Narbada species. If the species should eventually turn to be distinct, the name of
R. namadicus might be appHed to it : the name is provisionally retained in the list
of species of Rhinoceros given below, in order to mark the existence of a second
Narbada species.

SiwALiK Bhinocerotid^. — Since the publication of the second fasciculus of
this volume, in which the teeth of the Indian fossil species of Rhinoceros were
treated of, as far as my materials then went. Professor Brandff has published
a synopsis of the living and fossil species of Rhinoceros, in which he has arrived at
conclusions, which appear to me unaccountable, in regard to Falconer’s three species
of Siwalik Rhinoceros. On page 39 of that memoir, R. sivalensis and R. palcein-
dicus are considered as being specifically identical with R. indicus {unicornis) ;
no reasons, however, are given for this union of the three species, except a vague
remark of the late Mr. Blyth, to the effect that there is a great resemblance in
the form of the skulls of the three forms. Professor Brandt appears to have
entirely overlooked the characters of the molars of the three species. The molars
of R. mdicus are characterized by their complex structure, that is to say, they are
furnished with a -‘combing-plate” and a “ crochet,” or, in other words, the second
main internal column is connected with the outer wall of the tooth by a continuous
ridge of enamel, and when the tooth becomes worn there are at least three islands,
or f ossettes, on the crown : further, the external surface of the tooth runs nearly
parallel to the long axis of the crown and is not produced into a buttress at the
antero-external angle. The three fossettes on the worn crown are situated nearly

‘ A single upper molar in the Indian Museum, collected by Mr. Hacket in the Narbada deposits, is indistinguish-
able from the corresponding tooth of the living Cervus (Rucervus) duvaucellii, indicating the existence of another
living mammal in the pleistocene.

^ I regret that the execution of the figures of the Narbada specimens is so defective, not giving at all a fair idea
of their form.

3 Mem. de I’Acad. Imp. des. Sci. de St. Pet., Ser. VII. Vol. XXVI, No. 5.

PEERAGE .

in the same line, and parallel to the antero-posterior axis of the crown. In MMno-
ceros sivalemis, on the other hand (as is well shown in the upper molar figured
in Plate Y, fig. 5 of this volume), there is no combing-plate, and the crochet
is quite unconnected with the outer wall of the tooth ; consequently when the tooth
is worn down, there are normally not more than two fossettes^ on the crown (as is
well shown in fig. 6 of Plate LXXIV of the “ Eauna Antiqua Sivalensis ”), while
the antero-external angle of the tooth is produced into a large buttress (shown
well in my figure). The teeth of JR. sumatrensis are of the same type, but that
species is distinguished from JR. sivalemis by having two horns in place of one horn.

The true molars and premolars of JR. palceindicus (I am here doing little
more than repeating the matter given in the text of this volume), likewise never
have a combing-plate (though from specimens lately acquired by the Indian
Museum, as well as from the young skuU represented in fig. 1 of Plate LXXIY of
the “ Eauna Antiqua Sivalensis ”), it appears that the milk-molars are, at all events
sometimes, furnished with one. In the adult dentition (as is shown in fig. 2« of
Plate LXXIV of the “ Eauna Antiqua Sivalensis”) there are normally three fossettes
on the worn crown of each tooth, at some period of its wear : these three fossettes are
not, however, placed in the same antero-posterior line, as in R. indicus, but the
middle one is placed somewhat externally to the other two (“ Eauna Antiqua Siva-
lensis”, Plate LXXIY, fig. 2a, Plate LXXY, fig. 1), being cut off from the end
of the main valley, and not from the hinder side of it. Einally, there is no distinct
buttress in the true molars of JR. palceindicus (“ Eauna Antiqua Sivalensis,” Plate
LXXIV, fig. 2a) as in R. sivalensis, the outer surface of each molar being nearly
flat in the former species. In treating of Rhinoceros palceindicus at page 24s of
this volume, I noticed a young cranium of that species, which is figured in the
‘‘Eauna Antiqua Sivalensis ” (Plate LXXIY, fig. 1), and of which there is a cast
in the Indian Museum. In noticing this cranium in the description of the plate.
Dr. Ealconer observes : “ Very perfect specimen of cranium with both zygomatic
arches entire. Shows two molars and two posterior premolars on either side. The
third molar is still in germ.” Erom the mention of premolars by Dr. Ealconer, it
is quite evident that he considered the specimen as showing the permanent dentition^
and in my notice of the cranium, I naturally followed this identification. If, how-
ever, we refer to the figure of the cranium in question in the “ Eauna Antiqua
Sivalensis,” we shall see that the last tooth is less worn than the thud, or any of
the preceding teeth. On Ealconer’s supposition, the third tooth, being the first true
molar, should have been more worn than the second tooth, or last premolar. Again,
had the second tooth been the last premolar, and being as much worn as it is, the
last true molar would have been in use; further, the first tooth is quite unlike
a second premolar. Erom the above, it will be quite evident that the four teeth
in Ealconer’s cranium really are the four milk-molars. There is therefore no abnor-

* A broken skull, probably belonging to this species, in the Indian Museum shows three fossettes on the two
middle premolars, but none on the true molars or last premolar.

PREFACE.

mality in the dental series oP this species, as was the case according to Dr. Ealconer’s
supposition, and the permanent dental formula given on prge 24 must now he
made to stand as follows^ : —

0— ? C ~ P — M —

J- 1_1 0—0 3—3 3-3

The table of measurements of the molars will also require the alteration of
their names, and will read as follows : —

Length of 1st milk-molar
„ of 2nd „

„ of 3rd „

„ of 4th „

Width of 1st milk-molar
,, of 2nd „

„ of 3rd „

„ of 4th „

In.

. 1-10
. 1-65
. 1-90
. ^2-20
. ‘o-89
. 1-50
. 1-8.5
. 1-90

Professor Brandt, in identifying H. sivalensis with E. palcRindicus, overlooks
the above differences in the dentition, as vi'ell as the differences in the crania.
E. sivalensis (“ Eauna Antiqua Sivalensis,” Plate LXXIII, fig. 2a) has the summit
of the occiput and parietals raised to a much higher point than in E. palceindicus
{Ibid., fig. la) ; the interval between the nasals and maxillae is much narrower in
the former than in the latter ; E. 'palceindicus is further distinguished by its broader
forehead. Both species agree in being unicorn.

On page 44 of the same memoir, Professor Brandt classes the Siwalik E. platy-
rhinus with E. sv.matrensis, in the genus, or sub-genus Ceratorhinus of Gray, ap-
parently merely on the grounds that both species were furnished with two horns.
Now, the molars of E. platyrlnnus are of the complex type (as is explained in the
text) of E. indicus and E. tichorJiinus, and not of the simple type of E. sumatrensis.
The descriptions given in the text vdll amply show the distinctness of the two
species above mentioned, and there are now additional materials in the Indian
Museum which still further illustrate the dentition of the fossil species, and which
I hope to bring to notice on a future occasion. As evidence apart from the upper
molars, the occurrence of three distinct forms of mandible of Ehinoceros (apart from
Acerotherium perimense, which is also represented in the Indian Museum) in the
Siwaliks, proves the existence there of three species, although their references to
the three named species by Ealconer (which I have accepted in the text) is
very probably open to doubt. I have shown that E. platyrhinus, as regards its
dentition, is nearest to E. indicus, and E. sivalensis to E. sumatrensis, which is
precisely the reverse of the, what I cannot but call, arbitrary identifications of
Prof essor Brandt. ‘If characters like those of the teeth described in the sequel are to
be completely ignored, both Zoology and Palseontology together would be impossible,
and all the species of a genus might as Avell receive a single name. Professor Brandt

’ The first milk-molar, which often persists, is not counted here in the permanent series.

Xll

PEEPACE.

makes no mention of the somewhat startling instance (according to his identifica-
tions) of a Miocene mammal (for R. sivalensis lived in the Miocene period in
Sind) being identical with a living species, or, in other words, that R. indicus is a
Miocene species, and was a contemporary of the long extinct Binotlierium, Hyopota-
mus, and Antliracotherium ; R. sumatrensis being also, according to the same author,
at least a Pliocene species. As far as I am aware, there are hardly any instances of
newer Pliocene mammals being identical with living species, and even by far the
greater number of the Pleistocene forms are extinct. I should be inclined to look
very doubtfully on such pedigrees for the Indian rhinocerotes even, were they sup-
ported by strong evidence, which in the present case is conspicuous by its absence.

The ‘small Rhinoceros tooth described on page 46 and drawn in fig, 10 of
Plate VI of this volume, as a premolar of an undetermined species, I now think, in
all probability, is an anterior milk-molar of R, platyrhinus.

In a recently published memoir on the fossil species of Rhinoceros and the
allied families,^ Professor C ope has placed Rhinoceros sivalensis in a distinct genus
under the name of Zalabis.

This generic distinction is made on the strength of the statement originally
made by Falconer, that this species was ‘‘ hexaprotodont.”^ I have already shown
in the text of this volume (p. 63), that the specimens figured in the “Fauna
Antiqua Sivalensis ” do not support this statement. In that work, there are figured
three forms of lower jaws of Siwalik Rhinoceros, referred to the three named species,
none of which are hexaprotodont, and of which the one referred to R. sivalensis
lias no incisors. None of tlie skulls of R. sivalensis with which I am acquainted
show any upper incisors. I cannot, therefore, see that there is any evidence on
which Professor Cope’s new genus can be supported.^

The very different conclusions arrived at by Professors Brandt and Cope in
regard to Rhinoceros sivalensis, afford subject for the most serious reflexions as to
the present conditions under which palaeontological research is carried on. In this
case we find two eminent palaeontologists, with precisely the same materials before
them, arriving at the most opposite conclusions; Professor Brandt identifying
R. sivalensis with a living species, and Professor Cope referring it to an entirely
new genus ! The former writer appears to have arrived at his conclusions from
neglecting to notice the specific differences pointed out by other workers, while
the latter has relied upon alleged differences which have been shown to be un-
supported by any kind of tangible evidence. In both cases it was incumbent on the
writers to have decidedly refuted all points which militate against their own con-
clusions, before instituting the sweeping changes which such conclusions involve.

’ Bull. U. S. Geol. Geog. Surv., Vol. V, p. 232.

2 Professor Cope reckons tke outer lower tusk of Rhinoceros as a canine.

® I may mention that on page 229 of his above quoted memoir, Professor Cope omits Acerotherium jperimense •
from his list of that genus, Rhinoceros iravadicus from the genus Rhinoceros, and R. deccanensis from the genus
Atelodus ; Professor Cope also alludes to the Siwaliks as being undoubtedly of upper Miocene age !

PREPACE.

Xlll

If the lower jaw assigned to R. simlensis in the “ Eauna Antiqua Sivalensis ”
he rightly assigned, and if we admit the sub- divisions into which the old genus
Rhinoceros is split up by many modern naturalists, R. simlensis would seem from
this point of view to belong to the genus Atelodus characterized by the symphysis
of the mandible of the adult being edentulous, as in Rhinoceros simus. The skull
of R. sivalensis is, however, unicorn, and, therefore, differs from that of R. simus.
The species in fact, if the remains are rightly correlated, will not fit into any of
the modern sub-divisions of the genus.

Rhinoceuos ieavadictjs. — The fragment of a right maxilla of a species of
Rhinoceros with two teeth described on page 45, and figured in Plate V, fig. 4,
of this volume, which was not specifically determined, and the teeth in which were
considered to be premolars, I now find to belong to a young individual with the milk-
molar dentition. It appears probable that these milk-molars may have belonged to
a young individual of Rhinoceros iravadicus, the permanent molars of which species
are figured on the same plate ; in having a combing plate they are more complex
than the true molars.

Aceeotherium perimense {Rhinoceros planidens). — As I have already men-
tioned in the “ Records,”^ the two imperfect upper molars of a rliinoceros figured
on Plate IV, figs. 7 and 9, and described on page 41 of this volume, as belonging
to a new species of Rhinoceros, under the name of R. planidens, really belong to
Acer other ium perimense of Ealconer and Cautley. The upper teeth of that species
figured on Plate VI, figs. 2 and 5, as upper true molars, really are premolars, and
the unnamed specimen represented in fig. 6 of the same plate is likewise an upper
premolar of the same species.^ The name of R. planidens must accordingly be
erased from the list of Asiatic species of Rhmoceros given on page 52, and the
description of its upper molars be read as those of A. perimense. In a subsequent
volume I shall hope to illustrate more fully the dentition and craniology of the
last named species : a cranium is now in the collection of the Indian Museum. As
the teeth of this species, described on page 51 as molars, are really premolars, the
statement as to the difference in shape of the molars of this species from the molars
of Rhinoceros will consequently not stand.

It is at present unknown whether Acerotherium perimense was furnished witli
three or four digits to the forelimb, and from the condition in which Siwalik fossils
usually occur, it is very improbable that this point will ever be determined. It is,
therefore, impossible to say whether the species really belongs to Acerotherium or to
the new genus Aphelops of Professor Cope,® differing from Acerotherium in having
only three anterior digits. I prefer provisionally to retain the species in the older
genus. •

* Vol. XII, p. 47.

^ On page 44 I mentioned that I thought it possible this tooth should he referred to R. 'planidens (A. perimense).
The cingulum is remarkably developed in this tooth, and causes it to resemble the premolars of R. deccanensis, as
noticed in the description.

® Bull. U. S. Geol. Geog. Surv., Vol. V, p. 236.

XIV

PEEPACE.

Species oe Asiatic Ehinoceuotid^. — I append a few notes on the list of
species of Rhinoceros given on page 52. Professor Brandt, in his memoir quoted
above, admits Rhinoceros inermis of Lesson as a distinct species, which should, there-
fore, be added to the list. Professor Cope, in the above quoted memoir, suggests
tliat this species should be referred to the genus Aphelops. Professor Brandt includes
under Rhinoceros javanicus {sondaicus) both R. nasalis and R. flowed of Gray
(the latter species classed as a synonym in my list). By the same writer R. sieno-
cephalus of Gray is considered to have been founded on a young individual of R.
indicus i(u,nicornis'). Again, under Rhinoceros {Ceratorhinus) sumatrensis,VvoiQ^^ov
Brandt includes R. crossii, R. hlythii, and R. nig er of Gray; the two former
so-called species were omitted from my list, as being probably synonyms.

The following list of the recent and fossil species of Rhinocerotidcc of South-
Eastern Asia is given to replace the one given on page 52, as embodying the more
recent views. The synonomy of the existing species is given on the authority of
Professor Brandt, and is taken from the memoir already cited. The name by which
each species has been referred to in this volume is taken as the name of the species.
The sub-generic (or generic) divisions of Rhinoceros (except Acerotherium) have
been ignored, as they are, in many cases, inapplicable to the fossils, and the species
have been arranged in alphabetical order; synonyms are in italics.

1. Acerotherium perimense (Falc. and Caut.) Mio-Pliocene. India and Burma.

Rhinoceros pe7'imensis (Falc. arid Caut.)

Rhinoceros planidens {olim nobis).

2. Rhinoceros deccanensis (Foote). Pleistocene. India.

3. Rhinoceros indicus (Cuv.) Recent and Pleistocene. India.

R. asiaticus (Blum.)

R. iiamadieiis {olim nobis)^

R. stenocephalus (Gray).

R. unicornis (Linne).

4. Rhinoceros inermis (Lesson). Recent. India.

5. Rhinoceros iravadicus {nobis).^ Mio-Pliocene. Burma.

6. Rhinoceros javanicus (Gray). Recent. South-Eastern Asia.

R. Jiotoeri (Gray).

,R. javanus (Cuv.)

R. nasalis (Gray).

R. sondaicus (Horsfield).

7. Rhinoceros lasiotis (Sclater). Recent, Mallacca.

? 8. Rhinoceros namadicus (Falc. and Caut.) Pleistocene. India.

9. Rhinoceros palseindicus. (Falc. and Caut.) Mio-Pliocene. India.

10. Rhinoceros platyrhinus (Falc. and Caut.) Mio-Pliocene. India.

1 1 . Rhinoceros sinensis (Owen). ? Pliocene. China.

12. Rhinoceros sivalensis (Falc. and Caut.) Mio-Pliocene. India.

R, indicus fossilis (Baker and Durand).

13. Rhinoceros sumatrensis (Cuv.) Recent. South-Eastern Asia.

R. blythii (Gray).

R. crossii (Gray).

R. niger (Gray).

R. sumatranus, (Raffles).

1 The word nohis in this volume always refers to myself, and not to Mr. Foote.

PREPACE.

SiWALiK Artiodactyla DESCRIBED IN 2nd PART. — Since the publication of
the second fasciculus of this volume, the hst of Siwalik suine AHiodactyla has
been considerably increased, and the list given on page 78 is consequently incom-
plete : to that list must now he added —

Syotherimn sindiense.

Sivameryx.

JBLemimeryx.

Chceromeryx silistrensis has also turned out to he distinct from Anthracotherium
silistrense}

Pal^ortx. — Two upper molars in the Indian Museum, from the Siwaliks,
seem to me to he in all probability generically identical with Falceoryx of the
Pikermi beds of Attica.^ I cannot, ' however, at present he quite certain of this
determination, owing to the extreme difficulty of distinguishing the molars of many
genera of ruminants.

PoRTAX. — Some upper and lower jaws with ruminant teeth, from the Siwaliks,
appear to me generically indistinguishable from those of Fortax, and indicate a
Siwalik representative of that genus.

PoRTAX NAMADicus (Riit.). When noticing the additions made by Professor
Riitimeyer to the Indian fossil Ruminants,® I had not observed the new Narbada
species of For tax (P. namadicus) named by the Professor, from the hinder part
of a skull in the British Museum.^ The addition of this species to the Narbada fauna
is of great importance, as connecting the living and Siwalik species.

An atlas of a ruminant in the Narbada collection of the Indian Museum corre-
sponds so exactly with the atlas of Fortax picUis (except in being slightly larger)
that I think it probably belongs to Fortax namadicus. Two left molars of a rum-
inant, formerly in the Museum of the Asiatic Society of Bengal, and entered on
page 255 of Ealcoher’s catalogue of that collection, as being from the Narbada,

* I am now uncertain of the generic distinctness of Ilippofotamodon.

^ “ Animaux fossiles et Geologie de I’Attique,” Plate XLVII.

3 pp. 178-180.

■* “ Die Binder der Tertiar-Epoche, &c.,” p. 89, Plate VI, figs. 7 and 8. I hope Professor Eutimeyer will pardon me
if I mention the inconvenience which arises from first mentioning the name of a new species in the midst of a paragraph ,
as he has done in this case. The description of every new species ought to have a distinct heading, hy which it at once
catches the eye. In the case of Fortax namadicus not at first having time to read Professor Biitimeyer’s memoir
through, the name of this species, from appearing in the middle of a sentence, did not catch my eye until I had time
for a more leisurely perusal of the work. Dr. Gray uses the name Fortax ficta for the living species, which also
occurs in a former part of this volume ; the Greek word Ilo'pra^, a young hovine animal, is, however, either masculine
or feminine, and the former gender should have the preference. Hence Forta.x pictas, as given by Jerdon
and F. namadicus, by Eutimeyer, are correct.

XVI

PEEFACE.

and doubtfully referred to the genus Germs (No. N, 68), I have carefully compared
with the corresponding teeth of Fortax pictus, and find that the two are generically
indistinguishable, and I therefore come to the conclusion that the fossil teeth
probably belong to Fortax namadicus of Professor Eutimeyer. Similar teeth have
been obtained by Mr. Hacket from the Narbada, and by Mr. Eedden from the
Pem-ganga.

Bovine. — Among Mr. Theobald’s Siwalik collection, I have lately determined
two specimens of the axis vertebrae of a bovine, which, from their large size, must,
I think, undoubtedly belong to Bos acutifrons. I have compared these vertebrae
with a perfect specimen of the axis of Bos primigenius in the Indian Museum, and
with another specimen figured by Professor Eutimeyer,^ and I find that although
the Indian and European vertebrae are similar in general plan, yet there are
many points of detail in which they differ considerably, thus confirming my con-
clusions drawn from the skulls.

It may be well to mention here that at the time of writing the third fasciculus
of this volume,^ I was unacquainted u ith the fact that some varieties of Bos primi-
genius have horn-cores with an elliptical cross section at the base, as the specimen
drawn in fig. 3 of Plate II of the above- quoted memoir of Professor Eutimeyer.
In this respect, therefore, there is a closer relationship between Bos primigenius
and B. planifrons and B. acutifrons than I have indicated in the text.

It should also be observed that Professor Boyd Dawkins drops the name
Bos primigenius, and identifies that animal with the prehistoric and historic Bos
urus, now represented by the cattle of Chillingham Park. By Professor Eutimeyer,
in his last published memoir,^ the name Bos primigenius is retained for the
Pleistocene form, and the name Bos taurus adopted (as a race name) for the living
form.

With regard to the bisons. Professor Eutimeyer has come to the conclusion
that the bones of Bison priscus are indistinguishable from those of Bison ameri-
canus, while, on the other hand. Bison priscm and B. europceus are also indistin-
guishable specifically.® There is thus established an intimate connection between
the now widely different European and American bisons. In his latest work, how-
ever, Professor Eutimeyer ® retains the three above mentioned specific names, and
places B. sivalensis as the earliest representative of the group.

The genus Bubalus of other writers is split up by Professor Eutimeyer into two
genera,*' which are termed Bubalus and Buffelus ; the former includes the European
Pleistocene B. antiquus and the living African B. coffer and B. hracliyceros > the
latter includes B. platyceros {sivalensis), B. palceindicus, and B. pallasii in the
Pleistocene. In the recent period Professor Eutimeyer names the living Indian

> Nov, Mem. Soc. Helv., Vol. XIX, PI. IV, figs. 1, 2-
- See page 112.

* Quar. Journ. Geol. Soc., Lond., Vol. 22, p. 394.

'* “ Die Kinder der Tertiar-Epoche, etc.,” p. 189.
® See “ Prehistoric Times,” 2nd ed., p. 296.

" loc, cit.

PREPACE.

xvn

buffaloe, Bu-ffelus mdicus : this name is, of course, synonymous with the name Bubalus
buff elm adopted by Gray,^ and B. arm, adopted in tbis volume, after Jerdon.^

Professor Riitimeyer also mentions another, apparently Indian species, under
the name of Buffelus sondaicus.

The genus Bibos, according to Professor Riitimeyer, includes the domestic In-
dian cattle, B. indicus, the Gour, B. gaurus, a doubtful species named B. gavceus,
and B. sondaicus {banting of Gray and this work). Bibos frontalis (the Mithun or
Gayal) of Lambert and Hodgson is omitted from the list, unless it be B. gavceus.
Whether or no the Mithun occurs now in the wild state or not, it appears to me to
be a distinct species, though there are some skulls in the Indian Museum which
bridge over the gap between this animal and the typical Gour.

Cee-vtjs. — A palatal portion of a skull containing teeth similar to those
figured on Plate VIII, fig. 3, under the name of Cervus simpUoidens, seems to
confirm the generic identification of those teeth. Of Cervus trijplidens (figs. I, 2,
Plate VIII) I have not obtained any additional specimens, but I think the generic
identification is here also correct. One of the Siwahk deer had flattened and
branching antlers much like those of Cervus duvaucellii, indicating the high state
of evolution of the group in Siwahk times.

The Indian Museum has lately obtained a complete maxilla, containing molars
like the one tooth drawn in figs. 7 and 10 of Plate VIII. I had some hesitation,
as noticed in the text, in referring the figured tooth to Cervus, and I still am not
quite sure whether such reference is correct, though I think it very possibly is.
The amount of variation in the form of the molars of the Gervidee is so great that
it is very difficult to say, in the case of isolated teeth, how far this variability may
extend. I have, on account of the possibility of doubt, not included Cervus latidens
in the list of ruminants given on page ISO. I hope eventually to obtain materials
which will decisively indicate the genus of the teeth in question.

The lower molars of a Cervus drawn in fig. 5 of Plate VIII, which I thought
might possibly belong to G. simplicidens, I now find, from the character of their
enamel, belong, in all probability, to a new species, which I propose, on a subse-
quent occasion, to call G. sivalensis.

Pal^omertx. — A single tooth of a ruminant from the Siwaliks, lately presented
to the Indian Museum by the Roorkee (Rurki) Museum, seems to belong to
Balceomeryx, and to have been about the size of P. bojani : this tooth seems to
have come from the lower Siwaliks (Nahans), and a lower molar from probably
the same horizon in Sind, not impossibly belongs to the same genus.

Camelopardalis. — A more complete lower jaw of Camelopardalis sivalensis
than the one of which the molars are drawn in fig. 5 of Plate VII of this
volume has lately been acquired by the Indian Museum ; the new specimen is

Brit. Mus. Cat.

2 Mammals of India, p. 370,

XVlll

PREFACE.

rather smaller than the figured one. A large series of molars of this genus have
also been obtained which seem to indicate the existence of other Siwalik species.

ViSHNUTHEmuM.— Two associated upper molars of a sivatheroid from the
Siwaliks of the Punjab, present characters which distinguish them from the molars
of Slvatherium, Bramatherium, and HydaspitJierkmi ; from the size, and certain
details of the form of these teeth, I have thought it not impossible that they may
belong to Vishnutherimn iravadicum, of which the lower molars are figm?ed in
Plate VII, figs. 1 and 2.

Hydaspitheritjm megacephaltjm. — It is possible that some critic may say
that the sj)ecific name of tins species is a barbarism, and, therefore, should be
replaced. The name megaceplialum is undoubtedly a barbarism, hut the termi-
nation appears to have gained general acceptation as a convenient adjectival form
for scientific names. The Greek substantive Ke<^aX7] if translated into Latin,
should probably he Cepliala, and the only direct adjectival form is Ke<^aXctJT09 ;
by naturalists, however, a latinised adjectival termination, Cep]ialus,-a~mi, is in
common use ; e. g., Plesiosaurus hrachycephalus, Halcyon leueocephala, Btycliozoon
liomaloceplialum. The term megaceplialum is used for brevity in place of megaloce-
plialum, on the precedent of Megatherium for Megalotherium.

Helladotherium. — M. Gaudry^ states that Helladotlierium occurs in India. I
am not aware on what grounds this statement rests.

These additional specimens will aflbrd ample material for another memoir,
illustrative of Siwalik ruminants and their allies.

Tetraconojdon magnum. — Since publishing (p. 79) the description of the imper-
fect mandible figured on Plate X, I have discovered the two last molars belonging to
that specimen, which had previously been mixed up with soma other specimens.
As the execution of Plate X was very poor, I take the opportunity of re-issuing that
plate with one of the newly found teeth in its proper serial position. To the
measurements of the molars of the figured specimen given on page 80 must he
appended, “length of last molar 1-95 inches, width of ditto 1‘3 inches.” This tooth
shows that Falconer’s specimen belonged to the upper jaw.

Conclusion. — The above additions and corrections will, I hope, render the con-
tents of this volume correct, as far as my present knowledge goes. European palieon-
tologists will, I hope, pardon many shortcomings and redeterminations in my work
which have been in many cases almost unavoidable. The work of a student in ver-
tebrate palaeontology in India is one of peculiar ditficulty in many ways. He has first
of all the difficulty, common to workers in other countries, of having very frequently
exceedingly imperfect and scanty remains from which to determine the affinities of
an animal, and is consequently liable to false inferences from this som*ce. Secondly ^
he feels the want of a large collection of the remains of described European species
of vertebrates for comparison : it is true the Indian Museum in Calcutta possesses a

' “ Les Encbainements du Monde Animal, Mammiferes Tertiaries,” p. 79.

PREFACE.

XIX

considerable collection of casts and remains of non-Indian fossil vertebrates, but these
really comprehend only a few of the better known genera.^ Thirdly, although the
library of the Geological Survey is an extensive one, there are wanting a great
number of the older works on vertebrate palaeontology, many of which are now out
of print : in many cases, moreover, works, when ordered from England, arrive in India
too late for the purposes of the worker who required them. Fourthly, the student in
Indian vertebrate palaeontology at the present time labours under the great disad-
vantage of working without the possibility of appealing to other workers in the
same branch of study for assistance and advice in cases of doubt and difficulty.
There are also wanting in the zoological department of the Indian Museum skulls
of many genera of living mammals which are required for an exhaustive comparison
of their fossil congeners. Finally, in the case of Siwalik fossils, there occurs the
additional and special difficulty of taking up the work in the incomplete state it was
left at the premature death of Dr. Falconer, and of determining ill-defined, imper-
fectly described species in Calcutta, without access to the original specimens in
London. This last difficulty has been the cause of several errors in the specific
determinations which occiu’ in this volume.

On the above grounds, I venture to hope that any want of references or com-
j)arisons to European fossils, and consequent possible shortcomings in this work,
will be looked upon with a lenient eye by my European and American fellow
workers.

In conclusion, it is but fair to mention that although there has devolved upon
myself the task of describing the Siwalik fossils in the Indian Museuni of Calcutta,
yet that the far more onerous task of amassing and bringing together that, perhaps,
unrivalled series has devolved upon my colleagues in the Geological Survey, chief
among whom are Messrs. W. T. Blanford, F. Fedden, W. Theobald, and A. B.
Wynne. By far the largest share of this work has been executed through the
indefatigable energy and perseverance of Mr. Theobald, who, through heat and
cold, drought and rain, has traversed many a weary mile of the hills and plains of
the Punjab to attain his object : to him especially are due the thanks of all
interested in the history of the tertiary vertebrates of India.

Indian Museum,
Calcutta, January 1880.

]

R. LYDEKKER.

1 I have the authority of the Superintendent of the Geological Survey of India to offer named duplicate
specimens of described Siwalik fossils in exchange for named specimens of teeth of European or American tertiary .
mammals.

LIST OF PLATES

Plate.

I. — Uhinocbros deccanensis, (Foote.)

IL — Ditto.

III. — Ditto.

IV. — RhinocbroSj sp. var.

V. — Ditto.

VI. — Ditto.

VII. — Amphicyon^ Bbamathemum, Camelopardalis, DoucatheriuM; Vishnutherium.

VIII. — Cervus, Listbiodon, Manis.

IX. — Dinotheridm pentapotami.®, (Fale.,) Sanitherium schlagintweitiIj (Meyer.)

X. — Tetbaconodon magnum, (Falconer.)

XI.' — Bos wamadicus, (Falc. & Cant.)

XII. — Bos AcuTiFRONS, (Nobis,) Bos planifrons, (Nobis.)

XIII. — Bos ACUTiFBONSj (Nobis.)

XIV. Bos PLATYRHINUS, (Nobis.)

XV. — Bison sivalensis, (Falc. & Cant, sp.)

XVI. — Bos, sp. var.

XVII. — Bison sivalbnsis, (Falc. sp.) Bubalus pal®indicus, (Falc. & Cant.)

XVIII. — Bubalus platycbros, (Nobis.)

XIX. — 'Bubalus PAL®mDicus, (Falc. & Cant.) ’

XX. — 'Hemibos occipitalis, (Falconer sp.) g (Described in third fasciculus as
Perihos occipitalis).

XXI. — Hemibos occipitalis, (Falconer sp.) S (Same specimen as in last plate) .

Hemibos acuticornis, (Falconer sp.) 9 (Described in third fasciculus as
AmpJiihos) .

XXI A. — Hemibos occipitalis, (Falc. sp.) S
XXI B. — Hemibos acuticornis, (Falc. sp.) ?

XXII. — Hemibos acuticornis, (Falc. sp.) S (Described in third fasciculus as Hemibos
triquetriceros) .

XXIII.— Hemibos acuticornis, (Falc. sp.) $ (Same specimen as in last plate).
XXIIIA. — Hemibos acuticornis, (Falc. sp.) §

XXIV. — Hemibos occipitalis, (Falc. sp.) Z (Described in third fasciculus as H^
triquetriceros) .

XXV. — Antilope sivalensis, (Nobis,) A. patultcornis, (Nobis,) A. porrecticornis,
(Nobis.)

xxii

LIST OF PLATES.

Plate.

XXVI. — Hydaspitherium megacephalum, (Nobis.)

XXVIT. — Hydaspitherium megacephalum, (Nobis,) Sivatherium giganteum, (Falc. &

Cant.)

XXVIII. — Capra sivalensis, (Nobis,) C. perimensis, (Nobis,) Capra,. sp.

XXIX. — Dinotherium pentapotami^, (Falconer.)

XXX. — Ditto.

XXXI. — Dinotherium, sp. var.

XXXII. — Mastodon palconeri, (Nobis.)

XXXIII.— Ditto.

XXXIV. — Mastodon pandionis, (Falconer.)

XXXV.— Ditto.

XXXV A.— Ditto.

XXXVI.— Ditto.

XXXVII. — Mastodon latidens, (Clift.)

XXXVHI.— Ditto.

XXXIX.— Ditto.

XL. — Mastodon perimensis, (Falc. & Cant.)

XLI. — Mastodon perimensis, (Falc. & Cant.,) Mastodon sivalensis, (Falc. & Cant.)

XLII. — Mastodon perimensis, (Falc. & Cant.)

XLIII. — Ditto.

XLIV. — Mastodon sivalensis, (Falc. & Cant.) '

XLV. — Stegodon cliftii, (Falc. & Cant.,) S. bombifrons, (Falc. & Cant.,) S. insignis,

(Falc. & Cant.)

XLVI. — Stegodon bombifrons, (Falc. & Cant.,) S. insignis, (Falc. & Cant.)

LIST OF ABBREVIATIONS USED, AND WORKS AND MORE IMPORT-
ANT MEMOIRS QUOTED IN THIS VOLUME.

“ Abhand. der k. k. Geol. Eeich.”=: “ Abbandlungen der k. k. Geologiscben Reichsanstalt,” q, v.

“ Abhandlungen der k. k. Geologiscben Eeicbsanstalt,” Wien.

“ Abhandlungen der Schweizerischen palantologischen Gesellschaft,” Basel.

Adams, A. L. “ Dentition and Osteology of Elephas antiquus.” Palseont. Soc., 1877.

“ Dentition and Osteology of Elephas primigenius,” Palteont. Soc., 1879.

“ Maltese Eossil Elephants.” Trans. Zool. Soc., Lon., Vol. IX.

“ Addenda and Corrigenda to Paper on Tertiary Mammalia,” Lydekker, R. G. S. I., Vol. IX, p. 154.

“ Amer. Jour. Science and Art” = “ American Journal of Science and Art,” q. v.

“ American Journal of Science and Art,” New Haven.

“ Anatomy of Vertebrates,” Owen, London, 1866-68.

“ Ancient Euminant Eauna of Switzerland” = “ Die Eauna der Pfahlbauten in der Schweiz,” q. v.

“ Animaux Eossiles et Geologie de I’Attique,” Gaudry, Paris, 1862-67.

“ Ann. Mag. Nat. Hist.” = “ Annals and Magazine of Natural History,” q. v.

“ Annals aad Magazine of Natural History,” London. .

“ Arch, du Mus. d’Hist. Nat. de Lyon”= “ Archives du Museum d’Histoire Naturelle de Lyon.”

“ Archives du Museum d’Histoire Naturelle de Lyon,” Lyons.

“ Asiatic Eesearches,” Calcutta.

Baker and Durand : “ Memoir on sub-Himalayan fossils.” J. A. S. B., Vol. V, p. 490.

Bettington : “ Memoir on certain fossils, more particularly a new Euminant, found at the Isle of Perim, in the Gulf of
Cambay.” Jour. Eoy. As. Soc., Vol. VIII, p. 340.

Blainville : “ Osteographie des Mammiferes,” Paris, 1839-64.

Blanford and Medlicott : “ Manual of Geology of India,” Calcutta, 1879.

Blum. = Blumenbach.

Brandt : “ Tentamen Synopseos Ehinocerotidum viventium et fossilium.” Mem. de I’Acad. Imp, des, Sci. de St.
Pet., Ser. 7, Vol. XXVI, pt. 5.

“ British Fossil Mammals” = “ History of British Fossil Mammals and Birds,” q. v.

“ British Fossil Mammals and Birds.” = “ History of British Fossil Mammals and Birds,” q, v.

“British Museum Catalogue of Ungulates” = “ Catalogue of specimens of Mammalia in the British Museum, pt.
Ill, Ungulata,” q. v.

“ Bulletin of the United States Geological and Geographical Survey of the Territories.” Washington.

“ Bui. U. S. Geol. Geog. Surv.” = “ Bulletin of the United States Geological and Geographical Survey of the Terri-
tories,” q. V.

“ Catalogue of A. S. B.”= “ Catalogue of the Fossil Eemains of Vertebrata in the Museum of the Asiatic Society of
Bengal,”’ q. v.

“ Catalogue of specimens of Mammalia in the British Museum, pt. Ill, Ungulata,” Gray, London, 1852.

“ Catalogue of Euminants” = “ Hand-list of the Edentate, Thick-skinned, and Euminant Mammals in the British
Museum,” q.v.

“ Catalogue of the Fossil Eemains of Vertebrata in the Museum of the Asiatic Society of Bengal,’! Falconer and
Walker, Calcutta, 1859.

“ Catalogue of Ungulata” = “ Catalogue of specimens of Mammalia in the British Museum, pt. Ill, Ungulata,”

q. V.

“ Cave Hunting, W. B. Dawkins, London 1874.

XXIV

LIST OF ABBREVIATIONS, &c.

Clift, W. ; “ On the Fossil Eemains of two new species of Mastodon and of other vertebrated animals found on the
left hank of the Irawadi.” Trans. Geol. Soc., Lon., Ser. 2, Vol. II, p. 369.

“ Comparative Anatomy of Vertehrata” = “ Anatomy of Vertebrates,” q. v.

“ Compt. Rend.” = “ Comptes Rendus Hebdomadaires des Seances de I’Academie des Sciences,” q. v.

“ Comptes Rendus Hebdomadaires des Seances de lAcademie des Sciences,” Paris.

“Contributions to the Extinct Vertebrate Fauna of the Western Territories,” Leidy, U. S. Geol. Surv. Rep., Ser. 2,
Vol. II.

Cope : “ Extinct Vertehrata of New Mexico.” ^ United States Geographical Survey west of the 100th Meridian, Vol.
IV, pt. 2, Washington, 1877.

“ On the Extinct Species of Rhinoceridae of Rorth America, and their Allies.” Bui. U. S. Geol. Geog. Surv.,

Vol. V, p. 227.

Croiz = Croizet.

Cuv. = Cuvier.

Cuvier ; “ Recherches sur les Ossemens Fossiles,” Paris, ed. 1836.

Dawkins, W. B. : “ Cave-hunting,” London, 1874.

“ On the Dentition of Rhinoceros etruscus (Falc.) Q. J. G. S. L., Vol. XXIV, p. 207.

“ On the Dentition of Rhinoceros leptorhinus (Owen) :” Q. J. G. S. L., Vol. XXIII, p. 213.

— — “On the Dentition of Rhinoceros megarhinus.” Nat. Hist. Rev., 1865, p. 399,

“ On the Fossil British Oxen.” Pt. I. Q. J. G. S. L., Vol. XXII, p. 391.

— “ On the Molar Series of Rhinoceros tichorhinus,” Nat. Hist. Rev., 1863, p. 525.

“ Dentition and Osteology of Elephas antiquus” Adams. Palaeont. Soc., London, 1877.

“ Dentition of Rhinceroses (Rhinocerotes), and on the characters afforded by the skull,” Gray : Ann. Mag. Nat. Hist.,
Ser. 4, Vol. XI, p. 356.

“ Dentition of Rhinoceros etruscus" Dawkins : Q. J. G. S. L., Vol. XXIV, p. 207.

“Dentition of Rhinoceros megarhinus" Dawkins: Nat. Hist, Rev., 1865, p. 399.

“ Description of a cranium of Stegodon ganesa" Lydekker : R. G. S. I., Vol. IX, p. 42.

“ Die Fauna der Pfahlbauten in der Schweiz.” Rutimeyer. Zurich, 1862. Nov. Mem. Soc. Hel., Vol. XIX.

“ Die Rinder der Tertiar Epoche, &c.,” Rutimeyer : Ahhand. der Schweiz, pal. Gesell., Vols. IV, V, 1877-78.

“ Distribution of Animals” = “ Geographical distribution of Animals,” q. v.

“ Extinct animals of North America,” Flower, London, 1876.

“ Extinct Mammalian Fauna of Dakota and Nebraska,” Leidy : Jour. Acad. Nat. Sci. Phil., Ser. 2, Vol. II.

“Extinct Vertehrata of New Mexico,” Cope : U. S. Survey west of 100th Meridian, Vol. IV, pt. II, Washington,
1877.

“ Extinct Vertebrate Fauna of the Western Territories” = “ Contributions to the extinct vertebrate fauna of the
Western Territories,” q. v.

F. and C.= Falconer and Cautley.

Fal. or Falc.= Falconer.

Falconer and Cautley : “ Fauna Antiqua Sivalensis,” London, 1846-47.

■ Walker: “Catalogue of the Fossil Remains of Vertehrata in the Museum of the Asiatic Society of

Bengal,” Calcutta, 1859.

Falconer: “ Palseontological Memoirs of the late Hugh” — C. Murchison, London, 1868.

“ F. A. S.” = “ Fauna Antiqua Sivalensis,” q. v.

“ Fauna Antiqua Sivalensis,” Falconer and Cautley, London, 1846-47.

Flower, W. H. : “ Extinct Animals of North America,” London, 1876.

Forsyth : “ Highlands of Central India,” London, 1871.

“ Fossil British Oxen.” W. B. Dawkins. Q. J. G. S. L., Vol. XXII, p. 391.

“ Fossil Eemains of Mammals found in China,” Owen : Q. J. G. S. L., Vol. XXVI, p. 417.

“ Fossil Eemains of two new species of Mastodon and of other Vertebrated Animals found on the left bank of the
Irawadi,” Clift : Trans. Geol. Soc., Lon., Ser. 2, Vol. II, p. 369.

Further notice of Siwalik Mammals,” Lydekker : E. G. S. I., Vol. XII, p. 33.

Gaud. = Gaudry.

Gaudry : “ Animaux Fossiles et Geologie de TAttique,” Paris, 1862-67.

■ — “ Les Enchainements du Monde Animal dans les temps Geologiques : Mammiferes tertiaries.” Paris, 1878.

“ Genera of Indian Cattle” = “ Illustrations of the Genera of the Bovinse,” q. v.

LIST OL ABBREVIATIONS, c^c.

XXV

“ Geol. Mag.” = “ Geological Magazine,” q. v.

“ Geological Magazine,” London.

“ Geographical Distribution of Animals,” Wallace, London, 1876.

Gervais, P. : “ Paleontologie et Zoologie” = “ Zoologie et Paleontologie Generales,” q. v.

“ Zoologie et Paleontologie Generales,” Paris, 1867-69.

Gray, J. E. : “ Catalogue of specimens of Mammalia in the British Museum, Pt. Ill, Ungulata,” London, 1852;

“ Dentition of Rhinoceroses (Rhinocerotes), and on the characters afforded by the skull.” Ann. Mag. Nat.

Hist., Ser. 4, Vol. XI, p. 356.

“ Hand-list of the Edentate, Thick-skinned, and Ruminant Mammals in the British Museum,” London,

1873.

“ Hand-list of the Edentate, Thick-skinned, and Ruminant Mammals in the British Museum,” Gray, London, 1873.

“ Highlands of Central India,” Forsyth, London, 1871.

“ Himalayan Botany” = “ Illustrations of the Botany and other branches of the Natural History of the Himalaya
Mountains,” q. v.

“ History of British Fossil Mammals and Birds,” Owen, London, 1846.

Hodgson, B. : “ Genera of Indian Cattle” = “ Illustrations of the Genera of the Bovinse,” q.v.

“ Illustrations of the Genera of the Bovinse J. A. S. B., Vol. X, p. 449.

“ Illustrations of the Botany and other branches of the Natural History of the Himalaya Mountains,” Royle, London,
1839.

“ Illustrations of the Genera of the Bovinee,” Hodgson ; J. A. S. B., Vol. X, p. 449.

“ Introduction and succession of Vertebrate Life in America,” Marsh, New Haven, 1877.

Jerdon : “ The Mammals of India,” London, 1874.

“Jour. Acad. Nat. Sci., Philadelphia” = “ Journal of the Academy of Natural Sciences of Philadelphia,” q. v.

“ J. A. S. B.” =“ Journal of the Asiatic Society of Bengal,” q. v.

“ Journal of the Asiatic Society of Bengal,” Calcutta.

“ Journal of the Academy of Natural Sciences of Philadelphia,” Philadelphia.

“Journal of the Royal Asiatic Society,” London.

“Jour. Roy. As. Soc.”=“ Journal of the Royal Asiatic Society.”

Kaup : “ Ossements Fossiles des Mammiferes, qui se trouvent au Museum grand-ducal de Darmstadt, ” Darmstadt, 1832 .
Kowalewsky : “ Monographic der Gattung Anthracotherium. Cuv. etc.” “ Palseontographica,” Vol. XXII.
Lart.=Lartet.

Le Conte: “Manual of Geology,” Philadelphia, 1879.

Leidy, “ Contributions to the Extinct Vertebrate Fauna of the Western Territories.” U. S. Geol. Surv. Rep., Vol. I,
Washington, 1873.

“ Extinct Mammalian Fauna of Dakota and Nebraska.” Jour. Acad. Nat. Sci., Phil., 1869.

“ Les Enchainements du Monde Animal, dans les temps geologiques : Mammifdres tertiaries,” Gaudry, Paris, 1878.
Lortet and Chantre :” Mastodontes du Bassin du Rhone.” Arch, du Mus. d’Hist. Nat. de Lyon, Vol. II.

Lubbock : “ Prehistoric Times, etc.” 2nd Ed. London, 1 869.

Lyd. = Lydekker.

Lydekker, R. : “ Addenda and Corrigenda to Paper on Tertiary Mammals.” R. G. S. I., Vol. IX, p. 154.

“ Description of a cranium of Stegodon ganesa.” R. G. S. I., Vol. IX, p. 42.

“ Further notices of Siwalik Mammals.” R. G. S. I., Vol. XII, p. 33.

“ Notes on the Mammalian Faunae of India and Burma.” R. G. S. I., Vol. IX, p. 86.

“ Notices of new and other Vertebrata from Indian Tertiary and Secondary rocks.” R. G. S. T.,

Vol. X, p. 30.

“Notices of Siwalik Mammals.” R. G. S. I., Vol. XI, p. 64.

“ Maltese Fossil Elephants,” L. Adams. Trans. Zool. Soc., Lon., Vol. IX.

“ Mammals of India.” T. C. Jerdon. London, 1874.

“ Manual of Geology,” Le Conte, Philadelphia, 1879.

“ Manual of Geology of India,” Medlicott and Blanford, Calcutta, 1879.

“Manual of Palaeontology,” Nicholson, Edinburgh and London, eds. 1872-79.

“Marsh, 0. C. : “Introduction and Succession of Vertebrate Life in America,” New Haven, 1877.

“ Mastodontes du Bassin du Rhone,” Lortet and Chantre. Arch, de Mus. d’Hist. Nat. de Lyon, V’ol. II.

“ Materiaux pour la Paleontologie Suisse,” Pictet, Geneva, 1854-73.

XXVI

LIST OL ABBEEVIATIONS, &c.

Medlicott and Blaiiford: “Manual of Geology of India,” Calcutta, 1879.

“ Mem. de I’Acad. Imp. des Sci. de St. Pet.” “ Memoires de I’Academie Impeviale des Sciences de St. Petersburg.”

“ Mem. Geol. Surv. Ind.”— “ Memoirs of the Geological Survey of India.”

“ Memoir on certain Fossils, more particularly a new Ruminant, found at the Isle of Perim, in the gulf of Cambay,”
Bettington. Jour. Roy. As. Soc., Vol. VIII, p. 340.

“Memoires de I’Academie Imperiale des Sciences de St. Petersburg,” St. Petersburgh.

“ Memoirs of the Geological Survey of India,” Calcutta.

Meyer, H. : “ Studien liber der Genus Mastodon.” Palreontographica, Vol. XVII.

“ Ueber die fossilen Reste von Wirbelthieren, welche die Herren von Schlagintweit von ihren Reisen in Indien

und Hoch. Asien mitgebracht haben.” “ Palaeontographica,” Vol. XV.

“Molar series of Mhinoceros tichorliinus,” W. B. Dawkins. Nat. Hist. Rev., 1863, p. 525.

“ Monographie der Gattung Anthracotherium, Cuv. etc.” Kowalewsky. “ Palaeontographica,” Vol. XXII.

Murie: “ On the Systematic Position of Sivatherium giganteum ol Falconer and Cautley.” Geol. Mag., Vol, VIII,
p. 438.

“ Nat. Hist. Rev.”=“ Natural History Review.”

“ Natural History Review,” London, 1861-65.

“Nature,” London.

Nicholson ; “ Manual of Palaeontology,” Edinburgh and London, eds. 1872-79.

“ Notes on the Fossil Mammalian Faunae of India and Burma,” Lydekker. R. G. S. I., Vol. IX, p. 86.

“ Notices of new and other Vertebrata from Indian Tertiary and Secondary rocks,” Lydekker. R. G. S. I., Vol. X,

p. 30.

“ Notices of Siwalik Mammals,” Lydekker, R. G. S. I., Vol. XI, p. 64.

“ Nouv. Mem. Soc. Hel.”=“ Nouveaux Memoires de la Societe Helvetique.”

“ Nouveaux Memoires de la Societe, Helvetique,” Zurich.

“ Odontography,” Owen, London, 1840-46.

“ On the Extinct Species of Rhinoceridae of North America, and their Allies.” E. D. Cope. Bui. U. S. Geol. Geog.
Surv., Vol. V, p. 227.

“ Ossemens Fossiles”=“ Recherches sur les Ossemens Fossiles,” q. v.

“ Osteographie”=“ Osteographie des Mammiferes,” q.v.

“ Osteographie des Mammiferes,” Blainville, Paris, 1839-64.

Owen, R. : “ Anatomy of Vertebrates,” London, 1866-68.

' “Fossil remains of Mammals found in China.” Q. J. G. S. L., Vol. XXVI, p. 417.

“ History of British Fossil Mammals and Birds,” London, 1846.

“ Odontography,” London, 1840-46.

' “ Palaeontology,” 2nd ed., Edinburgh, 1861 .

“ Palaeontographica,” Cassel.

“ Palaeontological Memoirs”=“ Palaeontological Memoirs and Notes of the late Hugh Falconer,” q.v.

“ Palseontological Memoirs and Notes of the late Hugh Falconer,” Chas. Murchison, London, 1868.

“ Paleontology,” Owen, 2nd ed., Edinburgh, 1861.

“ Paleontology, Manual of” — Nicholson, London and Edinburgh, eds. 1872-79.

“ Paleontologie et Zoologie”=“ Zoologie et Paleontologie Generales,” q.v.

“ Paleontologie Suisse”=“ Materiaux pour la Paleontologie Suisse,” q.v.

“ Pal. Mem.’ — “ Paleontological Memoirs and Notes of the late Hugh Falconer,” q.v.

Pictet : “ Materiaux pour la Paleontologie Suisse,” Geneva, 1854-73.

“ Prehistoric Times, etc.” Lubbock, 2nd ed. London, 1869.

“ Proceedings of Zoological Society of London,” London.

“P. Z. S.”— ‘Proceedings of Zoological Society of London.”

“ Quarterly Journal of the Geological Society of London,” London.

“ Recherches sur les Ossemens Fossiles,” Cuvier, Paris, ed. 1836.

“ Records”=“ Records of the Geological Survey of India.”

“ Rec. Geol. Surv. Ind.=” “ Records of the Geological Survey of India.”

“ Records of the Geological Survey of India,” Calcutta.

Royle ; “ Illustrations of the Botany and other branches of the Natural History of the Himalaya Mountains,”
London, 1839.

LIST OF ABBREVIATIONS, &c.

XXVll

Eut.=“ Riitimeyer,

Rutimeyer : “ Die Fauna der Pfahlbauten in der Schweiz.'’ Zurich, 1862. Nov. Mem. Soc. Hel., Vol. XIX.

——“Die Rinder der Tertiar-Epoche, etc.” Ahhand. der. Schweiz, pal. Gesell., Vols. IV, V.

“ Studien iiber das Genus Mastodon,” Meyer, Paleeontographica, Vol. XVII.

“ Systematic Position of SivatJierium giganteum of Falconer and Cautley,” Murie. Geol. Mag., Vol. VIII, p. 438.
“Tertiary and Alluvial Deposits of the Nerbudda Valley,” Theobald. Mem. Geol. Surv. Ind., Vol. II, p. 279.
Theobald : “ On the Tertiary and Alluvial Deposits of the Nerbudda Valley.” Mem. Geol. Surv. Ind., Vol. II, p. 279.

Transactions of the Geological Society of London,” London.

“ Transactions of the Zoological Society of London,” London.

Trans. Geol, Soc. Lon.”= (Transactions of the Geological Society of London.)

Turner: “ On the characters of the Basis-cranii of Ruminants, as affording grounds for classification.” Proc. Zool.
Soc., Lon.,” 1848-49-50.

Uber Osteireichische Mastodonten und ihre Beziehungen zu den Mastodonarten Europas,” Vacek, Ahhand. der
k. k. Geol. Reich., Vol. VII, pt. IV.

“ Ueber die fossilen Reste von Wirbelthieren, welche die Herren von Schlagintweit von ihren Reisen in Indien und
Hoch Asien mitgebracht haben.” H. Meyer. “ Palseontographica,” Vol. XV.

Vacek : “ Uber Osteireichische Mastodonten und ihre Beziehungen zu den Mastodonarten Europas. Ahhand. der
k. k. Geol. Reich., Vol. VII, pt. IV.

Wallace: “ Geographical Distribution of Animals,” London, 1876.

Zoologie et Paleoutologie Generales,” Gervais, Paris, 1867-69.

CORRIGENDA AND ADDENDA.

Page 4, second line of note, for “ 1864 ” read “ 1865.”

„ 5, line 12 from top, for “ Davis ” read “ Davies.”

„ 15, „ 23 „ bottom, for “ Palceindicus ” read “ Palmindicus.”

„ 22, „ 9 „ top, for “ 2nd Ser., Vol. Ill, ” read “ 1863.”

„ „ „ 4 „ bottom, for “ fig. 4 ” read “ fig. 1,” and on next page.

„ 23, „ 4 „ top,for“Plate70, fig. 6, ’’read “Plate LXXIV, fig. 1.”

„ 24, „ 12 „ bottom, for “ least ” read “ last.”

„ 26, „ 11 „ top, for “ figs. 2 and 5” read “ fig. 5.”

„ 31, „ 13 „ bottom, after the word mandible add “ (Plate VI, fig. 4).”

„ 32, „ 10 „ top, for “ molar ” read “ premolar.”

„ 36, „ 1 „ „ for “bysodont,” read “hypsodont.”

„ 40, „ 10 „ bottom, for “ conven ” read “ convex.”

„ 41, „ 7 „ „ for “ Plate 2 ” read “ Plate 5.”

„ 43, ., 7 „ top, for “ swmatranus ” read “ sumatrensis."

„ 46, „ 19 „ „ for “ fig. 23 ” read “ fig. 26.”

„ 49, „ 20 „ „ for “ anis ” read “ axis.”

„ 51, „ 17 „ bottom, for “ valley ” read “ valleys.”

„ 52, „ 1 „ top, for “ molar ” read “ premolar.”

„ 54, „ 1 „ bottom, for “ bysodont ” read “ hypsodont.”

„ 55, „ 2 „ top, for “ bysodont ” read “ brachydont.”

„ 58, „ 11 „ „ for “ as ” read “ or.”

„ „ „ 17 „ bottom, for “became” read “become.”

„ 60, ,, 8 „ top, for “ bysodont ” read “ hypsodont.”

„ 61, after line 11 from top, add “ Genus Camelus.”

„ 63, line 19 from top, for “ elegans ” read “ naui.”

„ ,, „ 3 „ bottom, for “ casps ” read “ cusps.”

„ 65, „ 19 „ top, for “ bysodont ” read “ hypsodont.”

„ 69, „ 16 „ „ for “ elephas ” read “ elaphus."

,, 70, after line 17 from top, add “ Division Perissodactyla” (P)

„ 72, line 3 from top, for “ Selonadont ” read “ Selenodont.”

„ „ after line 11 from top, add “ Order Proboscidia.”

„ 76, above first line from top, add “ Order Ungueata. Division Artiodactyla.’
„ 77, line 15 from bottom, for “ Acotherium ” read “ Acotherulum."

„ 78, „ 8 „ „ for “ memoirs ” read “ memoir.”

„ 84, „ 8 „ top, for “ Pietet ” read “ Pictet.”

„ „ lines 19 and 31, from top, for “ Plate IV ” read “ Plate VII.”

„ 85, line 18 from bottom, for “ Sihursi ” read “ Subursi.”

„ „ „ 12 „ „ ior “palcendicus” iea.di“ palceindicus. ”

„ 86, „ 4 „ iog, £ov “ palcendicus” veaLdi“ palceindicus.’'

„ „ „ 7 „ „ ior “ kelveticas ” rea.i “ helveticus.”

„ 88, lines 4 & 5 from bottom, for (Plates G and S) read [Plates R and, I).

„ 92. This list of Ruminants will be found amplified on page 180.

„ 93, line 6 from bottom, for “just” read “ first.”

XXX

CORRIGENDA AND ADDENDA.

Page 94, line 20 from top, for “ Mammalia” read “Mammalian.”

„ 95, „ 14 „ „ transpose “ Section Selenodonta ” and “ Division Artiodactjla.”

„ 99, „ 1 „ „ for “vertebrate read “ Vertebrata.”

„ 102, „ 9 „ bottom, for “Perim-ganga,” read “ Pem-ganga.”

„ 114, „ 3 „ „ for “ the following page ” read “ page 66.”

„ 117, „ 1 „ „ for “ those ” read “ that.”

„ 126, „ 1 „ top, for “ Mammals of Nepal ” read “ Genera of Indian Cattle.”

„ „ „ 2 „ „ for “ urus ’’ read “priscus.”

„ 139, „ 1 „ „ for “ Perim-ganga” read “ Pem-ganga.”

„ 141. The species here described as Perihos occipitalis, on page 174 is renamed, Hemihos occipitalis.

„ 145. The skull here described under the name of Hemihos triquetriceros, on page 177 is renamed
Hemihos acuticornis,

„ 148, line 20 from top, for “ fig. 6 ” read “ fig. 2.”

,, 150, Of the two skulls described as Amphihos acuticornis, one (Plate XXI, fig. 1) is renamed on page 178
Hemihos acuticornis, and the other H. occipitalis.

„ 157, line 9 from top, after the word “ species” add a colon.

„ 161, Note, for “ page 44 ” read “ page 438.”

„ 167, line 6 from bottom, for “Severtgoff ” read “ Severtzoff,”

„ 173. Above “ Bubalus platyceeos, ” add “ Genus Bubalus.”

„ 193, line 11 from top, for “ left” read “ right.”

„ 258, „ 11 „ „ „ “ left ” read “ right.”

if. B.—The reader is requested to make the above corrections with pen and ink.

DIRECTIONS TO THE BINDER.

In fasciculus 1 (“ Rhinoceros deccanensis’^) substitute descriptions of Plates I, II
III, issued with fifth fasciculus (Prohoscidia) for the one sheet in the firs,
fasciculus ; destroy title page and preliminary notice.

In fasciculus 2 (“Molar Teeth and other Remains of Mammalia”), destroy
title page and preface, corrigenda and addenda, and appendix issued subsequently ;
also the page entitled “ Index to plates.” Eor Plate X in that fasciculus substitute
the corresponding plate issued with the fifth fasciculus, also the descriptions of
Plates IV, Y, and Vl.

In fasciculus 3 (“Crania of Ruminants”), destroy title page and preface,
substitute for the original descriptions of Plates XX to XXIY the new descriptions
issued with the fourth fasciculus, and intercalate Plates XXI A and B and XXIII A
(issued with the same fasciculus) in their proper serial place; substitute reissue
of Plates XXI and XXIV for original ones.

Bind title-page to volume, contents, list of plates, etc., at commencement, and
index and all the plates at end of volume.

R. L.

FAUNA

OF THE

INDIAN FLUVIATILE DEPOSITS.

RHINOCEROS DECCANENSIS,

A new Species discovered near Gokak, Belg-aum District, by R. B. FOOTE, F.G.S.,
Geological Survey of India.

In May 1871, while looking for sections in the hed of a small nullah, I came
upon some fragments of fossil bones and teeth, and amongst them part of an upper
molar of a Rhinoceros. I at once set to tracing out the source whence these
bones might have been derived, and, after closely examining the banks of the
nullah for a few dozen yards further up its course, found a row of large mammalian
teeth exposed at a height of 3 feet above the bed of the nullah and fully 8 feet or
more below the top of the bed of black clay which here forms the bank, and which
black clay passes up into the typical regur of this neighbourhood. The spot at
which this discovery was made lies about 3^ miles east north-east of Gokak (a
talook town in the Belgaum District, well-known from the proximity of the great
fall of the Gatparba River), and about f of a mile south south-east of the little
village of Chickdowlee, immediately west of which the small nullah falls into the
Gatparba River.

This small nullah* has cut deeply into and through the regur at that spot,
and has formed a small cliff on the face of which the row of teeth abovementioned
was exposed.

The rain-wash from the upper part of the little cliff had covered up every-
thing, the teeth excepted, but on removing it carefully, I found the teeth belonged
to the right ramus of the mandible of a Rhinoceros.

Beneath the rain- wash the black clay, though much broken up by sun-cracks,
was hard, and the angular fragments were so closely wedged together that it
required a good deal of time and trouble to remove those immediately surrounding
the bones without disturbing the latter, which were not only extremely brittle, but
also much comminuted in situ by the action of sun-cracks.

* The Chickdowlee nuUah is a small stream rising in the hiUs to the north-west of Buneechmurdee, and not as it
is represented on the map, the extension of the large nullah flowing past Kelvee, which reaUy falls into the great
Mumdapoor nuHah close to the viUage of Maldinee.

PAUNA OP THE INDIAN ELIJVIATILE DEPOSITS.

Had it been feasible I should very gladly have deferred the extraction of
these bones for a day or two to do it more leisurely than was then possible ; unfor-
tunately the weather was very unsettled, numerous heavy thunderstorms had taken
place within the previous week, and another was gathering at the time. Should
a flood come down the nullah the greater part of the bones laid bare would
certainly be swept away and lost. Then also the spot being between six and seven
miles from my camp, it was impossible to keep watch over the fossil remains thus
exposed, while the attention of several field labourers had been attracted by seeing
me busily removing the soil with my hands, and they in their cmaosity would, in
all probability, have utterly ruined this valuable specimen had I left it unguarded.
I decided, therefore, to take it at once as being the siu’est way of getting it as nearly
as possible entire.

A few inches below the surface the clay became damp, but was still extremely
tenacious, and it required great care and much patient labour to loosen the bones
without entirely crushing them.

The position occupied by the head was suggestive of its having been drifted into
its present resting-place, the heaviest part, the cranium, being undermost. The
head had, however, not been entirely overturned, but originally rested on the right
frontal bone and supraorbital ridge. The greater part of the right side of the head
had been broken away by flood action undermining the bank. Unfortunately most
of the bones thus detached and found loose in the nullah were too fragmentary to
be joined together.

The left maxilla and left ramus of the mandible were in perfect apposition
when freshly exposed, but the left side of the head had suffered severely prior
to its entombment, as the frontal and nasal bones were missing.

The position of the left ramus in apposition to the maxilla offers a strong
indication of the head, though much mutilated, not having been entirely deprived
of its external covering of flesh at the time it was buried in the black clay.

No indications of any bones but such as belonged to the head were met with in
situ, though the bank was excavated to some little depth after removing all that
remained of the cranium.

Among the bones found loose in the bed of the nullah only one or two frag-
ments appear to belong to the body of the animal, and they are somewhat doubtful.

The bones found imbedded were —

1. — The mandible, nearly perfect.

2. -— The left maxilla with jugal and lachrymal bones attached.

3. — The squamosal hone with meatus auditorius and post-tympanic process of the left side.

4. — Part of right frontal bone.

5. ' — Hyoid bones?

6. — Pterygoid bone (right side ?)

Lying loose in the sandy bed of the nullahs were parts of the right maxilla,
fragments of teeth, and two or three bones too fragmentary to be determined.

RHINOCEROS HECCANENSIS.

When found the bones were largely covered with minutely botryoidal calcare-
Bones very thickly encrusted with ous coucretious, forming an extremely hard and dense
two kinds of kunkur. crust irregularly distributed over the surface, in some

places in large and thick patches, at others in small wart-like excrescences. Erom
the brittle condition of the bones these were very difficult to remove without destroy-
ing the underlying surface of the bones. Many had to be gently rasped away — an
operation requiring much time and patience, because of the hardness of the material.

Besides this botryoidal encrustation the whole surface of nearly every bone
was covered by a thinner crust of a more earthy material less hard than the former.
The less dense parts of this crust assimilated to common earthy kunkur, but some
of the dense parts of it, which showed an arrangement in vermicular masses, though
less hard, were extremely tough and fully as difficult to remove as the other form
of encrustation. The removal of these crusts was, however, quite essential, as they
completely hid and altered the true shape of many parts of the bones. The encrusta-
tion was by no means confined to the surface — it had penetrated most of the
numerous fissures due to the expansion and contraction of the surrounding clay.
These intrusions of kunkur had in many cases, particularly where they affected
the teeth, given rise to total deformity of the parts by wedging them asunder,
sometimes to the extent of half an inch or more. Part of the surface of many of
the teeth has been eaten into, as if by a kanker, by the calcareous crust resting on
it, forming small shallow pits on the surface of the enamel.

The degree of alteration the bones have undergone is very various in different
parts ; especially is this the case with regard to the teeth, in many of which part
of the enamel is perfectly preserved, while closely adjacent parts have been greatly
changed and have lost all lustre and become quite mealy in texture. Excepting
external discoloration from contact with surrounding soil, the bones are but
slightly altered from their natural color, though a good deal of calcareous matter
became infiltrated in their cancellar tissue. The enamel of the upper molars is
rather browner in color than that of the lower ones. The dentine of all the broken
teeth is a good deal stained along the minute capillary tubules of a deep black,
apparently due to the presence of oxide of manganese. The cavities in the fangs of
the teeth left by the decay of the pulp are mostly lined with acicular crystals of
arragonite. The dentine was found to be traversed by innumerable minute cracks,
rendering the mass extremely brittle, especially near the base of the crowns and in
the fangs. Owing to this many of the teeth fell to pieces, and had to be built up by
fitting piece to piece— a very long and tedious task from the great number of tiny
fragments that had to be dealt with. But for the fact that the manganese stains of
the dentine tubules above alluded to often formed patterns on the broken surfaces,
the building up process of the dentinous parts would have been simply impossible.

Some change in the form of some of the bones has been caused by the pres-
sure of the mass they were imbedded in. The parts principally affected by such
pressure are the mandibles, the jugal arch, and the supradental portion of the left

4 EAUNA OE THE INDIAN ELHVIATILE DEPOSITS.

maxilla. The mandible is materially distorted, the right ramus especially having
been forced over to the left side very considerably, and the symphysis having been
twisted so that the left side of the symphysial prolongation is fully half an inch
higher in level than the right. The right ramus is also rather broken at the lower
edge about the middle of its length.

The upper and outer part of the left maxilla above the molar series is much
crushed out of shape, so much so that the jugal arch, instead of being parallel to
the side of the skull, has been turned over outwards, so that at present it has a
position nearly at right angles to its normal one.

The plane of the dental series of the maxilla has also been considerably more
curved than normal, the result being that the molars and premolars, instead of being
in close apposition to each other as in all other species of Ehinoceros, are divided by
spaces. This is especially observable between premolars 2 and 3 and molars 1 and
2 and 2 and 3. The bones appear to have been in a rather soft state when thus
affected, else they must have been far more extensively fractm’ed.

On comparison with all the other described species of Ehinoceros, both living
and fossil, the head discovered at Chickdowlee shows such marked differences that
it cannot be assigned to any one of them, and deserves, therefore, to be considered
as a distinct and hitherto undescribed species. As such I propose to call it Rli.
-Deccanensis ; and as the region in which it was found belongs distinctly to the
Deccan in the older and fuller meaning of the name, and most of the other Asiatic
Ehinoceroses, both recent and fossil, have been distinguished by geographical
specific names, the one now proposed appears quite suitable.

The head only of Hli. Deccanensis being known, comparisons could only be
instituted with corresponding parts of specimens of other species. Eight distinct
points of character came specially under comparison, and they were in order of
importance —

1. — The proportional height of the crowns of the teeth.

3. — The form of the symphysis of the mandible.

3. — The presence or absence of incisor teeth and their size.

4. — The special structure of the upper molar series.

5. — The form of the bones of the periotic region.

6. — The form and proportions of the zygomatic arch.

7 • — The relative size as compared with that of other species.

8. — The deciduous character of premolar 1.

In carrying out the comparison of the remains of this Ehinoceros with those
of other species, I have followed the methods adopted by the late Dr. Ealconer and
by Mr. Boyd Dawkins, e. r. s.,* and for the descriptive portion and plates have
adopted the terms (with two exceptions) and indicative letters employed by the
latter palaeontologist in his several very able papers on the dentition of the Ehino-
cerotes found fossil in Great Britain.

* See Falconer’s Palaeontological Memoirs and Mr. Boyd Dawkins’ papers in the Natural History Review, 1863
and 1864, and Quarterly Journal of the Geological Society, Vol. XXIII, 1867, and Vol, XXIV, 1868.

EHINOCEROS DECCANENSIS.

With regard to living species I have followed the latest enumeration given by
Dr. J. E. Grey in the “Annals and Magazine of Natural History”^ in continuation
of the list published in his catalogue of the Pachydermata in the British Museum.

Unfortunately some species included among those are founded on external
characters only, e. g., Bh. Oswellii and Bh. CrossU, and with such, of course, com-
parisons were impossible.

My specimen was compared with all known published descriptions and with the
fossil and recent specimens in the British Museum, the Boyal College of Surgeons,
London, the Jardin des Plantes, Paris, the Museum of the K. K. Geol. Beichs
Anstalt, Vienna, the Imperial Museum, Calcutta, the Madras Museum, and last,
but not least, the Geological Museum, Calcutta. And here I would express my
sincere thanks to Mr. Henry Woodward, v. r. s., to Mr. Davis of the Palaeonto-
logical Department of the British Museum, and to Professor A. Gaudryof the Jardin
des Plantes, for the true courtesy with which they gave me every assistance in
their power in carrying out such comparison with the specimens in their custody.

The head of Bh. Beccanensis indicates a smaller and slighter animal than
Bh. Indicus, but one larger in all probability than any of the other living Asiatic
species.

The head is that of a young adult animal whose permanent dentition is re-

Dentai formula presented by the formula H ? C|, P. M. Hi, M.

The teeth, which were not furnished with a cement
layer, are not much worn down by use; indeed, the last molars in each jaw had
only just begun to show signs of wear. The animal belonged very markedly to
the hypsodont section of the family.

The incisors are wanting in the mandible, and from the rather broken con-
dition of the incisive border of the symphysial portion, it is difdcult to be quite
positive whether alveoli had ever been developed there or not. The incisors, if
developed there, were extremely small and quite rudimentary.

The premaxillary bones are unfortunately wanting; hence the presence or
absence of incisors in the upper jaw cannot be determined, but the probability is
that they were extremely small or wanting. Two fragments of bone were found
loose, which present some resemblance to the anterior extremities of pre-maxillary
bones in other species, and, if they should really be such, their appearance certainly
disfavors the idea that the animal possessed upper incisors.

Professor Owen has pointed out that the development of the horns of the
Rhinocerotes is in the inverse proportion to the magnitude of the incisors. If this
law held gOod in Bh. JDeccanensis, it must have had a very large horn or pair of
horns. t Unfortunately, however, the nasal bones were not found ; so, this point
remains for the present undecided.

* Annals and Magazine of Natural History, 4th Series, Vol. XI, p. 356.
t Owen’s Comparative Anatomy of Vertehi’ata, Vol. Ill, p. 356.

PAUNA OE THE INDIAN ELUVIATILE DEPOSITS.

The mandible.

The anterior or symphysial part of the mandible is prolonged beyond the
premolars in a narrow beak-like projection unlike in
form or proportion to any species hitherto described
and figured. A correct idea of its form will be obtained by referring to figures 3
and 4, Plate II. The peculiarity of this symphysial prolongation consists in its
narrowness and in the sudden diminution in width in front of the permanent pre-
molars. Bh. Etruscus, Ealconer, which of fossil species most resembles it, shows
a less sudden constriction, if the term may be used, of the extension of the sym-
physis in front of premolar 2, and the extension is moreover wider in proportion. In
Bh. mger. Gray, the symphysis is narrow, but longer than in the Deccan species,
and ends in a sharp toothless incisive border. Many species of Rhinoceros, both
recent and fossil, ^show this extension of the symphysis, but instead of its being
narrow throughout its length and constricted immediately in front of the premo-
lars, it is wide at first, then narrows, and before reaching the incisive border spreads
out again in a spatulate form. In many species large or moderate sized incisors
project from the anterior border. In Bh. JDeccanensis the rather broken condition
of the incisive border («. b. fig. 4, PI. II) renders it a
The mclsois. little doubtful whether incisors had been developed there

or not. If the small and irregular cavities observable in that situation were true
alveoli, and probably they were so, the incisors must have been very small and
rudimentary, as has already been mentioned above. There is no trace of them
left, so they had probably been shed while the animal was alive.

There are no indications of large mentary foramina Kke those so strongly
developed in Bh. Etruscus.

As already mentioned {ante, p. 6) the symphysis has been somewhat twisted
by pressure, and the right ramus has been forced over considerably towards the
left one, and the central part of the lower edge rather broken. The left ramus
appears to be unaffected by the pressure, but has lost the coronoid process. The
condyle is relatively small, the transverse length of the
articular surface being only 3 inches, while the same part
in an otherwise much smaller mandible of Bh. Sondaicus* measm’ed 3^ inches across.
The curve of the ascending part of the ramus commences just below the median
groove of premolar 3, The height of the ascending part of the ramus is 9T5 inches
to the summit of the condyle, measured vertically from the smTace the mandible
rested on, or 11’75 inches, measured with a tape along the posterior edge from the
condyle to the angle. Erom the angle to the incisive border, measured along the
under side of the ramus, is a distance of 19 inches. The left ramus only was
measured, being much the more perfectly preserved. The posterior edge of the
ascending ramus rises almost straight, and is not notched below the condyle as in
Bh. Sondaicus and various other species.

The ascending portion of the ramus.

* In the collection of the Imperial Museum, Calcutta.

EHINOCEEOS DECCANENSIS.

The lower molar series is represented by six normally-shaped tall crowned teeth —
three premolars and three molars. Of these premolar
The lower molar series, 2, on either side, is perfect, except that the posterior

wall is rather broken.

Premolar 3 is imperfect on both sides, the inner being broken away. Premolar 4
is altogether wanting in the right ramus and wants the inner wall in the left ramus.

Of the true molars in the right ramus, molar 1 is rather imperfect ; molar 2 and
molar 3 are perfect in the crowns; these are figured in Plate III, figs. 2 and 3.
It will be observed that the posterior collis of premolar 3 had only lately begun to
come into wear. The corresponding teeth in the left ramus are less well preserved.
The lower molars of Rh. Deccanensis are not specialized, and offer no strikingly
characteristic differences from many other species. The guard is but sMghtly deve-
loped on the anterior and posterior walls, and does not show on either the inner or
outer walls in the lower molars. Molar 1 of the left side has been forced upwards
and backwards by an intrusion of the encrusting matter, as shown in fig. 3, Plate II.

The annexed measurements of the lower molars may be interesting for pur-
poses of comparison — they are as exact as the imperfect state of the specimen
admitted of their being. They were made in the following directions, after
Mr. Boyd Dawkins’s system, at the base of the crowns : —

1. — Antero posterior, along outside of crown.

2. — Antero transverse, across anterior collis.

3. - — Postero transverse, across posterior collis.

Tooth.

Side.

Premolar 2

Eight

1" 3"'

61'"

Ditto 3

Left

1" 3'"

Ditto 4

do.

1" 7'"

Molar 1

Eight

1" 8"'

i"2r

1" 4'"

Ditto 2

do.

I'll'"

1"3"

1" 3i"

Ditto 3

do.

2" 1"'

1"3'"

1" 2|"

The length of either row of lower molars on the right side is 11|- inches,
measured along the outer bases of the crowns.

Before passing on to the description of the much more complex upper molar
series, it will be better to give a key to the indicative letters used in the plates, and
which, as already mentioned, agree with those employed by Mr. Boyd Dawkins
in his several memoirs. As both upper and lower molars are recognized to be formed
on the same plan, though differing very greatly in the degree to which that plan was
developed, the letters apply to the homologous parts in both series. The teeth are
compared to a hill sub-divided by two valleys running down from the main ridge or
outer wall of the tooth. Besides the subordinate hills thus formed, there are
certain processes jutting from the walls of these hills into the area of the valleys,
certain prominent ridges on the outer wall, certain remarkable ledges of
enamel running round the sides of the walls, and certain grooves dividing the
outer wall into areas, which all have to be accounted for, as their form, or presence,
or absence are of great and often specific import. The comparison to a hill only

PAUNA OP THE INDIAN PLHVIATILE DEPOSITS.

holds good to a certain extent, but it would be very difficult to find any natural
object with which abetter comparison could be instituted, and it has been adopted in
part by so many eminent palseontologists that it is better to carry it on than introduce
another, though a much better one might be found if a fortification, such as a small
Indian fort or medigeval castle, were the object adopted for the comparison. The follow-
ing list shows the principal parts of the teeth and the letters they are indicated by

*a. Anterior valley .+

*b. Posterior valley. |1
*c. Accessory valley.

*d. Anterior collis.§

*e. Median collis.^

*f. Posterior collis.
ff. Anterior process=crista. •»

hi . Median process. ?• “ Combing plates ” of Boyd Dawkins.

712. Posterior process=crochet=uncus. ^
i. Median groove on outer wall of tooth.

Costse on the outer wall.

1. Outer wall (divided by “ i” into “ m” and “ n”). “ External lamina ” of Boyd Dawkins .f
m. Anterior area.

n. Posterior area.

0. Guard=cingulum=bourrelet=wulst.
p. Opening to anterior valley or “ pass.”

The left maxiUa was found in situ with its six teeth, and is figured in Plate I.

Of the right maxilla only fragments were found loose
in the bed of the nullah, and from these the teeth were

The maxilla and upper molar series.

missing, but fragments of some teeth were also found loose. The upper molars were
unfortunately more affected by decay (whatever may have been the cause), and
more penetrated by the veins of encrusting matter, while the great depth of the
valleys rendered them more fragile than the lower molars. Of the teeth in the
left maxilla p^'emolar 2 and molar 2 are much broken. Molar I and molar 3 are also
considerably damaged : still enough remains to give a fair idea of this most charac-
teristic part of the whole dentition. In both premolar 2 and 77iolar 2 the outer
wall is wanting.

The premolars of Bh. Deccanensis are specialized by the very great development
of the ledge of enamel, known as the cingulum or guard ( Boyd Dawkins ), which
occurs mostly on the anterior and inner walls of the teeth, and which is well shown
in both figs. I and 2 of Plate I. I think I am right in saying that in no other
species is this peculiar appendage of the premolars so strongly developed, and in this
respect it resembles some of the European miocene species. Premolar I was deciduous
and probably a very small tooth, as no sign of it can be made out on the edge of the
maxilla. In premolar 2 a small pit,* fig. 1, Plate I, shows in the enamel ledge or

* The equivalent terms in Latin are employed by Brandt {Bh, tichorldnus, Mem., Acad. St. Peter, 6e ser., tom. VII.)

J Vallon oblique, in upper molars, Cuvier.

11 Ecorchure au bord posterieur, Cuvier ; Fossette posterieur, Blainville.

§ Colline seconde, of upper molars, Cuvier.

^ “ La troisieme colline,” of upper molar, “ le bord posterieur de la dent.” Cuvier.

t Collis externus, Brandt. “ Colline pi’cmiere qui suit exactement le bord.” Cuvier.

EIIINOCEROS DECCANENSIS.

guard on the anterior wall of the tooth; this is the only part of the building up of the
teeth that I feel a shade of doubt about, because such a pit does not seem to be known
in other species; the fragmentary parts, however, fitted most perfectly together. The
corresponding portion of the right jaw premolar, is unfortunately unknown. This tooth
is a good deal more worn than premolar 3 and 4, and the pass p, leading from the inner
wall of the tooth to the anterior valley, is almost obliterated by wear. The edge of
posterior wall of the tooth has also been so much worn that the posterior collis
is separated from the posterior valley by a belt-like surface of dentine. In premolar 3,
on the contrary, much less wear has taken place, and the three colles stand up dis-
tinctly. The guard commences on the anterior wall, at about one-third of the length
of the wall from the anterior angle of the outer waU, and runs all round the inner
wall till it merges in the posterior collis. The pass p dividing the anterior and
median colles is deep and sharply defined. The external side of the anterior valley
(^. e., the side next the exterior wall of the tooth) is rather broken round the top.
A spur of enamel projects from the median collis very nearly across the anterior
valley, and shows that a very strongly marked crochet characterized this tooth.
The posterior valley is rather oval in shape, imperfectly so however, as the curve
on the posterior side of the major axis is much greater than that on the an-
terior side of the axis, which is parallel with the axis of the median colKs. The
posterior waU of the tooth descends but little from the posterior collis, and thus
shows signs of wear, and also gives the posterior valley a decidedly pit-Mke appear-
ance quite different from the bay-like appearance it presents in many other spe-
cies, owing to the posterior waU of the tooth being deeply notched by the posterior
vaUey. This character belongs also to the posterior valleys of premolar 2 and 4. The
outer wall of premolar 3 is characterized by the anterior angle or costa ki, forming an
acute angle ; k2 the second costa is well developed, but the whole wall of the tooth
is remarkable for its flatness.

Premolar 4 is, on the whole, very little different from premolar 3, though of
considerably larger size. The chief difference lies in the larger proportional size of the
anterior colUs, which is taller and more bulging a little below the present surface of
mastication. Premolar 4 is perhaps a little less worn down. The anterior angle of
premolar 4 is also rather more acute. The appearance of greater height in the anterior
collis of premolar 4 as compared with premolar 3 is in measure due to the lower posi-
tion occupied by the guard on the anterior half of the inner wall of the tooth. The
central part of the masticatory surface is rather broken, but there is a well marked
fold of enamel projecting from the median coUis, showing that a large crochet
would be present were the tooth unbroken. The outer wall of the tooth is, like that
of premolar 3, remarkable for its flatness, which greatly exceeds that of aU other
species I have been able to compare it with. This character will be apparent from a
comparison of flg. 1 in Plates I and II.

The true molars, as already stated, are unfortunately less well preserved, but
enough remains to recognize many of their chief characteristics. The true molars

PAUNA OP THE INDIAN PLHVIATILE DEPOSITS.

will be seen not to show the great development of guard wMch is so conspicuous
in tbe premolars. The guards in both molars is confined to the anterior wall of
the tooth, and there it occupies a much smaller space, extending less than half the
length of the wall from the inner anterior angle. Moreover, in molar 1 its position
is very different, as it lies very high, nearly level indeed with the crown surface of
the anterior collis, upon which it rather eneroaches and makes a small shelf, harp-
shaped in plan, instead of a jutting ledge. In molar 2 the guard forms a wide ledge,
sloping upward from the inner anterior angle nearly to the middle of the anterior ;
unlike m. 1, however, the guard springs from a point about half way down the side of
the anterior collis. The anterior collis of the true molars is of much larger propor-
tions than in the premolars, as compared with the whole size of the teeth, being very
broad and stout, especially in molar 1, where it forms much more than half of the
inner side of the tooth. By this increase of size in the anterior collis the anterior
valley is greatly narrowed, and the pass no longer occupies a median position, and has
become very narrow and much deeper than in the premolars. The crochet }i%. is very
large, and all but touches the posterior wall of the anterior collis. The outer wall of
the anterior valley also forms a projecting fold of enamel {Ji\, fig. 1, Plate I), which
projects forward and inward. The posterior valley is unfortunately wanting, that
part of the tooth being broken away, but judging from the fragment of molar 1 of
the right jaw, which was found loose in the nullah-bed, the posterior valley most likely
resembled that of premolar 4. This fragment, which is figured in Plate II, fig. 2,
shows a saddle-like slope descending from the posterior collis /into the posterior valley.
On the outer wall of the tooth it will be seen that the anterior angle is less promi-
nent than in premolar 4 ; the hinder part of the outer wall is much broken, but not
too much to show the peculiar flatness described as characteristic of the premolars.
The figure of the fragment of molar 1 of the right jaw above referred, was given to
show the great depth of the anterior valley, which, as before remarked, is one of the
special characters of the dentition of Deccanensis.

Molar 2, the largest of all the series, is unfortunately the least perfect, the an-
terior collis only remaining in tolerable preservation. The general character of this
tooth can, however, be traced in plan. The form of the guard has been referred to
already in the description of molar 1. The anterior collis {d) differs less from the ana-
logous parts in premolar 3 and 4 than does that of molar 1. It is less massive in form,
and does not encroach so much on the anterior valley. The median collis is, on the
contrary, actually and proportionately stouter than in molar 1. The pass {p) is a shade
less narrow, but would appear to have been quite as deep proportionately in the
unbroken tooth. The two crochets h\ and hi were, judging by the remaining lower
parts (regarding the tooth as in an inverted position,) much stouter than in molar 1,
but the angle at which they rise seems to indicate that they projected less into the area
of the valley.

The posterior valley is smaller at the same depth than that of premolar 4, and
forms a long ellipse in plan ; the major axis of the ellipse being nearly parallel to

EHINOCEEOS DECCANENSIS.

the Kne of the outer wall of the tooth, and far from parallel to the axis of the
median collis (e). The outer wall is entirely wanting, and so in fact is the masti-
catory surface of the entire tooth, that of the anterior collis only excepted.

Molar 3, though unhappily also much damaged, is more perfect than molar 2,
and exhibits the trihedral form seen in most of the Rhinocerotes. It had under-
gone hut little wear at the time when its owner was entombed. This is proved by
the great height of the anterior collis.

The guard occurs only on the anterior walls very much below the crown of the
collis. As in the other teeth the anterior vaUey is extremely deep, and was intruded
into a very stout and long crochet {hz ), the base of which is seen at some depth.
The posterior coUis (/) is represented by a little spur-like cusp (shown in figs. 1
and 2, Plate I) low down on the posterior angle of the tooth.

The teeth of both jaws are furnished with long fangs, but they were not
exposed sufficiently in any case to observe any peculiarities they may possibly
possess.

The thickness of the enamel varies greatly in different parts of the teeth,
being thickest in the walls of the median collis and thinnest in the walls of the
crochets and in the exterior walls of the anterior valleys. In the lower molars it
was thickest in the outer wall, and thinnest in the walls of the anterior collis.

The length of the upper molar series measured along the edge of the crowns
from the posterior angle of molar 3 to the anterior angle of premolar 2 is 10‘9 inches.
As before mentioned (page 6), the supradental part of the left maxillary bone
is much affected by crushing, so much so that when the malar bone is placed
in apposition its plane lies almost at a right angle to its normal position. The
distortion is greatest above the true molars. On account of this distortion the
zygomatic arch has not been figured. It forms a strong broad band connecting
the maxilla with the squamosal bone by a rather flat arch. The zygomatic process
is unfortunately rather broken at its base, and the connection of the zygoma with
the squamosal bone therefore incomplete.

The lachrymal bone shows a well marked post-orbital projection, and the malar
bone has a distinct protuberance on its upper edge opposite to the position the post-
orbital process would occupy if developed in this genus.

The squamosal bone is figm*ed in Plate III, fig. 1, in order to show the peculiar
form of the groove between the post-glenoid and post-tympanic processes in
which the meatus auditorius is situated, as the form of this part of the cranium in
Rh. Deccanensis differs very greatly from many other species, as will be shown
further on in enumerating the points of difference between this and other allied
species.

The area between the two processes is wide and shallow at its upper part,
rather deeper below the opening of the meatus ; in plan it is very nearly rhom-
boidal. The post-glenoid process terminates in a lobe about two-thirds of an inch
below its inferior junction with the post- tympanic. The latter process is greatly

FAUNA OF THE INDIAN FLUVIATILE DEPOSITS.

thickened and protuberant near its centre. The meatus occupies the lower half of
the area above defined, and is large.

The hinder edge of this mass of bone would appear to be part of the supra-
oecipital, possibly of the ex-occipital also, and if so, it proves that the occiput was
protuberant, not concave as in some species.

The other bones found imbedded were part of the right frontal bone, the
right pterygoid and two slim bones which lay between the rami of the mandible,
and have a resemblance to hyoid bones, but are not sufficiently well preserved to be
identified with certainty.

Rh. Reccanensis is distinguished specifically from all the brachydont miocene
species, and also from Bh. Etruscus, by its strongly marked hypsodont character
and by the non-persistency of the first premolar tooth, but it is allied to many of
them by the strong development of the guard in the molar series.

It is allied to the African forms of Ehinoceros by the rudimentary character
(or possible absence) of the incisors, but separated by the greatly elongated sym-
physis of the mandibles and the great development of the guard in the premolars.

There remain then the three Em’opean pleistocene species — megarMnus, Jiemi-
tceclms, and tichorhinus — one hypsodont miocene form from Pikermi in Greece, the
pliocene American form Bh. crassus, Leidy, and the living and fossil Asiatic
species with which to compare it.

Mr. Boyd Dawkins, r. n. s., in his very interesting paper on Bh. Etrnscm,*
when speaking of the division of the Bhinocerotes into two classes by the relative
heights of the unworn crowns of their teeth, reckons all the known living species
to the hypsodont division, as also all the Asiatic fossil species. It appears to me,
however, that three of the more recently established living Asiatic species — Bh.
Floweri, Gray, Bh. stenocephaUis, Gray, and Bh. {ceratorhinus) niger. Gray — show
such low crowned teeth that they approach more closely to the brachydont type,
and that the conclusion that this type had ceased must be modified.

Taking the different species to be compared seriatim, we find that the Deccan
species differs from Bh. megarhinus by the narrowness
of the extended symphysis, which is broad and spatu-
late in the latter; by the great development of the guard, which is slight in
megarhinus ; by the greater development of ( ^2 ) the second costa on the outer
wall; by the different form of the posterior valley, and by the absence of the
deep notch on the posterior edge of the rami of the mandible immediately below
the condyle.

From Bh. hemUceehus Bh. Reccanensis differs by the absence of the thick
layer of cement found in the molars of the former;
also by the different character of the molar series, for
in Bh. hemitcechus the anterior and median colles are very narrow and compressed,
and the posterior collis very low and small. In Reccanensis, on the contrary, the

* Quarterly Journal, Geol. Soc., Vol. XXIV, 1868, p. 214.

Eh. megarhinus.

Eh. hemitffichus.

HHINOCEEOS DECCANENSIS.

Eh. from Pikermi.

Eh. crassus.

anterior and median coUes are stout and broad, and the posterior collis, though
small, attains the same level as the median.

From Rh. tichorhinus we must separate Reccanensis^ because of its not
possessing the thick layer of cement on the molars ;
because of the absence of the accessary valley “c,”
and because the grinding surfaces of the molars are not flat as in tichorhinus^ but
deeply excavated.

From the hypsodont miocene species from Pikermi, near Athens, described,
but not named, by Professor A. Gaudry in his splendid
work — “Animaux Fossiles et Geologic de I’Attique” —
Rh. Reccanensis differs by its greatly smaller size ; by the position of the guard,
which is much higher up the side of the tooth than it is in the Pikermi species.
In the latter the posterior wall of the tooth in premolar 3, molar 1, and molar 2 is
deeply notched by the posterior valley, which is not the case in Rh. Reccanensis.

The pliocene American species, Rh. crassus^ Leidy, possesses large incisors
in the lower jaw, with a broad spatulate symphysis
strongly resembling Rh. Indicus, while the upper molar
series is characterized by the presence of four valleys (anterior, two median, and
posterior). Its specific diversity from Rh. Reccanensis is, therefore, abundantly
clear.

The rudimentary character (or possible absence) of the incisors at once sepa-
rates Rh. Reccanensis from Rh. Indicus^ Sondaieus^ Sumatranus, nasalis, Gray, and
stemcephaluSi Gray, but there are other distinctions also which will be pointed
out separately. If I am right in my conclusion that the brachydont type is
not yet extinct, Rh. Reccanensis would on that ground alone be separated from
the remaining living Asiatic species (of which the bones are known), namely, Rh.
Flowerii Gray, and Rh. {ceratorhinus) niger. Gray, as also from the fossil species
Rh. Sinensis^ Owen ; but there are other distinctions also which require their specific
separation.

Taking each species by itself, it will be seen that Rh. Reccanensis differs from
Rh. Indicus in having only rudimentary (or no) incisors,
instead of extremely large ones; also by the greater
development of the guard, by the much greater relative depth of the valleys, and
the much greater flatness of the outer walls of the upper molars. The rugosities
at the angle of the mandible, so conspicuous in Rh. Indicus^ are hardly at all
developed in Rh. Reccanensis, which likewise has not the deep notch on the
posterior edge of each ramus below the condyle. The broad spatulate extension of
the symphysis is quite unlike the narrow beak-like form it assumes in Rh.
Reccanensis. Then the auditory fossa on the squamosal bone is quite unlike, being
broadly rhomboidal in shape in Reccanensis and much taller and narrower in
Rh. Indicus. In size Rh. Reccanensis was certainly quite one-fourth less than the
average Rh. Indicns, if the size of the head offers a sufficient datum to go upon in

Eh. Indicus.

FAUNA OF THE INDIAN FLUVIATILE DEPOSITS.

Ell. Sondaicus.

Eh. Sumatranus.

making sucli an estimate. Many other minor but well marked differences might
be adduced, but the above seem sufficient.

From nil. Sondaicus Bh. Bencanensis is separated by the character of its
incisors, which are very good-sized in the former. In
Sondaicus the guard is only very moderately developed
in the premolars, while the teeth are much less high crowned and show relatively
much shallower valleys and very tumid outer walls instead of nearly flat ones like
those of Bh. Beccanensis.

In Bh. Sumatranus we find large strong incisors, a very slight development
of the guard in the premolars, and a totally different
form of the auditory fossa of the squamosal bone,
whereby to distinguish it specifically from the Deccan species. In Sumatranus,
owing to the very great curvature of the post-glenoid process, the apex of the fossa
lies far behind the opening of the meatus auditorius, whereas in Beccanensis it is
very nearly vertically over it. The posterior edges of the mandible are deeply
notched below the condyle in Sumatranus. The zygoma also is much stouter, more
cmwed in the vertical plane, and shows a crescent-shaped excavation on its upper
edge just behind the post-orbital angle — all characters absent from Bh. Beccanensis.

The difference between Bh. Beccanensis and Bh. nasalis. Gray, consists in the
possession by the latter species of good-sized incisors and
a persistent first premolar. In nasalis the guard is but
very slightly developed, and the rami are very much slighter than in Beccanensis.

In Bh. Floiceri, Gray, the teeth are decidedly less tall-crowned than in Bh.

Beccanensis. Premolar 1 is persistent instead of deci-
duous, and the molar series is characterized by a very
prominent development of the second costa ( ^2 ) of the outer wall. The zygoma
also is much more arched laterally than in the Deccan species, which was a con-
siderably larger animal than Bh. Moioeri.

A comparison of Rh. Beccanensis with Bh. niger. Gray, shows that the latter
has an extended mandibular symphysis, longer in
proportion than that of the former, which terminates in
a narrow incisive edge not furnished with teeth and not showing any signs of
alveoli. Bh. Beccanensis shows several small cavities on the incisive border, sup-
posed to be alveoli, from which the existence of rudimentary (though very likely
deciduous) incisors must be inferred. The mandible of Bh. niger is much slighter,
and the ascent of the ramus does not begin till well behind molar 3, whereas in Bh.
Beccanensis it commences at the middle of the outer wall of molar 3. The ascending
portion of the ramus of Bh. niger is also much slighter, and is incurved along the
posterior edge below the condyle instead of straight as in Beccanensis. The
zygoma shows a well marked lunate excavation on its upper edge not found in
my new species. The meatus auditorius of Bh. niger is narrow and leans slightly
forward, and is altogether unlike that of Bh. Beccanensis.

Eh. nasalis.

Rh. Floweri.

Eh. niffer.

EHINOCEEOS DECCANENSIS.

Eh. Sivalensis.

Eh. Perimensis.

Eh. Paloeindicus.

It has already been shown that Mh. Deccanensis agrees with the fossil Indian
Ehinoceroses hitherto described in being hypsodont. It differs, however, in many
points which will now be enumerated.

Bh. Sivalensis, Falconer, does not show the guard on the inner side of
the upper premolars, which is so marked a feature in
Deccanensis^ and both the anterior and posterior valleys
are relatively much shallower. The frontal bones of Sivalensis are deeply incurved:
in Bh. Deccanensis they are flat, or but very triflingly incurved. The zygoma of Bh.
Sivalensis is not so wide in proportion to its length as that of the Deccan species.

In Bh. Berimensis, Falconer, the guard is also absent from the upper pre-
molars. The anterior and median colles are much
more oblique and much narrower, and the anterior
valleys are much shallower and pointed at their anterior extremities, while the
posterior form deep notches on the posterior wall — all points of marked difference
from Bh. Deccanensis. The ascent of the rami of the mandible begins well behind
molar 3 in Bh. Berknensis, and not at the median groove on the outer wall of that
tooth as in Deccanensis. Bh. Berimensis was a larger animal than Deccanensis.

The most striking difference between Bh. Balceindicus, Falconer, and Bh.

Deccanensis lies in the form of the auditory fossa, which
in the latter forms a broad, shallow, roughly rhom-
boidal area, with the meatus opening into the lower half. The meatus itself is some-
what triangular in form. In Balceindicus the fossa is triangular and very small,
with a circular meatus opening centrically and filling nearly the whole space between
the post-glenoid and post-tympanic processes. Bh. Balceindicus had also good-sized
lower incisors, and the prolonged symphysis, although slightly constricted in front
of the premolars, expands further forward and becomes spatulate. The upper
premolars did not possess a guard, and the valleys of the whole molar series are
much shallower than in Bh. Deccanensis. The zygoma is more slender than in
Bh. Deccanensis.

Bh. platyrhinm, Falconer, differs from Bh. Deccanensis in possessing large
incisors and a broad spatulate symphysis. The auditory

Eh. plaha’hinus. i i i j i .

lossa IS also much narrower, and the zygoma much
narrower and more slender. Bh. platyrhinus does not show any ant-orbital wart-like
rugosities as does Bh. Deccanensis. The anterior and median colies in the molar
series in Bh. platyrhinus are more oblique, and the walls of the valleys much
more complicated by foldings of the enamel.

Bh. Sinensis, Owen, is very distinct from Bh. Deccanensis. It is much smaller
and distinctly brachydont ; the upper premolars do not
possess a guard ; the valleys are very shallow, and the
crochet a mere wave in the enamel wall of the median collis. Molar 3 is quadrate,
rather than trihedral, in plan, and the enamel walls of all the teeth are relatively
very much thicker than in Deccanensis.

Eh. Sinensis.

PAUNA OP THE INDIAN PLHVIATILE DEPOSITS.

At least three species of Rhinoceros appear to have been found fossil in
Burmah, but have not yet been named and determined : of these one was de-
scribed and figured by Mr. Clift in the Transactions of the Geological Society, 2nd
Series, Vol. II, “ but not named:” the other two are represented by specimens in the
Geological Museum, Calcutta. Mh. Deccanensis differs specifically from all three.
Of the animal described by Mr. Clift, two much worn left upper molars are there
shown, in both of which the posterior valleys form deep notches in the posterior
wall — a character not seen in Deccanensis. The crochets also are mere waves in
the enamel walls of the median collis. The anterior outer angle of the teeth is less
acute and less projecting than in Deccanensis. K2, the second costa on the outer
wall, is also less prominent than in my new species.

The second Burmese species is represented by a very large and fine left upper
molar,* which must have belonged to a very large animal, and far exceeds the
largest tooth of Bh. Deccanensis in size. It is less tall crowned ; has the guard
well developed anteriorly, but faintly only on the inner side. A fan-shaped
denticule stands at the mouth of the pass into the relatively much shallower
anterior valley, and a furrow in the enamel wall descends from the anterior coUis
into the valley just within the pass — a feature not met with in Bh. Deccanensis.
There is only one simple crochet. The posterior area of the outer wall is deeply
concave instead of being flat, and the posterior outer angle is much less acute than
in the Deccan species.

Two right upper premolars, also in the Geological Museum, Calcutta, derived
from an animal considerably smaller than Wi. Deccanensis^ represent the third
Burmese species. They are characterized by the excessive development of the second
costa of the outer wall ( A;2 ) in a very median position, and relatively much poste-
rior to the position of the moderate-sized second costa in Bh. Deccanensis. Two
straight spur-like crochets project into the very deep anterior valleys. The more
forward of the two crochets, which corresponds to h\ in my species (see Plate I,
figs. 1 and 2), is the larger, and juts out nearly at right angles to the outer wall
of the tooth. The anterior and median colies are narrower and more obliquely
placed than those in Bh. Deccanensis. It is only on the anterior side of the front
tooth that any guard is shown.

At the same time that I obtained the remains of Bh. Deccanensis, I found,
lying loose in the bed of the nullah, a fragment of

Associated remains. . , , i i i ,

a right maxilla with two teeth (probably molars 2
and 3) of a large bovine animal, allied to Bihos gaurus, which still lives in the
Syhadri range. f The condition of the specimen and character of the encrustation
covering it indicated that it came from a position corresponding to that of the
Rhinoceros now described. In the following season, 1872, I had an opportunity

* This tooth was presented to the Geological Museum by General Sir Arthur Phayre, K. c. s. i., late Chief
Commissioner of Burmah.

t The proper name of the so-called Western Ghats.

EHINOCEllOS DECCANENSIS.

of revisiting tlie locality for a few hours, and besides collecting a variety of frag-
ments, some belonging apparently to the individual I have described, and others
to a smaller but very similar Ehinoceros, I also had the good fortune to find a
great part of the skull and many vertebrse of a large bovine whose teeth are
identical with those obtained in 1871. As in the case of the Ehinoceros these
bones broke up a good deal during and after extraction, and I have not yet had
an opportunity of restoring them and determining the species of the animal.

The bovine remains were found in the bed of very dark brown clay underlying
the black clay in which the Ehinoceros occurred, and intermediate between these
two formations I found two thin beds of clayey grit containing numerous speci-
mens of TJnio These shells have all been identified by

my colleague, Mr. W. Theobald, as being of living species : the age of the formation
they occur in, and of the overlying bed containing the Rhinoceros, may, therefore,
be reasonably regarded as pleistocene. There is no record of the existence of
Ehinocerotes so far south in the Peninsula of India, nor, as far as I could ascertain,
does any tradition of their existence remain among the people. When the indi-
vidual in question inhabited that region, the principal geological features were
probably but little different from what they are now, but the general surface was
doubtless covered with vast forests and morasses. Many featm’es of the present
sm’face indicate that various lakes or jheels existed at intervals along the valley of
the Gatparba Eiver, formed by the damming back of the waters by several rocky
barriers, which have since been worn or broken through and the lakes consequently
drained.

The spot at which Rh. Becccmensis and the other bones were found lies
well within the area of the uppermost of these supposed lakes, which was
drained by the lowering of the rocky barrier, in this case of trap, which crossed
the Gatpurba Valley at Tegree (Tegedi) some ten miles north-east of Gokak. The
idea that this valley was occupied by a lake in former times had been arrived at
quite independently by my friend Mr. A. C. Palles, c. e., from the data he
obtained when carrying out a great series of levellings in connection with Govern-
ment irrigation schemes in the Gatpurba Valley.

The Ehinoceros lived no doubt among the swampy valleys at foot of the Gokak
hills, and its remains were drifted into the lake after its death.

4*!aWB."

> .-_

: ■« ‘-'t ; ' , ■.’

.'I i!a/?.,:

\ jrvrvi

■ ; f f’y-

•.:fi

/iltiist'Vl '■■:
. , XM.) ^

i?.' ' M'

r.’...

i;f ^ ■•

" v >'.*/••} r.-lfh .Jj/
^ ‘'i;. .

* ' :.v/ . \ ;

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

DESCRIPTIONS

OP THE

MOLAR TEETH AND OTHER REMAINS OF MAMMALIA.

By R. LYDEKKER, B.A.,

GEOIOGICAL SURVEY OF INDIA.

The present memoir is the second of a series intended, as far as possible, to com-
plete our knowledge of the Vertebrate Paunse of the tertiary and post-tertiary strata
of India, a work which was left unfinished at the death of Dr. Falconer ; since that
time many new forms of mammalia have been collected by different members of
the Geological Survey, and more especially by Mr. Theobald. The present part
chiefly contains descriptions of specimens selected from a very large series of
detached molar teeth, now in the collection of the Indian Museum. Many of these
belong to new species or genera, or have been hitherto known merely by slight
notices or small-sized figures in the “ Fauna Antiqua Sivalensis” and the “ Palaeonto-
logical Memoirs” of Dr. Falconer.

The first part of the paper treats of the molar teeth of all the species of fossil
Indian 'Rhinoceros at present known, and of which there are a large series in the
Indian Museum ; the second part describes the molars of other genera, chiefly
TJngulata^yyxxi also comprising Dinoiherimi, Manis, and Amphicyon ; this somewhat
heterogeneous mixture of genera can scarcely he avoided in taking up a subject
parts of which have already been completed ; in this paper I have described species
of which in most cases we possess little more than the molar teeth or solitary bones,
or species of which the crania have been previously figured, hut of which the teeth
have never been described.

( 19 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

The next part of the series will contain figures and descriptions of the crania
of a large number of new and undescribed Muminantia, while a third part will be
devoted to the description of the remains of Carnivora, of which order several species
were named, but only a few described by Dr. Ealconer ; at the end of this part I
shall add a classified synopsis, with references, of all the extinct Indian Mammalia.

Before proceeding to the description of specimens, a few facts may be noticed
as to the distribution of some of the species of mammals treated of in this part. Of
the genus Rhinoceros we have one new species, R. iravadicus, now for the first time
described, which is confined to Burma, while R. 'platyrhimis, R. sivalensis, and
R. imlceindicus are found throughout the typical Siwaliks to the east of the river
Sutlej; to the west of the Sutlej, the new species R. planidens seems to replace
R. platyrJiinus, as no remains of the latter species have been brought among a large
series of specimens by Mr. Theobald from the latter area, and the former species was
not known to Dr. Ealconer, whose specimens were chiefly obtained from the country
to the east of the Sutlej : Acerotherium perimense had a larger range than any other
mammal, extending from Perim Island along the Siwaliks to Burma. JDinotJierium
has hitherto been only found in Perim Island, Each, Sind, and the Punjab. Tapiroid
animals {Listriodon and Tapirus) have been found in Sind, the Punjab, Burma,
and China (Owen) ; Amphicyon has hitherto occurred only in the Punjab district,
in strata which are probably somewhat older than the typical Siwaliks, while
Borcatlierimn is found along the whole of the Sub-Himalayan Siwaliks and in Sind.
Ruminants of the genera Cervus and Antilope are much more common in the
Tertiaries of the Punjab than elsewhere, while Elephants, and more especially
Stegodon, are there much more rare than in the Siwaliks to the east. The remains
of Eippopotamus are found very abundantly in the strata on the banks of the
Markanda and Jhilam rivers ; far away from the present river-courses, the remains
of this genus do not seem so common ; if this be more fully confirmed, it will lend
support to Mr. Medlicott’s suggestion that the present river-com’ses existed in
Siwalik times.

It will perhaps be said by some that I have formed species on somewhat insuffi-
cient material ; to this I must reply that I have been very careful never to name any
species except on the evidence of characteristic molar teeth or of very characteris-
tic and unmistakeable bones. From the condition of the fossils of the Siwaliks the
occurrence of entire skulls of the larger species, except of the stout skulls of Elephants
and Bovines, is extremely rare in comparison to the vast number of specimens
discovered.

In the case of the Cervidce, I cannot find instances of any crania having been
found in India, either before or since Falconer’s time. We are therefore obliged to
depend solely upon the characters of the molar teeth, of which we possess a very
large collection in the Indian Museum, for specific determination- As there appears
but little chance of ever obtaining a skull of this family from the Siwaliks, these teeth
alone are important as giving us an idea of the numerous species of these Ruminants
( 20 )

MOLAU TEETH AND OTHER REMAINS OE MAMMALIA.

wMcli prevailed in former times ; should a skull with teeth he hereafter discovered,
there will he no difficulty in assigning it to its respective species. The same
remarks apply to other animals, of many of which we only possess single teeth.
The following species or genera are now for the first time made known or fully
described

Listriodon pentapotamim^ Ealc. sp,
Mhimoeros planidens^ nobis.
MMnoceros iramdicm, nobis.
Cerms triplidens^ nobis.

Cerms simplicidem^ nobis.

Cerms latidens^ nobis^

Dorcatherium majus, nobis.
Dorcatherium minus, nobis.
Vishnutherium iravadicum, nobis.
Dinotlierium pentapotamicB, Ealc.
Manis sindiensis, nobis.

AmpMcyon palceindiGus, Ealc. et nobis.

I have here to mention my obligations to Dr. Anderson, the Director of the
Zoological Department of the Indian Museum, for his kindness in furnishing me
with specimens of the osteology of living mammals for comparison with the fossil
forms.

Order: UNGULATA.

Div. PERISSODACTTLA, Genus Rhinoceros.

In the Eauna Antiqua Sivalensis/’ the late Dr. Ealconer distinguished three
well-marked species of Rhinoceros the remains of which had been found in the
strata of the Sub-Himalayan Siwaliks; these species were respectively named
R. sivalensis, R. palcBindicus, and R. platyrhinus ; aU. these species were founded
upon well-preserved crania. A fourth species, E. namadicm, from the Nerbudda
was named in manuscript, but only a few limb bones from this locality are figured
in the Eauna Antiqua Sivalensis.** Beyond a short description of the cranium of
R. simlensis, given by Messrs. Baker and Durand in the Journal of the Asiatic
Society of Bengal for 1836, no description of any of the above has appeared. A
fifth species of the allied genus Acerotheriura was also determined by Dr. Ealconer
on the evidence of molar teeth obtained from the ossiferous beds of Perim Island in
the Gulf of Cambay : the name of A. perimense was assigned to this species.

Since Dr. Ealconer’s death another species—i?. deccanensis, described by Mr.
Eoote in the first part of the present volume — ^has been added to the Indian fossil
fauna : this species was probably of Pleistocene age ; from Pliocene strata in
China, Professor Owen has described molar teeth of another species, R, sinensis,
allied to R. sumatrensis.

With the exception of a figure of an isolated upper molar tooth of R. platyrhinus,
the teeth of all the species of Rhinoceros figured in the “ Eauna Antiqua Sivalensis”
are drawn on so small a scale, and the specimens themselves are generally so
imperfect, that the figures, which have no accompanying description, are almost
useless for the specific determination of detached molar teeth.

( 21 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

In endeavouring to refer to their respective species a large series of molar teeth
of Rhinoceros now in the collection of the Indian Museum, I found the great want
of good figures and descriptions of the known fossil species with which new ones
might be compared. In the present memoir I have endeavoured to remedy this
want by giving figures and descriptions of the molar teeth of these old species,
together with those of others which could not he referred to any of them.

Eor the technical names assigned to the different portions of the molar teeth of
Rhinoceros^ I may refer to Mr. Boyd Dawkins’ paper on the molars of Rhinoceros
tichorhinus (Nat. Hist. Rev., 2nd Ser., Vol. 3, p. 526), where they will be found
fully explained. There is also a list of most of these terms given in Mr. Eoote’s
Memoir in the first part of this volume.

In the old Siwalik area it will be found that, including the new species here
described and the Perim Island species which has lately been discovered in the
Siwaliks by Mr. Theobald, there were five species inhabiting the Sub-Himalayan area,
though not perhaps all living in the same spot. The assemblage of such a large
number of forms of the same genus in a limited area is, to say the least, very
remarkable, and one is led to wonder how or for what purpose so many distinct
species were differentiated at the same time. The explanation is probably to be
found in the existence of an abundant supply of food suitable for the nourishment
of large herbivores, and in the presence of a large area for them to wander over.
Herbivores, such as the Elephants and Rhinoceros of the Siwaliks, were of too
large a size to he much molested by the attacks of Carnwora^ and they were also,
as far as we know, free from human persecution. Under such circumstances the
species of a genus might increase almost indefinitely in number. It appears to
me probable that when in any given area the conditions of life are peculiarly
suitable for a genus of animals, in that area one would expect to find a great
number of species of that genus ; the production of species being, according to my
ideas, merely an extension of the production of individuals. In a suitable area, the
number of individuals would clearly be large, and some of them would vary, and
so would originate a new species. When the physical conditions in the same area
became less favourable to the genus, the “ survival of the fittest ” would come into
play, and the less hardy or less modifiable forms would die out. In the gravels of
the Thames valley three species of Rhinoceros lived contemporaneously in the same
area,* being, I think, next to those of the Siwaliks, the largest number of species
in one area.

Rhinoceros Pal^indictjs, Falconer. PI. 4, figs. 3 and 4.

Of the upper molar teeth of this species of Siwalik Rhinoceros, we have no
very complete specimens in the collection of the Indian Museum ; I have therefore
been obliged to content myself with figuring the imperfect but characteristic

( 22 )

* Boyd Dawkins : Nat. Hist. Rev., 1865, p. 403.

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

fragment drawn in tlie accompanying plate, together with a figure of a complete
premolar tooth copied from the ‘^Eauna Antiqoa Sivalensis” {plate 75,^^. 4). A
fair idea of the general form of the upper molars may be obtained from the small-
scale figure of a cranium in the above work {plate 70, 6). My descriptions

are in part taken from other incomplete teeth in the Indian Museum.

The fragment drawn {fig. 4) is a portion of the inner half of, probably, the
second upper molar of the right side ; it exhibits the crochet, median valley, and
portions of the adjacent colies, which are the most characteristic portions of the
tooth. The entrance into the median valley is seen to be extremely narrow, with
a very low pass, and no tubercle. The crochet (projecting into the valley from
the right) is large, simple, and slightly recurved at its free extremity. In the
present state of the tooth (about half- worn) it almost completely blocks the median
valley ; were the tooth still more worn down, the valley would be completely
blocked and the crown would show three pits or fossettes : one of these would be
formed by the posterior valley, and the other two by the divisions of the median
valley. At a still later stage of abrasion the crown would show two fossettes only ;
this would be caused by the disappearance of the shallower portion of the median
valley, which is placed on the inner side gf the crochet, the outer extremity of
the median valley being much deeper than the inner.

The anterior collis {on the left side of the figure) forms a nearly symmetrical
blunt cone, without any " antecrochet,” projecting into the median valley opposite
to the crochet. This collis is larger than the median collis {ofiohich a portion is seen on
the right side of the figure)^ and the two do not overlap one another at the entrance
to the median valley. The anterior face of the tooth is nearly straight, having a
narrow cingulum along its internal half ; the anterior valley is scarcely defined.

The general outline of the crown is approximately square ; the dorsum is
almost flat, and parallel to the long axis of the tooth. This is the most character-
istic part of the tooth, by which it is at once distinguished from the teeth of
Rhinoceros sivalensis — the only species which it at all resembles : there is one pro-
minent costa on the dorsum* situated about half an inch behind the antero-external
angle ; the latter angle has another rounded costa, but is not produced into a
buttress — another characteristic point. The costa on the dorsum does not extend
down to the neck of the tooth ; the free edge of the dorsum rises into two very
slight prominences, — one at the termination o£ the costa, and another at an equal
distance from the postero-external angle. The posterior valley is large and
approximately circular.

The length of the dorsum of an imperfect specimen in the Indian Museum
is 1’95 inches ; this is the only measurement I can give, owing to the imperfect
state of our specimens.

The premolar of the same species ( plate 4, fig. 3) belongs to the right side :
the general characters of the tooth are the same as those of the molar, viz., the
narrow median valley, the large anterior collis {on the right of the figure) bending

( 23 )

MOLAE TEETH AND OTHEE EEMAINS OF MAMMALIA.

towards the smaller posterior collis (pn the left). The crochet is long, pointed, and
simple : there is no antecrochet or combing-plate. The posterior valley forms a
long narrow slit, cutting into the left side of the figure. There is a small infold
of enamel on the anterior side of the first collis ; there appears to be no cingulum
on the anterior surface.

The peculiar straight dorsum occurs, as in the molars ; but this is distinguished
in the premolars by the absence of any distinct costa, and by the antero-external

angle being sharper. The dimensions of the premolar are—

In.

External side ... ... ... ... ... ... ... 2*5

Internal side ... ... ... ... ... ... 1'8

Anterior side ... ... ... ... ... ... ... 2'6

Posterior side ... ... . ... ... ... ... 2*3

The molars of this species may be shortly distinguished by the following
points : —

“ Narrow median valley ; simple crochet ; straight dorsum ; absence of combing-
plate, of buttress at external angle, of antecrochet, and of tubercle at entrance
of median valley ; three fossettes on worn-crown.”

The persistent dentition of this species seems to be less complete than that
of any other species of Rhinoceros. In the young skull figured in the “ Fauna
Antiqua Sivalensis” {plate 1) the last molar has not pierced the gum, and

only two premolars are present : the first of these is a small, imperfect, triangular
tooth, and does not seem to have had any tooth in advance of it, which view
is borne out by the narrowness and smallness of the alveolar border of the maxilla
in front of this tooth. If still younger specimens are procured, it would be inter-
esting to make a section of the maxilla to see if there are any rudiments of the
anterior premolars “ in alveolo.” For the present the persistent dental formula
must be written as follows : —

I- 4^- c. ^ P. M. -lEl

The penultimate upper premolar, as before said, is a small and irregularly
triangular tooth with a smooth dorsum ; the least premolar is a much larger tooth,
with a straight dorsum having a single median costa opposite the median valley ;
the true molars have all two costae on the dorsum. The dimensions of the molars

in the above-mentioned cranium are as follows : —

In.

Length of penultimate premolar ... ... ... ... ... I-IO

Length of last ditto ... ... ... ... ... 1*65

Length of first molar ... ... ... ... ... ... 1*90

Length of second ditto ... ... ... ... ... ... 2*20

Width of penultimate premolar ... ... ... ... ... 0*89

Width of last ditto ... ... ... ... ... ... 1*50

Width of first molar ... ... ... ... ... ... 1*85

Width of second ditto ... ... ... ... ... ... 1*90

( 21 )

MOLAR TEETH AND OTHER REMAINS OF MAMMALIA.

The upper molars of this species somewhat resemble those of the European
R. leptorhinus, Ouv. {R. megarhinus, ChristoL), in having the dorsum nearly straight
and without a buttress at the antero-external angle. The molars of the Indian
species are, however, at once distinguished by having three fossettes on the worn-
crown, whereas those of the European species have only two ; further, the two
colies are more equal-sized in the former, and the median valley is much narrower
at the entrance than in the latter. The molars of R. leptorhinus are further
distinguished by a small combing-plate. R. hemit<schus, Falc., also has its upper
molars with a straight dorsum ; and the two colles are in close opposition, as in
R. palceindicus : the teeth of the former are, however, sufficiently distinguished
by never showing more than the two fossettes on the worn-crown, and by having
a small combing-plate.

From the molars of R. sumatrensis and R. javanicus those of the present
species are distinguished by having three fossettes on the crown-surface instead
of only two.

The upper molars both of this species and of R. indicus present three fossettes
on the worn-crown, but the fossettes are formed in a different manner : in the
present species the two fossettes in the median valley are formed by the valley
being divided by the crochet ; in R. indicus the two fossettes are formed by the
union of the crochet and the combing-plate ; the molars of R. palceindicus have
no combing-plate. The same remarks apply to the molars of R. tichorliinus, (Ouv.),
which are formed on the same plan as those of R. indicus.

A figure of the lower molars of this species will be found in the Fauna
Antiqua Sivalensis ” {plate 75, fig. 2) ; they are characterised by having the
posterior valley considerably larger and deeper than the anterior valley, so that
the latter becomes completely obliterated at an early period of w^ear. The median
collis is the larger of the three, and the limited dentine surfaces of this and the
anterior collis form a complete semicircle.

The lower molar figured in the accompanying plate (plate 6, fig. S) seems to
belong to this species ; it was brought by Mr. Fedden from the Manchhar (Siwalik)
beds of Sind. The tooth is from the left ramus of the mandible, and is about one-
third worn down. The posterior collis (left of figu^^e) is small, and has its dentine
surface placed obliquely to the long axis of the crown ; the inner extremity of this
collis is rounded and narrow ; the posterior valley is of considerable depth and
size with a wide open entrance; it would not be obliterated until the crown became
worn down almost to its base. The median collis (centre of figure) forms by far the
greater part of the worn-crown surface ; its inner wall is flat and vertical, and angulated
next the posterior valley ; the anterior boundary of the posterior valley forms a wall
of enamel running nearly at an angle of 45" to the long axis of the tooth ; on the
outer side there is a marked angulation in the enamel wall of the anterior moiety
of this tooth; the anterior collis (right of figure) is very small, though it extends
to the inner border of the tooth ; the anterior valley is very small and shallow, and

( 25 )

MOLAE TEETH AND OTHER REMAINS OE MAMMALIA.

would be soon obliterated. The whole inner border of the tooth forms a nearly
straight line; the dorsum has a single vertical groove; there is no cingulum on any
part of the crown ; the enamel has a nearly uniform thickness. The length of the
specimen is 2'2 inches and its greatest breadth I'l inches.

This tooth differs from Falconer’s figure by the inner wall of the median collis
being longer and flatter, and by the obliquity of the median enamel wall to the long
axis of the crown. The tooth is, however, nearer to those of this species than to
those of any other.

The mandible of this species is spatulate and provided with one large broad
incisor on each side.

Rhinocehos sivalensis, Falconer. Plate 4, figs. 2 & 8 ; and Plate 5, figs. 2 & 5.

Of this species I have figured a very splendid specimen of the upper second
molar of the left side, collected by Mr. Theobald in the Siwaliks of the Pot war
district. Figures of penultimate and ultimate molars of the right maxilla are
given in Messrs. Durand and Baker’s paper, noticed above ; these figures have been
copied in De Blainville’s “ Osteographie ” {Vol. IIT^ plate 4). I have again
copied the figure of the penultimate molar, as it presents slight differences from our
specimen. I have also figured {plate 4, Jig. 2) the ultimate molar of the right
side taken from a cranium in the Indian Museum, collected by Mr. Theobald ; this
tooth, though considerably worn down, still shows the general relations of its com-
ponent parts.

Taking first the specimen from the Potwar {Jig. 5 ), we find the general shape
of the crown is approximately square, with a protrusion at the antero-externajl angle.
The anterior collis ( left side of Jigure ) is considerably the larger of the two ; it is
a blunt cone in form ; the worn dentine surface of this collis runs obliquely towards
the antero-external angle of the crown ; there is a vertical hollow on the anterior
surface of the collis. The median valley ( centre of figure ) runs in approximately
the same direction as the dentine surface of the anterior collis ; it becomes deeper
as it passes, outwards and terminates in a triangular-shaped cavity, which extends deep
down into the crown. The median valley is entered by an exceedingly narrow pass,
with a sudden fall on either side ; there is no tubercle at the entrance to this
valley, the bottom of which forms a mere line between the colies. A single pointed
crochet extends three-fourths across the valley from the median collis. There is
no combing-plate or antecrochet in the median valley.

The median collis ( right side of figure ) forms a slender cone slightly twisted
on itself ; a small but distinct third or posterior collis is seen on the right side of the
former separated by a shallow cleft ; this posterior collis is continued outwards,
as a narrow wall, becoming lower as it passes outwards, which forms the boundary
to the posterior valley ( middle of right side of figure ) ; this valley is ovate at the
top, becoming circular at the base.

( 26 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA,

On the anterior side of the tooth {left of figure) a cingulum runs along the
greater part of the surface ; the cingulum is broadest on the inner half of the
anterior surface, and forms a distinct but shallow, and triangular anterior valley ; there
is a gap in the cingulum at the middle of this valley : there is no cingulum along
the inner surface of the tooth. The antero- external angle of the crown is produced
into a very prominent process, the worn surface presenting the section of two
arches separated by a valley, and with another valley on the inner side of the most
anterior of the two arches ; the postero-external angle forms an acute spur.

The dorsum of the tooth is placed obliquely to the long axis of the crown ; its
upper half curves over towards the inner side ; posteriorly to the buttress there is no
distinct costa on the dorsum. The worn-crown presents two fossettes only.

The characteristic points of the teeth of this species are as follows : —

“ Buttress at antero-external angle : single crochet ; curved dorsum and narrow
median valley ; absence of comhing-plate, of antecrochet, and of tubercle at en-
trance to median valley ; two fossettes on worn-crown.”

The measurements of this specimen are as follows : —

In.

Length of anterior surface ... ... ...

Ditto external surface ... ... ... 2-55

Ditto internal surface ... ... ... 1'85

Ditto posterior surface ... ... ... 1*95

The presence of the buttress, of only two fossettes on the crown, and the curved
line of the dorsum, at once distinguish the molars of this species from those of the
preceding species.

The molar tooth figured by Baker and Durand is remarkable for the unequal
development of the two colles, the anterior collis being placed much nearer to the
inner border of the crown than the median collis.

The ultimate upper molar drawn in plate 4 {fig> 2) is from the right side ;
it shows an anterior collis of large size ; the median valley is narrow at the en-
trance, expands and becomes deeper as it passes outwards ; the crochet is very small
and blunt, the median collis {on the left of the figure) is placed obliquely to the
anterior wall {right of figure) and is confluent with the outer dentine mass, as in all
ultimate molars. In one of Messrs. Baker and Durand’s figures of an ultimate molar
of this species the crochet is divided at the extremity.

In both this and the preceding species there are no combing-plates present in
the median valley, by which the molar teeth are at once distinguished from those
of the two next species. The presence of only two fossettes on the worn-crown
distinguishes the upper molars of B. sivalensis from those of B. jplatyrhinus, in
which there are three. The molars of this species have some resemblance to those
of B.jamnicus; the buttress at the antero-external angle is, however, more pro-
duced in the fossil species, and the dorsum in consequence is more curved : the
cingulum is also larger on the anterior surface in the fossil species : the crochet is

( 27 )

MOLAE TEETH AND OTHER REMAINS OE MAMAIALIA.

nmcb. larger and more pointed in the Siwalik form than in the living species ; the
molars of both species agree in having only two fossettes on the worn-crown.
The molars of M. sivalensis are distinguished from those of R. tichorliinm (Guv.), and
of R, incUcus by the presence of only two in place of three fossettes on the worn-crown
and by the absence of the combing-plate. They resemble those of R. siimatrensis in
having a large buttress at the antero-external angle ; the median collis is, however,
much smaller in proportion to the anterior collis in the fossil than in the living
species, and the latter lacks the large cingulum which occurs on the anterior sur-
face of the teeth of the former species.

Erom the molars of R. leptorliinus and R. hemitmcJius the very prominent
buttress at the antero-external angle, the absence of the combing-plate, and the
curved dorsum of the molars of the present species are sufficient distinctions.

The molars of R. etniscus (“ Eal. : Pal. Mem.^ Vol. II, ylate 25, jig. 7) some-
what resemble those of R. sivalensis, both having a buttress at the antero-external
angle, a long cingulum on the anterior surface, unequal colles, a simple pointed
crochet, no combing-plate, and presenting only two fossettes on the worn-crown
surface. The molars of R. sivalensis are, however, distinguished by the posterior
collis being more distinct and elongated, by the inner extremity of the median valley
being three-cornered instead of oblong, by the crown being relatively longer in
proportion to its width, and by the ridges of the buttress being placed farther apart,
and extending lower down on the surface of the crown.

The persistent dentition of this species seems to be as follows : —

I 0 (P)-0 (?*) p 0-0 p 4-4 3-3

0—0 '^•0—0 3—3 3—3

Eigures of very much worn lower molars of this species are given in plate 75
of the “ Eauna Antiqua Sivalensis.” The specimen figured here {plate 6, Jig. 1)
is the last molar of the left ramus of the mandible ; the crown is in an early state
of wear. Ths anterior collis {extreme left of figure) is remarkable for its very
small size ; its inner boundary not extending beyond the middle line of the crown ;
the anterior valley {left of figure) is narrower and shallower than the posterior valley
{right of figure), and extends farther on the outer side ; the worn dentine-surface of
the median collis {centre of figure) runs nearly at right angles to the antero-
posterior axis of the tooth, while that of the posterior collis is placed very obliquely
to the same axis. The posterior surface of the tooth is divided by a deep and
nearly vertical groove ; on the inner side of this groove there is a very small conical
tubercle ; there is no cingulum on any part of the tooth ; the enamel is smooth and
polished and of uniform thickness. Owing to its shallowness the anterior valley
becomes obliterated at an earlier period of wear than the posterior valley. The
length of the specimen is 2’4s inches, and its width 1*4 inches.

Erom Ealconer’s figure the species seems never to have developed the first
premolar in the lower jaw ; the second premolar is of relatively large size. The
( 28 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

mandible is but little produced in front of this tooth, and seems to have been un-
provided with incisors.

Rhinoceeos plattehinus, Falconer. Plate 4, fig. 4.

The subjoined description of the upper molars of this species is partly derived
from the restored figure of a penultimate upper molar of the right side given in
the “Eauna Antiqua Sivalensis” {plate 76, Jig. 11 a) : [this Jigu7'e is copied m the
pv'esent paper (plate 4, Jig. 4,) on a scale of one-half the natm'al size and partly
from a cast of the molars of the complete cranium of the species discovered by
Colonel Baker, and now in the British Museum. The species is confined to the
Siwaliks.

The specimen in the accompanying figure is the above-mentioned penulti-
mate right upper molar ; the transverse diameter of the anterior collis {on the right
of the figure) considerably exceeds that of the median collis {on the left) ; the
base of the median collis overlaps that of the anterior collis, so that, in the
position of the figure, no part of the base of the median valley externally to the
crochet is visible from the front. The anterior collis forms a thick depressed cone,
projecting on the posterior side into the median valley. The median collis is un-
symmetrically shaped, the posterior wall being nearly vertical, while the anterior
boundary forms an irregular convex line, blocking the entrance to the median valley.
The pass to this valley is sharp and well marked, somewhat external to the inner
boundary of the tooth, and as high as the level of the cingulum, with an abrupt
descent into the valley. There is no tubercle obstructing the entrance to the median
valley, which winds between the bases of the two colies as a narrow sinuous line.

The crochet is given off from the side of the median collis nearly at right
angles to the inner border of the tooth, and has a loop of enamel on its external side ;
in the angle formed between the internal wall of the crochet and the anterior wall
of the median collis there is a small fold of enamel ; the base of the crochet is but
slightly constricted ; the anterior wall of the crochet slopes towards the inner side
of the tooth, so that the bottom of the valley is very narrow at this point ; the
whole of the bottom of the valley, from the entrance as far as the crochet, forms,
therefore, a mere line betw^een the colles.

Externally to the crochet the median valley is divided into two parts by the
combing-plate {seen at the top of the median valley), which is small and blunt, and
forms an angle of 45° with the internal border of the tooth ; there is no ante-
crochet given off from the anterior collis ; the anterior boundary of the median
valley forms a line running at right angles to the inner border of the tooth, and
there is consequently no production of the valley in the direction of the antero-
external angle of the tooth.

The dorsum of the tooth is marked by a slight anterior costa ; the antero-
external angle forms a sharp edge, but is not produced into a buttress ; the postero-

( 29 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

external angle is produced backwards. A distinct cingulum runs along the whole of
the anterior surface and that surface only.

The anterior valley is very narrow, bounded by the straight anterior portion of
the cingulum ; the posterior valley {on the left side of the figure) is very small and
narrow ; the median collis in the specimen is worn down and united with the small
posterior collis.

Were the crown to be more worn down than in the figured specimen, the
combing-plate and the crochet would unite and form a second fossette in the
median valley, so that the crown would present three fossettes, two formed by the
median valley and the third by the posterior valley.

The dorsa of the molars can be best studied from Colonel Baker’s cranium : in the
premolars there are two vertical ridges placed opposite the two large colies; in the first
molar the hindmost of these ridges is much smaller than the other ; in the second
molar it has nearly, and in the last molar quite disappeared. The outer border of
the crown in all the teeth is nearly parallel to the long axis of the tooth, and wears
quite flat. The colies in the premolars are more equal sized and the pass into the
median valley is higher in the premolars than in the molars. The following are
the dimensions of the molar series in Colonel Baker’s cranium :-~

Length of six molars

Width of penultimate molar
Width of first premolar

In.

13-0

Do.

of outer surface

of last molar ...

... 2-75

Do,

do.

of penultimate molar

... 2-70

Do.

do.

of first do.

... 2-40

Do.

do.

of last premolar

... 2-10

Do.

do.

of penultimate do.

... 2-05

Do.

do.

of first do.

... 1-75

3-00

2-22

The dimensions of the figured upper molar of this species, as given by
Dr. Ealconer, are as follows : —

External side
Internal side
Anterior side
Posterior side

In.

4-0

2- 7

3- 0
2-5

A smaller specimen, also figured in the “ Eauna Antiqua Sivalensis ” {'glate 72?
fig. 6), has the following measurements : —

External side

•In.

... 3-6

Internal side

... 2-2

Anterior side

... 2-9

Posterior side

... 3-0

Height of crown . . .

... 3-1

( 30 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

The persistent dentition of this species is as follows -

y 0“ 0 ip) 0 0

'2^ "0=0'

P. ^ M.

3-3

The median lower incisors were very small and functionless ; the first upper
premolar was shed before the last molar pierced the gum. Falconer compared the
upper molars of this species to those of Rhinoceros tichorhinus of Cuvier : the molars
of the two species agree in presenting three fossettes on the crown-surface | hut the
molars of the European species are distinguished by the greater size of the posterior
valley, by the non-prolongation of the postero-external angle of the crown,
by the combing-plate being placed nearly at right angles to the external border,
by the absence of the cingulum on the anterior surface, and by the dentine-surface
of the anterior collis being curved instead of straight. The presence of three
fossettes on the worn-crown sufficiently distinguishes the upper molars of Rhinoceros
platyrhinus from those of R. hemitcechus, R. leptorhinus, and R. etruscm of Falconer.

From the molars of both the preceding Siwalik species, the presence of the
combing-plates in the molars of this species is a sufficient distinction.

The upper molars of R. platyrhinus agree with those of the living R. indicus
by having a combing-plate and three fossettes on the worn crown-surface ; those of
the latter species are, however, distinguished by the non-development of the postero-
external angle, and by the curvature of the dentine surface of the anterior collis,
together with the absence of the cingulum on the anterior surface. Irrespective of
the upper molars the two species are at once distinguished by R. indicus having
but one nasal horn, while R, platyrhinus had two ; both species agree in having
a somewhat spatulate mandible provided with large outer incisors.

The upper molars of R. sumatrensis and R. jamnicus are distinguished from
those of the present species by having only two fossettes on the worn-crown, and by
the absence of a combing-plate.

To show the form of the lower molars of this species I have figured a
detached premolar of the right ramus of the mandible. The specimen is but little
worn down, and corresponds in size with the last premolar of the mandible of this
species figured in the “ Fauna Antiqua Sivalensis ” {plate 76, fig. 10). The anterior
collis ( extreme left of figure ) is very narrow, and extends as far inwards as the other
two. The anterior valley ( left of figure ) is noticeable for its very small size and
narrowness ; the posterior valley, on the other hand, is very large and deep ; the
anterior valley would be therefore obliterated at an early stage of wear, while the
posterior would remain for a much longer period. The difference in the times of
disappearance of the two valleys is greater than in the lower molars of R. sivalensis.
The worn dentine-surface of the posterior collis ( right of figure ) is placed at right
angles to the long axis of the crown ; the dentine-surfaces of the other two colies
are also nearly at right angles to the same axis. There is a distinct cingulum
running upwards and backwards from the entrance of the posterior valley along the

1 31 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

inner surface of the posterior collis. The enamel is very thin on the anterior collis ;
over the whole of the tooth it is minutely reticulated. The length of the specimen
is 1*6 inches, and its width 1 inch.

A last lower molar of this species in the collection of the Indian Museum has
the following dimensions : —

Extreme length ...
Width

Height of crown

In.

2-95

1-55

1-4..2

The form of this tooth is precisely similar to the figured specimen.

The first lower molar of this species is never shown : the second is very small
and conical ; the mandible is produced into a somewhat spatulate-formed symphysis,
and was furnished with a pair of large triangular outer incisors, and a small central
pair of round functionless ones.

Rhinoceeos namadicus, Falconer, Plate 4, figs. 6 and 6.

Ealconer assigned the above specific name to certain limb-bones of a species
of 'Rhinoceros from the Nerbudda valley, but I believe no teeth had at that time
been obtained. I am not aware whether these limb-bones had ever been compared
with those of the other species of the genus. The upper molar teeth in the Indian
Museum from the Nerbudda valley closely resemble in form those of the Siwalik
B. platyrJiinus, though they are of much smaller size ; and it is only after consi-
derable hesitation that I have separated the two species ; my conclusions are partly
drawn from certain differences in the teeth, partly from Ealconer’s separation of the
two species, and partly from the fact that almost all the Nerbudda Mammals are
distinct from those of the Siwaliks.

The first specimen that I have to describe is an upper molar of the right side
{plate 4, fig. 6) ; the specimen is probably the second of the series ; it is consider-
ably smaller than the corresponding tooth of the last species. The fangs of the
specimen are embedded in a hard matrix ; the crown is about one-third worn down ;
the whole of the anterior half and the inner side are complete, but the postero-
external angle is wanting.

The transverse diameter of the anterior collis {right of figure) is slightly
greater than that of the median collis ; the base of the latter does not overlap that
of the former, so that the entrance to the median valley is nearly at right angles to
the inner border of the crown, and is but slightly curved ; v/hen the tooth is placed
in the position of the figure, the whole of the bottom of the median valley is seen
from the front. The anterior collis is of a regular conical form, the median collis
is concave on the anterior side ; there is scarcely any distinct pass leading into the
median valley, the bottom of which is of uniform depth up to the crochet ; conse-
( 32 )

MOLAR TEETH AND OTHER REMAINS OF MAMMALIA.

quently tlie entrance to the valley is remarkahly low^ almost on a level with the
base of the crown, and below that of the cingulum. ; there is no tubercle at the
entrance ; the basis of the two coUes are in contact along the bottom of the valley.

At a distance of one inch from the internal border of the crown the median
collis gives off a very large and thickened crochet, which is directed forwards and a
little inwards ; the neck of the crochet is constricted, and there is no process of
enamel in the angle formed between the. internal wall of the crochet and the anterior
wall of the median collis | the crochet thickens slightly in the middle, and again con-
tracts towards its free extremity ; at a level a little below that of the present surface
of the crown, the crochet extends completely across the median valley ; there is no
loop of enamel on its external wall | the moiety of the median valley situated on
the inner side of the crochet forms a triangle with curvilinear boundaries, the
whole of which is visible from the inner side of the tooth. Externally to the cro-
chet the valley is trefoil-shaped ; the leaves of the trefoil are divided by two pro-
cesses,'— firstly, a smaU wedge-shaped combing-plate, projecting into the valley from
the outer side of the tooth ; and, secondly, a similarly shaped antecrochet from the
anterior collis. The direction of the combing-plate is exactly at right angles to the
internal border of the crown of the tooth: when worn down the surface of the
crown would display four fossettes,' — one formed by the posterior valley ; a second
consisting of that portion of the median valley which is internal to the crochet.
This fossette would not be completely isolated until the tooth becomes worn down
almost to the base of the crown, owing to the low level of the entrance to the
median vaUey i the two other fossettes would be on the outer side of the crochet,
divided from each other by the combing-plate ; they would be isolated at an early
period of wear.

The dentine- surface of the anterior collis is directed at an angle of 45° to the
internal border of the crown, so that the outer extremity of the median valley is
produced towards the antero-external angle of the tooth.

The anterior side of the tooth {on the right of the figure) has an undulating
outline, prominent in the centre ; a wavy cingulum runs along the whole length of
this surface ; on the internal half of this side the cingulum diverges from the main
wall, and forms a well-marked triangular '' anterior valley,” extending up to the
antero-internal angle of the anterior collis. There is no trace of any cingulum
along the internal surface.

The posterior valley {on the extreme left of the figure) is of very large size,
and has a regular oval shape ; the longer axis of the oval is nearly parallel with the
antero-posterior axis of the crown; the external wall of this valley is nearly verti-
cal, while the internal wall slopes rapidly away towards the median collis ; the pass
into this valley is sharply defined, forming a ridge descending from the summit of
the median collis ; the pass leading into the posterior valley is considerable above
the level of that leading into the anterior valley, and above that of the cingulum,
so that this valley would be isolated at a very early period of wear.

( 33 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

The antero-external angle of the tooth forms a small, sharp, wedge-shaped
process ; there is one prominent costa on the dorsum of the tooth near this angle ;
it is not continued upwards as far as the neck. Unfortunately the postero-external
angle of the tooth is broken away, so that it is impossible to determine the form of
this portion of the crown : from an examination of the base of the crown, where
it is buried in the matrix, I consider that the whole of the outer side of the
tooth was nearly parallel to the inner side ; and that the antero-posterior angle was
not produced. The dimensions of this specimen are —

In.

External side

... 2-0

Internal side

... 1-3

Anterior side

... 2-0

Posterior side

... 1-5

The molars of both Rhinoceros jplatyrhiniis and R. namadicus are distin-
guished from those of R. sivalensis and R. palceindicus by the complex form of
the crochet, and by the presence of a comhing-plate. The points by which the
molars of the present species are distinguished from those of R. platyrhinus are
the following : —

Difference in form of the median valley and the base of the two codes.

Presence of an antecrochet.

Difference in size of combing'-plates.

Greater size of anterior valley and cingulum.

Relations of posterior valley to median eollis, and its relative size.

Smaller size.

The other specimen figured {plate 6) is the entire and perfect germ of a

third or .ultimate molar of the left maxilla from the Nerhudda valley ; the specimen
is in the Indian Museum. I refer the specimen to the same species as the preceding.
The figure is taken from the postero-internal aspect of the specimen, looking dhectly
into the median valley ; the anterior eollis occupying the left side of the valley and
the median eollis the right.

The anterior eollis forms by far the greater extent of the base of the tooth ;
its transverse section has the form of an isosceles triangle ; the summit is bent
over towards the inner side ; on the anterior side the cingulum commences on the
external edge, and forms a V-shaped line along this side of the tooth, running as
far as the base of the anterior eollis ; close to the anterior eollis there is a narrow
anterior valley of triangular form {seen on the extreme left of the figure) ; the lowest
point of the cingulum is rather more than half an inch above the neck of the
crown. The median vaUey is entered by a low pass, without any tubercle at the
entrance ; this valley becomes deeper as it passes outwards ; the level of the pass
is the same as that of the cingulum ; the entrance of the valley is wedge-shaped.
The crochet {the first projection across the valley) is of great size and extends com-
( 34 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

pletely across the valley ; the comhing-plate {the second 'projection, on the same side
as the last) is shorter than in the last specimen ; it is opposed by the large ante-
crochet {the projection from the opposite side to the two former ones) from the anterior
collis; the difference in the relative position of these processes of the median
valley from those of the last specimen is caused by the different relations of the
boundary walls of the two teeth. Were the crown of this specimen worn down
it would present three fossettes in the median valley, one on either side of the
crochet, and a third, formed at the outer extremity of the valley, by the combing-
plate and the antecrochet. The posterior valley, as in all ultimate molars, is not
present, and the median collis becomes confluent with the external wall of the
tooth : the antero-external angle is sharp and pointed ; there is a single vertical
costa on the dorsum, situated at a distance of 4 inches from the antero-external
angle. The measurements of this specimen are as follows : —

Anterior side
External side
Posterior side
Height of crown

In.

2-1

2-5

2-0

2-8

A figure of an ultimate upper molar of the right side of Bhinoceros platy-
rhinus is given in the “ Eauna Antiqua Sivalensis” ; in this it will be seen that there
is no antecrochet, as in the present specimen, and the crochet is of much smaller
size, not extending right across the valley, while the anterior valley is scarcely
distinct from the cingulum ; the dimensions given below are also much larger than
those of the present specimen ; tending to confine the distinctness of the two forms.

The measurements of the ultimate molar of B. platyrhimis given by Dr. Eal-
coner are as follows in inches : —

Length (external side) ... ... ... ... 3‘2

Breadth (anterior side) ... ... ... ... 2’8

Height of crown ... ... ... ... 3*1

The Nerbudda species of Bhinoceros must, therefore, have been considerably
smaller than B. platyrhinus, which, apart from other characters, would probably be
sufficient to establish its distinctness.

The last upper molar of this species is distinguished from the last molars of B,
sumatrensis and B. javanicus by the presence of a combing-plate and an ante-
crochet, and by the greater size of the crochet, and by the presence of three fossettes
instead of two on the worn crown-surface.

The last molar of B, indicus has a combing-plate, an antecrochet, and a large
crochet ; the combing-plate and antecrochet do not, however, unite, so as to divide
the median valley in B. indicus as they do in the present species i moreover, the
crochet of the former species is recurved at its extremity, and passes up the median
valley between the combing-plate and the antecrochet. The crown of the present

( 35 )

MOLAR TEETH AND OTHER REMAINS OF MAMMALIA.

specimen is considerably more ‘‘ bysodont” tban that of tbe last molar of uni-
cornisy being absolutely higher and also narrower at the base.

It is to be hoped that eventually the skull of this species will he obtained from
the Nerbudda valley, that we may have more complete evidence as to the distinct-
ness of the species. The lower molars are unknown.

Rhinoceros iravadicus, n. sr. nobis. Plate 5, figs. 1 and 2.

The species is founded on two upper molars from the mammaliferous strata of
Ava {plate 5, Hgs. 1, 2) ; both specimens are in the Indian Museum.

The most worn of the two molars {fig. 1) was discovered in a pagoda at Prome,
and presented to the Indian Museum by Colonel Phayre ; from its mineral condi-
tion there can be no doubt but that it is derived from the bone beds of the Irawadi
valley. The crown is complete, with the exception of a small triangular piece of
enamel chipped from the outer wall ; the “ fangs” have been broken off close to
the neck. The tooth has been in wear for a considerable period, the height of the
anterior collis above the neck being only *9 inch. The specimen is from the left
maxilla ; and from its large size and elongated shape is probably the penultimate
true molar. This is the specimen referred to by Mr. Foote in his Memoir on R.
deccanensis, in the first part of this volume (p. 16).

The anterior collis {on the left side of the figure) is of much greater size than
the median {on the right side of the figure), occupying nearly one-half of the in-
ternal face of the tooth ; on its inner side it slopes away slightly from its base
downwards to the outer side ; on its posterior side, it gives off a blunted antecro-
chet, projecting into the median valley {the centre of the figure) internally to the
crochet. The anterior surface of the tooth has a projecting angle {the centre of the
left border of the figure) in the middle of its course; the cingulum commences
at this angle, and is continued as a distinct ledge up to the antero-internal angle
of the tooth ; there is a very small and shallow anterior valley between the cin-
gulum and the anterior collis ; along the internal surface of the anterior collis the
cingulum may he traced as a slight wavy line on the enamel, and as a still fainter
line on the median collis.

The entrance to the median valley is blocked by a large ovate tubercle, ver-
tically grooved on its internal surface ; this tubercle is continued outwards along
the bottom of the median valley as a rounded ridge. The median valley runs,
from its commencement, forwards and outwards, becoming deeper as it advances ;
it is of great width throughout its extent ; its transverse diameter at the entrance
is *4 inch, and at its narrowest part, caused by tbe projection of the crochet,
“2 inch; throughout its length it preserves a considerable width along the bottom;
the bases of the colles being nowhere contiguous.

The walls of the median collis are more abrupt than those of the anterior ;
the crochet is short and blunt, not projecting more than half way across the valley;

( 36 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

the posterior valley {in the middle of the right side of the figure) is very deep, and
almost perfectly oval in form ; the pass leading into this vaUey from the posterior
side of the tooth is divided by a median fissure.

The antero-external angle of the tooth {^eft-hand top-corner of the figure) is
produced into a prominent vertical buttress (as in Bhmoceros jamnicus and R. siva-
lensis) divided by a median groove, so that the enamel in this part of the crown-
surface forms two contiguous arches. Owing to the depth of the outer end of the
anterior valley, the worn-crown would present an isolated fossette at this point,
or two fossettes in all.

On the dorsum, the free edge forms a nearly horizontal line ; the surface is
concave antero-posteriorly. The costae of the buttress do not extend as far as the
neck of the tooth. The measurements are as follows

External side ... ... ... ... ... 3-6

Internal side ... ... ... ... ... 1'9

Anterior side ... ... ... ... ... 2‘7

Posterior side ... ... ... ... ... 2'0

Thickness of enamel on outer side ... ... ... ... O’ 1

Ditto on sides of median valley ... ... ... 0’06

Width of base of anterior eollis ... ... ... ... I’O

Ditto median ditto ... ... ... ... 0’7

Height of summit of tubercle above neck ... ... ... 0’4

Height of lowest part of cingulum above neck ... ... ... 0’5

The ether specimen {plate 5, fig. 2) is from the bone-beds of the Irawadi ;
it is the second upper molar of the right side. The whole of the grinding surface
is complete, but a considerable portion of the base of the external wall and the
anterior angle has been broken away; the crown is much less worn than that
of the last specimen.

The tooth being of the same form as the last, it is only necessary to note a few
points.

The median eollis {on the left side of the figure) is somewhat twisted upon
itself, and constricted in the middle : the pass into the posterior valley {the middle
of the left side of the figure) is more deeply cleft than in the last specimen, the
outer half being higher than the inner.

The crochet becomes gradually thinner towards its summit, and is sharply bent
inwards on itself ; a small process of enamel is seen at the outer extremity of the
median valley, which may, perhaps, be the rudiment of a combing-plate. The
measurements of this specimen are as follows : —

Length of external surface
Ditto internal surface
Ditto anterior surface
Ditto posterior surface

... 2-6
... 1-9

... 2-3

... 2-0
( 37 )

MOLAK TEETH AND OTHER REMAINS OE MAMMALIA.

The characteristics of the upper molars of this species may he shortly sum-
marised as follows : —

‘^Buttress at antero -external angle; anterior eollis tlie larger; crochet simple and blunt;
median valley very wide; broad tubercle at entrance; blunt antecrocbet and anterior eollis. No
combing-plate, and only two fossettes on the worn- crown.”

The only one of Ealconer^s three species of Siwalik ’Rhinoceros to which the
molars of the present species have any resemblance in form is Rhinoceros siva-
lensis ; the teeth of both these species agree in having a buttress at the antero-
external angle, in the anterior eollis being larger than the median, in the absence
of a combing-plate, and in the presence of two fossettes only on the worn-crown.

They are distinguished by the following points : —

In jR. simlensis there is no antecrochet ; neither is there any tubercle at the
entrance to the median valley, nor any trace of a cingulum on the inner surface.
The median valley is very wide in R. iravadicus, and very narrow in R. simlensis.
The cingulum on the anterior surface of the latter is cleft, and forms a wall above
a distinct anterior valley ; in the former this cingulum is not cleft, and merely
forms a flat ledge, without any distinct anterior valley : the process at the postero-
external angle is much more produced in R. simlensis than in R. iravadicus.

Erom the molars of R. palceindicus the present specimens are distinguished by
the presence of only two fossettes on the worn-crown ; and from those of R. platy-
rhinus and R. namadiciis by the absence of the combing-plate and by the presence
of two fossettes only on the crown.

There is no marked resemblance between the molars of any of the European
species and these specimens i from the molars of R. tichorJiinus, R. leptorhinus,
Cuv., R. hemitcecus, Ealc., and R. Indicus, the present specimens are distinguished
by the absence of any combing-plate.

The presence of the large tubercle in the median valley distinguishes the teeth
of R. iravad'icus from those of R. etriiscus, Ealc., and R. deccanensis, Eoote.

Erom R, sinensis of Owen they are distinguished by most of the above charac-
ters, and by the fact of the median valley becoming deeper as it passes outwards,
instead of becoming shallower.

They approach nearest in form to the molars of the recent Rhinocerus stma-
irensis and R. javaniciis; all the species having the buttress at the antero-external
angle and the wide median valley, with occasionally a small tubercle at the en-
trance ; they are, however, distinguished by the nearly horizontal line formed by
the free boundary of the dorsum, in place of the angulated line of the recent
species ; the two colles are more equally sized in the recent species, the ante-
rior valley smaller, and the cingulum less marked : the crochet is larger and extends
further across the valley in the recent species ; and from the larger diameter of the
median eollis, the posterior valley is smaller. It is, however, quite probable that
the Rhinoceros of the Irawadi may have been the direct ancestor of the recent
( 38 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

Rhinoceros sumatrensis, now inhabiting the same regions of Asia. The measure-
ments of the penultimate upper molar of Rhinoceros sumatrensis are-—

In.

External side
Internal side
Anterior side
Posterior side

2-0

1- l

2- 0

1-6

These measurements indicate a tooth rather smaller, but having the same rela-
tive dimensions as that of the Irawadi species.

In addition to the molar teeth described above, I have discovered, in the Indian
Museum among some bones brought from Pegu, some years ago, portions of the
occiput of a species of Rhinoceros ; these have been placed in the their relative posi-
tions and drawn {plate 5, Jig. 3) with an approximate outline of tbe whole
occiput. Erom the large size of the fragments, and from the locality in which they
were found, as well as from the difference of their form from tbe occiputs of either
of Ealconer’s three species of Siwalik Rhinoceros, I have no doubt but that they
belonged to the large Rhinoceros iravadicus of Burma.

The fragments consist of two detached portions of the occiput, as shown in the
figure ; one of these comprises the left half of the lower portion of the occiput, with
the condyle, partly broken on its lower border, of the same side. The second frag-
ment consists of the central portion of the upper half of the occiput, showing the
crest or superior curved line, and the depressions for the attachment of the cranio-
cervical .muscles ; a small portion of the parietals between the temporal fossae, with
parts of the latter, are also attached to the specimen. I cannot be quite sure
whether the relative positions of the two fragments are rightly placed in the figure
or not ; I made the interval between the summit of the condyle and the summit
of the occiput equal to two and a half times the long diameter of the former,
this being the average interval between the two points in the skulls of such species
of Rhinoceros as I have had an opportunity of observing. As the specimen is in
too imperfect a condition for exact description, I have given its measurements below,
and then compared it with the figures of the occipital regions of the three Siwalik

species figured in the Eauna Sivalensis ; —

Long diameter of condyle (partly broken) ... ... ... 3'6

Short ditto ... ... ... ... ... ... 2*4

Width of occiput above condyles ... ... ... ... 13'5

Width of upper fragment ... ... ... ... 7'0

Height of ditto ' ... ... ... ... ... 5

Width of parietals 14 inches in advance of occipital crest ... ... 1*8

Height from lo\^er border of occipital foramen to crest (?) ... ... lO’lO

Between external angles of condyles ... ... ... ... 6*3

Height of foramen magnum ... ... ... ... 2 '4

Erom Rhinocet'os sivalensis and R. palceindicus, this specimen is at once dis-
tinguished by its greatly superior size ; the interval between the lower border of

( 39 )

MOLAR TEETH AND OTHER REMAINS OF MAMMALIA.

the foramen magnum and the summit of the occipital crest of the specimens of the «
above species figured in the “ Fauna Antiqua Sivalensis” is respectively 9 and 7‘7
inches ; while that of our specimen is at least lOTO inches and perhaps rather more;
the height of the occipital foramen in Falconers two species is 1'9 and 1*2 inches
respectively; and the interval between the external angles of the condyles 4'8
inches. These measurements are much smaller than those of the present specimen.
The dimensions of the condyles of the Siwalik species are not given.

The occiput of jR. palceindicus is further distinguished from the present speci-
men by the great width of the parietals between the temporal fosse, — nowhere less
than 3’4 inches ; further, the occiput of R. palceindiGus drawn in plate 74, fig.

2(? of the “ Fauna Antiqua Sivalensis” differs from the present specimen not only
by its much smaller size, but by the square form of the upper half of the occiput,
and by the absence of the unusually deep lateral fossse, and the large median
tubercle seen beneath the crest of the present specimen.

Of B. sivalensis there is no figure of the occipital crest given in the " Fauna
Antiqua Sivalensis ;” the parietals of B. sivalensis^ between the temporal fossse
though in a much smaller specimen, are wider than in our specimen ; the relative
size of the two is, however, a quite sufScient distinction.

Retween our specimen and the occiput of Bhinoceros 'platyrhinus, which is
the only one of the Siwalik species which approaches it in size, there are very wide
and well-marked differences. An excellent figure of the occiput of B. platyrhinus
is given in the “Fauna Antique Sivalensis” {plate 72, fig. 2). On comparing this
with the figure of the present specimen, the two will he seen to differ greatly in the
length of the vertical diameter ; this, however, cannot he relied upon, owing to the
conjectural restoration of the present specimen. The condyle of this specimen,
though broken, is slightly larger than that of B. platyrhinus^ and the breadth of the
occiput is also rather larger. The occipital crest, however, at once distinguishes the
two ; the superior border of this is broad and convex in B. iravadieus ; in B. platy-
rhinus it is narrow and with a median excavation ; so that the surface of the
parietals between the temporal fossse in the latter species, as is seen in the crest of
Colonel Baker’s cranium, is concave, and placed considerably below the outer borders
of the crest. In the present specimen this surface of the parietals is flat or slightly
conven, and forms the highest part of the cranium. The narrowest part of the
parietals in the Siwalik species is 3 '3 inches, while the narrowest part remaining in
the present specimen is only 1’8 inches. The whole of the upper part of the occiput
beneath the crest in B. platyrhinus is somewhat hollowed, but not deeply so, and
with no prominent median projection ; entirely wanting the wide and deep hollows
of the present specimen.

The above description or comparison, although, from the state of the materials,
necessarily incomplete and crude, serves nevertheless to confirm the conclusions
previously arrived at from the study of the upper molar teeth as to the specific
distinctness of the large fossil Bhinoceros of the valley of the Irawadi : it may he
( 40 )

IIOLAE TEETH AND OTHER REMAINS OF MAMMALIA.

expected that further researches will bring to light more perfect specimens for future
examination and description. The lower molars of this species are unknown, as we
have none of these teeth from Burma in the Indian Museum.

R-hinoceros planidens, n. Sp. nobis. Plate 4, figs. 7, 9.

The two upper molars of this species (plate figs. 'I and 9) were collected
by Mr. Theobald in the Siwaliks ; the specimen drawn in fig. 9 is from the
left side; that in figure 7 is from the right side. Eig. 9 shows the median
collis, and posterior and median valleys, with the adjoining half of the anterior
collis; figure 7 comprises the whole of the internal half of the crown; the
summit of the median collis is broken off ; unfortunately neither of the speci-
mens show the dorsum. The teeth present a good many points of resemblance
to the Burmese specimens, but also show other points of divergence, which, taken
together with the wide distance between the localities from which they were
obtained, fully justify specific distinction.

The median valley is wide at the entrance, and, as in the Burmese and the other
Siwalik species, becomes deeper as it passes outwards ; so that the worn-crown would
present an isolated enamel pit. The crochet is blunt and simple, and does not extend
more than half-way across the median valley. The antecrochet is much larger than
in the Burmese species, bounded by a deep vertical groove at its internal border.

The tubercle at the entrance to the median valley is much wider than in Rhino-
ceros iravadicus, and its inner border is nearly on a line with the inner border of the
tooth, instead of being considerably external to it ; the outer side of this tubercle
descends suddenly to the bottom of the valley ; externally to the antecrochet the
valley is continued as a mere line between the bases of the colles, which are conti-
guous, the broad floor of the valley of the Burmese specimens being absent. This
narrowness of the bottom of the valley is caused by the greater proportionate size
of the median collis, the two colles being nearly equal in diameter ; the diameter of
the base of the anterior collis is ‘9 inch, and that of the median collis ’8 inch.

The anterior valley ( seen on the right side of fig. 7 ) forms in this species a
distinct cavity, instead of a flat surface as in R. iravadicus ; the cavity is triangular
and bounded by the continuation of the cingulum ; externally to this valley the
enamel of the crown is sharply folded on itself, and is of the same thickness as on
the inner side of the tooth. On looking at the corresponding part of the Burmese
tooth (the centre of the left side of fig. 1, plate 2), the enamel is seen to have
almost entirely disappeared at this point, and to have no trace of the fold. The
folded portion of the enamel of the Siwalik specimen forms a thick isolated ledge on
the side of the tooth, at a point where there is no appreciable coating of enamel at
all in the Burmese specimen.

The cingulum commences in the middle ofi the anterior surface of the tooth,
and is continued as a distinct ledge round the whole of the anterior collis (right

( 41 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

side of fig. 7), joining on to the median tubercle, which thus appears to be merely
a portion of the cingulum, and is not isolated as in Mhinoceros iramdicus. The
cingulum is continued posteriosly for a very short distance along the internal face
of the median collis {left side of fig. 7 and right side of figure 9) ; it then stops
abruptly, and for a space of half an inch in length, this surface has no trace of a
cingulum ; posteriorly to this space the cingulum again suddenly commences as a
bold wide ledge {well seen in fig. 9) very different from the slight cingulum at
this part in the Burmese specimens {right side of fig. 1, plate 5). This ledge, two-
tenths of an inch in width, is continued along the posterior surface of the tooth till
it joins the outer wall of the posterior valley. The pass to the posterior valley is
very sharp (while it is broad and flat in R. iravadicus)^ and the valley itself long and
oval, and less deep than in the Burmese species.

The cingulum is placed much higher up on the crown than in R. iramdicus ;
its lowest portion is *8 inch above the neck of the tooth : the height of the summit
of the median tubercle is *7 above the neck. In consequence of the median tubercle
being situated externally to the entrance of the valley, instead of within it, as
in R. iravadicus^ the interval between its summit and the outermost portion of the
median valley is greater. This interval in the Siwalik specimens is respectively 1 *6
inches and 1*7 inches, while in the Burmese specimens the same interval is only
1‘2 inches. The interval between the centre of the pass leading into the posterior
valley, and the centre of the internal surface of the median collis, is in the Siwalik
specimens IT inches, and in the Burmese species 0’9 inch. The length of the
internal surface of the specimen drawn in figure 7 is 2’2 inches, or *3 inch longer
than the corresponding surface of the molar of R. iramdicus ; the antero-posterior
diameter of the anterior collis, taken through the antecrochet, in the specimen
drawn in figure 7, is 1’3 inches ; the corresponding line in the molar of R. iramdu
CHS is IT inches. Erom the parallelism of the dentine-surfaces of the two colles in
the molars of R. planidens, I am inclined to think that the antero-external angle
of the tooth was not produced into a buttress. The worn-crown would present two
permanent fossettes only, although for a very short period there might be a very
small third fossette, placed internally to the crochet, the median valley being very
shallow at the crochet.

The molars of this species are distinguished from those of R. sinensis by the
presence of the tubercle at the entrance of the median valley, and by the valley
becoming deeper as it passes outwards.

Erom the molars of R. sivalensis the present specimens are distinguished by
the presence of the large tubercle at the entrance to the median valley, and the
very prominent cingulum on the inner and anterior surfaces, as well as by the
presence of the large antecrochet and the temporary third fossette.

The presence of only two permanent fossettes on the crown distinguishes the
present specimens from the molars of R. Itidicus, R. platyrhinus, R. namadicus, and
R. palceindicus. The absence of any combing-plate distinguishes these teeth from
( 4.2 )

MOLAR TEETH AND OTHER REMAINS OF MAMMALIA.

those of R. etruscus, Ealc., R. tiGhorJiinus, and R. leptorhinus, Guv. ; the presence of
the tubercle at the entrance to the median valley distinguishes them from the upper
molars of R. etruscus and R. deccanensis.

The great width of the median valley, the large antecrochet, and the unusual
size of the tubercle at the entrance, together with the wide cingulum on the
posterior side, apart from other characters, are amply sufficient to distinguish these
specimens from the molars of R. sumatranm and B. javanicus.

Rhinoceros, sp. var.

In addition to the above species of Rhinoceros, we have in the Indian Museum,
various upper premolar and lower molar teeth which I am unable to refer to any of
the above species, but of which I do not think it desirable at present to establish
new species. The first of these specimens {plate 6, Jig. 6) is an upper molar, or pre-
molar tooth brought by Mr. Fedden from Sind : the specimen is from the left side
of the upper jaw, and is complete with the exception of the dorsum, which is broken
away ; from its small size it is probably a premolar.

The general shape of the crown is squared, with the postero-internal angle
rounded off ; the transverse diameter of the tooth, when complete, probably exceeded
the antero-posterior. The two colies are approximately equal in size, and their
worn dentine-surfaces are directed obliquely to the median transverse line of the
tooth, and run outwards and forwards in parallel directions ; there is a vertical groove
on the antero-internal surface of the anterior collis {right ofjigure),

A cingulum surrounds the greater portion of the crown ; this commences
at the posterior valley {hollow on left of figure) as a flat-topped ridge, reaching
nearly to the summit of the median collis {left of figure) ; it descends obliquely
along the front of this collis, forming a waving line on its anterior surface. Between
the two main colles {centre of base of figure) at the entrance to the median valley
the cingulum forms a wide ledge ; this becomes narrower on the inner surface of the
median collis, and again expands into a broad ledge along the whole of the anterior
surface of the tooth {right side of figure). There is no distinct anterior valley. The
median valley does not properly commence at the gap between the two colles above
the cingulum, as is generally the case. Interiorly to the cingulum the two colles
are connected nearly up to their summits ; on the outer side of this pass there is a
sudden descent into the median valley ; this valley is consequently very short and
becomes deeper as it passes outwards. A triangular crochet projects about three-
fourths of the way across this valley ; there is no combing-plate. The posterior
valley is somewhat heart-shaped at its summit, becoming more circular as it
descends. The greatest length of the tooth is 1’35 inches, and the height of the
crown ‘9 inch.

Following my usual rule, I have not made a distinct species upon the evidence
of this solitary premolar tooth, though I think it extremely probable that it is distinct
from all the species noticed above.

( 43 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

The presence of the surrounding cingulum and the position of the pass into
the median valley sufficiently distinguish it from Ealconer’s three species of Siwalik
Rhinoceros, and from the Nerbudda and Burmese species.

The only one of the above noticed species with which I think the present
specimen can have the least affinity is R. planidens of the Siwaliks. The general
form of the cingulum is, however, very different in the two : in the present specimen
this continues in a straight line along the whole of the anterior surface, whereas in
R. planidens it bends to form a distinct anterior valley, and does not extend more
than half-way across the anterior surface of the tooth ; the cingulum, moreover, in
the latter species does not extend completely across the inner surface of the median
coHis, as in the present specimen.

The greatest difference, however, is found at the entrance to the median valley
in the two specimens : in R. planidens, the cingulum forms a distinct tubercle at
the entrance to the median valley, and the top of this tubercle becomes the pass
into the valley, there being no contact of the base of the collis further in.

These differences are so great that I do not think the two specimens can belong
to the same species : at the same time it is not quite impossible, as considerable
variations do sometimes occur between the molar and premolar series of the same
species.

The only described species of Rhinoceros with which this specimen agrees at all
closely is R. deccanensis of Mr. Eoote. The premolars of that species {plate 1 of
the first part of this volume) are of exactly the same general form and type as the
present specimen. They both have a complete cingulum, are of the same size, and
have the same internal pass between the colles into the median valley. The main
differences I can detect between the two are the following : in R. deccanensis the
cingulum forms a more complete collar round the crown, it makes a distinct ledge
on the inner surface of the median collis, instead of only a waving line ; while on
the anterior surface it is raised up into'a sharp edge on its free border, and so
forms an anterior valley, instead of a flat ledge. The pass between the colles is still
further away from the cingullum in the present specimen than in R. deccanensis ;
and in the latter there is not the vertical groove on the antero-internal surface of
the anterior collis. Eurther, the lowest part of the cingulum in the premolars of R.
deccanensis is opposite to the pass between the two colles ; the cingulum rises con-
tinuously on either side from this lowest part ; on the anterior surface the cingulum
is higher than on any part of the internal surface of the anterior collis. In the
present specimen the cingulum has its lowest level along the anterior surface ; the
portion on the internal surface of the anterior collis is higher than that on the
anterior surface.

I hope that further specimens will subsequently be acquired which will further
elucidate the affinites of this specimen ; for the present I think that it is probably
a distinct species. Its relationship with the Deccan Rhinoceros, the two separated
far in time and space, is very interesting, as I pointed out in my papers on tertiary
( 41 )

MOLAE TEETH AND OTHER REMAINS OE MAMMALIA.

mammals (Rec. Geol. Surv. Ind., Vol. IX), the Deccan fossil species showed affinites
to older forms, especially in the presence of its large cingulum; the present
specimen shows another link in the chain which relates it to the extinct Rhinoce-
rotes and Acerotheria of the European Miocene beds.

The specimen drawn in plate 5, figure 4, is a portion of the right maxilla of a
species of Rhinoceros^ containing two teeth, from Burma, in the Indian Museum.
At the distal extremity of the bone (not shown in the figure), there is the alveolus
of a third tooth, containing portions of the “ fangs.” This tooth will probably be the
second premolar, as the first usually falls out at an early period ; consequently the
two teeth drawn will be respectively the third and fourth of the premolar series.
At the proximal extremity of the bone {also omitted in the figure) are the imbedded
fangs of a fourth much larger tooth, which, if this view be correct, wUl be the first
of the true molar series : these teeth are referred to by Mr. Eoote in the first part
of this volume (p. 16).

The penultimate premolar {on the right side of the figure) is complete and
uninjured ; a portion of the enamel of the anterior collis of the ultimate premolar
has been broken away, and a caste of the dentine only remains.

Taking first the penultimate premolar, we find the external surface or dorsum
of the tooth divided into two equal portions by a prominent vertical costa ; the
antero- external angle {the right upper angle of the figure) is wedge-shaped and
slightly produced ; the anterior side is angulated in the middle, and set obliquely
to the outer side. The anterior and median colies are of nearly equal diameter ;
the latter gives off a sharp wedge-shaped crochet, small and simple; it projects
about half-way across the median valley. A large combing-plate is given off from
the outer wall of the valley ; this plate is thicker at its free extremity than at its
origin : the free extremity is separated by a small interval from the crochet ; the
worn-crown would present three fossettes. The pass leading into the median valley
is very low, nearly at the base of the crown ; it is of considerable length, running
as a level line between the contiguous bases of the colles for a distance of rather
more than a quarter of an inch, at which point there is a sudden descent into the
expanded valley : there is no tubercle at the entrance of the valley.

The cingulum occupies almost the whole of the anterior side of the tooth, and
forms a distinct ledge ; there is no cingulum on the inner side ; the anterior valley
is scarcely distinguishable. The posterior collis {on the extreme left of the figure)
is small, though distinct, separated by a channel from the median collis ; the
posterior valley is large and nearly circular with a sharp narrow pass. The measure-
ments are as follows

In.

External side

1-4

Internal side

0-9

Anterior side

1’2

Posterior side

1-1

( 45 ,

MOLAU TEETH AND OTHER REMAINS OE MAMMALIA.

The ultimate premolar is larger than the other ; it resembles the penultimate,
with the exception that the costa on the dorsum is placed, not mesially, hut close
to the antero-external angle. Its measurements are —

External side

1-8

Internal side

1-3

Anterior side

1-7

Posterior side

As these teeth belong to the premolar series, which is subject to considerable
variations in the same species, it would not he safe to found a new species upon
their evidence alone, although I doubt if they can he referred to any of the
described forms : Mr. Eoote considers them as belonging to a second Burmese species.

The premolars of Rhinoceros iravadicus, as in the recent Rhinoceros suma-
trensis, probably had a buttress at the antero-external angle, were nearly quadrate
in form, and had a wide open valley, without any combing-plate.

A figure of the premolar of Rhinoceros palceindicus is given on plate 4, figure
8, which will be seen to have no resemblance to the present specimens ; the ulti-
mate premolar of that species has but one costa on the dorsum and no combing-
plate.

In plate 73, figure 26, of the “ Eauna Antiqua Sivalensis,” the premolars of
Rhinoceros sivalensis are shown : they are quite quadrangular in form, with no very
prominent costa, no combing-plate, a narrower valley, and only two fossettes on the
worn-crown.

With Rhinoceros platyrhinus the present specimens are connected by the
presence of a combing-plate : this, however, is the only point in common. The pre-
molars of Colonel Baker’s specimen of R. platyrhinns in the British Museum are
double the size of the present specimens, are more quadrangular in form : they have
two costse on the dorsa of the whole of the series, and their external border is quite
straight and never like that of the second of the present specimens ; while the comb-
ing-plate and crochet are of much larger proportionate size, and closely approxi-
mated.

The premolars of Rhinoceros deccanensis are distinguished by the cingulum
encircling the whole of the internal surface.

The same remarks apply to Rhinoceros planidens as to R. iravadicus.

We do not know the form of the premolars of Rhinoceros namadicus, though
they no doubt had a combing-plate, and were somewhat oblique. The specimens
might possibly belong to that species, though the distance of the localities and the
difference of the age of the beds in which the two specimens occur tells somewhat
against it.

The specimen drawn in plate 6, figure 10, is from Siwalik strata, and is in the
collection of the Asiatic Society of Bengal; it belongs to the right maxilla of a
species of Rhinoceros. As there is no disc of pressure on the anterior side it is pro-
( 46 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

bably the first, or perhaps second (the first disappearing), of the premolar series.
The anterior wall of a larger succeeding tooth is seen on the left side of the figure.
The specimen is quite complete ; it had only just come into use, the tip of the
anterior collis and the anterior part of the external wall being the only portions
flattened by abrasion.

The tooth is remarkable for the excessive prolongation of the antero-external
angle {the angle on the right of the figure) in a manner unlike that of any other
recent or fossil form that I have seen. > Erom the presence of a combing-plate the
tooth belongs to the type of B. platyrhinus and R, namadicus ; but differs entirely
in form from the premolars of the first of these species, which, as shown in Colonel
Baker’s cranium, are nearly square.

The dorsum has three costae, one at the posterior angle {left of figure) and two
others placed at equal distances in advance of this, the anterior one occupying the
middle of the dorsum; from this anterior costa the dorsum slopes a’way to the
antero-external angle. The anterior collis on the right of the figure is very small,
with a concave border towards the median valley ; the dentine surface proceeding
from this collis towards the outer wall of the tooth runs obliquely forwards so as to
be almost parallel with the posterior half of the outer wall.

The median collis {on the left of the figure) is larger than the anterior; it
gives off a long sinuous crochet, stretching far into the median valley : from the
external wall of the tooth there are two long and slender combing-plates, parallel
to one another; the posterior of these unites with the extremity of the crochet,
while the anterior joins the anterior crochet ; a third small process occupies the
angle between the posterior combing-plate and the outer wall.

The median valley is entered by a low pass, without any tubercle at the
entrance ; it extends far up towards the antero-external angle of the tooth, and
is of larger extent than in any other described tooth.

Along the anterior side there is a wavy cingulum, but little raised from the
surface ; there is no distinct anterior valley. The posterior valley is large and
nearly circular, entered by a high and narrow pass.

The measurements of this specimen are —

la.

External side

... 1-6

Internal side...

... 0-6

Anterior side . .

... 1-4

Posterior side

... 1-3

The specimen differs from the preceding Burmese specimens by the greater
production of the antero-external angle, by the small size of the anterior collis in
proportion to the median, by the double combing-plates, and by the slighter
cingulum, and by the absence of the anterior valley. It belongs to the hypsodont
type, though tending towards the brachydont. The only species to which it could
possibly belong is Rhinoceros namadicus ; and until a larger series of the molars

( 47 )

MOLAH TEETH AND OTHER REMAINS OE MAMMALIA.

of that species he discovered, it cannot he said that it does not. Possibly, however,
from its very aberrant form, it belongs to yet another nndescribed Siwalik species.
The premolars of R. platyrhimis have each two costae on the dorsum. The ultimate
premolar of B, palceindicus has a large median costa on the dorsum, but no
combing-plates. In Ealconer’s figure of a premolar of this species there is no
costa on the dorsum.

The tooth figured in Plate 6, Eigure 9, was collected by Mr. Theobald in the
Siwaliks ; it is probably either the milk molar or a premolar of the left maxilla of
a species of BMnoceros. The crown is perfect with the exception of a few chips
out of the enamel ; it is about a third worn down.

The configuration of the crown surface presents some very remarkable pecu-
liarities which I have not seen in any other specimens of BMnoceros molars. The
crown surface is approximately square, with a slightly produced antero-external
angle ; the external surface or dorsum is nearly plane, having only a very slight
costa half-an-inch behind the anterior border. On the anterior surface {the left
side of the figure) a slight cingulum occupies the internal half; between this and
the external side of the anterior collis {left lower angle of the figure) there is a
small oval anterior valley {the pit on the left side of the figure) placed more exter-
nally than usual. The anterior and median colles are of nearly the same diameter :
their internal walls are nearly vertical, with no trace of any cingulum on them.

The anterior collis {left of figure) has a somewhat triangular cross-section.
Instead of being connected by a column of dentine with the dentine of the outer
wall of the tooth, the anterior collis forms a completely isolated pillar, there
being a narrow pass on its outer side connecting the anterior and median valleys.
This pass is so low that it would not be touched by wear until the crown became
worn down to the level of the cingulum. I have seen no other specimen, either
recent or fossil, in which the anterior collis is isolated in a similar manner.

The entrance to the median valley {middle of lower border of figure) is by a
narrow pass, slightly higher than the level of the cingulum, and without any
tubercle at the entrance. Externally to the pass the descent into the valley is very
abrupt and steep ; the extremity of the valley is nearly half-an-inch below the
level of the pass.

The median collis {right lower angle of figure) is united to the external
dentine wall by a very narrow neck of dentine, running at first forwards and
outwards, and then bent back on itself. On its anterior side this neck gives off
a very small bifurcate crochet, projecting about half-way across the median valley.
On its posterior side the dentine neck gives off another bluntly conical process,
projecting into the posterior valley {the pit on the right side (f the figure) ; the
narrowness of this dentine neck and the posterior process are quite unusual.

In consequence of the narrowness of this neck, the posterior valley is of large
diameter : its external wall is nearly vertical, and its anterior and posterior walls
sloping : the pass into this valley is very sharp and abrupt.

( 48 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

The dimensions of this specimen are —

In.

Anterior side. . .

... 1-6

Internal side...

... 1-2

External side

... 1-6

Posterior side

... 1-7

The lower molar figured in Plate 6, Eigure 7, is the antepenultimate tooth of
a right mandible of a ’Rhinoceros from the Siwaliks, different from either of the
three species of Ealconer. The specimen is in the collection of the Asiatic Society
of Bengal ; there is also a detached portion of the same jaw containing the ultimate
and penultimate teeth. The state of wear of this specimen is about the same as that
of the last. Dr. Ealconer described this specimen as a milk molar, but he was, ap-
parently, not aware that the next specimen in the catalogue belonged to this jaw. The
tooth differs from the lower molars of any other species of Rhinoceros that we have
seen, in the great relative size of the anterior collis {left of figure) which projects
on the inner side in advance of the other colles. The anterior valley is large
{left of figure) and of equal depth with the posterior {right of figure) ; on the
worn-crown, therefore, the two valleys would remain, as notches on the inner side,
for an equal period. The outer end of the anterior valley is rounded, and that
of the posterior valley angulated ; the three worn dentine surfaces leading from
the colles are nearly parallel, and placed obliquely to the longer anis of the
tooth.

On the anterior side of the tooth, the enamel of the crown shows a transverse
section of two vertical costae ; these extend downwards to the base of the crown.
On the anterior side of the median collis {middle of figure) there is a slight infolding
of the enamel, and other crenulations may be seen on the outer wall. The posterior
collis {right of figure) has its inner extremity expanded antero-posteriorly. In
the middle of the inner border of the posterior valley {right of figure) there is
a simple ovate tubercle detached from the posterior collis. There is no cingulum
on the inner face of the tooth ; but a small, though distinct cingulum, occurs both
on the anterior and posterior surface. The length of this tooth is 1*8 inches, and
the width *9 inch.

The succeeding tooth of this jaw {A. S. R. No. is unworn; it has the
same form and the same crenulated enamel as the figured specimen ; the tubercle
in the posterior valley is, however, considerably larger. The length of the tooth
is 2 inches.

The large anterior collis, the crenulated enamel, and the tubercle in the
posterior valley, sufficiently distinguish the tooth from the figured Siwalik speci-
mens ; but it would be somewhat unsatisfactory to found a new species on a lower
molar only.

The specimen drawn in Plate 6, Eigure 3, is another right lower molar of a
species of Rhinoceros, different from any of those described above. The tooth is

( )

MOLAR TEETH AND OTHER REMAINS OF MAMMALIA.

from tertiary beds at Kusbalgliar, near Attock, the precise age of which has not yet
been determined. A portion of the posterior collis was broken, which has been
restored in the figure.

The crown is very unsymmetrically shaped, the posterior valley {right of
figure) being greatly larger than the anterior valley {left of figure) ; the posterior
half of the tooth is worn down a quarter of an inch below the anterior half. The
dentine surface joining the posterior collis with the outer wall of the tooth is
situated obliquely to the long axis of the crown : the median collis exposes the
largest dentine surface ; its postero-internal angle is acute ; the posterior collis
{extreme left of figure) is smaller than either of the others, its innermost border
not extending so far inwards as those of the other two colles.

The inner border of the tooth presents a nearly continuous wall for some
distance up on the crown, which takes away from the prominences of the colies,
usually so conspicuous in the lower molars of other species of Rhinoceros. The
anterior valley is entered by a small and narrow pass through this innermost wall.
The entrance to the posterior collis is blocked for half its length by the inner crown-
wall. Behind this there is an open space of rather more than a quarter of an inch
in width leading directly into the valley ; the inner wall is again continued on the
anterior aspect of the posterior collis. The narrow channel leading into the posterior
valley is slightly higher than the bottom of the outer extremity of the valley.
When the crown becomes worn down, the outer extremity of the posterior valley
would be left as an isolated fossette upon the crown surface, the outer extremity of
the valley disappearing last ; in the lower molars of all other species of Rhinoceros
which I have seen described, except those of Rhinoceros simus of Africa, which
have the same peculiarity as this specimen, the posterior valley is shallower on the
external than on the internal side, and consequently never forms an isolated
fossette on the crown, but merely forms a shelving notch on the inner border.

The whole of the anterior valley on this specimen would disappear at an early

stage of wear. There is no cingulum on any part of the crown. The enamel is
thinner on the inner than on the outer side of the tooth.

The flatness of the inner wall of the tooth, together with the peculiar shape of
the posterior valley, sufiiciently distinguish the specimen from the lower molars of
any of Falconer’s species of Siwalik Rhinoceros ; the dimensions are —

In-

Extreme length ... ... ... ...

Breadth at centre ... ... ... 1-2

A single lower premolar of Rhinoceros merhkii figured in Gervais’ “Faldon-
tologie et Zoologie ” ( Rlate XIV, Fig. 3 ) shows a fossette on the worn crown,
representing the posterior valley ; but there is none on the molars of the same
specimen.

I have but little doubt that this tooth belonged to another distinct species of
Rhinoceros, which must have had affinities with the African R. simus,

( 50 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

Genus Acerotherium, Kaup.

Of this genus we have one species in India ; the species is only known by its
molar teeth, which correspond in general form with those of the Miocene European
species, but we have no means of knowing whether the Indian form was hornless, or
whether the number of its digits were similar to those of the European genus.

Acerotherium Perimense, Ealconer. Flate 6, Figs. 2 and 5.

The specimen drawn in figure 6 is a portion of a left maxilla from the collec-
tion of the Asiatic Society of Bengal, which was obtained in Perim Island in the
Gulf of Cambay; it contains the last premolar and the first molar teeth. The
other specimen in the same plate ( Fig. 2 ) is a first molar of the left maxilla ; this
specimen with two others was collected by Mr. Theobald in the Siwaliks of the
Punjab. These are the first recorded specimens from that formation. The Siwalik
specimen is less worn than those from Perim Island, but the two agree precisely in
size and form.

The molars of this species are distinguished from those of Rhinoceros by their
transverse diameter exceeding their antero-posterior in length. The median collis
( right side of figures ) is slightly larger than the anterior collis ( left side of figures) .
A cingulum surrounds the whole of the inner half of the crown and is lowest on
the inner face. The median valley {middle of lower border of figures) is entered by
a high pass. The bases of the colles are in contact at the pass. A small and blunt
crochet is seen in Eigure 2. In the Perim Island specimens this has become oblite-
rated by wear ; the direction of the upper portion of the valley is different from that
of the bottom, so that the valley in the two specimens have contrary directions.
A very small enamel island would be formed on the crown were the tooth worn down
to the level of the cingulum. The posterior valley {on the right side of the figures)
is small and oval ; the inner wall slopes rapidly. The antero-external angle if eft top
angle of figures ) forms a small buttress with two costae ; this angle is more bevelled
away in the specimen drawn in Eigure 2 than in the specimen drawn in Eigure 5.

The measurements of the Perim Island molar are —

In.

External side
Anterior side
Internal side
Posterior side

1- 5

2- 5

1- 4

2- 8

The premolar is less worn than the molar, and is almost square in shape. The
cingulum has been partly broken away along the inner side. Its measurements are —

External side
Anterior side
Internal side
Posterior side

. 1-5

. 1-4

. 1-2

. 1-4

( 31 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

A figure of an unworn left molar of this species is given in the “ Eauna
Antiqua Sivalensis ” {Flate 75, Fig. 15) from Perim Island.

The molar and premolar teeth of a right maxilla of Rhinoceros brought from
Ava by Mr. Clift and figured in the Transactions of the Geological Society of London
{Second Series, Vol. II, Plate 5) belong to the present species. Clift’s figure is
copied in Dr. Blainville’s Osteography {Atlas, Vol. Ill, Plate 14). On the same
plate there is another figure of an upper molar of the right side, also from Burma,
belonging to the same species.

This species had therefore a very wide distribution, having been obtained from
Burma, from the Sub-Himalayan Siwaliks and from Perim Island.

On comparing the above specimens with the upper molars of Acerotherium
incisimm of Kaup,, the general form of the two, though sufficiently different for
specific distinction, is very similar, evidently belonging to the same type, and
leaving no doubt in my mind that the Indian specimens are rightly placed in the
sub-genus. In both species we note the excess of the transverse over the antero-
posterior diameter of the crown (carried to the greatest extent in the Indian species),
the wavy cingulum, surrounding the inner half of the tooth; the simple valley
and single crochet ; the direction of the median valley is the same in both species.
The European species is distinguished by an antecrochet, nearly blocking the median
valley on the inner side of the crochet.

The presence of Acerotherium among the Indian Tertian Rhinocerotes is
another link connecting the faunas of early India and Miocene Europe ; in the
presence of the cingulum, and the simple form of the crowns of the molars,
Acerotherium is a connecting link between the true Rhinocerotes and the
Palceotheria.

Including the above new species and the living forms, we have the following
large list of species of Rhinoceros (or its sub-genera), which have been described,
from South-Eastern Asia ; some of the modern species are, however, probably
only varieties : —

a. nasalis, Gray.

R. indicus.

R. javanicus.

R. sumatrensis,

R. stenocephalus , Gray.

R. sivalensis, Falc.

R. palmndicus, Falc.

li. platgrhimis, Falc.

In addition to these there seems to be evidence of another species from Burma,
another from Attock, and a third from Sind. The fossil Rhinocerotes of Asia there-
fore vie with the Prohoscidia in the number of species. In the above list nine of
the species are extinct, while the total number of living forms is only six. It must
( 52 )

R. niger, Gray.

R. namadious, Falc. et nobis.
R. planidens, nobis.

R. sinensis, Owen.

R. deccanensis, Foote.

R. {A) perimensis, Falc.

U. iravadicus, nobis.

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

also be remembered that many of the latter are distinguished simply on external
characters — a kind of evidence not available in the case of fossil species.

Of three of the above fossil species, viz., R. sivalensis, R. palceindicus, and
R. platyrhinus, the crania, in a more or less complete state, have been discovered.
The two first species were unicorn and the latter bicorn. The fossil forms do not
bear out the relationship between the number of horns and the lower incisors
which occur in the living species, and which Dr. Gray has taken as a character
affording a distinction of sub-generic value. Among the living species, in the
single-horned forms, the mandible has one pair of large outer incisors, and a smaller
median pair; while the two-horned forms have only the outer pair of incisors
present. Precisely the reverse of the above occurs among the Sivalik species :
the bicorn R. platyrlimus has a pair of large outer incisors in the mandible, and
a smaller central pair; while the unicorn R. palcemdicus has the outer pair only.
{See Fauna Antiqua Sivalensis,’^ Rlate 74, Fig. 4, and Rlate 75, Fig. 10.) Dr.
Gray’s distinction, therefore, founded, on this relationship, is not of real classifica-
tory value, as it is confined to the species of Rhinoceros of one period only.

A statement made on the authority of Dr. Ealconer regarding Siwalik Mam-
mals, must of course meet with general acceptance ; in relation, however, to our
present subject, there is one which does not appear to me to be borne out by the
facts at our command. In the “ Introduction to the Siwalik Eauna” (Ealconer’s
“ Ralceontological Memoirs^’ Vol. I, p. 21), it is stated that Rhinoceros sivalensis
was provided with six incisors in both the upper and lower jaws ; that this
statement was not a local error we infer from a passage in Owen’s “Odontography,”
{Vol. I, p. 589), where it is asserted, from the verbal authority of Dr. Ealconer, that
one of the Siwalik Rhinocerotes was “ hexaprotodont” in both jaws ; this species
can only be Rhinoceros sivalensis. None of the figures, however, in the “ Eauna
Antiqua Sivalensis ” bear out this statement, as none of the incisor teeth of
R. sivalensis are shown : a figure of the mandible of this species, however {plate
75, Fig. 6), shows the whole of the molar and premolar series, but no incisors. In
this figure, the anterior premolars extend almost up to the symphysis of the mandi-
ble, precisely in the same manner as in the mandibles of Rhinoceros leptorhinus
of Cuvier {Owen, British Fossil Mammals, Fig. 135), and of the African Rhinoceros
simus {De Blainville’s Osteographie, Vol. Ill, Rhinoceros, Rlate 4), in both of which
species there are no persistent incisors. In all recent Rhinocerotes, in which the lower
incisors are persistent, the symphysis of the mandible is prolonged in a spatulate form,
considerably in advance of the first premolar, and there is a very long diastema
between the two series of teeth {Owen’s “ Odontography,” Vol. I, p. 596). In
the mandible of Rhinoceros sivalensis there seems to be no room for six incisors,
even without a diastema, a condition quite unknown in any animals of this class.
It appears to me that until some conclusive evidence of the hexaprotodont charac-
ter of this species be forthcoming, we are quite justified in regarding it as being
without permanent incisors. It is, to say the least, very remarkable that if

( 53 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

Ealconer had a specimen showing these six . incisors, he did not give a figure of
it in the ‘‘ Eauna Antiqua Sivalensis.”

In the form of their complex molar teeth, furnished with a large crochet and
combing-plate, B. platyrhinus and R. namadious approach the living R. mdicus
of India ; the lower incisors of R. platyrhinm and R. indicm have also the
same form and arrangement. Dr. Ealconer, moreover, considered the upper molars
of R. platyrhinus to approach those of the European R. tichorhinus. Both these
species also agree in having been bicorn. The lower incisors of R. tichorhinus^
however, differ from those of R. platyrhinus in having been deciduous at an
early period. The three species, R. platyrhinus^ R. tichorhinus^ and R. indicus,
agree in having had spatulate mandibles, and combing-plates in the upper molar
teeth.

The molars of the unicorn R. sivalensis are formed on the same plan as those
of the European Rhinoceros etruscus (Ealc.), but the latter was a bicorn species,
and had a spatulate mandible without permanent incisors. The mandible of R.
sivalensis is like that of the African R. simus.

The form of the cranium of the unicorn R. palmlndicus is very like that of
the unicorn R. javanicus, and the mandibles of the two species are alike furnished
with large outer incisors ; but the form of the upper molars is different. The
upper molars of R. iravadicus and R. sinensis have their nearest representatives in
those of the living R. sumatrensis.

The upper molars of Acerotherium perimense and of R. deccanensis, together
with the upper premolar noticed above from Sind, are cingulated on the inner side
like those of Kaup’s typical forms of the sub-genus (or genus) Acerotherium.
Whether any of these Indian forms were hornless or provided with four toes on the
anterior limb, we have no means at present of knowing ; the mandible of R.
deccanensis was spatulate and edentulous in the adult state, as in R. etruscus.

The lower molar from Attock, described above, presents a peculiarity of struc-
ture which is only found in the living R. simus of Africa.

The afidnities of the Indian fossil species of Rhinoceros are therefore widely
spread, showing relationships to forms, both living and extinct, scattered over
Europe, Asia, and Africa. Unless the hexaprotodont character of R. sivalensis
can be proved, there seems to be no more generalised form among the fossil than
among the living species ; while the highly specialised outer and inner lower incisors
is a character common to some of both fossil and living forms.

The difficulty of determining with accuracy the lower molar teeth, in the
absence of the original type specimens, from the Siwaliks, has prevented me from
making new species, in any case, on the evidence of these teeth alone. In the
case of upper premolars I have followed the same rule, from the known liability of
these teeth to vary in different individuals of the same species.

The three species of fossil Asiatic Rhinoceros described by Dr. Ealconer belong
to the hysodont type of Mr. Boyd-Dawkins ; as does R. deccanensis and the three
( 54 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

new species described above ; the molars of R. iravadicus, however, present a
tendency towards the hysodont type, and they are characterised by the very open
median valley, which is a remnant of a primitive type of tooth ; the only true
brachydont species is A. perimense. As all the Miocene species of European
Mliinoceros belong to the brachydont type, while the Pliocene species and all exist-
ing species belong to the hypsodont or specialised type, the Siwalik species of
Rhinoceros evidently belong to a modern group, and, as far as they go, are another
argument for the Pliocene age of the deposits in which they occur; while the
Irawadi species sliows a tendency to an older type, and therefore confirms the con-
jecture as to the somewhat older date of these Irawadi beds.

I have lately seen a note in “Nature” (Oct. 1876, p. 572), in which an extract
is given from a recent paper by Professor Elower (P. Z. S. 1876, p. 443) on the
crania of Rhinoceros, in which the following interesting difference is pointed out
between the skulls of the single and double-horned living species ; in the foriner
group, “ the external auditory meatus is embraced below by the fusion of the post-
glenoid and post-temporal processes of the squamosal portion of the temporal bone,
whilst in the other these two processes remain separate.” On looking at the skulls
of our fossil Indian species, I find that in the crania of R, sivalensis and R. palce-
indicus — both single-horned species — these two processes are united; on examining
the cast of Colonel Baker’s cranium of R. platyidiinus (the original of which is in
the British Museum) — a double-horned species — I find that in this species also the
two processes are similarly united, the external auditory meatus forming a long
tubular funnel, looking almost directly upwards, precisely as in the single-horned
R. indicus {unicornis'). This shows that Professor Flower’s distinction between the
two groups will not hold for the fossil Indian species ; the union of the two processes
is another point, in addition to the form of the upper molars, which connects R. platy-
rhinus with R. indicus, and lends support to the idea that the one may be the
ancestor of the other.

Sub-Order: RUMINA.NTIA.

ViSHNUTHEEiUM IRAVADICUM, nov. gen. nohis, Burma. PI. 7, figs, 1 and 2.

A short notice of the specimen on which this genus is founded was given by
me in my paper on the Siwalik fauna {Rec. Oeol. Surv. Ind. vol. IX, pt. 3) ; I now
give a figure and a more detailed description of the specimen.

The specimen is a portion of the left ramus of the mandible ; it contains the
first and second teeth of the permanent molar series ; these have only been in use
for a short period ; the animal was scarcely adult at the time of its death. The
inner sides of the anterior barrel of the first molar, and of the posterior barrel

( 55 )

MOLAE TEETH AND OTHEE EEMAINS OF MAMMALIA.

of the second, are broken away : at the anterior extremity of the fragment is
seen the hindmost barrel of the last premolar, which has never been in func-
tional use.

The teeth are implanted obliquely in the jaw, so that a line drawn across the
teeth at right angles to its longer axis forms nearly an angle of 45° with the long
axis of the jaw. The outer surfaces of the barrels are sharp in the middle, and
slope away slightly from the base towards the inner side ; the median valley which
separates the two barrels is deep, and extends downwards almost to the base of the
crown.

The hindmost surfaces of the outer halves of the barrels are somewhat concave.
Along the whole of the outer portion of the teeth there is a well-defined and
conspicuous cingulum ; this is most marked on the anterior and posterior surfaces ;
it becomes slighter on the external surfaces of the barrels, and on the surfaces
between tlie two barrels ; on the anterior surfaces of the teeth the cingulum is
raised into a number of points. In the median valley, which divides the barrels,
there is a large pointed tubercle, blocking the entrance to the valley ; the valley
itself is narrow and becomes shallower as it passes inwards. The whole of the
enamel for the outer half of the crowns is corrugated and roughened like that of
Sivatherium.

On the crown surface the postero-internal angle of the anterior barrel {middle
of upper border figure) forms the most prominent point on the inner side of the
tooth ; while the antero-internal angle is the least prominent point on this border.
The central enamel folds of the crown are long and narrow, and they wmuld not
form complete islands until the tooth became considerably worn dowm. The outline
of the inner border of the teeth is the same as in tlie molars of Camelopardalis ;
and the late period at wdiich the folds of enamel on the crown become insulated
is also a point common to the teeth of the two genera.

On the inner surface (/z^. 1) we are enabled to see the forms of the costse from
the contiguous unbroken lobes of the two molars : on the anterior half of the second
lobe there are three costae, of which the median is the highest ; the posterior and
anterior costae are of nearly equal height, the posterior being the most prominent ;
the anterior costa leads downwards into a kind of cingulum, bordering the anterior
edge of the tooth ; a deep channel, not extending to the base of the crown, divides
the anterior from the posterior lobe of the inner surface.

The inner surface of the hinder lobe has likewise its highest point on the
middle line, but there is only a very slight costa on this line ; the posterior angle
shows the extremity of the central enamel fold of the crown, forming a fissure on the
inner surface of the tooth.

At the base of the internal surface of the molars the rugosity of the enamel
has almost disappeared ; higher up this rugosity reappears, but it is nowhere so
much developed as on the outer surface.

( 56 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

The jaw" is slender, arcuated, and tapering ; the alveoli for the teeth are wide,
the jaw contracting suddenly in width below the alveoli. The dimensions of the
specimen are as follows : —

Length of fragment ...

In.

... 4-8

Depth of jaw

... 1-9

Thickness of jaw

. 1-5

Length of two molars ...

... 3'8

Ditto of second molar

l'4c

Width of ditto ditto ...

... 1-0

The specimen is distinguished from the molars of Sivatherium, not only by
its very much smaller size, but also by the peculiar form of the inner border of the
crown surface, and by the presence of the cingulum and the tubercle at the entrance
to the median valley.

The molars approach nearest in form to those of Camelopardalis, which they
resemble in the obliquity of their position in the jaw, and in the prominence on the
inner border of the crown of the postero-internal angles of the barrels : they, how-
ever. differ from the molars of Camelopardalis by the presence of the cingulum,
and by the presence of the large pointed tubercle at the entrance to the median
valley in both the molars ; the molars of Camelopardalis have no cingulum, and the
first molar only has a small and blunt tubercle at the entrance to the median valley.
The costse on the inner side of the teeth have the same form in both the genera.
The molars of the present form are nearly half-an-inch longer than those of any
described species of Camelopardalis, and their enamel is much more corrugated.
The depth of the present jaw at the second molar is half-an-inch greater than in
Camelopardalis giraffa, but the jaw is of a more slender type than in the latter
genus ; in C. gira ffa, and C. sivalensis, the depth of the jaw at the second molar is
equal to the length of the second molar plus the length of the hinder barrel of the
first molar ; in the present specimen the corresponding tooth is one-third of an
inch longer than the depth of the jaw at the same point.

The molars are distinguished from those of Bramatherium by their smaller
size, by the presence of the cingulum and accessory tubercle, and by being placed
more obliquely in the jaw.

Till the cranium be discovered, it is difficult to say whether this extinct form
of Buminant was most nearly allied to Sivatherium or to Camelopardalis ; the pro-
portions and curve of the jaw are most like those of the former genus, and the teeth
are intermediate in size and form between those of both genera. The jaw is some-
what more slender than that of the 8iwalik C. giraffa, and thereby approaches to
the living species of the genus.

Genus Camelopaudalis.

Eossils of this genus have been described from the Siwaliks of India, from
Perim Island, from the Upper Miocene of Attica {see Compt. Rendus, vol. Lll,

( 57 )

MOLAE TEETH AND OTHER REMAINS OE MAMMALIA.

p. 1295) and doubtfully from Erauce {see Genais, “ Faleontologie et Zoologie
Francaises,p. 148) . My reasons for doubting the validity of Ealconer’s second Siwalik
species will be found in the Records of the Geological Survey of India (vol. IX,
pt. 3).

Camelopaedalis sivalensis, Falc. et nohis, Plate 7, figs. 14 and 15.

The three lower molars of this spe'cies, as extended by me, which are figured in
the accompanying plate, were found by Mr. Theobald in the Siwaliks of the Potwar
district ; the two teeth implanted in a fragment of the jaw are the first and second
permanent molars of the right mandible ; they are considerably worn, and some-
what imperfect, but they are important, because the corresponding teeth have never
been figured as described, and the first of the two is a very characteristic tooth in
the living C. giraffa. The detached tooth is the ultimate premolar belonging to
the same lower jaw ; a less worn specimen of the corresponding tooth of the opposite
side is figured in the “ Paleontological Memoirs” (vol. I, PI. 16, fig. 78) under the
name of C. affinis.

The molar teeth are implanted in the jaw in the same oblique manner as in
the existing species ; the enamel islands in the first molar have been almost obliter-
ated by wear, showing that the animal had nearly attained its full age ; in the
second molar the posterior fold of enamel is completely insulated, but the anterior
fold is still connected with the median infolding of the crown ; the condition of
these enamel folds in the second molar shows that the Siwalik species had the same
peculiarity as the living African species, in which the molars are distinguished from
those of all other living Muminants by the late period of life at which the enamel
folds of the crown became completely insulated.

The inner wall of the anterior barrel of the second molar is, unfortunately, the
only portion of this surface of the teeth which remains complete ; in this portion
the postero-internal angle forms the most prominent point on the inner border of
the crown, while the antero-internal angle is the least prominent of the three
ridges into which the side of the barrel is divided, precisely the same as in the
molars of the living species.

In the middle of the entrance to the median valley in the first molar, there is a
large blunt accessory tubercle (the dentine of which is exposed by wear) ; in the
second molar there is no corresponding tubercle in the median valley ; and on
turning to Ealconer’s figure of the ultimate lower molar of this species (“ Fal. 31em,^
vol. l,pl. \Q,jig. 7) no corresponding tubercle is found in the valley of that tooth.
M'e are therefore assured that in C. sivalensis the only one of the lower molars
provided with an accessory tubercle is the first; in the recent 0. giraffa, and in no
other living Ruminant, precisely the same condition occurs in the lower molars,
this instance indicates the permanency of apparently unimportant dental characters
throughout long periods of time, and in widely separated localities, and affords an
( 58 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

additional argument, if one be needed, of the value of the characters of the molar
teeth in determining the affinities of Fossil Mammalia.

The enamel is slightly more rugose than in the living species ; the dimensions
of the present specimen are given in the first column of the following table, and
those of the corresponding parts of the living species in the second column -

Length of fragment

In.

... 4*3

In.

Depth of jaw at second molar

...

... 2-4

Ditto at first ditto

... 2-2

1-65

Thickness of jaw ...

..t

... 1-2

Length of first molar

... 1-3

1-0

Breadth of ditto ...

... 10

0-9

Length of second molar

... 1-26

1*15

Breadth of ditto ...

... 1-0

0*9

From the above table of measurements it will be seen that the jaw of the fossil
Indian species is three-quarters of an inch deeper than that of the living African
species, and slightly thicker, while the molars are also of a slightly larger size.
The greater stoutness of the fossil jaw perhaps indicates an animal with a shorter
and stouter-formed head, approaching in form to the Sivatherium,

The right premolar drawn in figure 14 is rather larger than the corresponding
tooth of the African C. giraffa ; both the enamel folds have become insulated by
wear ; the valley on the outer side, and the antero-external angle are rather more
developed than in the premolar of the living species. The dimensions in the table
below are compared with those of the ultimate premolar of the African species ; the
first column contains the measurement of the fossil tooth —

In. In.

Length .. ... ... ... ... 1-15 0*9

Breadth... ... ... ... ... ... 0’9 0-8

The fossil specimen is, therefore, more oblong than the recent specimen, thereby
exhibiting a more generalised type, as the almost square premolar of the African
C. giroffa is a very unusual character among 'Ruminants.

Genus Bmamatherium.

This genus was only known to Falconer by a single species from the one
locality of Perim Island. Upper molars of the genus have been described and
figured by Falconer (“Pa/. Jfm.,” ml. J, 399) and shown to be very closely allied
to those of Camelopardalis : the lower molars have not hitherto been figured, and
are now for the first time described.

( 59 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

Bramatherium perimense, Falc., Plate 7, fig. 13.

The specimen drawn in the above figure is a portion of the right ramns of the
mandible, containing the first and second molars, the latter of which is broken on
the inner side ; the specimen is in the collection of the Asiatic Society of Bengal.
The teeth are well worn, although the enamel folds still extend completely across
the crown, exhibiting the same lateness in their time of insulation which charac-
terizes the molars of Camelopardalis.

The crowns of the molars are of a more “ hysodont” type than those of Camelo-
pardalis, while the inclination of their long axis to that of the jaw is somewhat
more. The postero-internal angles of both barrels form the most prominent points
on the inner border of the crown surface, while their antero-internal angles are the
least prominent. The outline of the outer half of the crown surface of the hinder
barrel forms a triangle with curvilinear boundaries, and the postero-external angle
of this portion of the crown surface is produced inwards, to appear as a sharp point
on the inner surface of the teeth, between the first and second molars ; both of the
above characters are also common to the molars of Camelopardalis,

On the inner surface the dorsum of the molars is divided into two equal
portions, placed parallel to one another, but obliquely to the long axis of the jaw
each of these segments is terminated by a median pointed summit, and is divided by
three slight nearly vertical costm.

On their outer surfaces the teeth are covered by a rugose enamel ; on the inner
surface the enamel is almost smooth.

The lower molars are distinguished from those of Camelopardalis by the
crown surface wearing less obliquely, and by the absence of the accessary tubercle
at the entrance to the median valley of the first molar; they are also larger
in size.

The distinctions between the molars of this genus and those of VisJinutlierium
have been already pointed out.

There appears to be no difference, except that of size, between the lower molars
of this genus and those of Swatherium ; the upper molars of the two genera have,
however, several points of difference {see “ Fal. Mem.^^ ml. I, p. 399). In the
above-mentioned note on this genus, Ealconer says that the premolars of this genus
are distinguished from those of Camelopardalis, by the fact that the premolars of the
latter genus are characterized by “ their great excess of width compared with their
length in the description, however, of the premolar of Camelopardalis sivalensis
given above, it was noticed that the squareness of crown characteristic of the living
species does not hold in the fossil species ; and that, therefore, the form of the crown
of the premolar of the latter is the same as in Bramatherium. This fact points to
the very close connection between the Siwalik forms of the two genera ; the long
neck and the square-crowned premolar of the living C. giraffa being only a recently-
acquired character.

( 60 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

The jaw is very thick at the molar alveoli ; it is curved along the inferior border,
and is much deeper than the jaw of Camelopardalis giraffa^ but only slightly deeper
than that of C. sivalensis. The measurements of the specimen are given below in
the first column, and the corresponding measurements of C. giraffa in the second
column

In.

Length of fragment

... 6“0

Length of first molar

... ‘l-35

1-0

Width of ditto

... I'O

0-9

Length of anterior barrel of second molar

... 0-8

0-6

Depth of jaw at second molar

... 2-2

1-6

Thickness of ditto at ditto

... 1-5

1-0

Camelus sivalensis, Falconer.

The skull and upper molars of this species have been shortly described by
Ealconer (“ Fal. vol. J, p. 231) and shown to be close in general structure to

those of the living camel.

The upper molars, indeed, of the fossil species, seem to be almost indistin-
guishable from those of the living species, except by being slightly smaller in size.
The lower molars, however, of Camelus sivalensis, although exhibiting the same
simplicity of structure which characterizes the teeth of the family, are readily dis-
tinguished from those of the living species.

The inner surface of the lower molars of the fossil species is divided into two
equal halves by a median vertical ridge or costa ; this ridge is somewhat depressed
below the general surface of the wall of the tooth, and has a narrow groove on
either side ; slight oblique ridges bound the anterior and posterior sides of this
surface.

In the living camel, on the other hand, the inner surface of the lower molars is
divided into two portions by a very shallow and broad groove, without any trace of
the median ridge ; the anterior and posterior bounding ridges are also much less
strongly marked than in the fossil species, so that the whole of the inner surface of
the tooth is nearly flat.

At the antero-external angle of the first barrel of the lower molars of Camelus
sivalensis there is a prominent vertical ridge, not found in the teeth of the living
species. This ridge may be seen in the figure of the lower molars of the Siwalik
camel given in the “Eauna Antiqua Sivalensis,” {Flafe S7,Jig. 5) ; a similar ridge
occurs in the lower molars of the American Auchenia (Owen’s “ Odontography^'
vol. J, p. 686), by which they are distinguished from those of the old-world
camels. The slenderness of the jaw of the Siwalik camel is also an approacli to the
form of that of Amhenia ; none of the specimens of the fossil species exhibit the
relationship of the premaxillse to the nasals ; it will be interesting to note, if more
perfect specimens subsequently are found, whether the premaxillse were separated

( 61 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

from tlie nasals by the maxillse, as in the living camel, or whether they reached up
to them as in Auchenia. The discovery of an important dental character common
to two such remote forms as C. simlensis and the American Auchenias is very inter-
esting, and affords proof of the common origin of the two genera.

Genus Dorcatherium, Kaup.

The genus was originally described from the miocene strata of Eppelsheim by
Professor Kaup ; the skull was furnished with long curved canines, and with bony
peduncles for the antlers, in both of which characters it agrees with the living
Muntjac of India, which may have been descended from the older form ; the cingu-
lated molars of Dorcatherium are, however, quite peculiar.

Erom the Siwaliks, specimens of the molar teeth were obtained by Ealconer,
and in a manuscript note the specific name of D. moscMnum was assigned to these
remains ; no figures or descriptions were, however, ever published, and so it is
impossible to know to which of the forms noticed below the name was originally
intended to apply. Eor this reason I have discarded the manuscript specific name
and substituted new ones of my own.

Specimens of the molar teeth of this genus have been obtained from Kushal-
ghar, near Attock, from the Siwaliks of the Potwar district by Mr. Theobald, and
from the Manchhar beds of Sind by Mr. Eedden.

The teeth from these districts are at once seen to belong to two species, — a
larger and a smaller ; besides this marked difference, other smaller varieties are
noticeable in the specimens from the different localities ; these varieties are very
probably of specific value, but for the present I have thought it better to describe
all the specimens under two specific heads till more complete specimens shall render
further specific distinction necessary.

Dorcatherium majus, n. sp., nohls, var. a. — Kuslialghar near Attock,

Plate 7, figs. 9, 10, 11.

The two specimens from the above locality are those upon which Dr. Falconer,
by some unaccountable oversight, founded the second species of Merycopotamus
(M. nanus) {see “ Pal. Mem.,'' vol. I, p. 416). The larger of the two {figs. 9
and 10) is the last upper molar of the right side ; it is untouched by wear, with the
exception of the summit of the outer half of the anterior barrel ; the tooth is com-
plete, with the exception of the posterior half of the cingulum, which has been
broken away.

The tooth has a nearly square base ; the whole of the inner half is surrounded
by a deep and projecting cingulum, which forms a small tubercle at the entrance to
the median valley ; the inner extremities of the barrels are blunt and rounded,
( 62 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA. 45

rising nearly vertically from base to summit. The central infoldings of enamel
extend deeply into the orowo, so that it would not be until a late period that they
would become complete islands; the anterior and posterior walls of the central
enamel fold of the first barrel are curved round towards the anterior side of the
tooth : the summits of the outer halves of the barrels are bent over towards tbe
inner side.

On the external surface {fig. 10) the tooth is divided into two unequal and
unsymmetrical portions by a vertical groove ; tbe posterior lobe of this surface has a
large median ridge curving slightly forwards, as it passes downwards to form the
summit of this part of the tooth ; on either side of the lateral borders of this lobe
there are other still more prominent cost^, the posterior one of which is of great
breadth, reflected towards the central costa, and produced into a sharp angle beyond
the side of the tooth ; neither of these cost® extend downwards to the free edge of
the external surface. The anterior lobe {on the right side of the figure) lias one very
prominent median costa, and a shorter and wider anterior costa which embraces
the antero-external angle of the crown. The outer half of the hinder barrel slightly
overlaps the fore barrel.

Except in its much greater size {larger than that of any of the ^European species)
the tooth closely resembles tbe corresponding molar of Eorcatherium elegans of
Kaup ; the cost® on the outer surface are, however, more developed, and the cingu-
lum more prominent in the Indian form. The dimensions of the two molars are
given below ; those of the European species in the second column : — ■

In.

la.

Length

... 0-75

0-45

Width at first barrel

... 0-8

0-55

Height of crown

... 0*6

0-3

The second tooth from this locality {fig. 11) is also an upper molar of the
right side ; it is somewhat smaller in size than the last specimen, and is worn down
nearly to the level of the cingulum. On the external border {top line of figure) tbe
boundary of tbe anterior barrel runs parallel with tbe long axis of the tooth ; the
outer surface of this barrel is marked by two cost®, one in tbe middle, and the
other at the anterior angle. The external border of the binder barrel {left of figure)
is set obliquely to tbe long axis of the tooth ; the outer surface of this barrel is
marked by three cost®.

Tbe cingulum occupies half the circumference of the tooth ; it is much shallower,
and more closely applied to the surface of the barrels than in tbe first specimen ;
along the whole of the inner surface of the anterior barrel the edge of the cingulum
is raised into points and oasps, and there are two distinct tubercles at the entrance
into the median valley. Both this and the preceding specimen are covered with a
finely rugose enamel.

( 63 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

This specimen differs from the last chiefly by the smaller size of the cingulum,
and its cuspidated character ; the difference may, perhaps, not be more than an
individual variety. The dimensions of the specimen are as follows ;• —

In.

Length ... ... ... ... ... ... 0'6

Width ... ... ... ... ... ... 0-8

Variety h—from the Siwalilcs of the Fotwar.

The specimen of this variety consists of a portion of a left upper jaw {Vlate 7,
fgs. 4 and 6) containing the second and third molars. On the inner side the
teeth differ from the Attock form only by the cingulum being more closely applied
to the surface of the barrels.

On the outer surface {fig. 6) there is a considerable difference in the two
forms. The Siwalik has much simpler and slighter costse than the Attock form ; on
the posterior lobe ( right side of fig. 6 ) the hind costa is a very slight and simple
ridge, without any reflection, totally different from the broad reflected costse of the
Attock form ( left side of fig. 9 ) ; the anterior costse on both lobes are also much
simpler in the Attock than in the Potwar form. The size of the two forms is very
similar. Till specimens of the crania are obtained, it is impossible to say whether
these differences indicate more than individual varieties ; all the specimens from the
Potwar, however, in the Indian Museum are similar to those figured.

Doecathekium mines, Nov. Sp. nobis, Siwalilcs. — Plate 7, figs 3 and 7.

This species chiefly differs from the preceding by the much smaller size of the
teeth; we have specimens in the Indian Museum showing both the molar and pre-
molar series, so the smaller size of the teeth cannot be due to their being the milk-
molar series of H. majus. The specimens figured are the penultimate and ultimate
molars of the right side of the upper jaw. The teeth present the same general
form as those of the larger species from the Siwaliks ; the costse are, however, rather
less prominent : —

In.

Length of two molars ...

... 0-8

Length of last molar

... 0-45

Width of last molar

... 0-4

Genus Ceevus.

Of this genus Falconer announced the discovery of two species from the
Siwaliks (“ Val. Mem.f vol. I, p. 23), but these were never named or described,
and I am not aware on what materials the species were founded. In the Indian
( 6^ )

MOLAE TEETH AND OTHER REMAINS OE MAMMALIA.

Museum we have a very large collection of the molar teeth of the genus, and a
few broken fragments of antlers, hut no crania ; from the characters of these upper
molar teeth alone I have determined three species of Cerms from the Siwaliks, and
I think, from the occurrence of some lower jaws which I cannot make to correspond
with any of the above, that very probably one or more additional species existed.
From the absence of crania and the fragmentary condition of the antlers, the
molar teeth afiord the only grounds for the determination of the different species.
The distinctions between these teeth may at first sight seem to be too slight and
trivial for affording specific distinctions ; but we must recollect that the molars of
all Ruminants are formed upon such a similar general plan that variations, which
in the teeth of other orders would be considered of only specific value, in this order
rank as generic distinctions. The slight variations which do distinguish the teeth
of one species of living Cervus from those of another are always found to be
constant, and therefore of high value in distinguishing between the different species.

In writing of the molars of fossil species of European Deer, Professor Lartet
remarks ( Compt. Rend. 1868, p. 1119) that the older forms have short-crowned
(brachydont) molars with the depressions on the surface so shallow that the bot-
tom is always visible ; while in the recent, pleistocene, and newer tertiary forms,
the crowns of the molars are longer (hysodont) and the cavities so deep that the
bottom can never be seen in any state of wear. The teeth of the Siwalik JDorca-
therimn belong to the former division ; while those of the Siwalik C. latidens (infra)
are intermediate between the two, the crown being moderately high and the bottom
of the cavity visible when half worn : the teeth of the other two species of Siwalik
Cervus belong to the second type. The teeth of the Siwalik Cervidce therefore
indicate that the fauna of these beds is intermediate between that of the miocene
and the upper tertiaries of Europe.

From the fragments of antlers which we possess in the Indian Museum, I
find that all the species of Siwalik and Narbudda Stags belong to the Rucervine
division of the group. All the specimens described below were collected by Mr.
Theobald, and the majority of them from the Siwaliks of the Pot war district : in this
district it appears that Cervus was the most common genus of Mammals in Siwalik
times, whereas in the Sutlej district it was comparatively rare ; Elephants seem to
have been rare in the Potwar district, and abundant in the Sutlej district.

CePvVUS latidens, n. sp. nobis. — Plate 8, figs. 7 and 10.

The present species was the largest of the Siwalik Cervidce^ its teeth equalling
in size those of the English Megaceros. The specimen drawn in figure 10 is the
last upper molar of the right side ; it is complete and about one-third worn down :
the form of the dorsum of these teeth differs somewhat from those of other species
of Cervus i but for the present, at least, I retain them in that genus.

( 65 )

MOLAE TEETH AND OTHER REMAINS OE MAMMALIA.

The crown of the tooth is somewhat oblong in shape, with a produced postero-
external angle. On the inner side the anterior barrel projects more inwardly than
the posterior barrel. The central enamel folds are large and very deep, the
posterior one has no infold from the hinder side ; they have just become insulated
by wear. In the valley between the two barrels, there is a long and slender
accessory column, the summit of which is slightly touched by wear ; it is quite
distinct from the general dentine surface of the crown.

The outer wall of the fore barrel is parallel to the long axis of the crown ; that of
the hind barrel is set obliquely ; on the outer surface {Jig. 7) the hind lobe presents
two conspicuous lateral costse, the posterior of which forms an angle beyond the
posterior surface ; between these two there is a slight trace of a central costa ; the
fore lobe shows two costse, one median, and one at the anterior border, the latter
produced into an angle ; the costse on both lobes run perpendicularly to the base of
the tooth ; the outer surface of the hind lobe becomes wider as it approaches the base
of the tooth; the lobes on this surface have a general flat appearance. The
dimensions of this specimen are —

In.

Length ... ... ... ... ... ... 1-33

Width ... ... ... ... ... ... 1-0

The peculiar flatness of the external surface, and the want of parallelism of
the outer surfaces of the two lobes, together with the presence of the large ac-
cessory column, distinguish this tooth from the molars of any living species of
Cermis with which I am acquainted : the great production of the postero-external
angle is also a very peculiar character of the specimen. In no living genera of
CervidcB does the accessory column rise to such a height as in this and the succeed-
ing species.*' This column is absent in Tarandus and Bama ; short and rudimentary
in Megaceros^ Alces and Cervus ; slightly taller, but very thin, in Rusa, Rucervus
and Ranolia ; and tallest and thickest of all in the Siwalik and Rhine Cervidce.

The slight elevation of the costse on the dorsum, and general squareness of this
surface, distinguish this tooth from those of the typical Stags. This flatness, together
with the distinctness of the accessory column from the barrels, and the absence of any
infold on the posterior side of the hinder enamel island, distinguish this tooth from
those of the Oxen, and of Hemibos and Amphihos ; it is also distinguished by being
shorter, and by rapidly narrowing from base to summit. In all the above genera
the dentine surface of the column becomes united with the general dentine surface
of the crown, whereas in Cervus it remains distinct.

The large lower molar of a species of Cervus figured in plate 8, figs. 4 and 6,
may, I think, have belonged to an individual of the same species as the above tooth,
with which it corresponds in size.

The tooth is from the right ramus of the mandible, and is probably the second
true molar ; it is distinguished from the lower molars of any recent Stag by the

* Cervus giganteus of the Rhine diluvium has a similarly large accessory column.

( 66 )

MOLAE TEETH AND OTHER REMAINS OE MAMMALIA.

great height of the crown, and by the peculiar arrangement of the flutings on the
inner surface {jig. 4), as well as by the great size of the accessory column.

Both of the central enamel folds of the crown have become completely insula-
ted by wear ; the inner border of the crown surface {tipper border of jig. 6) forms
a continuous wavy line ; the median summits of the inner surfaces of the barrels
form the most prominent points on the inner border of the crown surface. The
accessory column extends two-thirds up the crown, and is of large relative size, and
placed somewhat internally to the mouth of the median valley.

On the external surfaces the enamel is slightly rugose : on the internal surface
the enamel is generally smooth, marked here and there by dots and pits. The
length of the specimen is I'l inches, and the width 0’8 inch ; the height of the
crown being 1*2 inches.

Ceuvus triplidens, nov. sp. nobis. Plate 8, figs. 1, 2, 5.

This species from the Siwaliks is named from the unusually large size of the
accessory column of the upper molars. The specimens figured consist of the
second and third molars of the right maxilla ; the first of these is slightly worn,
while the second is merely a germ, as yet untouched by wear.

The teeth are noticeable for the great elevation of the crown, which is con-
siderably higher than in most living species of Deer ; the base of each tooth has
its antero-posterior diameter rather longer than the transverse. The inner surfaces
of the barrels are broad and rounded, sloping slightly from the inner to the outer
side ; a vertical groove runs down the inner half of the posterior side of the hind
barrel. Between the barrels there is a tall and large accessory column extending
two-thirds up the crown ; the accessory column presents a triangular cross-section,
and has a deep vertical groove on its anterior face ; the innermost edge of this
column projects on the inner side of the tooth in advance of either of the barrels.
On the anterior surface of the tooth there is a very small rndiment of a cingulum
{not seen in figure) ; these two last characters are constant in all our specimens.

The central enamel-folds on the crown-surface are narrow and shallow, and
would become completely insulated at an early period of wear. The innermost
dentine surfaces are unsymmetrical and elongated. The outer segments of the
barrels are set slightly oblic[uely to the antero-posterior axis of the tooth, their
anterior costse forming the most prominent points on the outer border of the
crown.

On the external surface {jig. 2) the hinder barrel is divided into three very
prominent Costse ; the median costa is by far the larger of the three, and becomes
confluent at its base with the other two ; as it descends from the neck of the tooth,
it rapidly diminishes in width and bends slightly forwards : deep grooves divide
the costse. The external lobe of each barrel becomes narrower as it ascends from
the summit of the crown to the neck. The posterior barrel shows only two costae

( 67 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

on its external surface, the anterior of which is the larger. A rugose enamel
covers the whole of the teeth : the dimensions are as follows : —

In.

Length of two molars ... ... ... ... ... 1’8

Length of second molar ... ... ... ... ... 0*8

Width of ditto ... ... ... ... ... 0-95

Height of ditto ... ... ... ... ... I’l

Width of third molar ... ... ... ... ... I'O

Height of ditto ... ... ... ... ... 1-25

Other specimens in the Museum are slightly larger.

The upper molars of the present species are distinguished from those of the
preceding by the absence of the produced postero-external angle ; by the presence
of the prominent costse and grooves ; by the outer surface of the lobes becoming
narrower at the base instead of wider ; by the presence of the small cingulum on
the anterior surface ; and by the presence of the groove on the accessory column.

The tall accessory column distinguishes the teeth from those of the European
Deer.

Erom the Indian Musa and Rucervus, the present teeth are distinguished by
the antero-posterior diameter being longer and the transverse diameter shorter :
by the greater height of the crown ; by the larger size of the accessory column ;
and by the larger size and curving of the median costae on the outer surfaces of
the lobes.

The lower molars figured in plate 8, fig* 5, may, I think, possibly belong
to this species ; the specimen contains the two last molars of the right side : both
of them are complete, and in an excellent state of preservation.

The ultimate molar, as usual, is composed of three barrels, the last of which
is of relatively large size, though still smaller than either of the anterior barrels ;
the inner boundary of the crown runs parallel to the long axis of the tooth, so
that in the barrels the median costa becomes the most prominent point along the
whole inner border of the tooth. The external surfaces of the tooth are bluntly
rounded off and rise almost vertically from the neck to the summit of the crown ;
the median valley between the barrels extends far into the centre of the crown ;
at the base of the entrance into the external valley there is a very small accessory
tubercle. The central enamel-folds are completely insulated* by wear ; the folds
are of large size, constructed in the middle and swelling out at either end. At
the antero-external angle of the first barrel there is a loop formed in the enamel,
and a consequent ridge and groove at this part of the external surface of the tooth ;
the anterior boundary of the crown-surface runs at right angles to the long axis
of the tooth. The third lobe is connected by a narrow isthmus with the second.

On the inner surface the barrels are separated from each other by a deep
vertical groove : the anterior barrel has a very prominent median costa, and a faintly
marked costa on either side ; the second lobe has the prominent and tall median costa,
( 68 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

and a second costa on tlie posterior border only : the third lobe has no dis-
tinct costa.

The first molar is similar to the first two lobes of the last tooth. The whole
of the surface of the teeth is coated with a finely rugose enamel, something like
that of Camelopardalis. The teeth are considerably larger than those of the
Samhar, with which their measurements are compared below ; those of the Samlar
being in the second column : —

, In. In.

Length of last molar ... ... ... ... 1'4 1*25

Length of penultimate molar ... ... ... ... 1*0 0*9 ^

Width of ditto ... ... ... ... 0*7 0*6

The teeth are distinguished from those of the Samhar and Barasingha by
their outer surface being perfectly parallel to their long axis, instead of each barrel
being set slightly obliquely to this ; they are further distinguished by the very
small size of the accessory tubercles and by the roughness of the enamel.

The parallelism of the outer surface to the long axis of the tooth is a character
these teeth possess in common with Cervus elepTias of Europe,* from which they
are distinguished by the bluntness of the inner surfaces of the barrels, the small
accessory tubercles, and the prominent ridge at the antero-external angle.

It is only in regard to size that these lower molars agree with the upper
molars of C. triplidens ; the absence in the lower molars of the accessory column
is a great point of difference, and they may eventually have to be referred to a
fourth species of Siwalik Cervus.

Cervus simplicidens, non. sp. nohis. Plate 8, fig. 3.

This third species of Siwalik Cervus is founded upon another type of upper
molars, of which two are given in the above figure. These teeth are the last two
molars of the left side, and have but just been touched by wear. The general
form of the teeth is very near to those of the living Rucervus duvaucelUi^ and
very different from those of both the species described above. The crowns are
tall and somewhat slender, the inner surfaces sloping away from the inner to the
outer side. The outer surfaces of both barrels are set nearly parallel to the
antero-posterior axis of the crown ; the antero-external angle of each barrel being
the most prominent points on the outer border, but neither of them more promi-
nent than the other. The innermost summits of the barrels are extended antero-
posteriorly.

The median valley on the inner side extends for a long distance between the
barrels ; at the entrance to this valley there is a very small accessory tubercle,
less than a quarter of an inch in height, and never forming a part of the worn-
crown surface ; it is placed in the valley considerably further in than the bases
of the barrels.

( 69 )

MOLAE TEETH AND OTHER REMAINS OF MAMMALIA.

On the anterior and posterior surfaces of the teeth there is a semi-vertical
fold or ridge of enamel, looking like the commencement of a cingulum ; on the
external surface the form of the teeth is similar to that of the tooth of Cervus
triplidens described above. The enamel is slightly rugose. The measurements of
the specimen are given below

In.

Length of two molars ... ... ... ... ... 1*7

Length of last molar ... ... ... ... ... 0*9

Width of ditto ... ... ... ... ... 0'9

The teeth are distinguished from those of Cervus triplidens by the small size
of the accessory tubercle and its position far within the median valley, by the
squareness of the crown, and by the smaller obliquity of the external halves of the
barrels to the antero-posterior axis of the tooth.

The present molars approach somewhat in general form to those of the living
Mucervus duvaucellii ; they are, however, distinguished by the smaller degree of
obliquity of the outer surfaces of the barrels to one another^ and to the antero-
posterior axis of the tooth ; they are also distinguished by the presence of the
enamel folds on the anterior and posterior surfaces.

Genus Listriodon, jH. von Meyer.

The occurrence of this genus in the Indian Tertiaries was first announced by
me in my paper on Indian Fossil Mammalia (“ Rec. Geol. Surv., Ind., vol. ZX,
pt. 3) ; the specimens on which this identification was made are now for the first
time figured and described ; the genus was originally founded on specimens of the
molars from the Molasse of Switzerland ; a description of these teeth is given in
Dunker’s “ Palseontographica” (1874, page and plate VII, fig. 71). The first
of the Indian specimens is the same as that referred by Falconer to the genus
Tapiriis (“ Ral. Mem.fi vol. I, p. 415), and to which he assigned the name of Tapirus
pentapotamice : on the authority of this specimen the genus Tapirus is introduced
by Mr. Wallace among the Fossil Siwalik Mammalia.*

The second specimen was brought from the Siwaliks of the Panjab by
Mr. Theobald, and was, from its similarity to Falconer’s wrongly-named specimen, at
first referred by me to Tapirus. The only fossil from India that has been described
as Tapirus, besides the above, is a portion of the symphysis of a mandible from
Burma {Clift : ‘‘ Trans. Geol. Soc., Lon.fi Sec. Ser., vol. II) ; this may very probably
have belonged to the present genus.

Listriodon pentapotamijd, Falc. sp. PI. 8, figs. 8, 9.

The first of the two specimens {fig. 8) is the above-mentioned specimen of
Falconer’s (now in the Indian Museum) from the Tertiary beds of Kushalghar near

( 70 )

* Distribution of Animals, vol. I, p. 1 22.

MOLAR TEETH AND OTHER REMAINS OF MAMMALIA.

Attock. The specimen is from the right side of the upper jaw, and is either the second
or third true molar. The crown of the tooth is nearly square ; the summits of the
ridges are slightly abraded by oblique surfaces of wear, and there is a disc of
pressure on both fore and hinder surfaces.

The crown is raised into two parallel transverse ridges of nearly equal size ;
these ridges are separated by a deep intervening valley, which extends completely
across the tooth, and which slopes on the outer side quite down to the base, with-
out any trace of a longitudinal wall connecting the two ridges. The posterior
surfaces of the ridges are concave, and the anterior ridges slightly convex ; the worn
surfaces on their summits slope towards the anterior extremity of the tooth. From
the internal summit of the anterior ridge a wide ledge curves across the anterior
surface of the tooth to the internal extremity of the ridge ; a smaller and nearly
horizontal ledge joins the former at the antero-internal angle of the crown ; another
waving and slightly cuspidate ledge, bounding a hollow, runs along the whole of the
posterior surface of the tooth, and there is a small and blunted tubercle at both
extremities of the median transverse valley ; these ledges and tubercles may be
considered as an incomplete cingulum. The measurements of the specimen are as
follows : —

In.

Length

... 0-82

Width of anterior surface

...

... 0-80

Ditto of posterior ditto

... 0-80

The form of the tooth, as pointed out by H. von Meyer, differs from that of
Tapirus and LopModon, in that the transverse valley extends completely across the
crown, and that, consequently, the longitudinal outer wall connecting the transverse
ridges of the molars of the latter genera is wanting. The teeth, were it not for
their square form, might be mistaken for the lower molars of Tapirus ; they are of
the simplest form of structure that occurs in any genus of this group of Mammalia.
In certain upper molars of Lophiodon, figured by De Blainville {Osteographie Atlas
vol. IV, plate Chceropotamus), the longitudinal outer wall is nearly absent, and the
teeth ought, perhaps, to be referred to the present genus.

The tooth differs from the upper molar of Listriodon from the Swiss Molasse,
in that the ledges or cingula are not crenulated, and that the anterior of these is
confined to the anterior surface, and does not extend round the outer extremity of
the ridge to the median valley as in the European form.

The upper molars of this genus resemble in form the upper molars of Dino~
therium, while those of Tapirus resemble the premolars of Dinotherium ; the latter
genus is remarkable for having its premolars more complex than its molars ; the
lower molars of Idstriodon are very like those of Tapirus.

The teeth of this genus may be considered as transitional forms between the
bunodont type of Tetraeondon (infra) and the lopJiodont type of Tapirus ; in the
present genus the two pairs of outer and inner cones of the tetraconodont form are

( 71 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

connected by transverse ridges, while in Tapirm and its allies, the two outer cones
are also connected by a longitudinal ridge. In this manner a transition is effected
from the simple bunodont form through the lopJiodont to the selonadont type of
the Ruminants.

The second upper molar of this species figured {dg. 9) is from the Potwar
district of the Panjab; like the preceding specimen, this also is from the right
maxilla ; the form of the two teeth is similar, but the present specimen is rather the
smaller of the two; its dimensions are as follows : —

In.

Length ... ... ... ... 0’81

Width at anterior end ... ... ... 0’75

Ditto posterior end ... ... ... 0*74-

Genus Dinotherium, Kaiip.

This genus was first introduced to the Indian Eossil Eauna by Ealconer from
a specimen of a portion of a first lower molar and of a lower jaw discovered in
Perim Island (“ JPalceontological Memoirs,^' vol. I, p. 396) ; the species was named
by Ealconer Dinotherium indicum. Subsequently, the same naturalist identified
two other molar teeth from near Attock (“ Pal. ml. 7, p. 414) as belonging

to the same genus ; in the note on these specimens in the “ Paleontological
Memoirs” no specific name was assigned to them, though they were considered to
be of too small a size to have belonged to B. {perimense) indicum; these specimens
are now in the Indian Museum ; they are ticketed with labels in Ealconer’s hand-
writing bearing the name of 7). pentapotamice, which name I have accordingly
adopted. Other specimens have been obtained by Mr. Eedden from Sind and Kach,
and are figured in the present paper. All these specimens seem to belong to the
same species, which was much smaller than D. indicum or any European form ;
it is unfortunate that the Perim Island species is only known by the above-
mentioned fragmentary tooth and lower jaw, in the latter of which the crowns of
the molars have all been broken off, so that comparison between the teeth of that
species and of 7). pentapotamice is impossible. In the collection sent to the Asiatic
Society of Bengal by Captain Eulljames from Perim, there were certain molar
teeth which were conjectured to belong to LopModon : Dr. Ealconer suggested these
might also belong to Dinotherium Pal. Mem.” vol. I, p. 397). Unfortunately, I
am quite unable to discover what has become of these fossils.

Dinotherium pentapotami^, Pale, et nobis, PI. 9, figs. 1 to 5.

The first of the Attock specimens {fig. 1) noticed by Dr. Ealconer is “ the
penultimate or first premolar, upper jaw, right side ; at least this is inferred from
its form and an obscure disc of pressure on the posterior side, and from there being
( 72 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

no disc of pressure in front.” Dr. Ealconer might have added that the present tooth
is at once distinguished from the second premolar by the outer ridge not being
cleft by a transverse valley, and by the posterior tubercle being separated by a deep
longitudinal valley from the outer ridge.

The crown of the tooth is approximately square ; the tooth was inserted into
the jaw by three fangs, two very large ones at the outer angles, and a smaller one
placed on the inner border at the angle between the two former. The outer side of
the crown is raised into a longitudinal ridge, rising vertically on the outer side, and
sloping gradually on the inner side ; the ridge is slightly convex antero-posteriorly ; its
enamel is rounded off on its summit, but is not cut through by wear. The inner
side of the crown has two mammilloid tubercles, of which the anterior is the larger
and higher ; the enamel of the latter is slightly cut through by wear ; the posterior
tubercle is only rounded off ; a wide longitudinal valley, which is deepest at the
posterior extremity, divides the two tubercles from the outer ridge ; the transverse
valley between the tubercles is very narrow. The cingulum surrounds three sides
of the crown ; it is entirely wanting on the outer side ; its margin is irregularly
crenulated ; it is much the widest on the anterior side ; on the inner side it becomes
very slight on the surfaces of the tubercles ; between the tubercles it forms a cleft
ridge blocking the entrance to the transverse valley.

The specimen is distinguished by its much smaller size from any of the
European species; from the first premolar of D. giganteum and D. cuvieri, it is
distinguished by the following points : the antero-posterior valley is very much
deeper and wider in the Indian form, rendering thereby the inner tubercles more
completely conical ; the posterior tubercle is mammilliform in the Indian form,
whereas it is elongated transversely in the European form ; in the latter the
cingulum forms a much more regular rim round the three sides of the crown ; there
is a distinct ledge instead of a mere wavy line on the inner surfaces of the tubercles,
and consequently the portion of the cingulum at the entrance to the transverse
valley does not project in advance of the rest, as in the Indian form ; further, the
centre of the outer ridge is slightly hollowed in the European form.

The measurements of the first premolar of D. giganteum are compared below

with those of the present specimen i—

European.

Indian.

Antero-posterior diameter, outer edge

3*4

2-3

Ditto ditto inner ditto

2-75

1-6

Transverse diameter

3-1

2-1

The second of the Attock specimens {jig. 2) is the first true molar of the right
side of the upper jaw ; it probably belonged to the same individual as the last
specimen ; its position in the upper molar series is determined by its possessing
three transverse ridges, whereas the succeeding molars are simple, having only two
ridges (a very rare condition in the teeth of mammalia) ; that the specimen belongs
to the upper jaw is inferred by its having a cingulum or talon ridge at both ends,

( 73 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

whereas the corresponding molar of the lower jaw has a talon-ridge at the posterior
end only.

The crown is oblong, hearing three nearly parallel transverse ridges ; the median
ridge has been somewhat broken ; the two anterior ridges have been considerably
worn, and show pyriform dentine surfaces; the large extremity of these surfaces is
on the inner side ; the hindmost ridge is only slightly touched by wear ; the plane
of wear slopes obliquely towards the anterior side of the tooth. All the three
ridges are concave posteriorly and convex anteriorly ; they are raised into promi-
nences at both inner and outer extremities. Along the whole of the anterior side of
the tooth, there is a wide talon-ridge or cingulum, with a valley at its hinder
border : on the posterior surface of the tooth there is a smaller crenulated cingulum,
occupying the concavity of the hindmost ridge : on the inner border of the tooth
there is a small tubercle at the entrance to the transverse valley between the first
and second ridges ; a similar tubercle, when the tooth was perfect, probably occupied
the corresponding space in the second valley ; there is no trace of any cingulum
on the inner surfaces of the ridges. Except in its much smaller size, the form of
the tooth is very similar to the first upper molar of D. giganteum ; the latter, how-
ever, has a slight cingulum on its inner surface. Below the dimensions of the
Indian and European specimens are compared : — ■

European. Indian.

Antero-posterior diameter ... ... ... 4iT 375

Transverse ditto first ridge... ... ... 3-3 3-4

Ditto ditto last ditto... ... ... 3-65 3T

Dr. Ealconer compared this specimen with the third milk molar of the European
species ; from its being associated with a premolar, there can, however, he no doubt
but that this specimen belongs to the permanent and not to the deciduous series ;
and therefore that it should be compared with the large permanent first molar of the
European series.

The broken specimen {Jig. 4) is also the first upper molar of the right side ; it
is from Kach, and is precisely similar to the Attock specimen ; the specimen shows
the two hinder ridges and the tubercle of the first valley.

The large tooth in the centre of the plate {Jig. 3) is also from Kach, and was
collected by Mr. Eedden ; it is the second upper molar of the right side ; from its
size I make no doubt but that it belonged to the same species as the preceding
specimens. The tooth has four fangs ; the crown is nearly square and carries two
transverse ridges convex anteriorly and concave posteriorly ; the anterior ridge has
its enamel worn through at both extremities, the posterior ridge has the enamel only
obliquely abraded by wear. The median transverse valley is partly blocked at its
outer extremity by a low wall connecting the two ridges ; at its inner extremity it
has a large tubercle, flat internally and convex externally ; at the anterior end of
the tooth there is a wide cingulum or talon-ridge, raised into prominences at either
( 74 )

MOLAE TEETH AND OTHER REMAINS OF MAMMALIA.

end ; at the posterior end there is a small cingulum occupying the base of the con-
cavity of the second ridge, and a second smaller prominence below the outer
extremity of the same ridge.

Besides its smaller size, the tooth is distinguished from the second molar of
J). giganteum by the greater concavity of the posterior surface of the second ridge,
and by the consequently greater prominence of the cingulum ; the tubercle in the
median valley is also larger ; the anterior cingulum is raised into prominences
instead of being level ; and the transverse valley is blocked at the outer side instead
of being quite free. The dimensions of the two teeth are given below : —

European.

Indian.

Antero-posterior diameter ...

3-90

2-70

Transverse diameter of first ridge ...

3-65

2-45

Ditto ditto of second ditto

3-55

2-30

Greatest height of crown

2-05

1-50

The last specimen {fig. 5) is a portion of a first molar of the left ramus of the
mandible. The specimen is from Sind, and contains the first and second ridges
only, the third having been broken off ; its position in the series is determined by
its being three-ridged, and it is seen to belong to the lower jaw by the absence of
any cingulum at the anterior end. The inner side of the second ridge has the
enamel worn through at one spot ; the valley between the first and second ridges is
very shallow, and deepest at its extremities ; the anterior ridge is very short and
raised into a tubercle at its centre.

From its small size there can he no doubt but that the tooth belongs to the
same species as the previous specimens. The dimensions of this specimen are
given below, viz. : —

In.

Length of fragment ... ... ... ... ... 2'05

Width of first ridge ... ... ... ... ... 1'60

Ditto of second ditto... ... ... ... ... 2-05

The width of the base of the crown of the corresponding molar of the Perini
Island jaw is 2*9 inches, showing that the teeth of that species were very consi-
derably larger than the present specimens. A last upper molar, a first lower molar,
and fragments of other molars of a species of Dinotherium are figured by Professor
H. von Meyer in the German “ Palseontographica” {vol. XV, pi. 3) ; these speci-
mens were collected by the Messrs. Schlagintweit at Kushalghar ; the teeth corre-
spond in form with the present specimens, and doubtless belonged to the same
species ; the last upper molar is, as would be expected, rather larger than the
penultimate tooth figured here; the lower molar corresponds in size with the
present specimens ; the cingula in the German specimens are rather more crenu-
lated than in ours.

( 75 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

Genus Sanitherium, E. von Meyer.

In my paper on Tertiary Mammals^' above quoted this genus was unfortunately
omitted from the list of Indian forms ; it was provisionally established by Professor
von Meyer in 1866 for the reception of some small molar teeth of a Suine animal
Falceontogra'phica,’’ vol. XV, 15) collected by the Messrs. Schlagintweit at
Kushalghar. On comparing these figures with the specimen of the lower molar of
Sus ^nisillus of Ealconer in the Indian Museum, I found the two were identical.
Although Ealconer’s specimen was named in manuscript before 1866, the name was
not published till 1868 in the “ Palaeontological Memoirs.” Professor H. von Meyer’s
specific name {S. schlagintweitii), therefore, must stand. I may here add that in
the same paper Professor von Meyer has described another species of Equus
{E. ‘primigenins) from the Kushalghar beds, and from the lower Siwaliks of N urpfir
in Chamba : the species is distinct from either E. sivalensis, or Eippotlierium antilo-
pinum ; the existence of this third species, common to European and American
tertiaries, is another proof of the distinctness of the Mammalian fauna of these beds,
from that of the upper typical Siwaliks of Ealconer; the Nurpiir beds from which
also Ampliicyon was derived being at the very base of the Siwaliks.

Sanitherium schlagintweitii, E. von Meyer. Plate 9, figs. 6 to 9.

Of this species I have copied two of Professor Von Meyer’s figures {^^Ealceonto-
grapMca,'^ vol. XV, pi. II, Jigs. 9, 10) (figs. 6 and 7), and have added two figures of
Ealconer’s specimen {Sus pusillus) (figs. 8 and 9). A translation of Professor von
Meyer’s description of his specimen is as follows “ The fragment represented
belongs to the left ramus of the mandible, and contains a complete tooth, probably the
penultimate molar, with the first lobe of the last molar. The former tooth has a length
of O’Oll mm., and has its anterior lobe considerably worn down ; its greatest width
is 0'008 mm. Erom each of the main tubercles of the transverse ridges of the
crown a projection runs down to its postero-internal angle ; this projection is well
seen on the anterior one of the outer tubercles. Between the transverse ridges there
is an accessory tubercle ; besides this there is a distinct talon-ridge on both anterior
and posterior surfaces, and on the outer surface a crenulated cingulum, while the
antero- external main tubercle is connected with the anterior talon-ridge ; this con-
dition is well seen on the fragment of the last molar. The animal was about half
the size of Sus hysudricus, and there is no mention of such an animal in the “ Eauna
Antiqua Sivalensis,” or elsewhere in papers on Asia.

“ The teeth remind us of those of Cheer opotamus, but are smaller ; they are also
related to those of Eyracotherium, and are of nearly the same size; the lower molars
of this genus are, however, not known to me ; besides, there are other Siiince, such as
Eeccari, Bahirusa, Eyolherium, 8fC., which, however, have very distinct accessory
tubercles, and which lack the crenulated cingulum on the lower molars.

( 76 )

Rec., Gcol. Surv., lud., vol. ix., pt. 3.

MOLAH TEETH AND OTHER REMAINS OE MAMMALIA.

“ These fragments may, therefore, be considered as belonging to a new species ;
to determine the genus with accuracy, these fragments are not suthcient : assuming
the genus to be new, I propose the name of Sanitlierium (Sani : an Indian goddess)
with the specific name of its discoverers.”

The specimen in the Indian Museum {jigs. 6 and 7) from the same locality as
the above shows the two hindmost lobes of the last molar of the right side of the
mandible ; the alveolus of the anterior lobe is also shown. The tooth is very slightly
worn ; it is constructed on the general type of the teeth of Suina, showing a narrow
and shallow antero-posterior valley, and deeper transverse valleys, so that the
summit of the crown forms transverse ridges divided into tubercles at the outer and
inner extremities ; there is an irregular accessory tubercle in the centre of the
transverse valleys ; and the extremities of the main tubercles are irregularly defined ;
the crenulated cingulum on the inner surface does not extend to the hindmost lobe ;
the outer tubercle of the first ridge has long projections from each of its internal
angles. The plane of wear slopes very obliquely from within outwards ; the enamel
is slightly wrinkled. The dimensions of the specimen are as follows : —

In.

Length of last molar ... ... ... ... 0’68

Width of second ridge ... ... ... ... 0-31

Probable depth of jaw ... ... ... ...0'75

The teeth of this animal are distinguished from those of Stis by the greater
simplicity and distinctness of the main tubercles, and consequently by the bound-
aries of the dentine surfaces of the worn-crown being less contorted ; by the
tubercle in the transverse valley being less distinct and smaller ; by the hinder lobe
being larger and taller in proportion to the others ; by the plane of wear being con-
siderably more oblique ; and, lastly, by the presence of the cingulum.

The lower molars of Ryracotherium are at once distinguished from the present
specimens by their higher tubercles, more open transverse valley, without any
central tubercle, and absence of cingulum.

The lower molars of Chcerotherium and Acotliermm have simple conical tuber-
cles with very wide and open valleys, no accessory tubercles, and no continuous
cingulum ; those of Hyotherium have the outer tubercles concave on the inner side,
and the inner tubercles simple, and with no cingulum.

The lower molars of Chceropotamus also have the outer tubercles concave on
their inner side ; and the two pairs of tubercles are separated by a much wider and
deeper longitudinal valley, so that the worn dentine surfaces would not be united
till the crown was nearly completely abraded away : finally, the cingulum is not
crenulated, and does not extend on to the anterior lobe.

The lower molars of Rhagatherium and Hyopotanms have an approach to the
form of the teeth of Ruminants, and are therefore widely different from the present
specimens.

In the living Reccari and Balnrusa the lower molars have no crenulated
cingulum, and the third lobe of the last molar is much smaller in proportion to the
others than in the present specimens.

( 77 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

E’rom the above comparisons there can, I think, be no doubt as to the validity
of Professor Meyer’s genus Sanitherium : I have not found among our Indian collec-
tions any specimens which I can refer to the upper molars of this genus. This genus,
with the addition of the next, gives the following list of Suina from the Tertiaries of
India, a list equal in extent to that of the fossil Proboscidian fauna of India, viz, : —

SUID^.

Sus giganteus.

Stis hysudricus.

Anthracothbrid^ .

Anthrac other um {Chceromeryx) silistrense.

Hippopotamidjs.

Hexaprotodon sivalense.

Hexaprotodon iravadicim.

Tetraconodontid^.

Teiraconodon magnum

HippoJiyus connects the Figs with Hippopotamus, and Sanitherium with the
A nthracotheridce : Merycopotmius is a link between Hippopotamus and Hyopotamus,
and so joins on with the Buminantia ; while Teiraconodon presents affinities in its
molar teeth with Anthracotherium, and diverges from all other genera in its gigantic
premolars ; it is a difficult question to decide whether Merycopotamus should be put
into the same family as Hyopotamus {Anthracotheridrs) , with which it is so closely
united by the form of the molars, or whether with the Hippopotamidm, with which
it is connected by many points of its anatomy, most notably by the peculiar form
of the lower jaw. Teiraconodon should, I think, without doubt be placed in a
distinct family.

Hippohyus sivalensis.
Sanitherium Schlagintweitii.

Merycopotamus dissimilis.

Hexaprotodon namadictm.
Tetraprotodon palaindieum.

Genus Teteaconodon, ’Falconer.

This genus of Hippopoiamoid Ungulates was formed by the late Dr. Falconer
on the evidence of a portion of a lower jaw containing the ultimate and penultimate
molars ; the specimen was found by Messrs. Baker and Durand in Siwalik strata
between the Markanda Pass and Pinjdr ; it was figured by them in the Asiatic
Researches” {ml, XIX, pi. 6, fig. 2) ; the figure has been copied in the “ Palae-
ontological Memoirs” {vol. l,p. 150,^^. 5) to illustrate Dr. Falconer’s memoirs on
the specimen ; in this memoir the specimen is described as being the right upper
jaw ; this is the more strange, as Dr. Falconer himself says the last tooth is known
to be the ultimate molar ‘‘ by having the accessory spur or process which charac-
terizes this tooth this large accessory spur only occurs in the lower molars of the
Ungulata, to which series the specimen undoubtedly belongs. Falconer’s specimen
appears to have been lost, and no other specimens have hitherto been known till
the discovery of the specimen described below.

( 78 )

MOLAE TEETH AND OTHEE EEMAINS OE MAMMALIA.

The discovery of that specimen fully carries out Ealconer's idea of the generic
distinctness of the form^ although it seems to have been generally neglected by
subsequent writers.

Tetraoonodon magnum, Falconer. Plate 10.

The unique specimen of the greater part of the right ramus of the mandible
of this species, of which three views are given in the accompanying plate, has
already been shortly described in the “ Eecords of the Geological Survey of India”
{vol. JX, p. 101), but is now for the first time figured; the present description is
mainly copied from the above notice.

The specimen was discovered by Mr. Theobald in the Upper Siwaliks of Asnot
in the Panjab ; it contains the two last premolars, and the first and second molars ;
the whole of the penultimate and a portion of the ultimate premolar, together with
the first molar of the left side of the mandible, were also found with the specimen,
and it is probable that the whole jaw existed in the rock, but was broken up by the
native collector in extracting it. The first molar tooth is somewhat broken on its
anterior half ; the rest of the teeth are complete.

The second molar has only just commenced to be touched by wear on its
anterior tubercles ; the second premolar has a small dentine surface exposed on its
summit ; the summit of the first premolar is only blunted ; both these teeth have
not been protruded to their full height above the jaw ; the animal was evidently
adolescent at the time of its death.

The crown of the second molar, being the most complete, this tooth is here
selected for description ; the shape of the crown is oblong, having four conical or
mastoid prominences at its angles ; a cruciform valley divides these prominences
or cones ; the transverse portion of this valley is considerably the wider and deeper
of the two ; the extremities of this transverse valley extend downwards to the base
of the crown ; in the central hollow between the four cones there is a flat bi-lobed
tubercle ; a large talon-tubercle occupies the hindmost portion of the antero-pos-
terior valley, while a flatter tubercle occupies the anterior extremity of the same
valley ; there is a very small and conical tubercle at the outer extremity of the
transverse valley. The plane of wear of the molars, as shown on the first molar,
slopes slightly towards the outer side ; there is no trace of a cingulum on any por-
tion of the molars.

Turning now to the premolar teeth, of which these are the only known speci-
mens in existence, we find them of a most abnormal and interesting character ;
these teeth are of a considerably larger size than the true molars, a character
which is, I believe, unknown among other Ungulates^ though it occurs in the dental
system of Fteropus. The premolars are placed in direct contact with the molar
series, and in form have a general resemblance to the corresponding teeth of Hip-
popotamus, Merycopotamus, and AntJiracotheriwm ; each tooth is inserted into the
jaw by two fangs ; the penultimate premolar does not present any disc of pressure

( 79 )

MOLAE TEETH AND OTHER REMAINS OE MAMMALIA.

on its anterior surface ; this tooth was therefore probably separated by a diastema
from the preceding tooth, as in Hippopotamus.

The ultimate premolar has a nearly square base, from which rises a simple
oblique and compressed cone ; the summit of the cone is directed backwards and
placed somewhat in advance of the hindmost border of the crown; the anterior
face of the cone is produced into a sharp sinuous ridge, which extends from the sum-
mit nearly to the base ; below this the ridge bifurcates and forms a slight- cingulum
which extends along the greater part of the base of the anterior surface. A small
tubercle is placed between the summit of the cone and the posterior border of the
crown ; this tubercle forms the summit of a second cingulum which extends across
the posterior surface of the tooth ; this posterior cingulum slopes from the central
tubercle on either side towards the base of the crown ; the outer edge of this cingu-
lum forms a very distinct ledge at the postero« external angle of the crown ; a
rounded notch cuts into the outer and the inner sides of the crown, between the
roots of the fangs the inner surface of the tooth rises nearly vertically from the
jaw, while the outer surface slopes considerably towards the inner side. The summit
of the cone is worn obliquely, the facet of .dentine being directed upwards and
backwards ; the facet of wear presents two planes, the hinder of which is the most
oblique to the axis of the cone ; the form of the facet is oval, the long diameter
being placed in the line of the jaw. The enamel of the crown is arranged in
irregular branching ridges, which radiate from the summit of the cone to the
periphery of its base ; these ridges are again marked by fine parallel transverse strise.

The penultimate premolar is somewhat smaller than the last of the series, its
base approaches a triangular form, constricted laterally between the fangs, and
blunted at the apex ; the summit of the cone is placed slightly behind the centre
of the crown, and has a prominent ridge running to the base of the crown, on both
anterior and posterior surfaces ; there is a well-defined cingulum on the posterior
surface, which also extends along the hinder half of the internal surface. Like the
other premolar, the cone is convex and sloping on the external surface, and nearly
vertical on the internal surface ; this latter surface is divided by a vertical groove.

The dimensions of this specimen are as follows in inches and tenths : —

Length, of two molars ... ... ... ... ... 2‘50

Ditto second molar ... ... ... ... 1‘45

Width of ditto ditto... ... ... ... ... 1'30

Height of ditto ditto... ... ..." ... ... 0'80

Length of ultimate premolar ... ... ... ... 3‘15

Width of ditto ditto ... ... ... ... 2' 10

Height of ditto ditto ... ... ... ... 1‘80

Length of penultimate premolar ... ... ... ... 2'05

Width of ditto ditto ... ... ... ... 1'80

Height of ditto ditto .. ... ... ... 1’65

Depth of jaw at ultimate premolar, . . ... ... ... 3‘10

Length of the fragment ... ... ... ... 8*00

( 80 )

MOLAH TEETH AND OTHER REMAINS OE MAMMALIA.

The dimensions of Ealconer’s specimen are as follows : —

Length of the frag-ment
Ditto of the last molar .. .

Ditto ^ ditto excluding the spur

Greatest width of ditto ...

Height of posterior cones of last molar
Length of the anterior molar
Width of ditto ditto

Anterior width of top of crown of last molar ...

3-38

2-05

1'68

148

0- 75

1- 20
1-30
1-20

Erom the two specimens, we are able to say that the dentition of the mandible
of this species comprised three true molars, of which the last had a third lobe ;
in front of these there were two conical premolars, in close opposition; then
came a diastema, in front of which there was probably one premolar as in
Anthracotlierium, or two as in Kippopotamus ; of tbe canines or incisors we know
nothing. I should, however, imagine they were large in size, and that the incisors
w'^ere probably three in number. Dr. Ealconer instituted a careful comparison be-
tween the molars of his specimen and those of other genera of Suma ; he, however,
compared them with tlie upper instead of with the lower molars. Apart from the
premolars, which are of themselves amply sufiScient to distinguish the genus from
all other forms of Suina, the lower molars are also very characteristic teeth ; they are
the simplest teeth of any of the group ; their four simple cones, with the absence of a
collection of semi-distinct tubercles, and the slight degree of obliquity of the worn
crown, distinguish them from the molars of Sits. Erom the molars of Hippopotamus
they are distinguished by the worn-dentine surfaces of the cones, being approxi-
mately circular in form, and never trefoil- shaped. The lower molars of Antliraco-
therium are distinguished from those of this species by having the inner side of
their outer cones concave, thereby approximating to the Huminant type, by having
an incomplete cingulum on the outer side, and by the third lobe of the last molar
being larger and undivided.

The gigantic size of the premolars of this genus appears to be an extension
backwards of the ultra development of the anterior teeth which occurs in Hippo-
potamus, Siis, Fhacochcenis and Babirusa ; in the living genera this ultra develop-
ment is confined to the canines and incisors, while in the fossil genus it extends
to the premolars. The premolars of this species, though formed on the same plan
as those of Hippopotamus and Antliracotlierium, are of a more simple type, their
summits being more regular cones, without accessary columns, and lacking the
complete cingulum which invests the premolars of several species of Hippo-
potamus.

The teeth of this genus belong to a very simple and generalized type ; the true
molars belong to the typical bumdont form of Professor Leidy, while the premolars
are single cones of the trichecodont form of the same author. The bimodont form
Professor Leidy regards as the earliest and least specialised type of Mammalian

( 81 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

molars, although, as in man, it is occasionally persistent ; the great developement of
the premolars of Tetraconodon is, of course, a specialised character ; by an infolding
of the cones of the molars, and by heightening of their crowns, the teeth of Tetra-
conodon could be easily modified into the molars of Mippopotamus.

Order: EDENTATA.

Genus Manis (Pangolin, Gray.)

The announcement of the addition of the remains of a species of Udentata to
the Indian fossil fauna has been already made by me on a previous occasion
JRec. Geol. Surv., Ind^^ vol. IX, p. 106) ; the specimen on which this determina-
tion was made is now for the first time figured and described. The presence of a
representative of this abnormally distributed order in the Tertiaries of India was
only what might reasonably have been expected, and afi^ords another proof of the
common origin of the Indian and Ethiopian fauna in tertiary times ; of the four
modern Ethiopian genera, viz., Striithio, Hippopotamus, Camelopardalis and Manis,
which lived in India during the tertiary period, the latter genus is the only one
which has survived in that country down to the present day. The genus Manis
{and Tangolin) is now confined to Africa and India ; it was represented in Europe
during the older Pliocene period by one species {Pangolin gigantesque, Cuv.,
Macrotherium, Lartet), which was perhaps subgenerically distinct. Besides the one
Indian specimen and the Em’opean species, no other fossil forms of the family
are known.

Manis Sindiensis, n. sp. nohis. PI. 8, figs. 11 to 14.

The specimen on which the species is founded was collected last season by
Mr. Eedden from the Manchhar (Siwalik) beds of Sind ; it is the second phalange
of the third or middle digit of the manus, and is a very characteristic and unmis-
takeahle bone : four views of the specimen are given in the accompanying plate.

The superior surface of the bone {jig. 14) is ovate in form, and is divided by
a strong median ridge, which expands at both its extremities, the lateral margins
of the surface are also slightly raised, forming elliptical hollows on either side of
the median ridge j from this we infer that the distal surface of the first phalange
had a large, deeply grooved trochlea. The distal extremity of the bone forms a
trochlea, which extends almost to the borders of the superior surface {jigs. 11 and
13) ; the trochlea is divided into two prominent ridges by a very deep and rounded
median groove ; the ridges are slightly oblique to the long axis of the bone, and
on the posterior aspect converge slightly superiorly {jig. 13) ; the anterior and
posterior extremities of the superior surface are raised into prominences, of which
the anterior is much the higher {jig. 11) ; there is a slight depression on the
anterior and posterior surfaces at the superior termination of the trochlea. The
( 82 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

lateral surfaces of the bone are nearly flat. Below, the dimensions of the bone are

compared with those of the corresponding phalange of the
{Pangolin) indica, viz. : —

living

Indian Manis

Living.

Fossil.

Length, of anterior surface

0-62

1-10

Transverse diameter of dist. extremity

0-30

0-69

Antero-posterior do. do. do. . . .

0-42

0-95

Transverse diameter of superior surface

0-34

0-70

Antero-posterior do. do. do.

0-50

1'05

Erom the above it will be seen that the fossil bone is very slightly wider in
proportion to its length than in the living species : the only other distinction
between the two specimens is that in the bone of the living species the depression
for muscular attachment above the trochlea on the posterior surface is somewhat
larger and deeper than in the fossil specimen. The dimensions of the fossil bone
are about double the size of those of the recent hone, and consequently indicate
that the fossil animal was about double the size of the living species. The length
of the second phalange in the living species is equal to one-seventh of the length
of the cranium, and the length of the cranium is about equal to one-tenth of the
total length of the body ; taking these proportions, the head of the fossil species
would be about 7*7 inches in length (that of the recent Indian species being 4*2),
and the length of the whole body would be about seventy-seven inches, whereas
that of the living Indian species never exceeds forty-three inches, Ilanis {Pan-
golin) gigantea* of Western Africa attains a length of seventy inches, or rather less
than the Siwalik species. The present specimen is distinguished from the corre-
sponding bone of the Armadillos and Orycteropus by being much shorter in propor-
tion to its length ; from that of Ilyrmeeophaga by not being enclosed in the
terminal claw. It has nearly the same form as that of MegaloQiyx, but is shorter
in proportion to its length. With the phalanges of other Edentata it has few
points in common.

In referring this bone to the genus 3Ianis, I assume that the animal to which
it belonged must have had the same general organization as the living species ;
the fossil species was clearly fossorial, and being so, the presumption is that it did
not differ far from the living animal, though subsequent discoveries may prove that
there were differences in the form of some part of the osteology of the fossil
species which will require that it should eventually be placed under a new genus ;
other remains of this interesting fossil form will, I hope, be subsequently obtained :
the locality from which this fossil was obtained has already yielded several little
known forms to Mr. Eedden’s search, and it is from these Sind and Panjab districts
alone that the majority of new Siwalik fossils in the Indian Museum have been obtained.

* The name Mania gigantea appears to have been applied to two ditferent animals ; it was first applied by Illiger in 1811
to the living African species, and the name is used for this species in Gray’s catalogue of the Edentata in the British Museum
(1873). Subsequently the same name was applied by Holl in 1830 to Cuvier’s 'Pangolin gigantesqiie, (the latter name dating
from 1812) ; the name therefore belongs to the African species ; the fossil species was subsequently named by Lartet
Macroiherium giganteum. I have not had an opportunity of comparing our specimen with the corresponding bone of the
African M, gigantea ; it is much smaller than that of Macrotlierium,

( 83 )

MOLAU TEETH AND OTHER REMAINS OE MAMMALIA.

Order: CARNIVORA.

Genus Amphicyon, Lartet.

Amphicton pal^indicus, n. sp. Falc. et nobis. Plate 7, figs. 5, 8, 12.

This genus of extinct Subursoid mammalia^ distinguished from their living
congeners by having been furnished with three instead of two tubercular molars
in the upper jaw, has hitherto been definitely known only from the miocene strata
of Europe, where four well-marked forms have been determined, viz.^ A. major,
A. minor, A. dominans, Meyer, and A. lielveticus, Pietet, and from the Tertiaries
of Nebraska, from which two species, A. vetus and A. gracilis, have been described
by Professor Leidy.

No specimens of the genus were known to Dr. Ealconer at the time of pub--
lishing the “ Eauna Antiqua Sivalensis” ; subsequently, however, among a collec-
tion of mammalian remains sent to him for determination by Dr. Oldham from
tertiary strata at Kushalghar below Attock, Dr. Ealconer {see “ Palceontological
Memoirsj^ vol. I, p>- 1^16) discovered a tubercular upper molar tooth, which he said
must have belonged to a carnivorous animal, as large as the Folar Bear, and
allied to the Ampliioyon. This tooth was, with the other specimens from the same
locality, subsequently returned to the Indian Museum, where it now remains ;
for the first time I give a figure and description of the specimen {Bl. IV, jig. 5).
N o specific name ajopears in the account of the specimen in the “ Palaeontological
Memoirs,” but in a manuscript of Dr. Ealconer’s on this specimen given to me
by Dr. Oldham, the species is named A. palceindicus. It will be observed that
in its general form the tooth corresponds with that of the European forms of the
genus, but that it presents certain modifications of detail by which it is distin-
guished : these modifications, however, do not appear to me to justify the placing
of the specimen in a separate genus or sub-genus, and I have therefore retained
it under Ampliicyon ; it is, however, I think, doubtful from Ealconer’s remark on
the specimen, as being “ nearly allied to Ampliicyonj’ whether he himself intended
to place it in that genus.

In addition to the above specimen, we have in the Indian Museum another
tooth of Ampliicyon {Bl. lY,jig8. 8, 12) : this second specimen is the “carnassial”
tooth of the lower jaw — it is of such a size that it seems too large to have been the
lower molar of the same species to which the upper molar belonged ; to prevent,
however, the unnecessary multiplication of species, I have, for the present at all
events, referred it to the same species as the former specimen : this tooth was
brought by Mr. Medlicott from the Siwalik strata of Nurpur.

Reverting to the upper molar, we find that the specimen is the second true
molar of the right maxilla ; it is quite perfect, and belonged to an adolescent
animal, as none of the points of the lobes have been touched by wear.

The outline of the crown is nearly in the form of an isosceles triangle, of which
the outer border forms the base : the apex of the triangle is rounded off. The
( 81 )

MOLAR TEETH AND OTHER REMAINS OF ’MAMMALIA.

crown-surface is divided into two portions, a blade and tubercular portion, the
former occupying the outer side. The blade is divided into two unequal lobes by
a transverse valley (by this unequal division of the blade the tooth is distinguished
from the first molar, where the two lobes are symmetrical) ; the two lobes form
regular blunted cones, the anterior of which is nearly double the size of the
posterior.

The tubercular portion of the crown, which commences at the inner base of
these lohes, has an approximately flat general surface ; the inner border of this
surface is surrounded by a flattened semi-circular rim about a quarter-of-an-inch
in width at its broadest part ; within the concavity of this rim, and separated from
it by a portion of the general flat surface of the crown, there is placed a very low
and blunt cone, slightly elongated antero-posteriorly.

A narrow cingulum embraces the outer borders of the blade. In the following
table the dimensions of this specimen in inches and tenths (in the first column)
are compared with those of the corresponding tooth of the European Amphicyon
major (in the second column)

Length of blade ...

... 0-76

0-87

Length at centre of tubercular portion

.. ... 0-55

0-75

Greatest width

... 1-13

1-3

Height of anterior lobe of blade

... 0-4

o-sa

From this it will be seen that the present specimen in all its dimensions is
of rather smaller size than the corresponding molar of AmpMcyon major ; taking
the latter species as a typical fornl of the genus, we may compare the teeth of the
two species. {A figure of the upper molars of A. major is given in J)e Blainmlle's
Osteographie^^ Atlas, vol. If Sibursi, pi. XIV.) The general form of the blade is
very similar in the two specimens ; the anterior cone, however, is rather larger in
proportion to the posterior in the Indian than in the European form, and in the
latter the two cones are divided by a deeper channel on the outer side. The
greatest difference in form occurs in the tubercular portions of the two teeth ; in
place of the simple cone situated within the raised rim of the innermost border
of the crown of A, palmidicus, we find in A. major a second semi-circular elevation
placed within the first ; this central semi-circle runs up to join the angles of the
lobes of the blade ; at the base of the hindermost of these lobes, the central semi-
circle is raised into another smaller cone ; finally the semi-circular rim on the
internal border of the crown is crenulated in the European form, while it is quite
simple in the Indian.

These differences give a very marked variety in the contour of the two teeth ;
if* however, we examine the corresponding tooth of another and smaller European
species — -A. dominans of Meyer — we find that the central semi-circular rim only
exists on the anterior half of the crown, the posterior half of which is quite level;
this anterior ridge terminates in a small blunted cone, occupying the same relative
position on the crown as does the isolated cone on the crown of the Indian specimen.

( 85 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

It will, therefore, be seen that there is a transition from the complete semi-
circular ridge, which occupies the centre of the crown of the tooth of A. major, to
the single cone occupying the place of the middle of this arc in the tooth of
A. palcBndicus.

The present specimen, besides other characters, is at once distinguished from
the other three European and the American species by its much larger size, from
A. helveticas {see Pictet JBaleontologie Suisse vol. V, p. 135) it is further
distinguished by the small size of the cingulum surrounding the blade.

The lower molar belonged to a somewhat larger animal than the upper molar.
The specimen comprises a portion of the right ramus of the mandible containing
two teeth ; the first of these is the last of the milk-molar series ; the summit of the
second lobe of the blade has been broken away ; the crown is considerably worn.
The second and much larger tooth is the first of the permanent molar series, —the
“carnassial” ; it is quite perfect; none of the casps have been touched by wear, and
the jaw has been partly chisselled away to expose the base of the crown.

Viewing the carnassial from the outer side {FI. 7, jig. 8) we may call the whole
of the outer surface the blade of the tooth; this blade is separated into three
portions, of which the middle one is the highest ; the anterior and posterior lobes
are nearly of the same height, by which character the carnassial of Ampliinyon is
distinguished from that of Canis, in which the third lobe of the blade is very low,
and almost forms a part of the tubercular portion of the crown. A narrow vertical
groove runs down the postero-external angle of the median lobe. The anterior and
median lobes are separated only by a slight notch on their upper border ; a deep
valley, extending down nearly to the base of the crown, separates the median from
the third lobe. The anterior lobe presents a simple trenchant edge ; the median
lobe is placed somewhat obliquely to the long axis of the crown, and has a narrow
trihedral accessory-column at its postero-internal angle.

The third lobe is a laterally-fiattened cone, with a fore-and-aft cutting edge ;
at the base of its internal surface there is a narrow flat ledge which forms the only
truly tubercular portion of the crown. The third lobe, the accessory-column of the
median lobe, and the tubercle, enclose a hollow triangular space between their
respective bases.

The last molar of the deciduous series presents in miniature the form of the
larger tooth ; it agrees precisely both in shape and size with the permanent car-
nassial of the European A. dominans.

The dimensions of the larger tooth of the Indian species (in the first column)
are compared with those of the lower carnassial of the European A. major (in the
second column) —

In.

Length of tooth

... 1-32

1-2

Width at hinder extremity

... 0-6

0-5

Height of central lobe

... 0-9

0-7

( 86 )

MOLAR TEETH AND OTHER REMAINS OE MAMMALIA.

The lower molar of the Indian form is, therefore, as much larger than the
corresponding tooth of A. major ^ as the upper is smaller ; assuming that the pro-
portion between the lower and upper teeth is the same in the Indian form as in the
European, the width of the upper molar corresponding to this specimen should be
1*4 inches, whereas the width of the Attock specimen is only 1‘12 inches ; from
this I think it quite probable that there 'may be two Indian species of the genus.

The present specimen is chiefly distinguished (in addition to its greater
absolute size) from the carnassial of A, major by the greater proportionate size and
thickness of the third lobe, and by the presence of the groove on the outer side of the
middle lobe. The depth of the jaw in this specimen is nearly the same as in A.
major. Erom all the other small European and American species, the large size of
the present specimen is a sufficient distinction.

( S7 )

:• . ''•ift'i’4^ f •■^^.^^: _• c;. ^ ‘ J''‘tV' ’ ^'jjU \'^v .n^Xtv

< .tny ,i((. V ...■ ■•,. < ■•. ■ v:niJ/H<i t i,>i* V \. ’ ; '.('i()>^i‘ ' :-f.7;'i’i ^

-’i;.i ,j.'':yK’:v> ■ ^ m; ;; :V>' 7 c, /^. -Mj-,--;/

. cm'^'^'I^i-U.’- ,Ai :.v ■ ■

tv '. t •.

■ '^■T’-'-'-^'^'' v'.ffeLaiLJvri::: , I.;.;'.,) hi, i-^-f q/l’i’ ' '

prv.(r ;■,■;'■•■ ■ , > iV

:/3.; ^

*^ * * . . .‘

W*'’

’.(!

CRANIA OF RUMINANTS FROM THE INDIAN TERTIARES.

Bt r. lydekkeb, b.a.

GEOEOGICAL SURVEY OF INDIA.

[WITH PLATES XI— XXVIII.]

Among a series of (chiefly) unpublished notes and plates,* intended for the com-
pletion of the “ Pauna Antiqua Sivalensis,” the late Dr. Dalconer has indicated
the existence of a large and varied Ruminant Dauna which inhabited India in
tertiary times ; this fauna must have been at least equal in extent in the number of
peculiar forms to the well-known Proboscidian Pauna of the same epoch. More or
less complete remains of certain of these Ruminants, such as SivatTierium, Brama-
thermm, Camehis, Camelopardalis, and the Bovina of the Nerhudda valley, have
aReady been figured and partly described in the “ Palaeontological Memoirs ”
and in the publications of the Asiatic Society of Bengal. Pigures of the crania of
several species of Bovina have been drawn or engraved among the unpublished plates
of the “ Pauna Antiqua Sivalensis.” A notice of all the species of Indian Possil
Ruminants known to Palconer will he found in the first volume of the “ Palaeonto-
logical Memoirs.”

In the present memoir I have given figures and descriptions of aU the Bovoid
Ruminants at present contained in the collection of the Indian Museum, identifying
as many of them as agree with such of Palconer’s species of which there is any
figure or description extant : many, however, of Palconer’s species exist only on the
authority of manuscript names without figures or descriptions : these species I have,
of course, been obliged to completely reject. The total number of species of Bovoid
Ruminants named by Dr. Palconer from the Siwahks was six, viz., Bison sivalensis.
Bos occipitalis, Semibos triquetriceros, Ampliihos acuticornis, AmpJiibos elatus, and
AmpJiibos antilopinus: figures of the crania of the third and fourth of these species
are given in the unpublished plates of the ‘‘ Pauna Antiqua Sivalensis ” {Blates G.
and JA.), while measijrements of the crania are also given in the accompanying
descriptions of the plates : of Semibos triquetriceros we have several crania in the
Indian Museum ; and several also of Ampliibos acuticornis. Of the cranium of Bison
sivalensis there is an outline figure in the drawings intended for the completion of

' Of these plates, the Indian Museum has recently acquired photographic copies, which have been executed
under the superintendence of the British Museum Authorities.

89—2 CRANIA OR RUMINANTS PROM THE INDIAN TERTIARIES.

the “ Pauna Antiqua Sivalensis ” {Plate A, sheet 10) ; of this figure I have obtained
a copy by the courtesy of the British Museum authorities, and find that it corre-
sponds in character with a cranium of Bison in the Indian Museum collected by Mr.
Theobald, and which I have accordingly referred to this species. Of Amphibos elatus
and Amphibos antilopinus I cannot find any descriptions or figures, and the names
must therefore be abandoned. Of Bos occipitalis we have likewise no certain means
of identification ; Dr. Murchison, the editor of the “ Palaeontological Memoirs,” has,
however, suggested in a marginal note {Vol. 1, p. 281) that this name may refer to
a cranium of a Bovine Ruminant contained in the Siwahk collection of the Asiatic
Society of Bengal, which was shortly described by Palconer in the Catalogue of the
Collection, but to which, he, at that time at least, applied no specific name. Whether
this name could have been applied to the specimen in question depends in a great
measure on the respective dates of the manuscript note in which the name occurs
and of the Museum Catalogue : the date of the latter is 1859 ; if the note were
written previously to that date, it is quite clear that the name did not apply to the
specimen, or it would have appeared in the Catalogue. As I have no means of
judging on this question, I have thought it would prevent confusion to follow
Dr. Murchison and to continue to apply the specific name to the cranium in
question; the generic name I have, however, changed to Beribos, as the species
cannot be referred to the restricted genus Bos.

Besides the Bovidce, Palconer indicated the existence of three species of
Siwalik Antelopes ; two of these species, viz., Antilope gyricornis and Antilope picta,
are known only by manuscript names and must consequently be abandoned ; the
cranium of the third species. Antilope palceindica, is, however, figured in the
“ Palaeontological Memoirs ” ( Vol. 1, pi. 23), and the species may consequently
stand.

The peculiar form, Kemibos triquetriceros, seems to have been a small Bovoid
animal, forming a connecting link between the Oxen and the Goats. Amphibos
acnticornis, on the other hand, in the form of its horn-cores, connected the Oxen and
the Antelopes ; it was perhaps allied to the Bovoid Anoa depressicornis of the
Island of Celebes.

Of the true Bovidce there are at the present day seven well-marked species
inhabiting South-eastern Asia ; these are, Bubalus arni inhabiting the Sal forests
of Assam, the Nepal Terai, parts of upper Bengal, the Sunderbans, and the high-
lands of Central India ; Bibos gaurus inhabiting the eastern Sub-Himalayas, the
Nerbudda valley and adjoining hills, and parts of Burma; Bibos frontalis inhabiting
the hilly districts to the eastward of the Brahmaputra ; Bibos banting inhabiting
Burma and Pegu ; the domesticated Bos indicus of the plains ; Bos chinensis, partly
domesticated over a great part of China ; and lastly. Bison (BoephagusJ grunniens
of Tibet.

There is no true taurine ox at the present time living anywhere in Asia, the
aberrant Bos indicus being the only representative in India of the genus Bos as

OEANIA OF EUMINANTS FEOM THE INDIAN TEETIAEIES. 3—90

restricted by Hodgson and Gray. With the exception of the Buffalo all the wild
cattle of India proper belong to the genus Bibos : I have not, however, in the
Indian Tertiaries found the remains of any fossil species which belongs to this
characteristic Indian group, though several of the fossil oxen approach this group in
many points of structure of the cranium.

Of the restricted genus Bos, I have described the crania of four fossil species,
one of which is from the valley of the Nerbudda, and the other three are from the
Sub-Himalayan Siwahks ; two species from the latter area, viz., Bos acutifrons and
Bos planifrons, will be found to differ to a certain extent in the form of the occipital
surface from the crania of any living species of oxen ; this divergence from the type
form will be found carried to a stiU. greater extent in the occiput of Bos nama-
dicus from the Nerbudda valley ; and the divergence is greatest of all in the living
Indian wild cattle composing the genus Bibos. It is an interesting fact to note
that the peculiarly formed occipital region of the living genus Bibos, at present
exclusively confined to the oriental region, is found foreshadowed among the
Indian fossil species alone of the many forms of the genus Bos ; in one species at
least (B. namadicus) this abnormal form of the occiput is correlated with small
premaxillse like those of Bibos, which are never found in any European ox.

It may therefore not be improbable that Bos namadicus may have been one of
the progenitors of the living wild Indian cattle.

Bos planifrons, a new species from the Siwaliks, presents some points of
resemblance to the Nerbudda species, but does not come very close to any living
species. The long-horned Bos acutifrons is widely different from all other recent
and fossil species, and presents certain affinities to Bubalm — the same remark will
apply to the third Siwalik species, Bos platyrhinus.

Bison sivalensis may probably be considered as the direct ancestor of the living
Himalayan Bison {Boephagus) grunniens.

The living Bubalus arni of India is without doubt the direct lineal descendant
of the Bubalus palcEindicm of the gravels of the Nerbudda, and of the top-

most beds of the Siwaliks. With regard to the propriety of making distinct species
out of these two closely allied forms, we may note that the living wild cattle of the
English parks and the aurochs of Lithuania are generally regarded as the direct and
but little altered descendants of the fossil Bos primigenim and Bison prisons, and
yet are frequently considered as distinct species in the modern modified acceptation
of the term. It appears to me to be the best course to give a distinct specific name
to every distinguishable form ; since it is only by describing the small varieties
which exist between allied forms of the same type of animals that we are enabled to
obtain evidence of the descent of one so-called species from another. It may here-
after be advisable to invent a title which shall comprehend in one group all those
so-called species which can be proved to be the lineal descendants of one original
form ; the term species may either be taken in this larger sense or as the name of
distinguishable forms of one period only.

91—4 CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES.

The second fossil Indian species of the genus, Buhalus platyceros, is from the
Siwaliks, and differs in the form of its cranium from that of any living species of
the genus ; this difference mainly consists in the long upward prolongation of the
frontals between the bases of the horn-cores and in the want of the high and
vertical occiput ; by these peculiarities it shows an affinity to the Antelopes which
Professor Rutimeyer regards as the progenitors of the true Oxen.

Among the fossils from the mammaliferous beds of the Irawadi River now
contained in the Collection of the Indian Museum, there are several fragments of
cylindrical horn-cores of small size, together with a single upper molar tooth ; these
specimens evidently belonged to a species of Bovoid Ruminant ; but they are of
course too fragmentary, even for generic distinction ; it wRl be. interesting if more
complete remains should hereafter be discovered to determine whether or no the
fossil ox of the Irawadi beds was at all related to Bibos banting now inhabiting
the same regions.

The crania of nearly all the species of fossil Indian cattle show certain modi-
fications of form which render them more or less aberrant from the living types of
their respective genera, making it a matter of considerable doubt in some cases
whether to refer them to the existing genera or to make new genera or sub-genera
for their reception. When possible, I have generally placed the fossil species under
the recent genus, if there were several important characters common to the living
and fossil types ; not unfrequently, however, the fossil species will not bear a rigid
interpretation of all the characters of the genus under wliich it has been placed ;
this will be found to be the case with Bos namadicus. Bos acutifrons, and Bubalus
platyceros. A difficulty of this kind must always be expected to arise when we
have to deal with a large series of recent and fossil forms, and is one which tends
to do away with some of the sub-genera of modern Zoologists which are only
founded on minute and unimportant points of detail.

The Siwalik being one of the largest fossil Bovine fauna hitherto described, it
would not be unreasonable to suppose that it would be also the most rich in the
number of genera ; this we find in fact to be the case ; the three chief groups of
living oxen, viz., Bos, Bison, and Bubalus, being each represented by one or more
species, and there are, besides, several extinct genera or sub-genera unknown in
other deposits. In no region of the world at the present time are the three above-
mentioned genera found living together, though there is the nearest approach to it
in modern India, where Bibos (representing Bos) and Bubalus are found in the
plains, and Bison represented by Boepliagus in the Himalayas, the latter separated,
however, completely from the genera of the plains.

The crania of the fossil Indian oxen tend to do away with the distinctions
which separate the genus Bibos from Bos ; the three genera. Bos, Bison, and
Bubalus, are found to have possessed their most important craniological distinctions
in Siwalik times, and to have preserved them with some slight modifications up to
the present day. This is quite in accordance with the discoveries of Palaeontologists

CEANIA or EUMINANTS EEOM THE INDIAN TEETIAEIES. 5—92

in other countries. Professor Eutimeyer, in his “ Ancient Euminant Eauna of
Switzerland,” considers the derivation of the three genera to have taken place in the
following order, commencing from the oldest, mg., Bubalus, Bison, and Bos ; the
high frontal ridge of the latter genus the Professor considers as a highly specialised
character, which did not make its appearance in Europe till the post-Pliocene
period ; in India this character must have been differentiated at an earlier period,
as the remains of Bos are found throughout the Siwalik, which we cannot regard as
being newer than older Phocene.

The remains of Sheep have hitherto been met with very rarely in the Indian
Tertiaries ; we have no remains of this group in the Indian Museum. Goats are
also rare, hut I have distinguished three species, one of which was probably the
ancestor of the living Markhoor, while another seems more closely allied to the
Nilghiri Semiti^agus hylocrinus ; the third species does not show afffnity to any living
Indian form.

Including the new species described in the present Memoir the list of Indian
Tertiary Euminants will be as follows : —

PECORA.

C Bos namadicus — Nerbudda.

I Bos planifrons—^i^aWk.

Bovid® ...“{ Bos acutifrons — Siwalik.

I Bos 'platyrhinus — Siwalik.

^.Buhalus platyceros — Siwalik.

Antilopid^ ... f Siwalik.

I Antilope patuUcornis — Siwalik.

„ f Sivatherium qiqanteum — Siwalik.

SIVATHEEID.2E ... 5 , . , . _

Vishmdhertum iraoadtcum — Burma.

C.vMELOPAEDALiDiE. Camelopardalis sivalensis — Siwalik,
c Capra sivalensis — Siwalik.

I Capra perimensis — Perim Island.

Ovis sp. — Siwalik (Blylh).
f Cervus latidens — Siwalik.

I Cervus triplidens^ — Siwalik.

I Cervus simplicidens — Siwalik.
{^Cervus sp. — Burma,

TYLOPODA.

Camelus Siwalik.

Buhalus 'palceindicus — Nerbudda and Up. Siwalik.
Bison sivalensis — Siwalik.

Peribos occipitalis — Siwalik.

Semihos triquetriceros — Siwalik.

AmpJiibos acuticornis — Siwalik.

Antilope sivalensis —Siwalik.

Antilope porrecticornis — Siwalik.

Bramatlierium perimense — Perim Island.
Sydaspitherium megacephalum~ii\\^&Y\k.

Capra sp. — Siwalik.

Cervus sp. — Nerbudda.

Dorcatherium minus — Siwalik and Mancbhars.
Dorcatlierium majus — Siwalik, and Sind and Attock
Beds.

CAPRica:

Ovid.®

Cervid.®

Camelid.®.

The three species of oxen. Bos acutifrons. Bos planifrons, and Buhalus platy-
ceros, and also the peculiar genus Hydaspidotherium or Kydaspitherium, have been
already named and noticed by me in the Eecords of the Geological Survey of India.'^
In regard to the distribution in time of the fossil Indian genera of Euminants,
we find that in Europe the remains of the restricted genus Bos are only found in
Pleistocene and recent strata : in Nicholson’s Palaeontology (p. 439) it is stated that

‘ The teeth of this species are so like those of Bos that it may be well to mention that they h.ave been found
attached to a part of an undoubted cervine skull.

2 Vol. IX, p. 154, and X, p. 30.

93—6 CEANIA OE E¥MINANTS EROM THE INDIAN TEETIARIES.

“ a few remains of Bovidce have been found in deposits of Pliocene age, but the
oxen are essentially post-PUocene and Becent the same remarks apply to the
genera Buhalus and Bison. Professor Dana,’ in noticing the Mammalia of the
Siwaliks, considers these deposits to he of upper Miocene age, but adds the significant
remark : “ Bos and the related genera probably occur nowhere earlier than the
Pliocene,” apparently not being aware that the remains of these genera are found
in the same deposits as oi Swatherium Acerotlierium ; the whole series

will consequently he of Pliocene age. As I have already stated elsewhere,® the
presence of a large assemblage of hovoid genera in the Siwaliks is a most important
argument for the Pliocene age of that formation ; they occur in the same beds with
Cludicotlieriiim and other Miocene forms, wdiich latter must have lived down to a
later period in India than in Europe. This probable survival of ChalicotJierium (and
what applies to one genus applies to another) in Asia down to times later than the
Miocene has been akeady noticed by Professor Owen^ in his description of certain
Eossil Chinese Mammalia. That author, who followed Dr. Ealconer in considering
tlie Siwaliks as of exclusively Miocene age, remarks : “ If the Anoplotherioid

molar had not been in the series, such series would have been referred, without
hesitation, to a geological period not older than the upper Pliocene, and with a
possibility of post-Pliocene age.

“ I accept the evidence of the majority of the fossils, with the older alternative,
and conclude that this particular Anoplotherioid Artiodactyle, which has departed
from the generalised character of the type-genus by the suppression of a premolar
on each side of both jaws, and the commencement of a diastema or break in the
dental series, continued to exist in China, until the Pliocene division of tertiary
time, perhaps to a late period of that division.”

The above view quite concurs with my interpretation of the value of the different
genera in determining the age of the Indian upper tertiaries ; at the same time it
seems not improbable that some Siwahk fossils, such as those of Sind and Kushalgar,
do really belong to the upper Miocene period, there being no lithological break in
the latter area between the Nummulitics and the Siwahks.

The genera, or sub-genera, Semibos, Ampliibos, and Beribos, are exclusively
confined to the Siwaliks ; they belong to more generalised types than the true oxen,
and therefore claim affinities with older forms. The genus Antilope or an allied
sub-genus just occurs in the Miocene strata of Attica and Montpelier, and has lived
down to the present time. The four allied genera composing the peculiar family of
the 8wat}ierid(B are confined to the Indian tertiaries. The genus Camelopardalis is
first known from the Miocene of Attica, and now exists in Africa only. The genera
Capra and Oris have in Europe been discovered in post-tertiary formations only,
and therefore, like the Bovidce, give an exceedingly modern facies to the fauna in

^ Manual of Geology, p. 520.

^ Records of the Geological Survey of India, Vol. IX, p. 98.
3 Quar. Journ. Geol. Soc., Vol. XXVI, p. 432.

CEANIA OE RUMINANTS FROM THE INDIAN TERTIARIES. 7—94

which they occur : the presence of these fossils in the Siwaliks in company with
Chalicotherium and Dorcatherium must induce us to believe that these genera
originated in India in some part of the Pliocene period, and did not reach Europe
until the post-tertiary period. The genus Genus seems to have been represented in
the upper Miocene of Europe, but did nob attain to importance until the Pleistocene
period ; it belongs to an essentially modern group ; the alHed genus Dorcathermm
in Europe is confined to the Miocene period, but lived in the Siwalik period of India.
The genus Gamelus is unknown in a fossil strata except in India, though allied
forms like Frocamelus occur in the upper tertiaries of America ; the genus is probably
an essentially modern one.

The fossil Ruminants of the Indian tertiaries therefore all point very strongly
to the modern age of the deposits, there being only one exclusively Miocene genus
among them. Were we to study the Indian tertiaries from this group alone, we
should be compelled to place the Siwaliks in the upper Pliocene period at the
earliest ; but this view must be modified by the greater number of Miocene forms
which we find in other groups of the Mammalia.

In a former paper on the Fossil Mammalia of India and Burma,' I stated
that the number of extinct genera of Mammalia in these beds (Siwaliks) is so
large” that on these grounds we might consider them to be of Miocene age: this
statement was made from the comparison of the Mammalia fauna of the Indian and
European later tertiaries, and is true if we confine ourselves to those regions ; if,
however, we extend our comparisons to the later tertiaries and post-tertiaries of
America and Australia, we shall find that in the latter countries a vast number
of extinct mammalian genera are found in strata of Pleistocene age, while in
Europe aU the genera except Megaeeros from the Pleistocene, and most of those
from the Pliocene, are still existing on the globe. Thus, in the Pleistocene of America
we have among others the extinct genera Megatherium, Mylodon, Meg along x,
Scelidotherium, Glyptodon, and Mastodon ; while in the corresponding strata of
Australia we have Diprotodon, Nototherium, and Thylacoleo ; again, to take an
instance from the class of birds, we have in very modern strata in New Zealand,
Dinornis, Talapteryx, and several others. The proportionate number of extinct
genera in the Siwaliks cannot therefore be considered as being of any great value in
determining their age ; the only evidence which tends to place these deposits in the
Miocene period is the presence of genera characteristic of this period in Europe ;
their occurrence, however, as I have before said, is to be explained by their later
survival in India, and its value as evidence is over-balanced by the large percentage
of strictly modern genera in the Siwaliks.

The plates in this part all bear the name of the native artist who has executed
the outlines and a considerable part of the shading ; in many instances, however,
the finishing touches have been given by Mr. Schaumburg, the Artist to the Geolo-
gical Survey of India.

Eec. Geol. Surv. India, Vol. IX, p. 97.

95—8 CRANIA OR RUMINANTS UROM THE INDIAN TERTIARIES.

Eefore proceeding to specific descriptions, it may be well to state that in the
present work I have used generic terms in contradistinction to sub-generic terms,
in all cases ; one reason for this is, that the crania of the different forms of extinct
oxen generally differ from one another, more than do the crania of JBos and Buhalus,
which latter are frequently considered as generic, and not as sub-generic divisions.
If sub- generic terms were to be used in any case among the oxen, it appears to me
that they should be used in all cases, and that consequently all the forms of oxen'
described in the present Memoir should then be considered as belonging to different
sub- genera of one large genus Bos. I, however, prefer to use the term Bovidce,
somewhat in the latter sense, and to place the distinct types, as I have done below,
in distinct genera. The reader can, if he prefers, of course consider the generic
terms of the Bovidce employed here as being merely of sub-generic value, without
interfering with the general scheme of the classification employed.

ORDER UNGULATA : SECTION SELENODONTA ; DIVISION ARTIODACTYLA.

Family : B Q VII) FE.

Genus : BOS.

The genus Bos may be defined from the characters of the cranium as follows : —

Horn-cores placed immediately over the plane of the occiput ; occipital crest
extending high up between the bases of the horn-cores, and the occiput generally
squared, and with very slight lateral indentations of the temporal fossae : forehead
flat or mesially ridged, longer than broad, and frontal longer than facial portion of
cranium ; superior border of horn-cores at first convex ; in typical species the inter-
cornual space is straight and the horn-cores are cylindrical; in some aberrant
varieties the horn-cores are compressed, and the intercornual space is somewhat
arcuated.

It will be observed that the above definition differs in several points from those
given by Hodgson^ and Gray^ ; these alterations have been rendered necessary by
tlie new species now introduced into the genus.

Bos NAMADicus, Fcilconer. Pis. 11 and 16, f. 1 & 3.

Of the cranium of this, one of the best known species of the fossil Indian oxen,
figures and an imperfect description have affeady appeared in the Journal of the
Asiatic Society of Bengal {Vol. X, FI. F,figs. a, 6), and figures have also been given
in the unpublished plates of the “Eauna Antiqua Sivalensis” {PI. G,Jigs. 1 a, b,c, 2

' With the exception, perhaps, of Amphibos.

^ Asiatic Society of Bengal, Vol. X, p. 452.

® Catalogue of Ungulata in the British iliiseum.

CEANIA OP EUMINANTS PEOM THE INDIAN TEETIARIES. 9—96

a, 5, c) ; copies of two of these figures have been reproduced in the “ Palseontological
Memoirs” of Dr. Palconer {Vol. I, JPl. XXII) ; short descriptions of other crania of
this species are also given by Dr. Palconer in the Catalogue of the Manunahan fossils in
the Collection of the Asiatic Society of Bengal; these descriptions have been likewise
copied in the “Palaeontological Memoirs” {Vol. I,p. 286).

In the accompanying plates I have given a figure of a very perfect cranium
{PI. XI) of this species obtained from the Nerhudda valley, and now in the Indian
Museum ; this specimen, though far more perfect, agrees exactly in all its characters
with the specimen of the cranium figured in Plate G, fig. 2 a, h, c, of the unpuljlished
plates of the “Pauna Antiqua Sivalensis,” and which I take as the type of the
species; an occi23ital view of the same specimen is given in Plate XVI, fig. 1. The
specimen of the frontlet and horn-cores of another sjiecies of ox of which a front
view is given in Plate XII, fig. 2, and an occijDital view in Plate XVI, fig. 4, was
obtained by Mr. Theobald from tlie Siwaliks ; this specimen, especially in the form of
its occi2)ital region, seems at first sight to differ very widely from the NerlDudda
species ; if, however, we refer to the second specimen figured in the “ Pauna Antiqua
Sivalensis” {PI. G,fig. 1 a, b, c), and of which the front view is given in the Palaeon-
tological Memoirs” {Vol. I, PI. XXII, fig. 4), while the occipital view is copied in
the 2)resent Memoir (P^. ATPr, ^^5^. 5), we shall find that that sjiecimen exhibits
characters intermediate between the Siwalik and the type Nerhudda species, and
forms, therefore, a connecting link between the two, though the form of the horn-
cores in the two is a sufiicient specific distinction.

I will first of all shortly describe the figured Nerhudda cranium, and then point
out in what respects it differs from those of other species of oxen. The forehead is
long and narrow, and equal in length to the face ; between the horn-cores it is
almost flat, while at the level of the orbits it is slightly concave in the middle line ;
the orbits are relatively of large size, their anterior borders are placed very slightly
behind the plane of the f rentals, and are approximately parallel to the long axis of
the cranium, the axis of the orbit consequently looking almost directly outwards .
The supra-orbital foramina ]Dierce the frontals at right angles, inferiorly they are
continued mto long and deej) sulci, which run nearly ]oarallel to each other, and
extend downwards as far as the inferior border of the orbit ; the surface of the
frontals is somewhat more elevated on the inner than on the outer side of these sulci,
which are upwards of six inches in length.

The maxElse run nearly parallel to each other for a length of about five inches
and then contract somewhat suddenly in width, a well-marked tuberosity being
placed at the point of contraction. The nasals are long and somewhat arched from
side to side ; between these bones and the maxilla there occurs a long triangular
vacant space ; the apex of the nasals extends upwards only as far as the lower border
of the orbit ; a long interval separates the latter from the base of the horn-core.

The ridge between the horn-cores has a large ovate tuberosity occupying its
middle third ; in some specimens this tuberosity is less marked than in the figured

97—10 CEANIA OE EUMINANTS EEOM THE INDIAN TEETIAEIES.

cranium, and the intercornual ridge consequently then forms an almost unbroken
arch. At the upper part of the forehead, the frontals slope slightly backwards to
join the intercornual ridge, which is placed rather behind the plane of the face ; this
ridge is produced ujowards and backwards so as to overarch to a considerable extent
that part of the occipital surface intended for the attachment of the vertebro- cranial
muscles ; from its highest point the intercornual ridge slopes gradually downwards
and backwards to join the occipital crest ; the surface of that part of the occiput
which is above the crest being in consequence placed posteriorly to that part which
is below the crest. The occipital crest itself forms a continuous and unbroken arch,
which is separated by a considerable interval from the intercornual ridge. The
paroccipital process of either side is separated by a deep groove from the occipital
condyle. The bases of the horn-cores do not extend more than one-and-a-half inches
below the highest point of the occipital crest or curved line ; and there is a wide
interval between the former and the lateral borders of the crest. The temporal
fossae form small triangular notches immediately below the bases of the horn-cores
on the occipital surface ; the plane of the latter forms an acute angle with the plane
of the frontals. The palate is in too imperfect a condition for accurate description ;
it is, however, but slightly produced backwards behind the molar series, and the
line of the latter is very concave internally ; the basioccipital is triangular in form
as in the typical oxen, and carries four tubercles for muscular attachment, of which
the posterior pair are the .larger and are separated by a wider interval than are the
anterior pair.

The horn-cores are placed on the liighest ridge of the skull, immediately above
the plane of the occiput ; in cross section {FI. XI, figs. 3 8f 4) they are nearly circu-
lar in the middle, hut have a slight ridge along their superior border ; at their base
they are somewhat slightly compressed from side to side, the cross section being pear-
shaped at this point ; they diminish in width very gradually ; their direction is at
first upwards, outwards, and slightly forwards, then nearly directly forwards, and
finally inwards and slightly upwards ; at their extremities they are placed at a dis-
tance of nearly two feet in advance of the plane of the frontals. The measurements
of the cranium are as follows : —

Length from foramen magnum to diastema

. 12-0

„ „ intercornual ridge to apex of nasals

. 13-0

Height from lower border of foramen magnum to occipital crest

. 5-8

Interval between occipital crest and intercornual ridge .

. 2-5

Width at inferior border of orbits ....

. 8’6

„ at superior border of orbits ....

. 10-0

„ below orbits ......

. 7-4

„ above „ ......

. 8-1

„ of occiput through petrosals ....

. 10-3

Height from surface of palate to frontals

. 6-6

Interval between outer angles of condyles

. 5-2

Length of temporal fossa .....

. 6-5

„ of intercornual ridge .....

. 9-0

CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES. 11—98

Height from palate to tip of nasals
Interval between orbit and base of horu-core
Transverse diameter of left orbit
Antero-posterior diameter of left orbit .

Length of line of five molars
Greatest width of palate

Interval between outer surfaces of second molars
Length of left horn-core, upper surface
„ „ „ lower „

Antero-posterior diameter of base of left horn-core
Circumference of ditto ....
Transverse diameter of ditto
Interval between tips of horn-cores
„ „ supra-orbital foramina

Width of nasals at widest part .

. 3-2

. 5-3

. 27

. 31

. 5-9

. 3-8

. 5-8

. 39-0
. 32-0
. 3-8

. 12-5
. 3-4

. 360
. 5-7

. 2-7

I will now notice another frontlet of the same species in the Indian Museum,
collected by Mr. Hacket in the Nerhudda valley ; it belonged to a rather larger
animal than the preceding specimen. The general form of the frontals of the
two specimens is similar, the horn-cores, however, of Mr. Hacket’s specimen,
instead of a nearly circular, present an elliptical cross section, while, continuous with
the intercornual ridge, a well-marked ridge runs along the upper border of the
horn-core; the form of the horn-core of this specimen indeed approaches very
closely to that of the male Bibos gaurus. The horn-cores are more massive and
taper less rapidly than in the preceding specimen, while the ridge between, the
base of the horn-cores is quite straight, and without any median prominence ; the
dimensions of the specimen are as follows : —

Length from intercornual ridge to apex of nasals

. 13-0

Interval between superior curved line of occiput and intercornual ridge

. 3'5

Width at superior border of orbits

. 10-4

„ at constriction above ditto

. 8-5

Length of intercornual ridge

. 9'0

Interval between orbit and base of horn-core

. 5-5

„ „ supra-orbital foramina

. 5-6

Antero-posterior diameter of base of horn-cores

, 5-2

Transverse .ditto ditto

. 4T

Circumference of ditto

. 160

Two other specimens of the frontlet of the Nerhudda ox, in the Collection of
the Indian Museum, have horn-cores of which the cross section is intermediate in
form between those of the horn-cores of the two last specimens. The following
measurements show the difference in the form of the horn-cores of fom’ individuals
of this species ; the specimen in the first line is Mr. Hacket’s above-mentioned
specimen, those in the two middle lines are the two frontlets here referred to, while
that in the last line is the cranium figured in Plate XI ; the measurements are

taken at the base of the horn-cores : —

Ant.-post. diam. Tran, diam, Diif. Girth.

Flat-horned variety

.5-2

4-1

1-1

Intermediate ditto

. 5T

4-2

0-9

2nd ditto ditto

. 4-7

3-8

0-9

13-5

Round-horned ditto

. 3-8

3-4

0'4

12’5

99—12 CEANIA OF EU MIN ANTS FEOM THE INDIAN TEETIAEIES.

Iii the Catalogue of the Fossil vertebrate of the Asiatic Society of Bengal,'
Dr. Falconer has noticed several fragments of crania and horn- cores which belong to
this species ; it seemed, however, that at that time he was somewhat doubtful of the
distinctness of the present species, since he in some cases refers to the remains
simply as belonging to a species of Bos and at others as belonging to Bos nama-
dicus. Owing to this uncertainty of nomenclature, it appears that Mr. TheobahF
was led into error in founding the so-called new species. Bos falconerianus ; he
appears to have thought that in the description of the horn-core marked 30 in the
Catalogue and there called simply Bos, Dr. Falconer had intended to argue for the
specific distinctness of that horn-core from both Buhalus palceindicus and Bos
ncmiadicus ; whereas in reality Dr. Falconer was only arguing for the distinctness
of the specimen from Bubcdus palcBindicus and the two living species of Bibos. A
more careful perusal of the Catalogue would have shown such to have been the view
of the author, because the next specimen in the Catalogue (No. 31) is named Bos
ncmiadicus, and Dr. Falconer m noticing it says that “ it presents the same charac-
ters as No. 30,” clearly showing that he considered the two identical; again, in
the description of No. 33 {infra) Dr. Falconer refers to No. 30 as having belonged
to a male of Bos ncmiadicus. An examination of the two specimens of horn-cores,
30 and 31, shows that they are quite similar in form, and that they also agree with
the horn-cores of the specimen of Bos namadicus figured in Plate XI of this
memoir. Bos fcdconerianus must therefore he considered as Bos namadicus only.
As the specimens of horn-cores of this sj)ecies in the Asiatic Society’s Collection are
of larger size than any of those here described, I will here quote Dr. Falconer’s
notes on them as given in the Catalogue. Speaking of the above-mentioned speci-
men of a horn-core (No. 30), Dr. Falconer uses the following words : Bos — Enor-
mous horn-core of bovine ruminant, encrusted with matrix, much curved in tlie out-
line, and nearly chcular in section. Girth five inches above base, 1 foot 7
inches ; length along convex curve 3 feet 8 inches ; diameter where tip is broken
off, 3 inches. It is much more circular than that of the Gour or Gayal, in
which respect it differs still more than the Bos {Bubalus) palcBincUcus, and it
would apj)ear to indicate a distinct fossil species now extinct.” This specimen is the
largest horn-core of Eos that I have seen; it agrees with that of our

figured specimen in belonging to the round-horned variety. The next specimen in
the Catalogue (No. 31) is entered by Dr. Falconer as belonging to Bos namadiciis,
and is referred to in the following words : “ Fragment of horn-core, left side,
detached, presenting the same characters as No. 30, — namely, with curved outhne
and nearly circular in section ; it is less covered with matrix, and at the upper and
convex edge shows a ridge similar to (the horn-core of) Bos {Boephagus) grunniens,

' “Catalogue of the Fossil Remains of Vertebrate from the Sewalik Hills,” etc., in the Museum of the Asiatic
Society of Bengal, Calcutta, 1859.

2 Mem. Geol. Surv. Ind., Vol. II, p. 293.

CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES. 13—1^0

which, it resembles also in circular section. Length 11 inches ; girth near base
11| inches ; at small end of specimen 13 inches.”

A portion of a cranium of Bos nmnadicus (No. 33) from near Jhansi Ghat on
the Nerbndda is described by Dr. Falconer in the following words : “Mutilated
cranium, presenting the occipital and sphenoidal region nearly entire, left occipital
condyle and right mastoid (process) complete ; occipital and frontal surfaces perfect
from superior margin of foramen magnum to the commencement of the nasals ;
cores of both horns broken off, on the left side within the base of the pedicle of the
core, and on the right side the fracture includes the base of the core, the margin
of the frontal and orbit, all of which are removed. A portion of the upper border
of left orbit remaining. The frontal plane longitudinally is slightly convex vith a
shallow concavity upw^ards between the commencement of the horn-cores. The
vertex projects posteriorly to a great extent so as to over-arch the plane of the
occipital condyles 2| inches. Occipital region concave from above downwards ;
in these two respects differing very notably from the skulls of the Gonr and Gayal.
Judging from the section of the core on the left side, it was more or less cylindrical
in form, in this respect also differing from those two species.

“Length of specimen from vertex to broken margin in front 13 inches;
height from posterior edge of body of sphenoid to vertex about 8| inches ; width
of skull at base of occiput 9| inches ; height of occipital surface from inferior
margin of foramen magnum to posterior border of vertex 7|- inches ; long diameter
of occipital condyle 2| inches ; short ditto 1'6 inch ; width of frontal at constric-
tion behind orbits 85- inches.

“ Dr. Spilsbnry says that tliis cranium and the specimen No. 31 were found
in the same place; they were marked Nos. 1 and 2. In conjunction with the huge
horn-core No. 30, they would appear to afford conclusive evidence of a fossil species
of Bos from the Nerbndda distinct both from Bos {Buhalus') palceindicus and the
Gour and Gayal, or any other described existing form. From the complete synostosis
of the two frontals, it is inferred that the animal was an aged adult, and the smaller
size of the horn as compared with specimen No. 30 would seem to indicate that it
was a female.”

We will now compare the specimen figured in Plate G, figs. 1 and 1 a, of the
“Fauna Antiqua Sivalensis” with the specimen figured in Plate XI of this
memoir ; as before said, a copy of the frontal view of Falconer’s specimen is given
in Plate XXII, fig. 4, of the “ Palseontological Memous,” and the occipital view
is given in the plates accompanying the present memoir {Blate XVI, fig. 3).

In the accompanying index to the plates. Dr. Falconer refers to the specimen
in the following words : “ Fragment of cranium, showing forehead, occiput, occipi-
tal condyles, and foramen magnum ; portion of right horn, and core of left horn.
The specimen shows well the flat square forehead, the height being about equal
to the breadth. The horns are attached to the extremities of the highest salient
line of the head ; the horn-cores spread out horizontally, with a slight arch upwards

101—14 CRANIA OP RUMINANTS PROM THE INDIAN TERTIARIES.

and concavity below. The section of the horn-core is much more circular than in
the Gour or Gayal.”

The frontal view of this specimen is distinguished from that of the cranium
figured in Plate XI by the intercornual ridge being perfectly straight, instead of
being raised into a median prominence, and by the contraction between the orbit
and the horn-core being deeper and shorter, the frontals being thereby more
nearly square ; the horn-core seems to be shorter and with a sharper curve
forwards.

On the occipital aspect {Plate XVI, figs. 1 and 3) the two specimens present
more striking differences ; in Palconer’s specimen the superior curved line extends
upwards between the base of the horn-cores, and the indentations of the temporal
fossae are consequently considerably below the summit of the curved line, instead of
nearly on a level with it, as in our specimen ; the intercornual arch is also nearly
straight in Palconer’s specimen and arched in ours ; finally, the vertex of Palconer’s
specimen very slightly overhangs the plane of the occipital condyles, while it does
so considerably in our specimen, though not to any great extent in the position in
which the specimen is figured. I consider that the difference in the form of these
two crania is chiefly due to age, the high and projecting intercornual ridge of the
specimen in the Indian Museum being caused by a great development of the
intertabular cancelli : the horn-cores, however, must originally have been placed
somewhat higher in relation to the curved line of the occiput than in the British
Museum specimen : as before said, the specimen figured in Plate G, fig. 2, of the
“Pauna Anti qua Sivalensis,” agrees exactly in form with the specimen figured
in Plate XI of the present memoir, so that the two varieties were classed together
by Dr. Palconer.

In addition to all the above-mentioned specimens, the Indian Museum also
contains several crania of much smaller oxen from the Nerbudda valley, which,
however, belonged to adult animals, and which have the same general characters as
the larger crania of Pos namadicus from the same deposits ; these smaller crania,
I think, belonged to female individuals of the above species, the skulls described
above being those of males ; although there is a very great difference in the size
of these skulls, yet I think the difference is very little, if at all greater, than occiu*s
between the crania of male and female individuals of the existing Bibos banting.
The measurements of the most perfect of these smaller crania are given below \
the proportion of the antero-posterior diameter of the forehead to the transverse
will be found in the table given below ; from this it will be seen that the pro-
portionate excess of the antero-posterior diameter over the transverse diameter is
nearly the same as in the larger crania ; the following are the measurements

W idth at superior border of orbits . . . . , . . . 5’8

„ at constriction below orbits ........ 6'3

„ at constriction above orbits . , . . . . . . 6'5

Length from intercornual ridge to ape.x of uasaLs . . . . . . 9'5

Antero-posterior diameter of left orbit 2'6

CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES. 15—102

Transvei-se diameter of left orbit ........ 2'2

Length of temporal fossa . . . . . . . . . 5'3

Interval between base of bom-core and orbit . . . . . . .34

Height fi'om lower border of foramen magnum to occipital crest .... 4 o

Interval between occipital crest and intercornual ridge ...... 2’0

Length of intercornual ridge ......... 6 2

Interval between exterior angles of occipital condyles . . . . . .38

Width of occiput between temporal fossae . . . . . . . 4'7

Antero-posterior diameter of base of born-core . . . . . . ' . 2 5

Transverse ditto ditto . . . . . . . 2'4

Circumference of ditto ........ 6'0

Width of palate ... ........ 3'8

Interval between outer surfaces of second molars ..... • 5‘2

Length of six molars .......... 5‘5

„ of last molar .......... 1'4

The frontals of the above specimen differ from those of the larger crania by
being rather more concave in the middle line, and by having a sharp ridge nearly
half an inch in height running along the frontal suture for the attachment of the
frontahs muscle. The intercornual ridge forms a high convex arch overhanging
the plane of the occiput, when the skull is placed horizontally ; the horn-cores
have an almost completely circular cross-section ; the general form of the whole
occipital surface agrees exactly with that of the large Nerbudda specimen described
above ; the line of the molar series is nearly straight ; the costse on the external
surfaces of the molars are prominent and of nearly equal height ; the accessory
column on the inner side of the molars is large, with a constricted neck. The
dimensions of a detached horn-core of the female from the Nerbudda valley are as
follows : —

Circumference at base .......... 9'5

Length along upper curvature ......... 10'4

This horn-core is perfeetly cylindrical, is greatly curved, and tapers very
rapidly from base to tip another fragment of a larger horn-core in the Indian
Museum has a deep and wide groove running along the concave curvature, and
continuing almost to the tip ; I have not seen a similar groove in any other
specimen.

Eragments of the cranium and teeth of this species have been brought by
Mr. Eedden from the older alluvium of the Perim-ganga River in the Hyderabad
distriet ; remains of the species were also determined by Dr. Ealconer from the
older alluvium of the Jumna ; both deposits being probably near the horizon of
the Nerbudda gravels.

Dr. Ealconer suggested that the cranium of the typical Bos namadicus bore a
considerable resemblance to that of the European Bos prbnigenius ; a comparison
of the two crania, however, shows considerable differences ; for this comparison we
may refer to a figure of the frontal aspect of the cranium of Bos primigenim
given in Professor Owen’s “British Eossil Mammals and Birds” ( p, 498, ed, 1847),

103—16 CEANIA OE EUMINANTS EROM THE INDIAN TEETIAEIES.

The crania of the two species are characterised hj haying long and flattened
frontals, which considerably exceed in length the facial portion of the skull ; in both
crania the nasals are long and tapering ; they are, moreover, slightly arched, and
are wider at their infra-orbital angles than elsewhere ; in both the horn-cores are
approximately cylindrical, and extend far in front of the face, while the orbit is»
separated by a long interval from the base of the horn-core ; the cranium of J^os
namadicus is, however, somewhat narrower above the orbits than that of £os prlmi-
genius, in consequence of being more deeply excavated laterally. In the figured
specimen of the cranium of B. namadicus the intercornual space is arcuated,
whereas it is straight in B. primigenius ; the horn-cores in the crania of both
species are placed on the frontals in a very similar manner, but those of the Indian
species are larger and longer, and their direction is more upwards ; moreover, tliey
have no double curve, and their terminal curve is sharper than in Bos primigenius.
The variety of Bos namadicus from the Nerbudda, in which the horn-cores have an
ellipsoidal cross section, difiiers still more from Bos primigenius and approaches to
Bibos. The nasals of B. primigenius articulate with the premaxillae, while those
of B. namadicus do not.

The occipital region of the specimen of Bos namadicus figured in Plate XI I
have compared with the corresponding surface of a cranium of Bos primigenius
in the Indian Museum, and have found very striking differences between the two.
The occiput of Bos primigenius form of that of a typical species of the

genus, being characterised by its general quadrangular outline and by the absence
of any indentation of the lateral boundaries of tliis surface by the extremities of
the temporal fossae ; the occipital crest or superior curved line is flattened at its
summit, and its lateral borders are approximately straight ; marked angles are
consequently formed at the junction of the lateral and superior borders of the
infra-cristal portion ; there is a large shield-like protuberance for the insertion of
the ligamentum nuchse, and the summit of the occipital crest reaches upwards as
far as the middle of the bases of the horn-cores; between the occipital crest and
the base of the horn-core there is a smooth hollow surface. The paroccipital
proeesses are large, angular, and prominent, and there is a wide space on their outer
surfaces for the attachment of the digastric muscles ; the mterval between the
external angles of the occipital condyles is equal to half the interval between the
posterior extremities of the temporal fossae, is much less than the interval between
the inferior border of the foramen magnum and the summit of the occipital crest,
and is less than half the width of the oeciput measured through the foramen
magnum ; finally, the supra-cristal portion of the oeciput does not overhang the
infra-cristal portion, and the plane of the occiput consequently is nearly straight,
or slightly convex.

In the eranium of Bos namadicus, on the other hand, we find the occipital
surface (Plate XVI, fig. 1) indented laterally by the extremities of the temporal
fossse, and the quadrangular outline of the occiput consequently lost ; the oceipital

CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES. 17—104

crest forms an unbroken regular curve ; its summit does not extend upwards to
the middle of the base of the horn-cores, and consequently the supra-cristal portion
of the occiput is of considerable height ; moreover, the latter portion projects
backwards so as to overhang the plane of the infra-cristal portion of the occiput,
and the plane of the whole occiput is consequently concave from above downwards.
The ligamentum nuchse seems to have been inserted on a fiat smTace and not to a
tubercle ; the paroccipital processes are broad and rounded, and the surface for the
digastric muscle is narrow. The interval between the external angles of the occi-
pital condyles is only slightly less than the interval between the inferior border of
the foramen magnum and the occipital crest, the same interval is greater than half
the width of the occiput measured through the foramen magnum, and is equal to
two-thirds of the length of the interval between the extremities of the temporal
fossse ; finally, the intercornual crest is often arcuated.

The cranium of Bos namacUcm therefore difiiers from the cranium of Bos
primigenius in the following points : the short premaxillse, which do not reach the
nasals; the low position of the occipital crest relatively to the horn-cores, the
arcuated intercornual ridge, the intrusion of the temporal fossae on to the occiput,
the concave plane of the latter, and the regular curve of the occipital crest ; in
almost all these points, in which the cranium of Bos namadicus differs from that of
Bos primigenius, it approaches to the crania of the genus Bibos.

The peculiar forward curve of the horn-cores of this species, as is well shorni
in the profile view, presents considerable resemblance to the curve of the horn- cores
of the yak j this, however, cannot be taken as of any importance in showing kinship
between the two animals, as the forms of the crania of the two are so entirely
different.

We have now to consider in what manner the cranium of the present species
differs from those of the two Indian species of Bibos. If we refer to the table of
frontal measurements given below, it will be found that the antero-posterior
diameter of the forehead of this species exceeds the transverse diameter by one and
a half inches, whereas in the forehead of Bibos the transverse diameter is the longer
of the two.

In typical crania of this species the transverse section of the horn-core is either
circular or approximately circular ; and the horn-cores are placed on the highest
ridge of the skull, which is situated immediately above the proper occipital surface ;
the base of the horn-core is separated by a long interval from the orbit ; and the skull
between these two points is considerably contracted in width. The length of the
frontal portion of the cranium considerably exceeds that of the facial portion, while
the orbits are salient and placed on a level with the frontals, and the nasals are long
and widest at their infra-orbital angle.

On the other hand, in the genus Bibos the horns of the males are laterally
compressed, and their uppermost border is concave (the same border of the horn-
cores of B. namadicus being convex close to the skull) : the bases of the horn-cores

105—18 CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES.

and the orbits are closely approximated ; the cranium of Bibos gaurus is hut slightly
contracted in width between these two points, while that of Bibos frontalis is not
contracted at all in the same place ; the orbits of the former species are non-salient,
while those of the latter are considerably depressed below the plane of the f rentals.
The nasals of both species are wider at their free distal extremities than at their
infra-orbital angles, while those of Bibos frontalis are fm’ther distinguished by their
extreme shortness : finally, the length of the facial portion of the cranium in both
species is equal to that of the frontal portion.

If now we compare together the occipital surfaces of the three crania, we shall
find still more striking and marked differences between that of Bos namadicus and
those of Bibos. Eigures of the occiput of Bos namadicus are given in the present
memoir {Blate XVI, figs. 1, 3) ; and figures of the same region of the crania of
Bibos gaurus and Bibos frontalis will be found in Mr. Hodgson’s paper on the
genera of Indian cattle.’

Turning firstly to the occiput of Bibos frontalis, we find this surface of the
skull divided into two distinct portions ; interiorly there is the proper occipital or
infra-cristal smfface roughened for muscular attachments ; and superiorly the smooth
parieto-frontal or supra-cristal surface situated between the bases of the horn-cores.
At the junction of these surfaces the occiput is deeply indented from either side
by the temporal fossae — so deeply that the intervening surface of the parietals is
but very slightly, if at aU, wider than the interval between the external angles of
the occipital condyles. These temporal fossae are wide, and a portion of their inner
termination on the occipital surface is situated above the summit of the occipital
crest ; while the latter is placed below the lower border of the base of the horn-cores.
The interval between the external angles of the occipital condyles is precisely
equal to that between the lower border of the foramen magnum and the summit of
the occipital crest, and is equal to one-half the width of the widest part of the
occipital smfface ; and, as stated above, is very nearly equal to the interval between
the temporal fossae. Below the summit of the occipital crest there is a large shield-
like protuberance for the attachment of the ligamentum nuchae. The intercornual
ridge is wide and straight.

The occiput of Bibos gaurus agrees in form and comparative measurements with
that of the last species, with the exception that the intercornual ridge is produced
forwards between the horns in a high arch, so that the vertical height of the sujjra-
cristal portion of the occiput is greater than that of the infra-cristal portion.

On the other hand, the occiput of Bos namadicus has a totally different form
from either of the above. The temporal fossae make but very small indentations on
to the occiput, and consequently this surface is not so distinctly divided into two
portions. The indentations of the temporal fossae being so small, the summit of the
occipital crest is placed considerably above them, and extends upwards bevond

^ Journal of the Asiatic Society of Bengal, Vol. X, p. 470.

CRANIA OP RUMINANTS PROM THE INDIAN TERTIARIES. 19—106

the lower borders of the horn-cores ; consequently the interval between the ex-
tremities of the temporal fossae is far larger than that between the external angles
of the condyles ; the same interval also is considerably less than that between the
lower border of the foramen magnum and the summit of the occipital crest, and is
equal to more than half the width of the widest part of the occiput. There is no
marked prominence below the occipital crest for the attachment of the ligamentum
nuchae; and the supra-cristal portion of the occiput considerably overhangs the
infra-cristal portion; the intercornual ridge is slightly arcuated. Pinally, as a
character of minor value, the directions of the horn-cores are quite different ; those of
Bos namadicus have a graceful yery open cmwe, the first tmn of which has the
superior border convex ; in Bibos the horn-cores have but one slight curve, and their
upper border is concave. The anterior border of the orbits in Bos namadicus is very
nearly parallel with the long axis of the cranium, and the axis of the orbit is directed
outwards ; in Bibos the same border of the orbit is placed obliquely to the long axis of
the cranium, and the axis of the orbit is directed downwards and outwards. There
can therefore be no doubt as to the distinctness of the Nerbudda ox from the genus
Bibos. In one point, however, the cranium of Bos namadicus differs from those of
more typical species of the genus, and thereby approaches to the cranium of Bibos ;
this point is the relation of the premaxillse to the other bones of the face. In his
definition of the restricted genus Bos, as given in the first Catalogue of the TJngulata
in the British Museum {1853), the late Dr. Gray says that the premaxillse are large
and always extend upwards into the triangular space between the maxillge and the
nasals, and consequently articulate with both these bones, in the same manner as in
the genus Bubalus ; in the genus Bibos, on the other hand, the premaxillse are small,
and are attached only to the distal extremities of the maxillge, and are separated by
a considerable interval from the nasals. In the second Catalogue of Ruminants in
the British Museum {1873), however. Dr. Gray has added a new species of Bos,
namely, the Bos chmensis of Mr. Swinhoe, which vitiates the above distinction
between Bos and Bibos, though no mention of this fact is made in the Catalogue.
We may refer to a figure of the cranium of the Chinese species, which is undoubt-
edly a true Bos, given in the Proceedings of the Zoological Society of London for
1870, and we shall then find that the premaxillae are small, and do not extend uj)-
wards to the nasals, precisely as in Bibos.

In our figured cranium of Bos namadicus, the distal portion of the face has
been in great part broken away, so that the relations of the different bones cannot be
well observed ; on the right side of the specimen, however, a minute portion of the
proximal extremity of the premaxilla still remains ; this is situated near the ex-
tremity of the maxilla, and is far removed from the nasals, precisely as in Bos chmensis
and in Bibos ; other less complete specimens of the cranium of Bos namadicus in
the Indian Museum show this relation more clearly. The above facts, therefore,
compel us to abandon one of Dr. Gray’s distinctions between Bos and Bibos as not
being applicable to all the species. It is noteworthy that the short premaxillge

107—20 CRANIA OE RUMINANTS EROM THE INDIAN TERTIARIES.

which occur in the Indian genus Bibos also occur in two Asiatic species of the
genus Bos and in no others.

The nasals of Bos namadicus are widest at the angle immediately helow the
orbit, while below this angle they diminish in width regularly to their distal ex-
tremity ; in Bibos, on the other hand, and in most species of Bubalus, the infra-orbital
angle of the nasals is not their widest part; in Bibos gaurus the nasals at first slightly
contract below this angle, and then spread out again to a greater width than at any
other part : in Bibos frontalis the extremely short nasals preserve an almost uniform
width throughout their lecgth. The figured cranium of Bos namadicus also agrees
with that of Bibos gaurus in having an arcuated intercornual space, but this
character is not universal in the former species.

In the Collection of the Indian Museum we have a large series of limb-bones
of oxen from the Nerbudda valley, but it is generally impossible to say whether
they belong to Bos namadicus or to Bubalus 'palccindicus : the only bones I will
mention here are some specimens of the metacarpals and metatarsals, which from
their resemblance to the bones of Bos primigenkis may, I think, have belonged to
the Nerbudda ox. The metacarpal is strong and stout ; on the anterior surface there
is a slight, though distinct, groove for the extensor tendons, which does not, however,
extend far above the nutrient foramen ; the middle of the anterior aspect presents
a smoothly rounded surface, without any distinct ridges ; the posterior surface is
quite fiat, with no median hollow for the flexor tendons ; the articular surface for
the unciform is of very small size, the division between this surface and that for the
trapezo-magnum being placed on the outer side of the median line. The dimen-
sions of the specimen are —

Extreme length ........... 9'5

Transverse diameter of carpal surface . . . . . . . .26

Antero-posterior diameter of surface for trapezo-magnum . . . . .16

Transverse diameter of two trochlese 2'8

Cii'cumference of shaft at middle ........ 5'0

The s]iecimen exactly agrees in length with the metacarpal of a large male
Gour in the Indian Museum ; the metacarpal of the latter species is, however, more
slender, and has a well-marked extensor groove throughout its length. The bone
is half an inch shorter than the metacarpal of Bos primigenius described by Profes-
sor Owen in his “British Eossil Mammals and Birds,” and is 1’3 inches less in cir-
cumference, indicating that the Nerbudda ox was of slighter build than the
English Pleistocene ox : in fact, intermediate in this character between the latter
and the Indian Gour.

The metatarsal which I refer to this sjiecies is characterised by presenting a
nearly square cross section at its centre : at its proximal extremity it becomes
slightly expanded in an antero-posterior direction, while at the distal extremity the
transverse diameter becomes considerably the longer of the two. On the anterior
surface there is a wide and deep groove for the extensor tendons. The trochlem are

CRANIA OR RUMINANTS FROM THE INDIAN TERTIARIES. 21—108

separated by a wide interval : on the posterior surface the bone is flattened and has
no distinct flexor groove ; the whole length of the bone is equal to four-and-a-
half times the transverse diameter of the two trochlege ; its dimensions are as
follows

Length . ......... 11-16

Transverse diameter of two trochlese ........ 2-60

Circumference at middle of shaft 4-80

This specimen would correspond in proportionate size with the metacarpal
noticed above : in Bos primigenius the metatarsal is 1-3 inches larger than the meta-
carpal ; and the former bone is very like this specimen in form and size ; it is slightly
longer than that of the Indian species.

The flat-horned variety of Bos namadicus, as I have before said, differs from
type species of the genus Bos, and agrees with the genus Bibos. We have
already noticed various other points of resemblance between the Nerbudda ox and
the latter genus, and since there is now no wild Taurine ox Living in India, and as
the genus Bibos must have lived almost immediately after the extinction of the
Nerbudda species, and is at present unknown dmdng the period in which that species
lived, it may, I think, be a fair inference that the one is the direct ancestor of the
other ; or that, in any case, both have sprung from one common stock. If I am right
in considering the next described species of flat-horned ox from the Siwaliks as the
ancestor of Bos namadicus, we are enabled to trace the probable ancestry of the Gout
and Gayal backwards to the Siwalik period, and to mark their gradual divergence
from the true Taurine type. I hope that Zoologists and Geologists in India
will give particular attention to collecting any remains of Mammals that may occur
in the upper Pleistocene and early recent beds of India, in order that we may be
enabled to trace the earlier species more directly up to the present species, and to
discover intermediate forms such as I presume must have existed between Bos
namadicus and Bibos gaurus.

The Nerbudda ox, owing to the discovery of a palaeolithic implement by Mr.
Hacket in the beds in which its remains occur, was at all events, during a certain portion
of the period in which it existed, a contemporary of the early human inhabitants
of India ; and the genus Bibos, if we adopt the view that it is descended directly
from B. namadicus, must also have acquired its characteristic modifications of
cranium within the same period. Whether man was instrumental in causing the
final disappearance of the Nerbudda ox, or whether it was unable to compete against
the modern oxen, or whether the disappearance of the sM forests from the Nerbudda
valley, which Captain Forsyth in the “ Highlands of Central India ” considers as
the cause of the disappearance of Bubalus palceindicus from that district, was
fatal to the existence of Bos namadicus, we are at present unable to decide ;
but I am inclined to think that the second of the three hypotheses is the more
probable.

109—22 CEAXIA OP EUMINANTS PEOM THE INDIAN TEETIAEIES.

Bos PLANiFRONS, n. sp. nohis. PI. 12, f. 2 ; PL 16, f. 4.

The specimen figured in the above-mentioned plates consists of an upper
part of the frontlet with the greater part of the horn-cores of a species of Bos
distinct from the last species ; the specimen was collected in the Siwaliks of the
Punjab by Mr. Theobald. On referring to the table of measm^ements on page 66,
it will be seen that the antero-posterior diameter of the forehead of this species is
longer than the transverse diameter, while if we turn to the view of the occipital
surface {Blate XVI, fig. 4), we shall see that the horn-cores are placed upon an ele-
vated ridge immediately over the occiput, that there is no distinct parietal band across
the back of the cranium, and that the frontals are nearly flat ; the transverse section
of the horn-cores is somewhat eUiptical, and the occiput is squared and but slightly
indented laterally by the terminations of the temporal fossse. All the above charac-
ters clearly show that the specimen should be referred to the restricted genus Bos,
although the elliptical cross section of the horn-cores differs from type species
of that group.

The above characters also sufficiently distinguish our specimen from the living
genus Bibos ; the direction of the horn-cores sufficiently distinguishes it from
Bos namadicus, since in the present specimen the horn-cores are directed almost
immediately outwards, with a slight tendency upwards and inwards, and they do not
curve forward as in Bos nmiadictis. In the latter species the tips of the horn-cores
of the figured specimen are about two feet in advance "of the plane of the frontals,
whereas in B. 'plcmifirons the horn- cores where broken are not more than three
inches in advance of the same plane. The occipital surfaces of the two type forms
are also slightly different, that of the Siwalik species being more strictly Taurine.

As is shown in the figure of the forehead, the frontals are longitudinally
fissured in our specimen, the parts on each side of the line of fracture being
somewhat laterally compressed, so that the median line of the frontals is tlirown
forwards in advance of its proper plane, thereby causing the forehead to appear
more convex than it naturally would be. In the unbroken condition of the
frontals these bones would be very nearly flat between the horn-cores ; below the
bases of the horn- cores, the median line of the frontals becomes somewhat more
prominent, and the lateral surfaces slope away on either side, so that the anterior
border of the orbit is plaeed somewhat below or behind the median line of the
frontals. The lateral surfaces of the frontals are marked by supra-orbital sulci
of great width and depth, which converge together as they descend : there are no
distinct supra-orbital foramina in the specimen ; this is, however, probably only an
individual peculiarity. The lateral contraction of the forehead between the base of
the horn-core and the orbit is distinct, though of no great length ; the re-entering
angle for the insertion of the apex of the nasals is seen on the lower border of the
figure ; it does not extend upwards as far as the centre of the orbits.

CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES. 23—110

The lower and left portions of the occipital surface have been broken away
in our specimen ; the occipital crest forms a very wide and flattened arch, somewhat
prominent and separated by an oblique groove from the base of the horn- core of
either side : the smooth surface above the occipital crest is concave from side to
side and slightly convex from before backwards ; it overhangs to a very slight
extent that part of the surface of the occiput below the crest. The portion of the
occiput below the crest or curved line is deeply marked by ridges and hollows for
muscular attachments ; a deep pit is seen on the right hand of the figure for the
attaclunent of the recti-capitis muscles. The bases of the horn^cores are separated
from each other by a considerable interval ; the ridge occupying this interval is
slightly convex when viewed from behind ; in a front view the same ridge is either
straight or very slightly concave, according to the position of the skull, but never
convex. The horn-cores themselves are at first directed outwards, upwards, and a
little forwards ; they are continued in a single unbroken curve as far as those of
our specimen extend ; they certainly when complete had not the double curve of
the horn-cores of Bos primigenms, neither were they extended forwards or bent
upwards towards the tip as in the Nerbudda species. At their junction with the
f rentals the horn-cores expand suddenly both above and below, the areas of their
transverse section being here greater than those of the portions of the f rentals to
which they are attached ; the summit of the curved line of the occiput extends far
upwards between the bases of the horns, and consequently is far removed from
the extremities of the temporal fossae ; these latter form wide but shallow fissures
on the occiput. The transverse section of the horns is laterally compressed. The

dimensions of the specimen are given below : —

Width at superior border of orbits . , . . . . . . . 9‘3

„ constriction above „ . . . . . . . . 8'8

Length from intercornual ridge to apex of nasals ...... 10‘2

„ of intercornual ridge ......... 7'0

Interval between occipital curved line and intercornual ridge ..... 2'0

Width of occiput between temporal fossse ....... 6'7

Circumference of base of right horn-core ....... 14'5

Length of remaining portion of „ ....... 19'0

Circumference of right horn-core at fracture ....... 9'0

Antero-posterior diameter of right horn-core . . . . . . . 4‘5

We have now to compare this cranium with our specimens of the crania of
Bos namadicus. We will first compare it with Ralconer’s specimen of the latter
species afieady so often referred to. In the figures of the occipital regions of the
two specimens {Flate XVI, figs. 3 and 4), the cranium of the Siwalik species {fig. 4)
has the occiput placed more vertically than in Ralconer’s specimen {fig. 3) ; in
consequence of this the crest or curved line of the occiput of the latter appears to
be placed lower down than in the former specimen.

On the frontal aspect the intercornual ridge of the Siwalik species is slightly
concave instead of straight or convex, while the lateral indentations of the skull

111—24 CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES.

between the hom-cores and the orbits are rather less deep, and the centre of the
forehead is rather more prominent : the horn-cores tend more directly outwards
with only a very slight cmve forwards. On the occipital aspect the chief difference
is that the vertical diameter of the infra-cristal portion is considerably greater in
proportion to the transverse diameter in the Siwalik species ; this gives the whole
occiput a more quadrate appearance, by which it approaches, as before said, more
closely in form to the occiput of the true Taurine oxen, like Bos primigenius. The
occiput of Falconer’s Nerbudda specimen forms a transition from the Siwalik
species to the occiput of our large Nerbudda specimen {Blate XVI, Jig. 1), in which
the long transverse diameter of the occiput and the separation of the infra-cristal
portion of the same surface is a character approaching to the occiput of Bibos-
The occipital crest of our specimens of the Nerbudda species {Blate XVI, Jigs. 1^3)
is completed in a single simple curve, whereas the proper occiput of the Siwalik
species has high lateral borders at first scarcely curving at all, and with marked
angles at their junction with this superior border ; the indentations of the temporal
fossge on to the occiput are very shallow and are placed much below the level of the
summit of the occipital crest or curved line. In the Nerbudda species, these in-
dentations are somewhat larger, and are placed nearer to the level of the occipital crest :
in all the above characters Falconer’s specimen of Bos namadicus is intermediate
between the Siwalik cranium and the typical Nerbudda crania.

In the Indian Museum there is another similar frontlet of an ox (No. 141),
also collected in the Siwaliks by Mr. Theobald, which I refer to the same species
as the last specimen, though it presents some slight differences in form. The speci-
men comprises only the intercornual portions of the frontals, with the upper part
of the occiput and the bases of the horn-cores. The dimensions are as follows : —

Length of intercornual ridge ......... 7‘2

Interval between occipital crest and intercornual ridge . . . . . . 1'8

Width of occiput between temporal fossae ....... 7 0

Circumference of base of born-core . . . . . . . • 14'5

These few measurements agree nearly with those of the last specimen : the
cross section of the horn-cores, however, exhibits a more nearly circular form,
though the horn-cores themselves are directed upwards and outwards, with no incli-
nation forwards. The occipital surfaces of this and the last specimen are almost
identical in form, the curved line or crest extending high up between the hom-
cores ; in the second specimen, however, the intercornual ridge is raised up into
a median prominence, as in the figured specimen of the Nerbudda species.

We may now compare the cranium of this Siwalik species with that of the
European Bos primigenius in order to see in what points it differs from that most
typical form of Taurine Oxen. Turning first to the occipital surfaces of the two
crania, we shall find that the occiput of the Siwalik species is much more close to the
European form than is the occiput of Bos namadicus. In both the European
and the Siwalik crania, the summit of the occipital crest extends high up between

CRANIA OR RUMINANTS PROM THE INDIAN TERTIARIES. 25—112

the bases of the horn-cores ; in both, the general outline of the occiput is approxi-
mately quadrangular, while the indentations of the temporal fossse on to the latter
surface are of very small extent in the Siwalik species, and are almost entirely
wanting in the European species. The occipital surfaces of the two forms are,
however, broadly distinguished by the greater relative length of the transverse
diameter in Bos primigenius. The summit of the occipital crest in the Indian
cranium is not flattened to the same extent as in the European cranium, and is
separated by only a small interval from the intercornual ridge. The foreheads of
the two species have nearly the same general proportions, hut that of the Indian
species is distinguished by being slightly convex. We are unable to tell whether
the relations of the premaxillse to the nasals in Bos planifrons are similar to those
which occur in Bos primigenius or to those which occur in Bos namadicus and
Bibos.

The horn-cores of the Siwalik species present an elliptical cross section, and
thereby differ very markedly from the European species, in which the horn-cores
are cylindrical ; in the same degree they approach the horn-cores of the flat-horned
variety of Bos namadicus and Bibos.

The proportions of the frontals in Bos planifrons, the general form of the
occiput, and the small size of the indentations of the temporal fossse on that surface,
show that tins species was nearer to the true Taurine Oxen than any other living or
fossil species of Indian Ox. As I have before said, however, the occiput of
Dr. Falconer’s so often quoted variety of Bos namadicus is intermediate in form
between the occiput of the typical variety and that of Bos 'planifrons : this,
together with the fact that the occurrence of flattened horns is a character common
to Bos planifrons and to some varieties of Bos namadicus, and from a certain
general resemblance in the form of the crania of the two species, induces me to
think that the latter may very possibly he the direct descendant of the former.

Bos ACUTIFRONS, u. sp. nobis. Plate 12, fig. 1 ; Plate 13 ; and Plate 16, fig. 2.

The cranium of this magnificent species has been already shortly noticed and
named by me in the tenth volume of the “ Records of the Geological Survey of
India.” The cranium in question was discovered in 1874 by Mr. Theobald in the
Siwaliks of the Punjab.

In the accompanying plates I have given two views of the frontal aspect of this
cranium, one on a larger scale showing the frontlet and a portion only of the horn-
cores, and the other the whole of the specimen on a much smaller scale. On Plate
16, figure 2, I have also given an occipital view of the same specimen. When
brought to the Indian Museum the horn-cores were broken into numerous fragments,
and the specimen was coated with the tightly-adhering grey Siwalik sandstone. The
matrix has now been cleared away, and the specimen mounted as shown in the

113—26 CHANIA OP RUMINANTS FROM THE INDIAN TERTIARIES.

smaller figure. The greater part of the horn-cores are complete, though they have
both lost their extremities ; almost the whole of the f rentals are also remaining and
are in excellent preservation ; the left orbit is entire, together with the right temporal
fossa, and the greater part of the zygomatic arch of the same side. The whole of
the lower part of the face below the apex of the nasals has unfortunately been
completely broken away ; the occipital region has been also considerably damaged, but
its characteristic upper boundaries are fortunately preserved ; the occipital condyles
have disappeared, and the borders of the foramen magnum and the surrounding
portions of the occiput are somewhat broken and crushed ; the basi-occipital and the
sphenoidal regions are uninjured.

In the general form of the forehead this cranium is quite unlike that of any
other described species of recent or fossil Oxen ; in its general characters, however, it
appears to approach nearest to the restricted genus Bos, although its differences are
such that we might very readily refer the species to a distinct sub-genus, were it not
that the multiplication of these smaller groups is extremely undesirable.

The frontals are produced mesially into a very prominent longitudinal ridge,
which extends from the vertex of the cranium to the apex of the nasals ; from the
presence of this very characteristic ridge, I have assigned to the specimen its specific
name of acutifrons. From this central ridge the frontals slope away on either side,
backwards and outwards, like the two sides of a roof, and the forehead has in
consequence two distinct planes. From this peculiar configuration of the frontalsj
the anterior borders of the orbits are necessarily placed greatly below or beliind
the median line of the forehead, so that if a horizontal rod be laid across the frontals
above the centre of the orbits, the anterior borders of the latter will be found to be
more than two inches below or behind this rod, whereas in no recent species of
Oxen would the same border of the orbit be more than half an inch below a
similarly situated rod.

The above-mentioned median frontal ridge, commencing at the highest point of
the cranium, is continued directly downwards as a single and sharp line as far as the
superior border of the orbits, at which level it expands into a broader ridge, which
embraces the apex of the nasals ; the orbit of either side is separated from this
median ridge by a wide v-shaped channel, which expands as it deseends on the
face ; this channel is a downwards continuation of the supra-orbital sulcus, and is
quite unlike the sulcus of any other species of Oxen; the broad notch for the
insertion of the apex of the nasals is seen on the lower border of the figure ; this
notch does not extend higher up on the skull than the level of the inferior border of
the orbits.

The horn-cores are placed on the highest ridge of the cranium, immediately above
the occipital surface ; they are set very obliquely on the forehead, so that the interval
between their inferior angles is nearly twice the length of the interval between
their superior angles ; in a transverse section throughout half their extent the
horn-cores are pyriform, the apex of the pear being upwards ; in consequence of

CRANIA OE RUMINANTS EROM THE INDIAN TERTIARIES. 27—114

this form the superior border of the hom-eore is produced into a distinct and sharp
ridge ; several deep grooves, to receive ridges on the inner surface of the horn, run
parallel to the superior ridge. The second half of the horn-core {JPl. 12, jig. 1, a)
is nearly cylindrical ; throughout its length the inferior surface is rounded and smooth ;
. there is a slight “burr” at the junction of the horn-cores with the f rentals. The
direction of the long horn-cores is at first upwards, outwards, and backwards ; towards
the middle of their course their direction becomes backwards and downwards, while
towards their extremities it becomes forwards and downwards ; they never extend far
in advance of the plane of the frontal-ridge ; throughout their length their upper
border is convex, while their lower border is concave.

The anterior surface of the base of the horn-core is continued outwards and
backwards as an extension of the surface of the frontal of the same side, both
having the same slope ; in consequence of this the central point of the upper half
of the f rentals is the most prominent point of the whole cranium.

A long interval separates the orbit from the base of the horn-core of the same
side ; between these two points the cranium is considerably contracted in width ;
the orbits themselves are laterally salient, and quite circular; their' anterior
border is parallel with the long axis of the cranium, as seen on the right side of tlie
figure ; the direction of their axis seems to be directly outwards.

Turning to the view of the occipital surface {FI. 16, jig. 2) we find this
portion of the skull very dilferent from that of any other Bovine ; which difference
might be taken as a reason for placing the species under a new sub-genus. The
occipital crest or superior curved line forms a very high narrow arch, which
extends upwards to within a short distance of the vertex cranii, so that the supra-
cristal portion of the occipital region is reduced to a very narrow band ; this supra-
cristal portion is not carried upwards directly in the plane of the supra-cristal portion,
but is rounded off in front to join the f rentals. A smooth, slightly concave surface,
separates the lateral boundaries of the occipital crest from the bases of the horn-
cores ; the occipital crest overhangs the true occipital surface to an extent of nearly
an inch. The whole of the occipital surface below the crest forms a rude triangle
with curvilinear boundaries. On the occipital surface the temporal fossae form
narrow triangular slits beneath the horn-cores : these slits are directed upwards
and inwards, and are situated far below the summit of the occipital crest. The
bases of the occipital Qondyles, the condyles themselves being broken away, are seen
at the lower part of the figure : the upper part of the foramen magnum has also been
broken away, and therefore the foramen appears to be placed higher on the occiput
than it naturally should be. The paroccipital processes are large and recurved ;
the space for the attachment of the digastric muscle, external to the paroccipital, is
narrow and long.

Turning to the table of frontal measurements on the following page, we find that
the antero-posterior diameter of the forehead exceeds the transverse by a considerable
length. The plane of the occiput forms an acute angle with that of the centre

115—28 CEANIA OE EUMINANTS EEOM THE INDIAN TEETIAEIES,

of the frontals. The hasi-occipital is triangular, and characterised by the very
large size of the posterior pair of tubercles, and the small size of the anterior pair.
The dimensions of the specimen are given below : —

Interval between occipital crest and intercomual ridge

1-8

Width at superior border of orbits ....

11-0

„ at constriction above orbits ....

8-7

Height from lower border of foramen magnum to occipital crest

5-8

Distance from occipital crest to apex of nasals .

130

Length of intercomual ridge ....

5*0

Width of occiput through petrosals ....

11-4

Interval between extremities of temporal fossae . . .

6-2

„ between orbit and base of hom-core

4-0

Antero-posterior diameter of left orbit ....

2-6

Transverse „ „ . . . .

2-4

Length of temporal fossae

5-8

Antero-posterior diameter of left horn-core at base

6-0

Transverse „ „ „ . .

Circumference „ „ „ . .

170

„ „ „ at broken extremity .

7 0

Length of ditto along upper curvature ....

49-0

„ „ lower ditto ....

39-0

Interval between broken extremities ....

87-0

Erom the above description it will be gathered that the present cranium
presents a combination of characters which render it difidcult to assign it to any one
of the recent genera of Bovina, without infringing on their definitions. The form
of the occiput is, however, sufficient to distinguish it at once from Bihos and Bubalus
by the high position of the occipital crest and the small size of the indentations of
the temporal fossse ; it is also separated from the latter genus by the acute angle
which the plane of the occiput makes with that of the frontals. In the following
points the present specimen agrees with the cranium of Bos : —

The acute angle formed by the planes of the frontals and oeciput.

The position of the horn-cores on the highest ridge of the skull, directly above
the plane of the occiput.

The excess of the antero-posterior over the transverse diameter of the forehead.

The small intrusion of the temporal fossae on to the surface of the occiput.

The long interval between the base of the horn-core and the orbit.

The parallelism of the anterior border of the orbit with the long axis of the
cranium.

The convexity of the upper eurve of the horn-cores.

On the other hand, it differs from all other species of Bos in the peculiar con-
vexity of the frontals. Professor Owen made the flatness of the forehead a distinct-
ive character of the genus ; this, however, is not taken as a character by Dr. Gray
in his Museum catalogue, since the forehead of the Indian Ox is slighly convex, and
is yet admitted into the genus Bos. Moreover, in the above-mentioned Siwalik species
of O^, Bos planifrons,i}iQ frontals are also slightly convex, although in other charac-

CRANIA OR RUMINANTS FROM THE INDIAN TERTIARIES. 29— IIG

ters that species agrees very closely with typical forms of Bos, and cannot be separated
from that genus. From the convexity of the frontals in that species, the orbits were
of necessity somewhat depressed below the plane of the median line, though not to
•the same extent as in the present species. With regard, therefore, to the convexity
of the frontals, I cannot consider this character if taken alone, as sufficient cause for
subgeneric distinction. Again, with regard to the form of tlie occipital surface,
Mr. Hodgson' in his Memoir on the genera of Indian cattle, describes the occiput
of the genus Bos as not being indented by the temporal fossae ; this, however, as we
have seen in the instance of the Nerbudda Bos namadicus, is not universally the case,
and therefore not of generie value. The small size of the indentations, however,
and their position, does appear to be qf value. Looking at the position of the
occipital crest and the indentations of these fossae in the present specimen, and
comparing them with the crania of other figured species of the Oxen, we can only find
any resemblance to the occiput of our species among the crania of the restricted
genus Bos. At the same time, in no other species does the erest extend so high
up on the oceipital surface, or possess the same triangular shape, as in the present
species ; if, however, the summit of the occipital crest of Bos namadicus formed a
somewhat higher arch, and were consequently narrower, then the form of the occiput
in the two species would not be so very dissimilar. The occipital crest of this speeies
overhangs the infra-eristal surface of the occiput in a manner only found among the
species of the restricted genus Bos ; the rounding-off of the supra-eristal portion of
the oeciput is a character approaching the Bubaline type, but the other characters
do not bear this out. The basi-occipital is very similar to that of Bos taurus.

The approximation of the superior angles of the bases of the horn-cores is a
eharaeter which, though not belonging to the genus Bos, is equally divergent from
the same part in any other genus, though there is an approach to it in the fossil
Buhalus platyceros deseribed below ; this, however, is not a eharaeter of much
value in generic distinctions, being one whieh varies considerably with age even in
individuals of the same species. The position of the apex of the nasals at the lower
border of the orbit is a minor characteristic of the eranium of Bos.

Taking the interval between the external angles of the oecipital condyles of
the typical form of Bos namadicus as a standard of comparison, we find that this
interval falls considerably short of half the width of the occiput measured through
the foramen magnum, and is also less than the interval between the lower border of
the foramen magnum and the summit of the occipital crest, showing that the occiput
is both narrower and higher than in other forms. On comparing the measurements
of this speeimen with those of Bos namadicus, given above, we find that the
interval between the inter-cornual ridge and the apex of the nasals is the same
in both ; the forehead of this species is, however, one inch wider at the orbits, and
is probably the largest cranium of all the fossil species.

* Journal Asiatic Society of Bengal, Vol. X, p. 449.

117—30 CEANIA OP RUMINANTS PROM THE INDIAN TERTIARIES.

The orbit is remarkable not only for its position so much below the median line
of the forehead, but for its very small diameter and nearly circular form ; its longer
diameter is only 2* 6 inches, and the difference between the two diameters is *2 inch.
In Bos namadicus the longer diameter of the orbit is 3T inches, and the difference
between the two is A inch. In the much smaller cranium of Bibos gaurus the
longer diameter of the orbit measures 2’7 inches. Prom the occiput being placed far
beneath the overhanging frontals the temporal fossae are shorter than in Bos nama-
dicus. The ovate section of the horn-cores is undoubtedly a character of Bibos, but,
as noticed above, some forms of Bos namadicus have the same form of horn-cores,
and there can be no doubt as to the genus of that species.

Pinally, I regard this cranium as presenting characters considerably divergent
from those of the type forms of Bos, but not so divergent as to afford valid grounds
for placing it in a distinct sub-genus, as we find approximations to those diverg-
ing characters in other species of Oxen, which undoubtedly belong to that restricted
genus.

In addition to the above specimen, there are portions of two horn-cores in the
Indian Museum which belong to the same species. These specimens also were
collected by Mr. Theobald in the Siwaliks.

The interval between the tips of the horn-cores of this species could not have
been less than nine feet, and when covered by the horn-sheaths their span must have
been at least ten feet. Taking the interval between the tips of the horn-cores as
nine feet, this would be equal to ten times the breadth of the cranium between the
superior border of the orbits. In Bos namadicus the interval between the most
distant parts of the horn-cores (rather below the tips) is only equal to four and a half
times the width of the cranium between the orbits. In the European Bos
primigenius, the interval between the tips of the horn-cores is slightly more than
three and a half times the width of the cranium between the orbits. The large horns
of Bubalus arni, figured in Plate I7l of the “ Ossemens PosslLes,” have an interval of
rather less than seven feet between their tips, equal to seven and a half times the
width of the skull between the orbits ; the interval between the horn-cores alone
would be at least a foot less than the above. Specimens of Bubalus arni, I am
informed, have been found reaching to twelve feet between the tips of the horns, but
these are very rare. As it is not improbable that other specimens of the present
species were larger than this specimen, and as I have made a small estimate for the
distance between the horn-tips, individuals of this species were probably quite as
large as any described recent or fossil form. The circumference of the base of the
horn-core is rather less than in Bubalus palccindicus. The present species will some-
what interfere with Cuvier’s definitions of the distinctions between the form of the
frontals in the genera Bos, Bubalus, and Bison. In the sixth volume of the
“ Ossemens Eossiles ” (p. 220) he writes, — Le front du bocuf est plat et meme un
peu concave; celui de V aurochs est bombe, quoi-qu'un peu moins que dans le bufjle.^'
The present skull is of course distinguished from those of the Aurochs by the excess

CRANIA OE RUMINANTS EROM THE INDIAN TERTIARIES. 31—118

of tlie longitudinal over the transverse diameter of the frontals, and by the vertical
position of the intercornual ridge above the occiput.

The cranium is distinguished from that of Bos planifrons by the form of tlie
occiput, by the convexity of the forehead, by the horn-cores being more pyriform
in cross section, and by their form and direction. In Bos planifrons the centre of the
frontals is placed somewhat behind the anterior surface of the base of the horn-cores,
whereas in the present cranium the corresponding part of the forehead, as before
said, is very considerably in advance of the anterior surface of the base of the horn-
cores. The differences in form between this cranium and that of Bos namadicm are
so striking as not to require pointing out.

I do not know any existing species of oxen to which the cranium of this species
'shows any marked relationship ; the flattened horn-cores, however, with the ridge
along their superior border, have an affinity to those of the genus Bibos, which is
confined toindia and Burma, and also to Bos planifrons. It is noteworthy that almost
all the Indian Oxen have flattened horns, — the Zebu, the Yak, and the typical
variety of Bos namadicus being the only species with cylindrical horns.

The Indian Museum, besides crania, also possesses a large series of limb-bones
of various species of oxen, some of which are of unusually large size, and probably
belonged to the present gigantic species ; it is, however, impossible to be certain in
any identifications of the limb-bones of Ruminants, since those of different species
and genera resemble one another so closely. The only bone which I will refer to is
what I consider to be the metacarpus of this species. We have several of these
large-sized ruminant metacarpals in the Indian Museum, and there is also a similar
specimen mentioned by Dr. Ealooner in the catalogue of the Collection of the Asiatic
Society of Bengal (No. S. 346) ; this specimen was conjecturally referred by
Dr. Ealconer to a female Sivatherium, but it is very much smaller than the type
metacarpals of that animal, and, as I shall show, is not of too large a size to have
belonged to the present species.

The anterior surface of the bone is perfectly plane without any longitudinal
grooving ; the troohleae of the distal extremity are bold and prominent, and there is a
considerable interval between the two ; the trochlear ridges are placed obliquely to
one another. The shaft of the bone is much flattened and expanded laterally.
Immediately above the trochleee, the bone is perfectly flat, on the posterior surface ;
higher up the lateral boundaries of this surface are raised into sharp and prominent
ridges. The length of the bone is equal to about three and a half times the breadth
of the distal extremity.

The dimensions of the specimen as given by Dr. Ealconer are as follows : —

Length of bone .......... 13'0

Transverse diameter of both condyles , . . . . . . . 4'3

„ „ of carpal surface 4'3

The bone is considerably smaller than a specimen of the metacarpal of
Sivatherium, described by Dr. Ealconer (No. S. 345), in which the transverse

119—32 CRANIA OP RUMINANTS PROM THE INDIAN TERTIARIES.

diameter of the carpal surface is 4-8 inches ; the form of all the specimens is the
same ; and Dr. Palconer says that specimen No. S. 346 corresponds very closely
with the same hone in Bos, differing only in its greater size.

As there are a considerable number of metacarpal and metatarsal bones in our
collections, differing slightly in size one from another, it becomes a question whether
they should all be referred to Sivatlierium, or whether some of them may not really
belong to Bos. In favoim of the latter view of the case it may be stated that, with
the exception of these metacarpal and metatarsal bones, the bones of Sivatherium
are of somewhat rare occurrence, while those of Bos are much commoner.

If some of these bones could therefore be referred to the latter genus, it would
render the proportionate number of these bones of the two genera more consonant
with the number of crania. To decide this question it is necessary to find some
correlation between the dimensions of the metacarpus and the cranium ; such a
correlation I am at present unable to discover, which wMl prevail in all cases, the size
of any particular portion of the cranium of the Bovina not always increasing in
the same ratio with its general dimensions.

In some cases, however, the width of the occiput does seem to be a fair
criterion of size ; the diameter of this surface in Bihos gaurus is 8’1 inches, while
the transverse diameter of the condyles of the metacarpus is 2‘8 inches ; in Bibos
frontalis these two diameters are respectively 7’4 and 2'4 inches. The transverse
diameter of the occiput of Bos aoutifrons is 11'4 inches, while the transverse diameter
of the trochlear condyles of the present metacarpal is 4’4 inches. The proportion in
all the above cases is therefore rather more than three to one ; and if the metacarpus
varies in size in the same proportion as the occiput, the present metacarpus would
not be too large to belong to Bos aoutifrons, under which head, as above said, I am
disposed to place it. The bone is an inch wider and two inches longer than the
metacarpal of Bos namadicus, and, if I am right in the position assigned to it, indicates
that Bos aoutifrons is the largest known species of fossil ox.

Bos PLATYRHINUS, n. sp., nobis. Plate 14.

This species is founded upon the lower half of a cranium figured in the above
plate. The specimen was collected by Mr. Theobald in the Siwahks of the Kangra
district, and, though less complete than might have been desired, is yet of
so peculiar a form that it cannot be referred to any of the foregoing species of
Siwalik oxen. Unfortunately we do not possess the lower portions of the skulls of
the above-described species, so that an exact comparison between them and our
present species is impossible. The only portion that is common to the skulls of the
three species is the portion about the orbits; that portion, however, is amply
sufficient to distinguish them from each other. The specimen is broken off across
the lower border of the orbits, apparently by a very recent fracture ; the portion
which remains is very nearly perfect.

CRANIA or RUMINANTS PROM THE INDIAN TERTIARIES. 33—120

The cranium is at once noticeable on account of the extreme breadth and flat-
ness of the nasal bones, which are as much as three and a half inches in width at
their distal extremity ; this width is three-quarters of an inch greater than the widtli
of the nasals in Bubalus arni, in which the nasals are wider than in any other
described Bovine, either recent or fossil. The nasals, moreover, extend very low
down, reaching to within two and a half inches of the free extremity of the maxillae ;
in no living species of oxen is the interval between the nasals and the maxillae so
small ; the general form of the nasals is the same as in the Arni, their width at the
infra-orbital angle being less than at their distal extremity.

The premaxillae are long and extend upwards to articulate with the nasals as in
Bubalus and typical species of Bos — a character which at once distinguishes the
species from Bos namadicus ; a very prominent ridge connects the apex of the nasals
and the inferior angle of the orbit ; the planes of the frontals and of the lachrymal
are nearly at right angles to each other, being divided by tliis ridge. A somewhat
similar ridge occurs in the crania of Bos and Bibos, although not developed to such
an extent ; the presence of this ridge serves to distinguish the species from Bubalus,
to which genus the broad nasals approximate it. In the skull of the latter genus
there is a deep and uninterrupted channel leading gradually down from the forehead
to the sides of the face between the orbits and nasals, and no ridge or angularity.
The skull is further separated from that of Bubalus by the free posterior edge of the
palatines being entire, and unbroken by the intrusion of the vomer, as well as by
the shortness of the face below the orbits.

The skull differs from all recent forms in the excessive vertical height from
the crowns of the molars to the surface of the frontals ; from the crown of the first
molar to the anterior border of the orbit the vertical height is upwards of nine inches.
In the skulls of the Gayal, the Gaur, and Bos namadicus, the corresponding measure-
ment is about six inches ; and in the Arni seven and a half inches. The lower
borders of the orbits are not depressed below the level of the plane of the frontals,
as in Bubalus, but are nearly continuous with this, as in Bos namadicus.

The skull is distinguished from the skulls of Bibos gaurus and B. frontalis by
the large premaxillse, and by the great size of the nasals.

The apex of the nasals does not reach to within half an inch of the lower border
of the orbit ; in Bubalus the nasals reach to the centre of the orbit. The palate is
produced posteriorly but a short distance behind the last molars, and is undivided,
as in Bos. The inner border of the molar series is but slightly curved : the molars
have a large median accessory column on the inner side, which is but slightly con-
stricted at its junction with the main dentine mass of the crown ; the enamel of
the second enamel island is deeply infolded from the posterior side, as in Bubalus.
The measurements of this specimen are given below : —

Width at lower border of orbit ....... 7 8-3

„ of nasals, at widest part , , . . . . 3‘5

Length of „ » • • • • • ■ • . 7'8

121—34 CRANIA OP RUMINANTS PROM THE INDIAN TERTIARIES.

Height from palate to centre of nasals . . . . . ... .5-3

„ from posterior extremity of palate to f rentals . . . . . . 7‘8

Interval between outer surfaces of second molars ...... 6 0

Width of palate at second molars . . . . . . . .36

Length from posterior extremity of palate to muzzle . . . ... . 12 0

„ of six molars .......... 6'0

Width of maxillae at diastema ......... 4‘6

The characters by which the skull agrees with Bos are as follows : —

The flatness of the frontals between the orbits, and the presence of the
ridge between the frontals and the nasals.

The shortness of the nasals.

The unbroken posterior border of the palate.

The low position of the apex of the nasals.

The Bubaline characters are —

The width of the nasals at their lower end, (being wider below than above).

The height from the palate to the frontals.

Comparing the figure of this cranium with that of Bos namadicus, the present
specimen is distinguished from the latter by the premaxillse reaching up to the
nasals, and by the apex of the nasals not reaching to the lower border of the orbit ;
the nasals of the two are very different ; the widest part of the nasals of Bos
namadicus is at their infra-orbital angle, while those of Bos platyrhimis are far wider
at their distal extremity than at this angle. The angle between the orbit and the
nasals is more marked in the present species than in Bos namadicus, wliile, as noticed
above, the height from the palate to the frontals is much greater in this species.

Comj)aring the present figure with the skull of Bos acutifrons on Plate 13, we
find only a small portion of the cranium common to the two, viz., that immediately
between and below the orbits ; this is, how^ever, amply sufiicient to distinguish
between the two. The apex of the nasals reaches to the lower border of the orbit in
Bos acutifrons, and the median line of these and the frontals is elevated far above
the margin of the orbits, the lateral surfaces sloping away from this line ; moreover,
the supra-orbital sulci have their lower portion extending down between the orbit
and the nasals, and there would be no prominent angle between the frontals and the
lachrymals. In Bos platyrhinus, on the other hand, the apex of the nasals does not
reach upwards to the orbit, consequently the supra-orbital sulcus lies altogether
above the nasals ; the frontals, between the lower angles of the orbits, are perfectly
flat, the margins of the orbits being on a level with the plane of the frontals : and
there is a most marked angle between the frontals and the lachrymals ; and no trace
of the deep and wide channel between the orbits and nasals that occurs in Bos
acutifrons.

In Bos planifrons, again, the apex of the nasals reaches upwards to the middle
of the orbits, which last are somewhat depressed below the level of the frontals ;
the median line of the frontals being very prominent. The inclination of the frontals
to the basis of the cranium in this species and in Bos acutifrons is so sharp that the

CRANIA OR RUMINANTS FROM THE INDIAN TERTIARIES. 35—122

vertical lieigEt from the palate to the frontals would have been much less than is the
case in Bos platyrhinus, probably not much greater than in Bibos gaurus : a side
view of those crania would, I imagine, he somewhat like that of Bos namadicus.

It is much to he regretted that the upper portion of this skuU was not obtained,
(the fracture is quite recent) that we might have been enabled to compare the
characteristic upper part with those of the other species. I think the above com-
parisons, however, quite sufficient to establish the distinctness of the species, though
I am not qnite certain whether, when the occiput and horn-cores are found, we may
not have to place the species under the genus Bubalus as an intermediate form
between that genus and Bos. The form of the nasals is distinctly Bubaline, while
their point of insertion and the inter-orbital region is like that of Bos. From the
prominence of their inferior borders, the orbits probably looked almost directly out-
wards, as in Bos.

Genus: BISON.

The genus Bison may be defined from the characters of the cranium as
follows : —

“ Skull less massive than in Bos or Bibos, facial portion longer and more finely
tapering; superior portion of forehead transversely arched; intercornual space
centrally elevated ; viewed anteriorly, this portion is a truncated cone ; posterior
aspect of skull triangular — more extensive than in Bos, but greatly less so than in
Bibos.'' — Hodgson.

To this it may be added that the transverse diameter of the forehead of
Bison exceeds the antero-posterior diameter; that the horn-cores are sub-cylindrical,
and that their upper border is concave.

Bison Sivalensis, Falconer, sp. Pis. 15 and 17, fig. 1.

The single cranium, which, as stated in the introduction, I have referred to
this species, is another specimen from among Mr. Theobald’s numerous Siwalik
collection. The greater part of the upper half of the specimen is fairly complete
and perfect, although in several places the brain-case has been subjected to severe
crushing. The left horn-core is broken off close to the forehead, while little more
than half of the right one remains. The whole of the lower extremity of the face
is broken away at a short distance below the orbits ; both the temporal fossae and
the orbits are almost uninjured ; it was, however, found to be impossible to com-
pletely clear the former from the hard matrix with which they are filled ; the
greater portion of the occipital region is complete, the hinder half of the palate,
together with the last molars, have been cleared from matrix, but the occipito-
sphenoidal region and the fore part of the palate is still buried in closely adhering
sandstone.

123—36 CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES.

On referring to the table of the frontal measurements of the skulls of the
Indian Oxen, given below, it will he seen that the transverse diameter of the
forehead of this specimen exceeds the antero-posterior diameter by three and a
half inches — a proportion which makes the supra-nasal portion of the cranium
considerably shorter than the nasal and infra-nasal portions, and gives to the upper
part of the skull a great preponderance in wddth over the lower part.

The horn-cores are widely separated at them bases, and are placed on a ridge
which is considerably below the highest point of the cranium ; the superior border
of this intercornual ridge is concave ; when seen from in front, the summit of the
superior curved line or crest of the occiput projects above and between the bases
of the horn-cores, whereas in a similar view of the crania of the genera Bos, Bihos,
and Bubalus, no portion of the occiput will he seen. The whole of the above characters
show that the cranium belongs to the genus Bison.

The horn-cores are compressed antero-posteriorly ; their anterior surface is flat
from above downwards and concave from within outwards ; their posterior surface is
convex in both directions ; each horn-core is marked by a longitudinal groove along
the superior border ; a similar groove is found in the horn-core of Bison {PoepJia-
gus) grunniens into which is fitted a projecting ridge on the inner surface of the
horn ; this ridge and groove arrangement doubtless affords additional strength to
the union of the horn with the horn-core, and renders the former less liable to he
knocked off by any sudden blow. The direction of the fragment of horn-core
which remains attached to the skull is at first upwards, backwards, and outwards,
gradually curving until its direction becomes upwards, forwards, and outwards ; the
whole direction of the horn-cores was probably very similar to that of the horn-cores
of the living Yak, though in the fossil species the inchnation forwards was probably
somewhat less ; the horn-cores diminish gradually and regularly in diameter ; their
anterior surface is concave, and their posterior surface convex.

The frontals are somewhat crushed on the left side of the cranium ; they are
slightly hollowed between the upper angles of the bases of the horn-cores, and
below the intercornual ridge become still more hollowed out ; this frontal hollow
is somewhat in the form of the letter H, having one transverse portion and two
lateral vertical portions as seen in the figure ; the median portion of the lower half of
the frontals immediately above the apex of the nasals, is prominent and swelling ;
a ridge, concave on its outer side, descends from the outer angle of the base of the
horn-core to the superior border of the orbit. The supra-orhital foramina are
widely separated and are situated on the narrowest line of the frontals ; they
pierce the bone at right angles to its surface ; the sulci are entirely below the
foramina, and are nearly vertical and of considerable depth and width as far down
as the superior border of the orbit ; below tliis point they become much shallower
and converge. Between the orbit and the base of the horn-core, the cranium
})ecomes considerably narrower from side to side, and the orbit and the base of the
horn-core are separated only by a short interval. The orbit is salient, of small size.

CRANIA OR RUMINANTS FROM THE INDIAN TERTIARIES. 37—124

nearly circular, and its axis looks obliquely outwards and forwards ; the apex of
the nasals extends upwards as far as the first third of the diameter of the orbit ;
the lower border of the orbit stands out very prominently and almost at right angles
to the surface of the maxilla, the cranium contracting very suddenly and sharply in
width beneath the orbit. Between the orbit and the nasals the bone is sharply angu-
lated, so that the plane of the frontals and that of the laohrymo-maxillary bones are
at right angles to each other ; the ridge between the orbit and the nasals is obliquely
grooved. The border of the orbit is slightly depressed below the plane of the frontals.

The vertical height from the palate to the frontals is but small ; the temporal
fossse are remarkably short, and in consequence of the backward direction of the
plane of the horn-cores very narrow ; the superior rim of the orbit is an unusually
broad plate of bone placed at an angle of forty-five degrees with the jugal, the
latter being entirely below the orbit.

Turning now to the occipital surface, we find that the occipital crest is very
prominent and of great thickness, particularly at its summit ; it forms a low and wide
arch with a short flattened portion at the summit ; at the boundaries of the tem-
poral foss0B this arch on either side splits into two branches to enclose the fossse, —
the upper branches join the base of the horn-cores. Between the occipital crest
and the intercornual ridge, there is a depressed lozenge-shaped surface, higher in
the centre than at the sides ; the parietals and frontals gradually rise from the central
depression to the vertex cranii ; the plane of the occiput forms very nearly a
right angle with that of the frontals. Below the occipital crest there is a promi-
nent tuberosity for the attachment of the ligamentum nuchse, and a ridge is con-
tinued downwards from this tuberosity to the superior border of the foramen
magnum, which divides the surfaces for the attachment of the recti-capitis muscles.
The paroccipital processes are large and recurved, and the surfaces external to them
for the digastric muscles are narrow, but long. The temporal fossae make large
incisions into the surface of the occiput ; these incisions are directed obliquely
upwards and inwards, and their extremities are nearly on a level with the summit
of the occipital crest ; the summit of this crest extends nearly as high up as the
median line of the bases of the horn-cores. The most elevated point of the frontals
is slightly below the level of the bases of the horn-cores ; there is no ‘ burr ’ at the
bases of the latter. The interval between the external angles of the occipital
condyles is slightly less than that between the lower border of the foramen magnum
and the summit of the occipital crest, and somewhat more than half the width of
the occiput at its widest part ; the occipital condyles are placed very low on the
occiput, and are separated by a deep fissure from the paroccipital processes.

The palate exhibits the crowns of the last three molars, which are considerably
worn down, and which prove that the cranium belonged to a fully adult animal ;
the palate is unusually wide, and the two rows of molars run nearly parallel ;
the palatines are produced to the distance of about an inch behind the last molar,
and their free border is not divided posteriorly by the vomer.

125—88 CEANIA OE EUMINANTS EEOM THE INDIAN TEETIAEIES.

The crowns of the molars have wide central enamel islands, the second being
infolded on the posterior side ; the median accessory column on the inner side
is large, with a contracted neck. The dimensions of this skull are given
below : —

Width at superior border of orbits
„ at contraction above „

„ at „ below „ ...

„ of frontals between upper angles of horn-cores
„ „ lower „ „

Greatest width of occipital crest
Interval between external angles of occipital condyles
Width of occiput between temporal fossae
Height from lower border of foramen magnum to occipital cres
Distance between occipital crest and vertex cranii
Antero-posterior diameter of right orbit
Transverse „ „ . .

Width of palate at 3rd molars
Interval between outer surfaces of'Srd molars .

Distance from inferior border of foramen magnum to last mola
Length of last molar ....

„ of two molars ....

„ from vertex cranii to apex of nasals
„ of temporal fossae . ^ .

Circumference of base of right horn-core
Antero-posterior diameter of „

Transverse „ „

Antero-posterior diameter at fracture of ditto
Length of fragment of ditto

9-5

8-7

5- 6
8-2

10-3

6- 8

3- 9

4- 0
4-8
3-4
2-2
2-2
3-4

1- 4
27
5-5
5-2

100

3-4

2- 6
31
5-6

In the shape of the horn-cores, in their position below the highest point of the
cranium, in the projection of the occipital crest between them, and in the form of
the broad frontals and the salient orbits, the cranium of this species is at once seen
to be widely separated from aU the preceding species, and equally widely from all
recent species of the genera Bubakis, Bibos, and Bos ; the same characters also shew
its agreement with the crania of the genus Bison and the genus or sub-genus
Boephagus, under the former of which I have placed the species, for reasons which
I will immediately shew.

Unfortunately I have been unable to obtain a perfect cranium of the Yak
{Bison or Boephagus grunniens) to compare with this specimen ; the Indian Museum
has no specimen of the cranium of the latter genus, with the exception of one
imperfect skull of a calf, wanting the occiput, and in which the characteristic
points are not developed, and the frontlet of a cross-bred animal between the Yak
and the Indian Ox ; my comparisons, therefore, cannot be so accurate as I could have
wished. A lateral and a front view of the cranium of Boephagus grunniens are
given in Dr. Gray’s Catalogue of the Ungulataof the British Museum, 1853 {Blate 4),
together with corresponding figures of the cranium of Bison americanus, both on
a small scale. Figures of the frontal and occipital regions of the cranium of

CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES. 39—126

P. grunniens are given in Mr. Hodgson’s Memoir on the Mammals of Nepal a figure
of the cranium of Bison urns is given in Owen’s “ British Fossil Mammals and
Birds” (p. 491).

From the table of frontal measurements given below, it will he seen that
the diflerences between the two diameters of the forehead of this species and that of
Poephagus grunniens are identical ; hut allowing for the smaller size of the present
specimen, the breadth is stiU greater in proportion to the length than in the Yak. The
orbits are rather more prominent than in the Yak ; the interval between the orbits
and horn-cores are nearly the same in the two species ; and the form and direction
of the horn-cores are also approximately similar in both species. The prominence of
the orbits in the Siwalik cranium is like the orbits of Bison americanus. In a front
view of our specimen, the summit of the occipital crest, as seen between the horn-
cores, occupies only the middle two-thirds of the intercornual interval, as in the Yak,
and does not occupy nearly the whole of this interval as in Bison prisons and B.
americanus. The concave upper boundary of the intercornual ridge of our specimen
differs from the straight boundary of the same in Bison americanus and Poephagus
grunniens ; the supra-orbital foramina and sulci have the same form in all the
species. The frontals of the Siwahk species are slightly concave, whereas in the
others they are as slightly convex. The occipital crest of the Siwalik species does
not extend quite so high up on the plane of the occiput as does that of the Yak,
in which the crest forms a higher and narrower arch than in this species.

Comparing the occiput of this species with that of the fossil Bison prisons of
Europe, we find the occipital crest of the latter considerably wider and flatter at its
summit, though situated at the same relative height on the occipital surface, and
having the same relation to the temporal fossae ; in consequence of the greater width
of the summit of this crest, the interval between the extremities of the temporal
fossae is somewhat greater than in the Indian fossil species ; the shghtly concave in-
tercornual line, and the great interval between the bases of the horn-cores, is a
character common to the two species. In both, the occipital crest splits at the ex-
tremity of the temporal fossa, so as to inclose the upper half of the fossa on both
sides with a prominent ridge of bone. The occipital crest in both species is of great
thickness, and is rounded off.

I have retained the Siwalik species in the genus Bison, and not placed it in the
genus Poephagus, chiefly because the distinction between the two genera or sub-genera
is founded mainly on external characteristics, of slight classificatory value, and not
on important cranial differences. We have, of course, no means of knowing the ex-
ternal characters of the Indian fossil species, and I have therefore retained it in the
original or type genus. On distributional grounds, it might perhaps he considered
right to refer it to the sub -genus Poephagus if that genus is really sufficiently dis-
tinguished from Bison ; the characters of the Siwalik cranium seem to be interme-
diate between the true Bison and Poephagus.

^ Jourual of the Asiatic Society of Bengal. Vol. X, P- 471.

127—40 CEANIA OE EUMINANTS PROM THE INDIAN TERTIARIES.

The living genus Poephagus is now confined to the highlands of Tibet, and is
seldom found much below an elevation of fifteen thousand feet. It becomes an
interesting question to consider whether the allied fossil species was an inhabitant of
the plains or of elevated land. I am inclined to think that it most probably was
a dweller on the plains or low hills, because, with the exception of a few species of
goats and sheep, all the Siwalik mammalia were plain-dwelling forms, and it is
improbable that there was at any rate much elevated country within the old
Siwalik area ; moreover, the Bison of America and the Bison of Europe are both
j)lain-dwelling forms. A very probable theory is that the fossil Bison of the
Siwaliks dwelt on or near the plains, and from some cause or other in later times was
induced to migrate into more and more hilly country, until finally it gave rise to
the modern Yak, which cannot exist except in the rarified atmosphere of highland
Tibet ; at the same time the occurrence of a true Ibex, as recorded by the late
Mr. Blyth,' in the Siwaliks, might give us grounds for taking an exactly opposite
view of the case.

Genus BUBALUS.

Tlie cranium of this genus may be defined as follows : — Horn-cores placed
below the plane of the occiput, frequently triangular, forehead convex — broader
than long, nasals very large and wide, occiput rounded superiorly, and with distinct
indentations of the temporal fossge ; the superior border of the horn-cores concave.

Bubalus platyceros, n. sp. nohis. Plate 18.

It is with some difiidence that I refer the present species to the genus Bubahis,
the cranium on which it is founded being aberrant in many points from any modern
type forms. I rely, however, chiefly on the form of the occipital surface and the
shape of the horn-cores, both of which are clearly Buhaline in the present speci-
men. The form of the occipital surface and its relation to the intercornual ridge
appears to afford the most dependable characters in dividing the Bovina.

The cranium of which I have given two views is another specimen from
Mr. Theobald’s Siwalik collection. The greater part of the forehead and the upper
half of the face are tolerably perfect ; the zygomatic arches, the lower portions of the
orbits, the distal half of the nasals, the premaxillse, and the whole of the spheno-
palatine regions of the skull, are wanting ; the occipital region shews only the
smooth supra-cristal portion, with a small part of the inferior surface for muscular
attachments and the temporal fossae ; this remaining portion is, however, fortunately
characteristic. The greater portion of the right horn-core remains, and a large
detached fragment of the upper half of the left core was found with the specimen ;
this has been put into its proper position in the figure.

• Ann.- Mag. Nat. Hist. Ser. I, Vol. XI, p. 78.

CRAKIA OP EUMINANTS PEOM THE INDIAN TEETIAEIES. 41—128

The frontals are of great size and width, particularly between the inferior
angles of the horn-cores ; as seen from the table of cranial measurements given on
page 66, the transverse diameter of the forehead exceeds the antero-posterior diameter
by rather more than one inch. The plane of the frontals is nearly flat; the
median frontal suture forms a slightly elevated ridge, while a longitudinal hollow
runs along its either side. The upper border of the orbit is salient, while a well-
marked ridge is continued upwards from the orbit to the antero-inf erior angle of the
base of the horn-core. The lateral contraction of the frontals above the orbits is
well marked ; the latter are closely approximated to the horn-cores. Between the
orbits and the nasals, the fronto-maxillary bones are concave from above down-
wards, and there is no trace of any ridge leading from the nasals to the orbit. The
anterior border of the orbit is placed obliquely to the long axis of the cranium, so
that the axis of the orbit is directed downwards as well as outwards. The supra-
orbital foramina are large, and the sinuses below them are deep and converge as they
descend ; the foramina themselves are directed obliquely upwards, their superior
border overhanging them and being produced into a conical bony projection; this
overhanging projection and the direction and situation of tlie foramina are precisely
the same as in Bubaliis; in Bos and Bihos, on the other hand, the supra-orbital fora-
mina are situated some distance below the upper extremities of the sulci, and arc
directed inwards, at right angles, to the plane of the frontals ; consequently the fora-
mina are not overhung by any bony roofs. Since the sulci do not extend above the
foramina, they are necessarily very short, being only two and a quarter inches in
length; they do not extend below the orbits; the supra-orbital sulci of Biibalus arni
are of precisely the same form and length as those of the present specimen, and
agree very closely.

The nasals extend very high up on the forehead, reaching to a point al)Ove the
level of the centre of the orbits ; at their apex the nasals form an acute angle,
and they are wider at their infra-orbital angle than at any other point, diminish-
ing in width rapidly below this angle. The frontals are rounded off superiorly,
to join the occipital surface, so that the plane of the supra-cristal portion of the
occiput forms an obtuse angle with the plane of the forehead. Erom the great
width of the forehead, and its comparative shortness, and from the upward exten-
sion of the nasals, it is probable that the length of the facial portion of the skull
exceeded that of the frontal. The lower portion of the nasals is much arched
and vaulted, the lateral surfaces of the maxillaB sloping outwards rapidly from the
median line.

The temporal fossae are long and narrow ; their inner walls do not slope
inwardly posteriorly ; what remains of the posterior border of the orbit, shows it to
have been unusually wide and strong.

The occipital surface is divided by the large occipital crest or superior curved
line into two semi-distinct portions ; the smooth supra-cristal portion, and the rough
infra-cristal portion for muscular attachments ; the supra-cristal portion of the

129—42 CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES.

occiput is not continued upwards in tlie plane of the infra-cristal portion, but
slopes forwards to join the f rentals by a gentle curve, so that there is no well-
marked and angular intercornual ridge overhanging the occiput; between the superior
angles of the horn-cores, the surface of the frontals is concave from side to side.

The occipital crest or curved line itself forms a very wide and flattened arch ;
the extremities of the indentations of the temporal fossae are situated entirely above
the summit of this arch ; and the horn-cores, again, are above these, so that there
is a considerable vacant space between the true occipital surface and the horn-
cores ; the temporal fossae cut deeply into the occiput on either side, so that the
width of the parietals between them is probably not greater than the interval
between the external angles of the occipital condyles. The proper muscular
surface of the occiput is deeply marked by ridges and grooves ; a circular depres-
sion occurs immediately above the summit of the crest ; at the highest point of the
frontals the outline of the intercornual ridge is convex.

The horn-cores are triangular in cross-section and are set obliqiiely on the
frontals ; their superior angle is somewhat below or in advance of the plane of the
true occipital surface, so that they are not situated on the highest portion of thfe-
cranium. The cross-section of the base of the horn-cores has the form of an
isosceles triangle, the inferior surface forming the base of the triangle ; the plane
of the anterior surface is continuous with the plane of the frontals ; their antero-
inferior angle is round and blunted, wliile their superior angle is remarkably sharp,
and looks directly upwards. Then’ inferior surface is situated nearly at right angles
to the plane of the frontals; their postero-inferior angle is the most obtuse of
the three. The horn-cores diminish somewhat rapidly in diameter, and are directed
outwards, upwards, and somewhat forwards, forming one regular curve, the upper
border being concave, and the lower convex throughout their length ; at a distance
of nine inches from the median line of the skull, the anterior surface of the horn-
core is placed one and a half inches in advance of the plane of the frontals ; there
is a small “burr ” at the junction of the horn-cores with the frontals.

Erom the imperfection of the lower portion of the face it is impossible to
determine the relations of the premaxillse to the other bones.

The measurements of this specimen are given below in inches

Width of cranium at superior border of orbits

. 9'3

„ at constriction above orbits

. 8-4

Intei'val between occipital crest and superior angle of born-cores .

. 3-2

Length from superior angles of horn-cores to apex of nasals

. 80

Width of parietals between temporal fossae .

. 50

Interval between superior angles of bases of horn-cores

. 4-0

„ „ antero-inferior angles of ditto .

. 90

5, „ orbit and base of horn-core

. 24

Antero-posterior diameter of right orbit

. 2-9

Interval between supra-occipital foramina . .

. 4-5

Length of temporal fossa . . . .

. 7-0

Width of nasals at widest part .

. 2-6

CRANIA OR RUMINANTS PROM THE INDIAN TERTIARIES. 43—130

Width of anterior surface of base of right horn-core . . . . . 6‘1

„ inferior „ „ . . . . . 3'3

„ posterior „ „ . . • • ,55

Circumference of base of horn-core . . . . . . 16‘5

Length along convexity of right horn-core (broken) ..... lo'O
„ concavity „ . . . . . 120

Longest diameter of broken tip of ditto ...... 2'7

Tlie dimensions of a detached and broken horn-core of this species collected
by Major Godwin- Ansten in the Siwaliks, near the Markanda River, are as follows :

Greatest diameter of base
Length along inferior convex border
„ „ superior concave border

Greatest diameter at broken tip

Inches,

. 5-9
. 22-0
. 170
. 2-7

The form of tliis horn-core is the same as those of the specimen described
above, though it is of rather larger dimensions.

In the collection of the Asiatic Society of Bengal, there is a cast of a cranium
of this species, from the Siwaliks, the original of wliich is, I presume, in the
British Museum ; the horn-cores of that specimen are rather longer than in the
described cranium, but not much larger than the detached specimen collected by
Major Godwin-Austen.

The cranium is distinguished from the crania of the genus J5os by the follow-
ing characters : —

The direction and triangular cross section of the horn-cores.

The rounding off of the fronto-parietal region between the horn-cores.

The occipital crest being placed below both the indentations of the temporal fossae and
the lower border of the horn-cores.

The short interval between the horn-core and the orbit, and the salieney of the latter.

The excess of the transverse over the antero-posterior diameter of the forehead of the
cranium.

The upward extension of the apex of the nasals.

The excess of length of the facial over the frontal portion of the skull.

The obliquity of the anterior border of the orbit and the concavity between this and the
nasals.

The position and direction of the supra-orbital foramina and sulci.

In all the characters in which this cranium diifers from £os it approaches to
Biihalus.

The nasals differ from those of Bubalus arni and Bubalus caffei\ by being
wider at their infra-orbital angle than at any other part, but this is a character
which is found in the nasals of Bubalus palcemdicus, as will be noticed below.

The cranium differs from that of Bubalus arni in that the horn-cores are
placed much closer together on the frontals ; they, however, occupy a similar posi-
tion in the cranium of Bubalus caffer. If the cranium of Bubalus arni be placed

131—44 CEANIA OE EEMINANTS EEOM THE INDIAN TEETIAEIES.

horizontally, the centre of the frontals will he the highest point of the whole ; the
plane of the face, that of the parietals, and the planes of the anterior surfaces of the
horn-cores, sloj)ing away on all four sides from this central point. In the cranium
of the present species, however, only the planes of the frontals and of the parietals
slope away in this manner ; the horn-cores, instead of being directed upwards and
backwards, as in the Arni, are directed upwards and forwards, so that their bases are
in advance of the central frontal plane ; this position of the horn-cores prevents the
present species from having the rounded and prominent forehead, characteristic of
Biibalus arni. The frontals of JBubalm caffer, however, in the young state, are
somewhat concave between the bases of the enormous horn-cores ; while in JBubalm
braeliyceros the whole of the frontals are flat, and the horn-cores somewhat approxi-
mated at the vertex crania. The convexity of the frontals cannot therefore be
taken as a valid generic character, as is done by Mr. Hodgson in his above-quoted
paper on the genera of Indian cattle.

The present cranium is distinguished from those of all species of the genus Bibos
by the direction and triangular cross-section of the horn-cores, as W'ell as by the
rounded fronto-parietal surface, and by the shape and position of the nasals and supra-
orbital foramina. The concavity of the horn-cores is, however, similarly placed
upwards in both genera. In the present species there is no aj)proach either to the
long and straight intercornual ridge oi Bibos frontalis, placed immediately above the
true occipital surface, or to the highly arcuated intercornual ridge of Bibos gaurus.

On comj)ariug the figure of the occiput of the present species with Mr. Hodgson’s
figure of that of Bibos frontalis, it will be seen that the occipital crest of the latter
extends higher up on the occiput than in the former ; consequently in Bibos the
inner extremities of the temporal fossae are not placed entirely above the occipital
crest, whereas they are so placed in the present species, in which the interval
between the fossae is greater. In both there is a large shield-shaped protuberance
for the nuchal ligament, and wdde smooth surfaces for the recti muscles. The
summit of the occipital crest forms a very high arch in Bibos, whereas it is nearly
flat in the present species ; the supra-cristal portions of the two are so different
as not to require a detailed comparison.

Turning again to the occijmt oi Bub aim arni and comparing it with the present
specimen, we find that in both species the inner extremities of the temporal fossae are
entirely above the occipital crest ; and the interval between the foss£e has the same
length in the tw^o ; the crest is, however, flatter and placed lower down in the fossil
species. The supra-cristal portion in both slopes forwards, though in the one it is
extremely wide and in the other narrow, the width of the cranium of Bubalus arni
being greater than the interval between the temporal fossae, whereas in Bubalus platy-
ceros the reverse is the case. The point of attachment of the ligamentum nuchae in
the former species is continued downwards as a bony septum between the surfaces for
the attachment of the recti muscles. There is a somewhat similar flattened depres-
sion above the summit of the occipital crest in both, which is not found in other

CEANIA OP EUMINANTS PEOM THE INDIAN TEETIAEIES. 45—132

forms. If the line of the occipital crest in the present specimen were perfect,
the interval between this and the lower border of the base of the horn-cores would
he far greater than in any other species. The intercornual portion of the occiput
is depressed between the bases of the horn-cores in Buhalus platyceros, and raised
above them in Buhalus ami.

The greater separation of the plane of the true occiput from the plane of the
parietals and the post-cornual portion of the f rentals in this species than in B. ami,
is an approach to the Antelopine form of skull, and a wide departure from the high
intercornual crest of Bos.

The above comparisons show that this cranium has most points of resemblance
to the crania of the genus Buhalus, and we are therefore justified in classing it
with that genus ; the characteristic points of the cranium have been already noticed
in the Eecords of the Geological Survey of India, ‘

Bubalus paljeindicus, Falconer. PI. 17, f. 2, and PL 19.

Of this magnificent species of fossil buffalo no complete description of the
cranium has ever appeared, though short notices of it will be found in Dr. Ealconer’s
Catalogue of the Vertebrate Eossils in the collection of the Asiatic Society of Bengal,
and a figure and notes are given by Dr. Spilsbury in the Journal of the Asiatic
Society of Bengal.^ A figure of a very complete, though partially restored, cranium
with the horn-cores attached will be found in Plate XXII of the “ Palaeontological
Memoirs this figure is copied from Plate G of the unpublished plates of the “ Eauna
Antiqua Sivalensis,” where other imperfeet crania are also figured.

The Indian Museum possesses a cranium of tliis species, of which two views are
given in the accompanying plates (XVII, fig. 2, and XIX), and which exceeds in
size any other specimen which I have seen. The specimen is from the Nerbudda
valley, and were it not for the loss of the greater part of the horn-cores it would
be in a splendid state of preservation.

The greater part of the right horn-core and almost the whole of the left horn-
core are absent from the specimen figured ; the distal portion of the nasal bones,
with a small part of the extremity of the maxiUae and the whole of the premaxillge
are also wanting ; the right pterygoid and the pterygoid process of the same side have
also been broken away. The crowns of the whole of the premolar series are broken
off even with their alveoli ; the two last molars of the right side and the whole
three of the left side are, however, quite complete. The bone is firm, although it
adheres strongly to the tongue ; with the exception of the fronto-nasal sutures all the
cranial sutures are completely obliterated ; the molars are also well worn, so that the
skull belonged to a fully adult individual, which was in all probability a male.

1 Vol. X, p. 131.

2 Yols. Ill, VllI, IX, X, and XIIL

133—46 CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES.

The cranium is immensely larger and more massive in all its proportions than
that of the largest male Arni I could obtain for comparison. All the ridges and
hollows for muscular attachment are much more prominent, indicating an animal
of immense strength and power. Size alone, however, would afford no ground for
specific distinction ; since, as is well known to sportsmen, the recent buffaloes of
Assam are diminishing steadily in the size of their skulls and horns under the con-
stant persecution to which they are subject ; in early times they were probably of
much larger size than any specimens now existing.

As the resemblance between the crania of the living and fossil Indian buffaloes
is so close, it will be best to commence by comparing the two together; on the
frontal aspect the cranium of Buhalus 'palcBindicus is considerably broader in pro-
portion to its length, and the infra-orbital region is less contracted laterally ; the
frontals are more elevated between the inferior angles of the horn-cores, in con-
sequence of which there is a sharp descent from this part of the forehead to the
orbits, in place of the regular and even slope which we find in the forehead of the
Arni.

In the figured specimen the superior border of the orbit is shorter and thicker and
stands out more abruptly from the lateral surface of the frontal than in the Arni ;
the anterior border of the orbit is produced laterally almost to the same extent as
the posterior border, in consequence of wRich in the frontal aspect of the cranium
only a very narrow segment of the orbit comes into view ; the axis of the orbit
being directed almost immediately outwards. In the cranium of Buhalus arni, on
the other hand, the anterior border of the orbit is much less produced than the
posterior border, and consequently the greater portion of the orbit is visible from
the front; the axis being directed outwards and forwards. In Dr. Ealconer’s
restored figure of the cranium of Buhalus palceindicus, however, the orbits are
drawn of the same shape and direction as those of Buhalus arni, and the direction
of the anterior border of the orbit in the Arni occasionally varies in different
individuals ; so that the above cannot be taken as a constant character.

Between the orbit and the supra-orbital foramen there is a very prominent ridge
of bone, forming the external border of the supra-orbital sulcus, this external border
being much higher than the internal border ; in the figured specimen there is no supra-
orbital foramen on the right side, although the sulcus is present. The anterior
border of the orbit is entire, and not notched as in the Arni. Above the orbit the
cranium has a much greater depth than occurs in the cranium of the’ Arni; tliis is
probably due to the greater development of the frontal sinuses. The interval
between the longer border of the frontal, overhanging the temporal fossa, and the
vertex cranii, is 4-9 inches, while in the Arni it is only 3-6 inches.

In a profile view of the two crania, that of the Arni is seen to slope rapidly and
continuously away from the vertex cranii to the extremity of the nasals, so that the
anterior surface of the nasals, at a distance of two inches below the orbit, is situated
one and a half inches below the plane of the superior border of the orbit. (The crania

CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES. 47—134

are supposed to be placed on a horizontal surface, resting on the paroccipital processes
and anterior molars.) On the other hand, in the cranium of JBubalus palceivdicus
there is a very abrupt and sharp slope downwards from the vertex as far as the
orbits ; and then the nasals are continued downwards with a less steep slope, so that
at a corresponding distance below the orbits, the anterior surface of the nasals is
in the same plane with the superior border of the orbit : this difference may, how-
ever, be brought about by a greater or lesser development of the frontal sinuses.

Below the orbits there is a considerable difference in the form of the two crania ;
in B. palceinclicus there is less lateral contraction of the maxilla below the orbit,
and the external surface of this bone slopes mucli more rapidly towards the median
line than in the recent species ; this is caused by the very different form of the
nasals in the two species ; in Bubalus arni these bones are very broad and flattened,
with no great V'idth at their infra-orbital angles, and becoming wider towards their
distal extremities ; the infra-orbital angle is placed considerably below the inferior
border of the orbit : from the wide and slightly-arched form of the nasals the
transverse section of the nasal cavity presents a low and wide arch, broader than
it is high ; the transverse diameter of the nasals at their infra-orbital angle is one and
a half inches, and at a point three inches below the orbit, two and a half inches. In
Bubalus palcBindicus the nasals are much wider at their infra- orbital angle than
at any other point, and become rapidly narrower below this point ; the specimen does
not, however, show their distal extremity ; instead of being nearly flat in the middle
of their course, they are greatly rounded and arched, so that the lateral surfaces
become very nearly at right angles to the plane of the median surface ; the infra-
orbital angles reach almost as high as the inferior border of the orbits : the
transverse diameter of these bones at the infra-orbital angle is two and a half inches,
and at a distance of three inches below the orbit two inches, — the latter diameter
being considerably less than the corresponding one of the smaller-sized cranium of
the Arni. The nasal cavity in transverse section forms a vaulted arch, higher than
broad, its boundaries forming an isosceles triangle, whereas the corresponding bound-
aries of this part of the Ami’s cranium approximate very closely to an equilateral
triangle.

The length of the two crania from the inferior border of the foramen magnum
to the anterior extremity of the true molar series is also considerably different,
being upwards of three inches longer in the fossil than in the recent species ; this
is correlated by the enormous width of the occipital region of the former, exceeding
that of the latter by two and a half inches.

The spheno-palatine region of the cranium of Bubalus palmndicus is at once
distinguished from that of the Arni by the relative directions of the bones com-
posing the basi-cranial axis and the surface of the palate : in the cranium of the
former species, a straight rod resting on the palatal surface will also rest on the
whole of the basi-sphenoid and basi-occipital ; whereas in the latter species a
similarly placed rod would only touch the basi-occipital close to the foramen

135—48 CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES.

magnum. In tlie cranium of tlie adult Arni this obliquity of the sphenoid to the
plane of the palate is less than in young individuals, hut in no specimen that I have
seen is there any approach to the identity of the planes of the two surfaces, seen in
all the sjoecimens of the cranium of Bubalus palmindicus which have come under
my observation.

Again, if a transverse line he drawn across the frontals through the centres of
the orbits of the two crania, the space between this and the intercornual ridge will
form an almost exact square in Bubalus palceindicus ; while in Bubalus arni the
antero-posterior diameter of a similar space will he one and a half inches longer than
the transverse diameter ; this difference being partly due to the different forms of
the anterior borders of the orbits described above.

In Bubalus palceindicus the surface of the palate is produced backwards behind
the last molar to nearly double the distance that it is in Bubalus arni ; in the former
the length of the palate behind the last molar is 2’6 inches, and in the latter only
1'3 inches ; the posterior nares are consequently placed much further hack in the
fossil species. In both species the free edge of the palatines is mesially bisected
by the vomer.

The maxillo-palatine suture can just he distinguished in the figured specimen,
and is placed entirely behind the posterior palatine foramina ; whereas in the recent
species the suture bisects these foramina ; in both species the suture commences at
the interval between the penultimate and ultimate molars, and at the median line is
on a level with the interval between the antepenultimate and penultimate molars.

The pterygoid processes of the alisphenoid are much larger and stouter in the
fossil species, and extend lower down below the plane of the molar alveoli : the
surface of the palate itself is also more arched from side to side.

The horn-cores are placed on the frontals similarly in both species ; they are of
larger size and directed more outwards in the fossil species ; specimens of the recent
species from Assam and Cachar most nearly resemble the fossil form in this par-
ticular.

On comparing the occipital surfaces of the two crania, we find more differences
than in the frontal aspects ; the occipital surface of Bubalus palcsindicus forms a
very wide and flattened arch, with the crest distinctly and clearly marked, and its
lateral borders convex ; this convexity causes the temporal fossae (perfect on the
right side of the figure) to become very low and narrow ; on the occqDital surface
the indentations of the fossae appear as pointed triangular slits on either side of the
crest; the protuberance for the nuchal ligament is slight, and is not continued
downwards as a sharp ridge to the foramen magnum ; this narrowness of the temporal
fossae. Dr. Falconer says, is constant in all the specimens examined by him. The
occipital crest of Bubalus arni forms a higher and more irregular arch, being
flattened at the summit, and less prominently and sharply defined ; its lateral
boundaries are somewhat concave, and the temporal fossae are consequently of con-
siderable width throughout their length, and terminate superiorly in a blunt rounded

CRANIA OP RUMINANTS PROM THE INDIAN TERTIARIES. 49—136

extremity ; the tuberosity for the nuchal ligament is strongly marked, and superiorly
confluent rrith the supra-cristal portion of the occipital surface ; inferiorly it is pro-
duced into a strong vertical ridge, affording attachment to the septum between the
occipital muscles. The occipital condyles are less obliquely placed, and their
external angles somewhat less sharp in the recent than in the fossil species.

The basi-occipital of B. palceindicus is distinguished from that of B. arni
by its much greater relative width : 4’1 inches in place of 2 ’9 inches across the
posterior muscular tuberosities ; these tuberosities are also nearly double the size of
those of the Arni.

The pattern of the crown of the molars of the two species is very similars
though the teeth can be distinguished. The enamel islands in the fossil species are
unusually large, and their free surface much corrugated and rugose : on the internal
side the median accesory tubercle is of very large size. The ultimate molar is
nearly half an inch longer than the corresponding tooth of the Arni ; this extension
in length is chiefly caused by the great development of the postero-external angle,
which almost forms a separate third lobe, being divided from the main body of the
tooth by an infolding of enamel from either side ; the infold of enamel on the
posterior side of the second enamel island, in all the molars, is much deeper in the
fossil species, and consequently remains on the worn crown for a much longer period
than in the recent species. Below, the measurements of the figured specimen are
compared with those of a cranium of .an adult male Arni : —

J5. palwindieas. B. arni-

Interval between foramen magnum and occipital crest .

4-2

3-7

„ „ occipital crest and vertex cranii

4-5

4-4

Width at superior border of orbits ....

ire

10-0

„ at constriction above orbits ....

110

9-4

„ „ below „ ....

8-4

6-5

From occipital crest to tip of nasals {B. 'palceindicus, broken) .

200

21-0

Width of occipital through petrosals ....

10-0

8-5

Height from surface of palate to frontals

7-8

6-6

Length from foramen magnum to last premolar

14-5

11-8

Interval between external angles of occipital condyles .

6-2

„ „ occipital crest and base of horn-core .

2-9

2-2

Length of temporal fossa .....

7-5

„ of intercornual ridge .....

7-3

Height from palate to broken extremity of nasals

6-0

Interval between base of horn-core and orbit

2-5

2-4

Antero-posterior diameter of left orbit

3-0

2-8

Transverse „ „ ....

2-6

2-5

Length of five molars . .

6-3

5-8

Width of palate at second molars ....

4-1

4-0

Interval between outer surfaces of ditto

6-4

5-9

„ „ supra-orbital foramina

G-5

5-3

„ „ occipital crest and line connecting mid-orbits

120

10-0

Length of outer surface of second molar

1-6

1*3

Transverse diameter of „ ....

1-2

1-0

Length of outer surface of ultimate molar

1-8

1-4

Interval between extremities of paroccipital processes

6-1

5*0

137—50 CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES.

Interval between outer surfaces of zygomatic arcbes

JB. palceindicus,

. 11-0

B. arni.

lO'O

Length from foramen magnum to free extremity of palatines .

. 7-6

6-9

Antero-posterior diameter of right born-core at base

. 7-3

5'2

Transverse „ „ „ . .

. 2-8

2-5

Circumference of „ „ . .

. 18-6

130

Width of occiput between temporal fossae

. 7-0

Another specimen of the cranium of Bubahis palceindicus in the Indian
Museum, also from the Nerhudda valley, and consisting of the upper part of the
cranium and the bases of the horn-cores only, has the following dimensions : —

Width at constriction above orbits ........ 9’5

„ of occipitals tbrougb petrosals ....... ll'O

Interval between external angles of occipital cond^des ...... 6'1

From occipital crest to vertex cranii ........ 4'2

Width of occipital between temporal fossae ....... 6'2

Interval between foramen magnum and occipital crest . . . . . .4 3

,, „ supra-orbital foramina ....... 5'4

„ „ inferior angles of born-cores . . . . . .ll'O

Antero-posterior diameter of base of left born-core ...... 6'4

The frontals of this specimen are remarkable for their extreme convexity
between the bases of the horn-cores, and from their rough and corrugated surface :
the form of the occiput agrees in every detail with that of the last specimen.
I may here add a few notes and measurements of the crania of this species taken
from Dr. Falconer’s Catalogue of the Fossil Vertehrata in the collection of the
Asiatic Society of Bengal. The finest specimen of the cranium of this species,
marked No. N. 18, was obtained from near Sagouni in the Nerhudda valley; its
dimensions are as follows ; —

Inches.

Length of right horn-core ......... 33'0

Breadth of „ at base ........ 6'5

Thickness „ „ ........ 4'5

Length of left born-core (broken) . . . . . . . . 23'0

Breadth of skull at base of occiput ........ 13'0

Height from condyles to plane of frontals ....... lO'O

“ The horn-cores spread out more horizontally and with a less inclination upwards
than in the existing wild buffalo, slightly concave anteriorly, and convex behind.
A cord stretched between the tij)s would subtend the plane of the cranium behind
the frontals ; in all these respects differing from the existing wild buffalo, and, so
far as their horizontal offset is concerned, approximating to the Gayal, from which,
however, they differ in the flattened form of the horns and in every other respect.
The posterior border of the horn-cores encroaches much upon the temporal fossa,
which is narrow. These characters are so constant, as shown by still better speci-
mens in the British Museum, that there can he little doubt but that the species is
distinct from the existing wild buffalo.”

CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES. 51—138

It will be observed from the above that the narrowness of the temporal fossa
is characteristic of Dr. Ealconer’s specimen as of our figured specimen ; the outward
direction of the horn-cores does not, however, appear to be of much value, as the
female of the Assam variety of the existing species possesses nearly horizontally
directed horns, as in the fossil species.

A specimen (No. N. 20), which Dr. Ealconer regarded as probably belonging
to a female individual of this species, has the following dimensions : —

Inches.

Length . . . . . , . . . . . 17 50

Height from sphenoid to frontal ........ 8‘00

Width at constriction between orbit and horn-core ...... 8'25

„ between middle of orbits . . . . . . . . 7'50

Length from vertex to upper margin of orbit ....... 7‘50

„ of line of molars ......... 6’50

Greatest width of left b orn-core ........ 5'00

Thickness of ...... ... 3’00

A large detached horn-core of a male individual (No. N 32) has
dimensions : —

Length of fragment .........

Girth near base . . . .

Long diameter .........

Short „ .........

the following

Inches.

11-0

197

7-0

4-5

Erom the above comparisons and measurements we find that the cranium
of the fossil Indian buffalo differs in the following points from the cranium
of the existing wild buffalo, mz., in its larger size, in the form of the infra-
cristal portion of the occipital surface, and in the narrowness of the temporal
fossae ; in the form of the orbits, nasals, and nasal cavity, and in that of the ultimate
molar ; in the degree of the inclination of the basi-cranial axis to the plane of the
palate, and in the backward prolongation of the palatines; the most important of
these differences appear to me to be those relating to the nasals, the temporal
fossae, and the basi-cranial axis and palate. The whole of the differences are but
slight, and yet they are quite sufficient to distinguish the skull of the fossil from
that of the recent form, and I therefore think it best to continue to consider the
two forms as distinct species in the modified acceptation of the term, although
there can be no doubt but that the one species is the direct ancestor of the other.
In opposition to this view, however, I see that Mr. Eoyd-Dawkins ‘ considers the
fossil form (misnamed in his notice Bubalus namadicus) as specifically identical with
the Living Bubalus arni ; this determination was apparently made on the evidence
of Dr. Ealconer’s comparison of the two crania.

The fossil species ranged throughout a great portion of the central Nerbudda
vaUey, and its remains have been noticed by Dr. Ealconer from the older alluvium
of the Jamna valley ; bones of this species have also been obtained by Mr. Eedden

* Cave Hunting, p. 428.

139—52 CRANIA OE RUMINANTS EROM THE INDIAN TERTIARIES.

from the older alluvia of the Godavari and Perim-Ganga vaUeys, and I think by
Mr. Eoote from Madras ; all the above deposits undoubtedly belong to the upper
Pliocene or Pleistocene periods. Lately, however, as I have already previously
noticed,^ two crania of this species have been obtained by Mr. Theobald from the top-
most clay beds of the Siwahks near the town of Bubhor ; these crania were found in
company with the remains of Camelm sivalensis, and the beds are certainly newer
than those from which the great mass of Siwalik bones are obtained; I have
thought it very probable from the occurrence of this Bubalus in these beds, and not
in the beds below, that these beds are not far removed in time from the ossiferous
gravels of the Nerbudda valley. Mr. Medlicott, however,^ on stratigraphical consid-
erations, is inclined to think that there is a long gap between these topmost Siwahks
and the Nerbudda deposits ; the old high level terraces of the Sub-Himalayas
being more nearly contemporaneous with the latter deposits, and there is no doubt
considerable probability in this supposition. It is, however, very unhkely that a
species like Bubalus palcBindicus (for, as we shall see below, except in the matter
of size, I cannot distinguish between the upper Siwalik and Nerbudda crania)
which has been considered by some as identical with a living species, should have
existed at a period much older than that of the gravels of the Nerbudda. There
is no doubt, however, but that the Bubhor Siwaliks are serially continuous with,
and are not separated by, any (geologically) long period from the subjacent mamma-
liferous beds, and the occurrence of Bubalus palceindicus in the former is another
link in the chain of arguments for not placing the age of the great mammahferous
series of the Siwaliks below the older Pliocene.

To return to the Siwalik crania, we find that both of them are of considerably
smaller size than the average of the specimens from the Nerbudda valley ; the
larger of the two is, however, certainly an adolescent animal, the crowns of the last
molars having only just come into wear, and most of the cranial sutures being
still distinctly visible ; the skull is slightly larger in its dimensions than the female
skull from the Nerbudda noticed above, and it is therefore probable that both the
Siwalik specimens belonged to female individuals.

The extremity of the larger skull has been broken off between the premolar
and molar series, both zygomatic arches are wanting, together with considerable
portions of the boundaries of the orbits ; part of the right frontal with its horn-
core, and the greater part of the left horn-core, are also wanting ; the sphenoid
region is considerably injured, both the paroccipital and one of the pterygoid pro-
cesses having been broken off; in the right maxilla there are crowns of three
molars remaining, a portion of their outer walls being, however, broken away ;
on the left side there are only portions of the fractured crowns of the two
ultimate molars.

^ Kec. Geol. Snrv. India, Vol. IX, p. 88.

^ Kec. Geol. Surv. India, Vol. IX, p. 57.

CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES. 53—140

This specimen agrees with the Nerbuclda crania in having narrow and arched
nasals, witli the highly vaulted nasal cavity ; the anterior border of the orbit, though
fractured in our specimens, seems also to have been produced in a similar manner ;
the occipital surface agrees precisely in form with the Nerbudda specimens. The
proximal extremity of the nasals extends upwards on the face as far as the centre of
the orbits ; the fronto-nasal suture is distinct, and terminates in a simple unbroken
angle, there being no process of the f rentals descending between the two apices
of the nasals as in the Arni ; the palate is produced still further back than in
the Nerbudda specimens, and therefore differs still more Avidely from that of the
Arni : the portion of the horn-core Avhich remains is directed outwards as in the
Nerbudda specimens. The dimensions of the specimen are as follows : —

Interval between occipital crest and foramen magnum .

Inches.

„ „ and vertex

4-4

Width at constriction above orbits

9-J

„ below „ . . .

7-6

Length from occipital crest to tip of nasals (broken) .

17-0

AVidth of occiput through petrosals

8'8

Height from palate to frontals

6-2

Interval between foramen magnum and last premolar .

13-0

„ „ external angles of condyles .

5'3

Length of intercornual ridge

6-0

Interval between orbit and horn-core .

1-8

Antero-posterior diameter of orbit

2-4

Length of three molai-s ....

4-2

AVidth of palate at second molars

3-4

Interval between outer surfaces of second molars

5-8

„ „ supra-orbital foramina

3-6

AVidth of occiput betweeir hom-cores

4-4

Antero-posterior diameter of base of horn-core

6*4

Transverse „ „ »

The second Siwalik specimen, though of somcAvliat smaller size, agrees in form
with the described specimen.

I have elsewhere stated^ that another species of buffalo, Buhalus antiques,
has been described by Professor Gervais^ from newer Tertiary strata in Algeria,
Avhich in the form of its occiput, of its horn-cores, and of its frontals, approaches
much more closely to the Indian than to any of the African buffaloes ; it has, how-
ever, certain points of affinity to the African B. hrachyceros ; the existence of this
species is important, as shoAving a former intermingling of the Mammalian Tertiary
Eaunas of India and Africa.

In the valleys of the Godavari and Nerbudda, Bubahis iJalceindimis was un-
doubtedly a contemporary of man, since stone implements have been found in
the ossiferous beds of localities from wliich the remains of the extinct buffalo
are obtained.

Eec. Geol. Surv, India, Vol. IX, p. 99.

^ Zoologie et Paleontologie Generales : prem. ser. pi. XIX.

141—54 CRANIA OP RUMINANTS PROM THE INDIAN TERTIARIES.

Genus : PERIBOS, n. gen. noUs.

The name of this new genus, or, perhaps, snb-genus, inadvertently appeared
without any descriptive notice in my paper on Indian Tertiary Mammalia published
in the “Records of the Geological Smwey of India.” ^

The genus is founded upon the single cranium described below, and from cranial
characters, may be defined as follows : — Prontals fl:at, and broader than long ; horn-
cores pyriform in cross section, short and curved, closely approximated at their
bases, and situated on a ridge of the frontals, which is somewhat below the plane
of the occiput ; the true occipital surface bounded by its crest, and the latter
approximated to the horn-cores ; teeth and basi-occipital of the Bovine type.

Peribos occipitalis, Falc. sp. (?). Pis. 20 and 21, f. 2.

The doubt as to the original application of the above specific name to the
figured cranium has been already noticed in the introduction, and need not be
repeated here. The specimen is in the Siwalik collection of the Asiatic Society of
Bengal; it is numbered in the Catalogue of the Society’s collection S 562, and has
been there described by Dr. Ealconer ; it was obtained from the Siwaliks of Gana-
wur, and has been referred to in the “Journal of the Asiatic Society of Bengal.”*^
Dr. Ealconer’s description of the specimen is as follows : —

“ Eine skull of a Bovine Buminant, nearly perfect from the occiput to the
diastema, showing the zygomatic arches, temporal fossae, and the wEole of the
sj)heno-palatine region, together with two lines of molars in situ ; the crowns of
those on the right side broken off ; the three posterior molars on the left side nearly
entire ; a horn-core is present on the left side, absent on the right, through a fracture
which has carried it off below the base. The cranial part of the skull differs
remarkably from all known Bovine Bmninants in this respect, that the occipital
bone appears to terminate at the occipital crest, or close to it, and that no part of
the parietals enters into the occipital plane.

“ The horns are pyriform in section, with a very sharp edge behind, and a
broad surface in front ; they are closely approximated on the brow, and start
upwards and outwards, but curve forwards towards their tip. The plane of the
frontals is flat between the commencement of the horns, and then descends in a
sudden curve between the horn-cores to meet the plane of the occipital at an obtuse
angle. The occipital crest is very prominent ; orbital rim also prominent ; the
lachrymal bones present rough tuberosities at the orbital margins, as in the Bovine
group ; there is also no lachrymal fissure ; the two supra-orbital foramina large.

> Vol. IX, p. 90.

2 Vol. V, p. 184.

CEANIA OP RUMINANTS PROM THE INDIAN TERTIARIES. 56—142

The nasals are received into an angular fissure of the frontals, then’ apices ascending
nearly to a line with the anterior (inferior) border of the orbits.

“ The orbits differ in a remarkable manner from those of the ordinary Bovine
Bmiinants in having their greatest diameter in the vertical direction instead of
antero-posteriorly. Between the supra-orbital foramina there is a raised portion of
the surface of the frontals of a horse-shoe shape, about a line in thickness, and
with a rugous and fimbriated margin about two inches in breadth ; the sinus at the
posterior (superior) part passing gradually into the surface of the frontals. This
rugous disc is unknown in other Ruminants. The maxillaries contract sharply in
front of the orbits ; their tuberosities are very prominent.

“ Teeth. — The three posterior molars on the left side are well worn, showing
the animal to have been fully adult, and a large accessory pillar, narrow and com-
pressed, but of considerable depth, is seen in the sinus between the two barrels of
each molar on its inner side as in the Bovine Ruminants. The palate is broad.”

Most of the following dimensions were also given by Dr. Palconer : —

Inches.

Lengtli from occipital condjles to diastema

„ summit of occipital crest to apex of nasals
Width at superior border of orbits

„ at constriction below „ . . .

„ at inferior border of „

„ at constriction above „ . . .

From base of occipital to summit of occipital crest
Width of occipital through petrosals
Height from surface of palate to frontals
Interval between external angles of condyles .
Distance between occipital crest and base of born-core .
Length of temporal fossa • . . .

Interval between bases of born-cores
Height from palate to broken extremity of nasals
Interval between base of horn-core and orbit .
Transverse diameter of left orbit
Antero-posterior diameter of left orbit .

Transverse diameter of right orbit

Antero-posterior diameter of right orbit

Length of five right molars ....

Greatest width of palate ....

Length of fragment of left horn-core .

Antero-posterior diameter at base

Thickness of ditto .....
Antero-posterior diameter at fracture
Transverse „ "

Height from surface of palate to frontals below orbit .
Interval between supra-orbital foramina
Distance between outer surface of first molars .

Width of nasals at infra-orbital angle .

Interval between temporal fossse above horn-cores

12-6

8-6

7-7

5'5

6'0

8- 5
7-0

3- 8
IT
5'5
2-0
4‘f5

1- 9

2- 7

1- 9
30

2- 0

4- 9

3- 3

9- 0
50
3-0

3- 4
2-7
60

4- 7
5T
2-6
2-9

“ The dimensions of the two orbits differ a little in consequence of crusliing
on the right side. The principal distinctive marks are : The occipital not rising

143—56 CRANIA OE RUMINANTS EROM THE INDIAN TERTIARIES.

above the occipital crest. The great narrowness of the parietal region between the
liorn-cores and the occipital crest. The close aj^proxiination of the horn-cores.
The short interval between the horn-core and the orbit. The projection of the orbit
above the plane of the f rentals.”

To the above description I add the following remarks : On referring to the
table of the frontal measurements on the crania of Indian oxen given below’, the
transverse diameter of the forehead of this specimen will be found to be consider-
ably longer than the antero-posterior diameter. The nasals are very highly arched,
and are widest at their infra-orbital angle, diminishing very rapidly in width below
this point ; there is a ridge between the infra-orbital angle of the nasals and the
orbit. The supra-orbital foramina pierce the bone obliquely, and are placed at the
extremities of the sulci ; the sulci do not extend downwards as far as the lower
border of the orbit ; the orbits have their jDOsterior border prominent at the inferior
angle, their axis looks obliquely forwards and outwards. The constriction of the
skull above the orbits is of no great extent.

As noticed above by Dr. Ealconer, no portion of the parietals of this cranium
enters into the plane of the occipital surface ; that surface is consequently bounded
superiorly by the occipital crest, and is composed solely of tlie proper occipital
bones which do not extend above the crest. The parietals are in a different plane from
the occiput, sloping upwards to the highest ridge of the cranium, which is in
consequence placed considerably in advance of the plane of the occipital, instead
of being directly over it, as in Bos ; this form of the hinder part of the cranium
is an approach to the still more aberrant form of the cranium of the next genus
Uemihos ; owing to the obliteration of the sutures, the boundaries of the parietals
cannot be seen.

The whole of the temporal fossae may be seen in a back view of the cranium ;
they cannot be said to intrude on the plane of the oceipital, but rather to bound
the lateral portions of this surface. The summit of the occipital crest is flattened,
and its lateral borders somewhat concave superiorly ; there is a large tuberosity
for the attachment of the nuchal ligament ; the intercornual line is slightly concave;
the superior border of the horn-cores is convex, and the inferior concave, as in the
genus Bos.

The upper border of the temporal fossa is very nearly parallel to the lower or
zygomatic border; the descending post-orbital process of the frontal is nearly
vertical, forming almost a right angle both with the upper and lower borders of the
temporal fossa ; the inner waU of the temporal fossa is somewhat concave from
above downwards, but is straight from before backwards. The maxiUse do not slope
inwards to any great extent, so that the lateral surfaces of the nasals are nearly
at right angles to the anterior surface. The skull is of great vertical depth from
the orbit to the palate.

On the spheno-palatine surface the basi-occipital extends forwards for a short
distance in the plane of the palate ; at the anterior tubercles, however, it is bent

CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES. 57—144

sharply on itself ; the anterior tubercles are small ; the interval between the posterior
tubercles is not greater than that between the anterior pair. The posterior tubercles
are the larger of the two, and the basi-occipital is triangular as in the oxen, but
the tubercles are larger and with a more distinct median groove, as in the antelopes.

The palate is long and narrow, and but sKglitly arched ; the palatines are
produced backwards for a distance of one and a quarter inches behind the last
molar, before their division. The molar series is much curved from before back-
wards ; in front of the premolars the maxiUse are deeply grooved longitudinally.
The molars are characterised l3y a very large accessory tubercle on the iimer side ;
by the enamel islands being elongated posteriorly, and by the island of the posterior
lobe being deeply indented from the same side by an enamel fold ; the costm on the
dorsa of the teeth are slight, and the whole covered with a very thick coating of
cement. The length of the last molar is I'G inches and its width 1 inch.

The maxillo-palatine suture is placed entirely in advance of the posterior
palatine foramina ; on the outer side it commences at the level of the penultimate
molars, and passes inwards and forwards, so that in the median Ime of the palate
it is on a level with the interval between the first and second molars. The posterior
free border of the palatines is entire, and not bisected by the vomer.

A fragment of the proximal extremity of the premaxilla of the right side is
seen occupying the angle between the maxillae and the nasals, showdng that this
bone was of the long type, as in typical species of Bos and in Buhalus.

The present cranium is distinguished from the crania of all species of Bos and
Bibos by having the horn-cores placed on a ridge, which is situated below the plane
of the occipital surface, instead of directly over it ; by the peculiar form of the
cross-section of the horn-cores ; and by the plane of the occiput being entirely dis-
tinct from the plane of the parietals ; from the former of the above genera tlie
present cranium is further distinguished by the transverse diameter of the forehead
being longer than the antero-posterior diameter, and by the facial portion being
consequently longer than the frontal portion; from the latter genus it is also
distinguished by the convexity of the upper border of the horn-cores.

From the genus Buhalus this cranium is distinguished by the frontals being
slightly concave instead of convex, by the superior border of the horn-cores being
convex instead of concave, and by the form of the nasals and of the occipital
surface.

The cranium agrees with the crania of the genus Bison in having the horn-cores
placed below the plane of the occiput ; it is, however, distinguished from the crania
,of that genus by the approximation of the horn-cores at their base, by their pyriform
cross section, by the fact of no portion of the occipital crest being visible in a front
view of the skull, and by the convexity of the upper border of the horn-cores.

It is quite clear, therefore, that this species cannot be included in any of the
above genera : the only species with which it seems to have any affinity is Bos
acutifrons, the horn-cores of the two having a pyriform cross section.

145—58 CEANIA OE EUMINANTS EEOM THE INDIAN TERTIAEIES.

In the approximation of the bases of the horn-cores, and in the separation
of the occipital and parietal planes, the skull makes some approach to the skulls of
the goats ; the position of the highest point of the frontals, almost midway between
the orbit and the occiput, is an antelopine character ; the large size and distinctness
of the posterior tubercles of the basi-occipital is also a character of that group ;
the general form of the basi-occipital, however, and the structure of the teeth,
is essentially Bovine.

Genus: HEMIBOS, Falconer.

This genus was made by Dr. Ealconer for the reception of an aberrant Bovine
cranium from the Siwaliks ; the name, however, only has been published, no
description having ever appeared.

The genus may be shortly defined as follows, from the characters of the
cranium : —

Erontals concave, broader than long : horn-cores triangular in cross-section, short
and straight, situated below the occiput on a high ridge of the frontals ; occipital
and parietal planes distinct : facial longer than frontal portion : orbit and horn-core
approximated. Teeth and basi-occipital of the Bovine type.

Hemibos TRiQUETiiiCEiios, Falconev. Pis. 22 and 23.

Eigures of the cranium of this species have been engraved among the un-
published plates of the “Eauna Antiqua Sivalensis” {Plate 11), and a few measure-
ments are given in the accompanying index ; another figure of the cranium will be
found in Boyle’s “ Himalayan Botany ” {Plate VI, Jig. 5): no detailed description
has, however, apj)eared.

The cranium figured here {Plates XXII and XXIII ) was collected by Major
Godwin- Austen in the Siwaliks of the Markanda Biver ; the greater portion of the
specimen is complete, though somewhat weather-worn ; about one-half of the left
horn-core remains, while the whole of that of the right side is broken away : the
distal portion of the maxillae, together with the whole of the premaxillse and the
nasals, are also wanting. The palate is nearly free from matrix, and shows the roots
of the molar series of either side, the crowns of all the teeth having been broken off
close to their alveoli; the spheno-palatine region is somewdiat imperfect. The
cranium is that of a fully adult animal, nearly all the sutures having been
obliterated.

If we turn to the profile view of the cranium given in the accompanying
lithographs {Plate XXIII), we find this aspect to be totally different from the
corresponding aspect of the crania of any of the Bovines described above, and, as
will be more fully noticed below, approaching the form of the crania of the goats.

CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES. 69—146

The horn-cores are very closely approximated at their base, and are placed on the
summit of a high ridge, formed by the junction of the very distinct planes of the
parietals and frontals ; the former bones apparently form a distinct and broad band
across the cranium behind the horn-cores, as in the sheep and antelopes, and are con-
sequently much larger than in the oxen ; the intercornual ridge is very far removed
from the plane of the occipital surface ; the horn-cores themselves have a triangular
transverse section, which forms very nearly an equilateral triangle ; their internal
surface is placed nearly antero-posteriorly, and the external border immediately
above the orbit ; their direction is upwards, outwards, and a little backwards, dimin-
ishing regularly and somewhat rapidly in diameter, without any curve or twisting ;
their anterior border is continuous with the plane of the frontals : the base of the
horn-cores is situated directly over the anterior half of the temporal fossa, leaving
the posterior half uncovered above ; there is a very short interval between the base
of the horn-core and the adjacent border of the orbit. The orbits are somewdiat
worn and broken in this specimen, but naturally, as shown in other specimens, are
non-salient, somewhat depressed below the general plane of the frontals, and having
their antero-posterior diameter rather longer than their transverse diameter. From
behind the orbit the temporal fossae gradually increase in depth, and curve round on
to the plane of the parietals, so that they are closely approximated at this point, the
parietals being proportionately narrow. From the very forw^ard position of the
horn-cores and intercornual ridge, the antero-posterior diameter of the frontals is
very short, considerably less than the transverse diameter {see table of measurements
given on page 66). The cranium is but very slightly constricted in width, either above
or below the orbits, the front view showing an almost complete parallelogram ; the
intercornual ridge is deeply concave; this concavity extends downwards to the
middle of the frontals ; there is a high rounded prominence on the inner side of the
orbit : below this there is an oblique channel running outwards and downwards from
the frontal hollow, while the apex of the nasals again rises as a prominent pyriform
swelling ; the apex of the nasals extends upwards on the forehead, as high as the
lower thud of the orbit. On referring to Falconer’s figures of the cranium of this
species in the unpublished plates of the “Fauna Antiqua Sivalensis,” we find that
the facial portion is double the length of the frontal portion ; the nasals are wddest
at their infra-orbital angle, and the premaxillse are of unusual length, and extend
upwards between the nasals and the maxillae.

There is a prominent angle leading from the orbit to the nasals ; and the lateral
surfaces of the maxillae are nearly parallel and of great depth.

The anterior angle of the horn-core runs downwards as a kind of buttress on to
the surface of the frontals ; the base of the horn-core is burred.

Turning to the parietal and occipital regions of the cranium, w^e find the two
occupying entirely separate and distinct planes, the occipital plane being bounded
superiorly by the occipital crest ; the plane of the parietals forms an obtuse angle
with the planes of the frontals and the occipital.

147—60 CEANIA OE EUMINANTS EEOM THE INDIAN TEETIAEIES.

Owing to the obliteration of the sutures in all the specimens, the exact boundaries
of the parietals cannot he determined ; the anterior boundary must, however, have
been somewhere close to the posterior angles of the horn-cores, so that the parietals
formed, as I have said, a broad transverse band across the skull, as in the antelopes
and goats, the frontals being similarly bent on themselves at the intercornual
ridge.

The occipital crest forms a wide arch, somewhat flattened above, for a distance
of half the width of the cranium ; the tuberosity for the attachment of the liga-
mentum nuchse is large and distinct ; there is a median ridge, dividing the part of the
occipital surface above the foramen magnum, for the attachment of the intermuscular
septum of the recti and other occipito-vertebral muscles. The occipital condyles are
of large size ; the space for the attachment of the digastric muscle is very narrow ;
the form of the paroccipital processes is not shown in our specimens.

The supra-orbital foramina are situated far apart from each other, and pierce
the frontals at right angles ; there is scarcely any sulcus below them ; there are no
infra-orbital sinusus or fissures as in the true antelopes.

The palate is highly arched from side to side, and but slightly produced behind
the last molars ; the molar series are placed nearly parallel to each other ; the
posterior palatine foramina are situated at the level of the second molars, enthely
behind the maxillo-palatine suture.

On the basi-occipital {F. A. S. Plate H, Jig. 6) the posterior tubercles are
placed immediately in front of the occipital condyles, without any intervening
groove, and are far apart ; the interval between their internal borders is very nearly
equal to that between the external borders of the anterior tubercles : the anterior
tubercles are the smaller of the two ; the posterior are rounded ; the basi-occipital
is triangular in form, as in the oxen.

The dimensions of the specimen figured in Plate XXII are given below : —

From occipital crest to intercornual ridge
Width at superior border of orbits
„ contraction below orbits

„ „ above „

From lower border of foramen magnum to occipital crest
Width of occipital through petrosals
Height from palate to frontals at orbit
Interval between external angles of occipital condyles

„ „ occipital crest and posterior angle of born-co

„ „ base of horn-core and orbit

Antero-posterior diameter of left orbit
Transverse „ „

Length of five molars
Greatest width of palate

Interval between outer surfaces of first molars
„ „ foramen magnum and last molar

Length of left born-core fragment along anterior border
„ internal surface of base of left born-core
„ anterior „ „ „

„ posterior „ „ „

Inches.

5- 4

6- 7
5T

7- 8
5-3
40
31
1-8
2-7
2-5
50

2- 7
5-4
7-8
5T

3- 5
2-9
3“4

CRANIA OP RUMINANTS PROM THE INDIAN TERTIARIES. 61—148

Widtli of narrowest portion of parietals
Interval between supra-orbital foramina
Length of temporal fossa
„ intercornual ridge

Inches.

2-4

3'1

3*0

The measurements of two other specimens of cranium of this species collected
by Mr. Theobald from Siwalik strata are as follows : —

Width at superior border of orbits

8-0

7-9

„ constriction above orbits

7-2

0-0

Interval between occipital crest and intercornual ridge

6'3

Width of narrowest part of parietals

2-4

2-8

„ occipital through petrosals

7-2

7'6

Height from lower border of foramen magnum to crest

.4-8

4-5

Interval between external angles of condyles

. . . . 4-5

5-2

Length of intercornual ridge ....

1-7

2-2

Prom the above measurements it will be i

seen that the latter of

these two

specimens differ from the first specimen in having

a wider and lower occipital region

and larger occipital condyles ; the occipital crest forms a lower and flatter arch.
The parietals between the temporal fossae are also wider, and the horn-cores are of
smaller dimensions, and directed more outwards.

The specimen drawn in Plate H, fig. 6, of the unpublished plates of the “ Pauna
Anti qua Sivalensis” has the bases of the horn-cores set much farther apart than in
any of the above specimens, the interval between them being nearly six inches in
length.

The teeth of all the specimens in the Indian Museum are in a very imperfect
and damaged condition, but a few of them show that they were provided with
the accessory median tubercle so characteristic of the oxen.

We may now consider iu what points the cranium of the present peculiar
genus is related to allied groups of Ruminants ; firstly, we find that the jiosition
of the horn-cores, which are placed on a ridge of the frontals far in advance of or
below the plane of the true occipital surface, and almost directly above the orbits,
together with the distinctness of the planes of the occiput and the parietals,
distinguishes the genus from all the true oxen, both recent and fossil, and approxi-
mates it to the antelopes and goats. Prom the true antelopes, however, the
cranium of Semibos is distinguished by the form of its horn-cores ; no recent
antelope has triangular horns, although those of Antilo'pe cordieri of the Miocene
Tertiaries of Montpelier are angulated in front ; it is also distinguished from this
group by the non-depression of the facial surface of the lachrymals and by the
absence of any trace of the infra-orbital vacuity.

Among the aberrant antelopes and thefi allies we find that the cranium of
Semihos at first sight presents a considerable general resemblance to that of
Tragocerus amaltlieus from the Miocene Tertiary of Attica in both genera the
planes of the frontals and the parietals form an obtuse angle with each other, and

' Gaudry : “ Animauz Fossiles et Geologie. de 1’ Attique.” PI. XLIX.

149—62 CEANIA OP RUMINANTS PROM THE INDIAN TERTIARIES.

in both the molars have an accessory median column ; the skull of Tragonerus,
hovrever, is distinguished by having the horn-cores placed more immediately over
the orbits, by their cross-section being ellipsoidal, and by the general shortness of
the face.

Turning now to the true goats and comparing the crania of the two species of
Hemitragus with the cranium of Hemihos, we find that in both genera the plane of
the occiput forms an acute angle with the plane of the frontals ; in both the inter-
cornual ridge is situated above the superior border of the orbit ; while in both
the horn-cores are triangular in cross section, and have their anterior border con-
tinuous with the plane of the frontals, and the parietals form a distinct transverse
band across the skull behind the horn-cores. In both the horn-cores are approxi-
mated to the orbits, and the su]3ra-orbital foramina have the same general position
and relations ; the horn-cores of Semihos are, however, much wider apart at their
bases than in any living goat, and are also distinguished by being straight.

The cranium of Semihos^ again, differs from that of any existing species
of goat, in that the plane of that part of the frontals which is below the horn-
cores forms an obtuse angle with the plane of the parietals and of that portion
of the frontals which is behind or above the horn-cores ; whereas in the goats and
their alhes these two planes are placed at right angles to each other.

Again, the presence of the median accessory tubercle in the molars of Semihos
separates the cranium of that genus from the crania of all recent goats ; this acces-
sory tubercle occurs in the molars of all the oxen, and in those of the larger deer ; it
is generally absent in the antelopes, but is found in the uj)per molars of the
Gnu, and in those of Antilope hoodon, and of Tragocerus amaltheus of the European
Miocene ; it is entirely absent in all species of goats and sheep.

Turning now to the basi-occipital of Semihos, we find that this bone is wedge-
shaped, and therefore is more Bovine that Antilopine or Caprine in character ; the
posterior paff of tubercles are considerably the larger of the two, and are separated by
only a small interval from the occipital condyles, as in the oxen ; the median groove,
however, between the tubercles of either side is deeper than in the oxen, and more
hke the same part in the antelopes ; in the latter group, however, and in the goats,
the basi-occipital is wide and quadrate in shape, both pairs of tubercles are well
marked, but the anterior pair are generally the larger of the two, and the interval
between the two tubercles of opposite sides is nearly uniform ; in the goats there
is a considerable interval between the hinder pair of tubercles and the occipital
condyles.

The nasals of Semihos are narrow inferiorly, and taper to a point as in the
goats and antelopes.

On the whole, it appears that Semihos presents characters intermediate between
the oxen on the one hand and the goats and the antelopes on the other, but that
the majority of its affinities are with the oxen ; while in the form of the horn-cores
it is distinct from all.

CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES. 63—150

Genus : AMPHIBOS, Falconer.

This genus, as I have already stated in the introduetion, was formed by
Dr. Ealconer upon the evidence of the crania of three species of oxen-like animals
from the Siwaliks, of which species only one can be now recognised. As in the case
of the preceding genus, no definition of Amphihos was ever given by Dr. Ealconer ;
from the characters of the cranium it may be shortly defined as follows : —

Erontals flat or slightly hollow, broader than long : horn-cores rounded in front,
angulated behind, long and porrect at first, placed below the plane of the occiput on
a slight elevation of the frontals ; parietals shorter than in last genus ; facial nearly
as long as the frontal portion ; teeth of the Bovine type.

Amphibos acuticobnis, Falconer. PI. 21, f . 1 ; PI. 24.

Eigures of the cranium of this species have been given in one of the unpub-
lished plates of the “ Eauna Antiqua Sivalensis,” and a few notes are added in the
accompanying index. We possess in the Indian Museum several crania which agree
in form and size with Ealconer’s figures of this species, and which I have accord-
ingly referred to it ; the specimen of which a frontal view is given (Plate XXIV)
was collected by Mr. Theobald from the Siwaliks of the Kangra district ; with
the exception of the greater part of the premaxillse, which have been broken away,
the whole of the facial and frontal portions of the cranium are complete ; both
horn-cores have been broken away near to their origin ; the fragment of the horn-
core remaining on the left side of the specimen has been damaged on its anterior
surface, so that its true form is not shown ; the occipital and spheno-palatine regions
could not be cleared from the exceedingly hard matrix in which they are embedded ;
most of the cranial sutures are obliterated, and the permanent molars are much
worn, showing that the cranium belonged to an adult animal.

A profile view of another specimen also from the Siwaliks is given on Plate XXI,
Eig. 1. In the front view of the cranium the frontal and facial portions of the
cranium are seen to be of nearly equal length ; the frontals are somewhat broader than
long, and depressed in the middle line ; the orbits are salient and separated by a deeply
excavated but short interval from the horn-cores ; inf eriorly the maxillae contract
suddenly below the orbits (in some specimens, however, this contraction is not so
marked, and it may probably be a sexual character). The nasals are long and
rounded ; at their proximal extremity they reach about a third up the orbit ; they
are widest at their infra- orbital angles ; they contract below this and again expand
at their distal extremity. The horn-cores are situated on a ridge of the frontals
above and behind the orbits, but considerably in advance of the plane of the occi-
put ; they are separated by a moderate interval at their bases, the median line of
the frontals being hollowed between them ; from the fragments remaining in the

151—64 CHANIA OF RUMINANTS FROM THE INDIAN TERTIARIES.

figured and other specimens the horn-cores appear to have been directed upwards,
backwards, and outwards without cmwature ; in transverse section they are rounded
anteriorly with a ridge posteriorly ; their widest diameter is placed antero-poste-
riorly ; a slight angulation forms their anterior border, which is continued upwards
in the line of the face. Behind the intercornual ridge the frontals and parietals
slope backwards to the occipital crest ; the ante-cornual and post-cornual planes of
the frontals form an obtuse angle at their junction ; the post-cornual frontal plane
and the occipital plane also form an obtuse angle at then’ junction.

The temporal fossse are long and narrow, having a rounded distal border ; the
orbit is large, and has its antero -posterior diameter rather the larger of the two ;
from the antero-inferior border of the orbit there is a well marked ridge running
to the maxillary tubercle above the second premolar.

The parietals, as in the last genus, form a distinct hand across the cranium
behind the horn-cores ; the temporal fossse extend slightly on to this upper surface
of the parietals. The occiput, as in the antelopes, is entirely distinct from the
parietals, no part of the latter extending on to the plane of the former. The
occipital crest is- well marked, and considerably overhangs the general plane of the
occiput ; the occiput is expanded laterally, and the crest and condyles are con-
sequently approximated ; the paroccipital processes are long and placed unusually
close to the condyles.

On the palatal surface the lines of molars are placed rather far apart, and are
nearly straight ; the palate is vaulted and extends backwards considerably behind
the last molars ; the molars have a large accessory column. The hasi-occipital is
wedge-shaped, as in the typical oxen ; the posterior tubercles are large, rounded, and
scarcely separated from the condyles ; the anterior tubercles are situated consider-
ably in advance of the posterior pair; they are very small, elongated antero-
posteriorly, and approximated.

The dimensions of the specimen drawn in Plate XXIV are as follows : —

Length from foramen magnum to diastema
„ „ occipital crest to apex of nasals

„ of nasals .....

Width at superior border of orbit

„ constriction below „ ...

„ „ above „ ...

From lower border of occipital foramen to occipital crest
Width of occipital through petrosals
Height from surface of palate to frontals
Interval between external angles of occipital condyles .

„ „ occipital crest and superior angle of horn-core

Length of temporal fossa ....

Height fi'om distal extremity of palate to surface of frontals
Interval between horn-core and orbit
Antero-posterior diameter of right orbit
Transverse „ „ . . .

Length of six molars of right side

Inches.
. 12-0

. 8-0

. 7-8

. 5-4

. 6-5

. 5-0

. 8-3

. 6-1

. 3-9

. 1-2

. 5-5

. 4-4

. 2-2

. ^ 2-4

. ' 2-0

. 6-4

CRANIA OR RUMINANTS PROM THE INDIAN TERTIARIES. 65-152

Greatest width of palate

Antero-posterior diameter of base of right horn-core
Transverse „ „ „

Distance between outer surfaces of first molars

Inches.

. 30

. 3-6

. 2-4

. 5T

Tlie specimen is distinguished from the last genus hy the smaller diameter of
the horn-cores ; by the absence of a triangular cross-section of the horn-cores and
by then more depressed direction ; by the less backward projection of the occipital
condyles ; by the wider interval between the bases of the horn-cores ; and the less
excavated median line of the f rentals between them; by the smaller interval
between the intercornual ridge and the occipital crest ; by the greater proportionate
width of the transverse diameter of the frontals, and by the smaller elevation of
the intercornual ridge above the base of the skull ; the continuation of the line of
the base of the molars cuts the occipital condyles in the present species, while in
jff. triquetriceros a similar line would cut the occiput considerably above its centre ;
the form and position of the horn-cores of Amphihos suj0B.ciently distinguishes it
from Feribos.

Dr. Ealconer’s remarks on the specimen of the occiput and horn-cores of this
species figured in Plate I, figs. 2 and 2c^, of the “ Eauna Antiqua Sivalensis” are as
follows : “ The direction of the horns is more upright than in Semibos triquetri-

ceros ; the horns are rounded on their anterior surface and flattened behind and
taper to a point.” The measurements of the specimen figmred in fig. 3 of the same
plate are as follows

Inches.

Length of fragment ....

From occipital condyles to first molar

Length of molar series ....

„ three molars ....

Inteiwal between external angles of occipital condyles
Height of occipital crest from foramen magnum
Width of skull below horn-cores

„ between maxillary prominences
Breadth opposite sub-orbital foramina
Least width of nasals ....

Greater diameter of orbit ....

. 17-0
. 13-7
. 5-4

. 3-3

. 4-5

. 4-3

. 3-7 (?) or 6-7 (?)

. 5-8

. 4-1

. 1-4

. 2-8

The above measurements correspond in the main with those of the specimen
figured above ; in the specimen figured in Plate I, {lettered series) fig. 2, of the
“ Eauna Antiqua Sivalensis,” the horn-cores are complete and measure upwards
of twenty-seven inches in length ; this specimen belonged probably to a male.

In the distinctness of the parietal and occipital planes the genus makes an
approach to the antelopes ; the general form and dnection of the horn-cores is also
considerably more Antilopine than Bovine ; the horn-cores are, however, further
removed from the orbit than in the antelopes; the basi-occipital and teeth are
distinctly Bovine, and the lachrymal is not depressed as in the latter group. Some
specimens of crania, probably those of aged males, develope a protuberance of

153—66 CRANIA OR RUMINANTS PROM THE INDIAN TERTIARIES.

the frontals below the bases of the horn-cores ; this gives a somewhat convex profile
to the frontals, the horn-cores, and the lower part of the frontals, sloping away
on either side from the infra-cornual protuberance ; in the figured specimen the
profile is nearly straight.

Both this species and Semibos triquetriceros are probably to he regarded as
links between the true oxen and the Bovoid antelopes like Fortax, Oreas, etc. ;
the latter group is, however, among other characters distinguished from the oxen by
the absence of the accessory column in the molar teeth which the Siwalik genera
possess ; this group of animals is confined to India and Africa, only one species,
however, Fortax picta, now remaining in the former continent; in the Indian
Tertiaries, besides the antelopoid oxen above described, there existed a true antelope,
A . palaindica, which also presented some points of affinity to the African Bovoid
antelopes ; the above facts all tend to confirm the former common origin of the
Etliiopian and Indian faunas ; the more the Tertiary fauna of India is explored,
the more forms do we find to he common to or to represent one another in the
faunas of the two continents.

Table of Ti^ontal Measurements of the Bovidce.

The subjoined table shows the relative proportions of the antero-posterior and
transverse diameters of the frontals of the crania of aU the recent and fossil Indian
bovoid animals. The transverse diameter is taken across the widest part of the
cranium between the orbits, and the antero-posterior diameter is the interval
between a transverse line connecting the centres of the orbits and the intercornual
ridge. The corresponding diameters of the cranium of the European Bos primi-
genius are prefixed for comparison : —

Ant.-post.

Excess of

diam.

trans. diam.

ant.-post. diam.

Inches.

Hos primigenius ......

9-4

10-5

0-9

Bos namadicus J .

9-5

110

Bos namadicus $ .

7-5

8-4

0-9

Bos acutifrons ......

10-7

11-4

0-7

Bos planifrons ......

9-3

100

0-7

Bibos frontalis ......

110

9-5

i-s

...

Bibos gaums ......

100

8-5

1-5

Bison {Poephagus) gmnniens ....

100

6-5

3-5

Bison sivalensis ......

9-3

5-8

3-5

Bubalus platyceros ......

9-2

8-0

Bubalus arni ......

8-5

7-5

Bubalus palccindieus .....

111

9-9

1-2

Semibos triquetriceros

3-5

4'5

Peribos occipitalis ......

7-9

5-0

2-4

Amphibos acuticornis .....

6-4

4-4

2-0

The measurements of the crania of the living species in the above table are
taken from Mr. Hodgson’s paper on the Genera of the Indian oxen in the J oumal
of the Asiatic Society of Bengal (Vol. X, p. 453).

CRANIA OP RUMINANTS PROM THE INDIAN TERTIARIES. 67—154

Fmiily, — ANTIL OFIFJE.

Genus: ANTILOPE.

As stated in the introduction to the present memou% Dr. Palconer determined
several species of antelopes from the Siwaliks, hut only one of these, mz.. Antilope
palceindica, is known by anything more than its manuscript name. Of Antilope
palceindica there are figures of the cranium given in the “ Palaeontological Memoirs”
{Vol. I, Plate XXIII), and in the “Journal of the Asiatic Society of Bengal”
{Vol. IV, Plate XLVIII, figs. 40, 41, and Vol. XII, p. 769, and Jigs. 1, 2, of accom-
panying plate) ', the latter figure is accompanied by a short description from' the
pen of the late Colonel Baker. The characteristic points of the cranium of this
species are the great elongation of the face, the concavity of the profile, the
absence or small development of the infra-orbital sinus, and the small size of the
supra-orbital foramina. The horn-cores are approximated at their bases and directed
upwards, backwards, and outwards in a simple curve, without any twisting; they
are slightly compressed laterally ; the species seems to be allied to the African
Hartebeast.

In addition to the above species, I am now able to describe three new species
from the Siwaliks, founded upon specimens of the horn-cores or crania which are
in the collection of the Indian Museum ; the figured specimens were all obtained
by Mr. Theobald.

Several of the crania of the Siwalik antelopes differ considerably from the
type forms of the genus, and should probably be placed in distinct genera or sub-
genera ; I have thought it best, however, for the present to retain them under the
type genera.

Antilope Sivalensis, n. sp. PI. 25, fs. 1 & 2.

This species is founded on a cranium obtained from the Siwaliks of the Kangra
district ; it corresponds with the two less complete specimens in the collection of
the Asiatic Society of Bengal catalogued by Dr. Palconer (Nos. S. 569, 573).

This cranium is in a somewhat cracked and shattered condition, as is so com-
monly the case with Siwalik fossils ; the two horn-cores are broken off shortly
above their origin from the frontals ; the greater part of the nasals is also broken
away, and the borders of both orbits are crushed ; on either side the whole extent
of the zygomatic arch has been destroyed. The occipital condyles, together with
the greater part of the boundary of the foramen magnum, are likewise absent ;
the premaxillge and the extremities of the maxillae are abruptly broken off ; they
have been restored in outline in the profile view. The crowns of the ultimate molar
teeth are broken off close to their alveoli.

155—68 CEANIA OP EIJMINANTS PEOM THE INDIAN TEETIAEIES.

The cranium is slighter larger than that of the living Indian Antilope
ceroicapra, a figure of the cranium of which species will he found in the “ British
Museum Catalogue of Ungulates,” 1853 {Flate VIII, fig. 3). The horn-cores of
Antilope sivalensis, if, as is almost certainly the case, the cranium (No. S.573) in
the Asiatic Society’s collection belongs to this species, were not spirally twisted ;
otherwise the general form of the crania of A. sivalensis and A. cervicapra is
almost identical, and the two animals were no doubt very closely allied. Most of
the sutures in the present specimen are still visible, and the outline of the com-
ponent bones can therefore be determined with exactness.

The inclination of the basi-cranial axis in the cranium of Antilope sivalensis
forms a rather smaller angle with the plane of the palate than in the recent species ;
and consequently the whole of the cranial portion of the skull is placed more below
the level of the palatal plane. In Antilope cervicapra, if the line of the outer border
of the molar alveolus be produced backwards, it will be found to cut the lower
extremity of the occipital condyles ; in the cranium of Antilope sivalensis a similar
line will cut the occiput immediately above the foramen magnum. In the same
manner the apertm’e of the meatus auditorius externus in the recent species is
placed very slightly below the dental border of the orbit, while in the fossil
species the same aperture is placed very considerably below the corresponding
orbital boundary.

The occiput in both species is identical in form, being bounded above by the very
prominent and overhanging crest or superior curved line, and having a large
tuberosity for the attachment of the ligamentum nuchse ; the supra-occipital is,
however, rather narrower in the fossil species, the indentations of the temporal
fossae extending to a rather larger extent on to the superior surface of the
cerebral chamber. The parieto-frontal sutm’e in both species forms a doubly curved
line across the skull, immediately behind the horn-cores. Externally to the horn-
cores the superior surface of the orbit in the recent species forms a flat surface
placed at right angles to the outer surface of the horn-core ; in the fossil species
the two surfaces slope imperceptibly one into the other, without any angulation
at the junction. As far as can be judged from its broken condition, the orbit of
the fossil species seems to have possessed the same straight dental border which
is characteristic of the recent species.

The concavity which occurs between the orbits, immediately below the horn -cores
in Antilope cervicapra, is wanting in Antilope sivalensis, this surface being plane,
and without any ridge running obliquely from the horn-core to the orbit. The
horn-cores of A. sivalensis rise nearly vertically from the f rentals, with a small
interval between their bases, the supra-orbital foramina pierce the bases of the horn-
cores, and have no large sinus at their mouth. The nasals extend upwards
nearly to the middle of the orbits ; the plane of the frontal is nearly at right angles
to that of the laclmymal ; there is a notch in the anterior border of the orbit at the
commencement of the lachrymo-frontal suture : the lower extremity of the frontal

CRANIA OP RTJAIINANTS PROM THE INDIAN TERTIARIES. 69—156

forms a triangular process fitting in between tlie nasals and frontals ; at tlie junction
of the lachrymo-frontal, naso-frontal, and naso-lachryinal sutures there is a small
elongated fissure representing the larger infra-orbital vacuity of other antelopes.
The facial surface of the lachrymal is much depressed. The maxillse have a
large tuberosity above the second molar. With the exception of the direction of the
horn-cores and the smaller size of the supra- orbital sinuses, the whole of the above
characters exactly correspond with those of the cranium of Antilope cervicapra.

On the inferior aspect the basi-occipital is rather different from that of the
recent species : the posterior tubercles are nearly similar in both ; the anterior pair
of tubercles in the recent species are very large, and nearly three quarters of an
inch in height, and are separated by a long interval from the posterior pair ; in tlie
fossil species the two pairs of tubercles are more closely approximated, and the
anterior pau are but very slightly raised above the surface of the basi-occipital : in
consequence of their small size the whole basi-occipital becomes somewhat triangular
in form. By the approximation of the two pairs of tubercles the tympanic bulla
appears to be placed more anteriorly than in the recent species, being partly in
advance of the anterior border of the first pair of tubercles, whereas in A. cerm-
capra it does not extend farther forward than the middle of these tubercles.

The palate of the fossil species is considerably wider in proportion to its length
than in the recent species ; the embedded crowns of the molar teeth also show a
greater width in proportion to their length, the grinding surface being probably
nearly square instead of oblong. Whether the maxillse presented the peculiar
sickle-hke edges on the palatal surface, which are found in the recent species, we
are unable to say from the present specimen.

The measurements of this cranium are given with those compared below with
those of a cranium of an adult specimen of Antilope cervicapra : —

A. cervicapra. A. sivalensis.

Width at superior border of orbits ....

Inches.

. 4-30

Inches.

5-20

„ inferior „ „ .

. 3-40

4-10

Interval between outer sui'faces of born-cores at base

. 3-40

4-10

„ „ bases of horn-cores

. 1-00

0-90

„ „ superior angles of supra-orbital foramina .

. 1-70

2T0

Direct distance between extremities of inter-frontal suture .

. 2-60

3-50

Interval between middle of fronto-parietal and occipito-parietal sutures

. 2-00

2‘20

„ „ inferior border of foramen magnum and summit of occipital crest . 1'80

2-49

„ „ occipital crest and occipito-parietal suture .

. 0-70

0-90

Greatest width between temporal fossae

. 2-70

3-20

Width of occipital through petrosals ....

. 3-40

3-80

Interval between meatus auditorius and orbit

. 1-70

2T0

Antero-posterior diameter of right orbit

. 1-70

2T0

Transverse „ „ ...

. 1-50

1-70

Interval between inferior border of foramen magnum and palate

. 3-60

4-30

Width of palate between last molars ....

. 1-50

1-80

Interval between outer sides of last molars

. 2'30

3-40

Length of three last molars .....

. 1-60

2-20

„ last molar .....

. 0-65

0-80

Breadth .......

. 0-45

0-65

Circumference of base of right horn-core

. 3-80

4-90

157—70 CRANIA OR RUMINANTS FROM THE INDIAN TERTIARIES.

It wiU be seen from the above that the proportions of the crania of the two
species are very similar, but the differences noticed above are sufficient to distinguish
between the two.

The fossil cranium differs very markedly from that of Antilope palceindiaa of
Falconer ; it is much smaller in size, the horns differ in form, position, and direction,
while the facial and nasal portion are much smaller in proportion to the frontal ; the
profile of A.palceindica is concave, while it is nearly straight in the new species ; the
orbit of the former is entirely behind the molar series, and above it in the latter
species the supra-orbital foramina are very minute, and without a sinus in A. palce-
indica, while in A. swalensis they are large. It is quite probable that the latter
may have been the progenitor of the living Indian antelope, the twisted horns of
the living species being a recently acquired character.

ANTihOPE PATULicoRNis, n. sp. noUs. PI. 25, f. 3.

This the third fossil Indian species of the genus is founded on the imperfect
frontlet figured in Plate XXV, fig. 3, collected by Mr. Theobald in the Siwaliks.
Though in a very imperfect state, the specimen is very distinct from either of
the other two Siwalik antelopes ; indeed it is more than probable that it should
not be referred to the restricted genus Antilope at all, but should rather be placed
in a separate genus ; its affinities, however, are mostly Antilopine, and I have
therefore placed it in the type genus, until more complete specimens enable us to
define its position more exactly.

The specimen comprises the greater portion of the frontal bones with a portion
of the horn-core of the left side ; on the posterior surface the greater part of the
orbital cavities and the anterior extremity of the cerebral cavity are preserved.

The horn-cores are placed at a short distance above the orbits, and have a
considerable interval between their bases ; the frontal suture forms a somewhat
elevated median line, with a longitudinal hollow on either side. At a short dis-
tance above the border of the orbit there is a very large supra-orbital fossa,
triangular in shape ; at the superior angle of this fossa, the frontal is pierced by a
supra-orbital foramen, circular, and of unusually large size ; it passes directly
through the frontal into the cavity of the orbit.

The horn-cores present an ellipsoidal cross-section ; they are longitudinally
marked by parallel ridges and grooves : their general direction is upwards, back-
wards, and outwards, in a single gentle curve ; their anterior surface is convex
and the posterior concav^e ; there is a faint trace of a spiral arrangement of the
grooves on the cores ; the diameter scarcely diminishes at all, in the fragment pre-
served, towards the extremity. The frontals are of considerable thickness, with
finely cancellated structure between the outer and inner tables.

On the posterior aspect the specimen shows the anterior part of the cerebral
cavity, as far down as the commencement of the cribriform plate of the ethmoid ;

CEANIA OE EUMINANTS EEOM THE INDIAN TEETIAEIES. 71—158

this part of the cavity is triangular in shape with a slightly vaulted roof, with well-
marked ‘Hmpressiones digitatce the orbital cavity is smooth and rounded, and
pierced by the supra-orbital foramen. The dimensions are as follows : —

Inches.

Interval between base of horn-core and orbit ...... 0’9

„ „ bases of born-cores ........ 2 0

Thickness of frontal .......... I'O

Antero-posterior diameter of left horn-core ....... 1'8

Transverse „ „ . . . . . . . 1’6

The presence of the large supra-orbital fossa and foramen, the latter passing
directly from the surface of the frontal into the orbit, at once separate the specimen
from the sheep and goats, in which the foramen is small and has a sinuous course,
and there is no fossa present : the horn-cores also taper much less than in any
goat or sheep of the same size.

The concavity of the frontals between the horn-cores and the raised frontal
suture is an Antilopine character ; the horn-cores are placed on the frontals at a
considerable distance apart, but not more so than in Saiga tartarica ; from their
nearly uniform diameter they were probably of considerable length ; their ultimate
direction cannot be determined from this specimen ; if continued outwards in the
same manner as the part remaining, they must have differed considerably from those
of any existing species of antelope ; possibly they were suddenly bent inwards after
a certain distance, as in Alcelaplius.

Antilope poppecticobnis, n. sp. nohis. Plate 25, f. 4.

A large number of frontal bones and horn-cores of a species of antelope
different from any of the above have been collected by Mr. Theobald from the
Siwaliks of the Potwar district ; no complete crania have hitherto been discovered ;
for this species I propose the name of A . porrecticornis.

The specimen figured is the frontal bone, and the greater portion of the horn-
core of the left side ; the right side has been restored in outline. The frontal is
strongly arched, and excavated longitudinally between the prominent suture and
the base of the horn-core. The supra-orbital foramen is enclosed by a large pyri-
form sulcus, its extremity reaching almost up to the base of the horn-core. The
horn-core is placed immediately above the orbit, and is separated by only a short
interval ; the diameter of the horn-core is at first upwards, outwards, and backwards,
and then curves inwards towards its extremity ; the transverse section is oval, the
horn-core being compressed laterally. The measurements of the specimen are as
follows : —

Length of frontal suture

Inches.

2-40

Width of frontal „ , .

1-85

Length of fragment of horn-core

4-60

Antero-posterior diameter of base of horn-core

160

Transverse „ „ „

116

159—72 CRANIA OR RUMINANTS PROM THE INDIAN TERTIARIES.

The compressed horn-core and the large supra-orhital sinus approximate this
species to the Gazelles ; it is, however, distinguished from the living species of that
genus hy the inner border of the horn-core being concave instead of convex, by
the frontal being proportionately shorter, and hy the supra-orhital sinus being wider
and more open at the base.

The species is distinguished from Antilope palceindica hy the horn-cores being
placed wider apart at the base, by their curvature and direction, and by the presence
of the large supra-orhital sinus and foramen.

With the frontlet of Antilope patuUcornis the present specimen presents no
points of resemblance.

Erom Antilope sivalensis the specimen is distinguished by its smaller size, and
by the direction of the horn-cores, which rise vertically from the frontals in that
sj)ecies ; further, the supra-orhital sulcus and foramen are entirely distinct from the
horn-core in the present specimen, while in A. sivalensis, they are situated on the
base of the horn-core ; the sulcus also is much larger, and the frontal shorter in the
present species.

If the complete cranium be ever diseovered, I imagine that it will be found to
correspond in general character with the crania of the Gazelles, among which group
I think the species should probably be placed.

Family,— SIVA THFBID^.

Genus : HYDASPITHERIUM, nov. gen. noUs.

This genus of Sivatheroid animals was originally erroneously named Hydaspi-
dotlierkim the characters of the genus will be gathered from the description of the
cranium on which it is founded. It may not be out of place to state here that
several lower jaws and detached molars have recently been obtained from the Punjab,
which indicate the former existenee of other species.

Hydaspitherium megacephalum, nov. sp. noUs. Pis. 26 & 27.

The magnificent cranium figured in the accompanying plates is perhaps one of
the most interesting of the fossils discovered by Mr. Theobald in the Siwaliks of the
Punjab ; it belongs to that strange group of animals represented in the Tertiary
period of India by Sivatlierium, BramatJierium, and Vislinutherium, a group which
connects the still living and isolated genus Camelopardalis with other families
of the order. The distribution of this group in India, as far as we can at
present judge, is peculiar : Sivatherium seems to have ranged from the Debra Dun
district as far as the eastern side of the Punjab ; in the Western Punjab its remains

1 Rec. Geol. Surv. Ind., Vol. IX, p. 154.

CRANIA OR RUMINANTS PROM THE INDIAN TERTIARIES. 73—160

seem to be unknown : Kydaspithepium has been found only in the Potwar district
of the Punjab, while Bramatlierium is confined to Perim Island, and the smaller
Vislmutherium to Burma ; Camelopardalis has been found throughout the greater
part of the Sub-Himalayan Siwaliks, and it also occurs in Perim Island. Of course
these lines of distribution may by subsequent discoveries have to be modified, but
they seem to point out that the head- quarters of the different genera had each its
separate area, and that it was chiefly stragglers that intruded into the area of
another genus.

Before describing the present specimen, it is necessary to say a few words regard-
ing the allied genus BramatJierium ; this genus was founded by the late Dr. Ealconer
upon the evidence of certain upper molar teeth of a large ruminant from the
mammaliferous beds of Perim Island in the Gulf of Cambay; these teeth were
figm’ed and described in the Quarterly Journal of the Geological Society of London
for 1815 ( Vol. I, p. 356) ; these figures, together with the description, have been
copied in the “ Palseontological Memoirs” of Dr. Ealconer, {Vol. I, p. 389); the
lower molars of the same species have been figured and described in the foregoing
part of the present Memoir.^ These molar teeth are the only remains which
have been described under the name of BramatJierium ; there is, however, a cranium
of a large ruminant from Perim Island, allied to SwatJierium and Camelopardalis
described and figured by its discoverer, Mr, Bettington, in the Journal of the Royal
Asiatic Society for 1845 ( Vol. VIII, p. 340) ; an additional note on the specimen by
Professor Owen will be found in the same volume {p. 417) ; no name was assigned
to the specimen either by its discoverer or by Professor Owen, though the latter
seemed inclined to place it in the genus SivatJierium. At the time of writing
his own Memoir on the teeth of Bramallierium, Dr. Ealconer had not seen
Mr. Bettington’s Memoir, both memoirs being published during the same year ; in
a postscript, however. Dr. Ealconer mentions having seen an abstract of Mr.
Bettington’s Memoir, and suggests that the specimen there described might possibly
belong to the then new genus BramatJierium. Unfortunately, Mr. Bettington has
not given the measurements of the molar teeth of his specimen, so that it is not
possible to compare these teeth exactly with Dr. Ealconer’s specimens ; there is,
however, in Mr. Bettington’s Memoir a half-sized figure of the molars of his speci-
men, and on comparing this figure with the figures of Dr. Ealconer’s specimen, I
find that the two correspond exactly both in size and form. On this ground,
and as both specimens are from the same locality, I have no doubt but that
Mr. Bettington’s specimen is really BramatJierium perimense, and I shall so consider
it in the present Memoir. That cranium is characterised by being provided with
two pafis of horn-cores, the posterior pair of which arose from the sides of the
occipital crest, on distinct bases, and were directed outwards and backwards ; while
the anterior pair took their origin from a common base situated a short distance

‘ Pp. 42—60, Plate 7, fig 13.

161—74 CEANIA OF RUMINANTS FROM THE INDIAN TERTIARIES.

above the orbits ; these horn-cores were directed upwards and outwards ; their com-
mon base was constricted at its origin : the frontals were laterally contracted above
the orbits, and the face presented a slightly convex profile : the molar teeth want the
crenulated folding of the enamel of the inner semi- circle of the islands found in
the teeth of Sivatherium, and have a minute accessory tubercle in the valley dividing
the inner sides of the two barrels.

To return to our own specimen. We find that the cranium itself is almost com-
plete, though the horn-cores have been broken off at their base ; the extremities of
the nasals, maxiUse, and premaxillse are also wanting. The crowns of the first pre-
molars have been broken off at then’ bases.

The cranium is that of a fully adult animal, most of the cranial sutures having
become obliterated, and the molar teeth being about one-third worn down. Like
the crania of the allied genera, the present specimen is remarkable for its general
massiveness and bulk : at the line of fracture the frontals are produced into an
enormous mass of bony structure, which formed a common base for one or more
pairs of horn-cores ; this base measures as much as thirty-four inches in circumfer-
ence ; the profile of the face is very markedly concave, with an obtuse re-entering
angle at the junction of the frontals and nasals ; there are no horns immediately
over the orbits. The orbit, as in the allied genera, is placed very low down on the
skull, far removed from the plane of the face ; its antero-posterior diameter is longer
than its transverse diameter, and its central axis is directed outwards and forwards,
so that its zygomatic border is considerably more prominent than its frontal border ;
a complete bony ring surrounds the orbit, which is small in comparison to the size of
the cranium. The nasals are short and straight, extending as far upwards as the lower
border of the orbit; they seem to have been narrower than those of Sivathermm, and
have no trace of the arching found in those of that genus ; in the specimen the greater
part of the nasals are broken away, but a cast of their inner surface remains ; from
this fracture and from the obliteration of the cranial sutures, it is not easy to determine
the precise relations of the nasals to the surrounding bones ; it seems, however, to be
quite clear that the nasals articulated with the maxillse, though probably not with
the premaxillae. There does not seem to be any resemblance in the form of the nasals
and frontals to those of Saiga, with which Dr. Mmle* has sought to affiliate Siva-
therium, chiefly on account of the form and relations of the nasals : in the present
cranium the anterior nasal orifice is comparatively small, and bears no resemblance
to the enormous orifice of Saiga, neither do the nasals reach to the median line of
the orbits as in the latter genus, but, on the contrary, are completely below the orbits.
The superior surface of the nasals forms an obtuse angle with the frontal plane ; this
angle is larger than in Sivatherium ; the anterior nasal orifice presents an approxi-
mately circular transverse section, the transverse diameter being rather the larger of
the two ; the foramen for the maxillary branch of the fifth nerve is of rather large

‘ Geol. Mag., Yol. VIII, p. 44.

CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES. 75—162

size, and pierces the maxilla immediately above the fii’st premolar : the position of
this foramen on the anterior and not on the lateral surface of the maxilla is like what
occurs in Camelopardalis ; between the nasals and the orbit there is a large vacuity
shaped like an isosceles triangle, of which the apex is directed upwards and outwards,
while the base is nearly horizontal ; there is no depression of the surface of the lachry-
mals below the orbit as so commonly occurs in the antelopes ; in the absence of this
depression the cranium agrees with Camelopardalis ; the exact relations of the lachry-
mal to the surrounding bones cannot be determined, but the former seems to have
been of considerable relative size. Below the orbits the skull gradually contracts in
width down to the molar alveoli ; the maxillae are of great depth from the orbits to
the molars ; this character is more Bovine than Antilopine or Giraffine : the lateral
surfaces of the maxillae are nearly even, with the exception of a large conical tuber-
osity above the penultimate premolar. Above the base of the nasals the frontals
rise nearly vertically upwards, with a smooth facial surface, which gradually widens
above ; immediately over the orbits the whole skull is contracted in width for a
length of several inches ; at the point of fracture of the specimen the skull has again
expanded nearly to its original width : the mass of cancellous bone which formed
the base of the horn-cores does not show any sign of a median fissure, and the horn-
cores probably did not diverge from their common base until a considerable distance
above the plane of the frontals, as in Bramatliermm.

On the lateral smTace of the cranium we find a deep and wide temporal fossa
bounded superiorly by a well-marked ridge which extends backwards from the orbit
along the base of the horn-cores ; a short distance behind the orbit there is a small
tuberosity on this ridge ; the temporal fossa is open from above. The zygomatic-
process of the squamosal is indented near its origin, and then suddenly bends out-
wards to form the large glenoid cavity, the tympanic is compressed and without a
bulla ; the meatus auditorius externus is directed outwards, and closely approxi-
mated to the paroccipital process ; in the above characters this skull agrees with
that of Camelopardalis and differs from the skulls of the Antelopes : the junction
of the jugal with the zygomatic process of the squamosal is more prominent than
any portion of the orbit.

The occiput is remarkable for its great size and flatness, affording a large sur-
face for the attachment of the muscles necessary to support so large a cranium ; the
plane of this surface forms an obtuse angle with the plane of the frontals ; in shape
the occiput is rudely oblong, with its superior border arched ; the transverse diameter
is longer than the vertical diameter ; above the foramen magnum there is a deep
and nearly circular hollow of large size for the attachment of the ligamentum
nuchse ; this hollow is divided by a vertical median ridge : strong ridges directed
upwards and outwards run from the foramen magnum below the ligamental fossa to
the periphery of the occiput. The foramen magnum is of large size, and extends
upwards considerably above the level of the condyles ; its axis is almost perpendi-
cular to the plane of the occiput, and its vertical diameter is the larger of the two :

163—76 CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES.

the occipital condyles are placed high up on the occiput, and their long axes are
nearly horizontal. The paroccipital processes are of enormous width, forming
flat bony plates extending from the foramen magnum to the lateral borders of the
occiput. Above the occipital crest the common base of the liorn-cores rise.s
almost vertically, somewhat after the manner of the intercornual ridge of the oxen.
It is impossible to say how much of this portion of the cranium is formed by the
parietals and how much by the frontals, but I am inclined to think that in the
middle line the parietals formed a very narrow strip as in the true oxen ; there is
no trace of any horn-cores joining the occiput as occurs in the cranium of Brama-
tlierium.

On its palatal aspect the cranium, like that of Sivatlierium, is remarkable for the
great relative width of the portion behind the molar series ; tlie outline of this as-
pect of the skull presents the form of a blunt wedge, instead of the elongated wedge
which usually occurs in this aspect of the. skulls of the other genera of Ruminants.
In consequence of the direction of the foramen magnum, a palatal view of the con-
dyles shows that these are separated only by a very narrow notch, scarcely any part
of the foramen magnum itself being visible from below. The shape of the basi-occi-
pital and sphenoid is that of an elongated wedge ; the former bone bears two pairs
of tubercles ; the posterior pair are very small, closely approximated to the occipital
condyles, and separated from each other by a considerable interval : the anterior
pair are of much larger size ; they are narrow, approximated to each other, confluent
anteriorly, and have their free extremities directed backwards ; between these tuber-
cles and the glenoid cavity on either side is the foramen ovale in its normal posi-
tion ; most of the other foramina are concealed by matrix, though the foramen lace-
rum posterius may be distinguished at the base of the paroccipital process. The
plane of the basi-cranial axis is unbroken, and is almost parallel to the plane of the
palate ; in this important character the skull resembles the skulls of Camelopardalis
and the deer, and differs widely from the skulls of the antelopes, sheep, goats, and
oxen, in which the basi-cranial axis forms a considerable angle with the plane of
the palate. The free margin of the palatines forms in the middle hue a wide semi-
circular notch, which extends as far forwards as the penultimate molar, a character
very unusual among Ruminants ; the palatine foramina are placed on the same line
as the fore barrel of the penultimate molar ; the palatines are more like those of the
Giraffe than any other Ruminant : whether they terminated on the median line in
a point, as in the antelopes and sheep, or whether in a re-entering angle as in the
Giraffe, is not shown in the present specimen. The palate is relatively wide, and the
lines of the molars are nearly straight ; behind the last molars the maxillae and
jugals diverge rapidly outwards : the glenoid cavity is wide, and has a large and
slightly recurved post- glenoid process.

The teeth of this genus have the same general form as those of the allied
genera, and the enamel has the same general rugose character, though the sculp-
turing is rather finer; apart, however, from their smaller size these teeth have

CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES. 77—164

certain distinctive characters of their own. The specimen figured {Plate .27, Jig. 4)
is the ultimate upper molar of the left side ; I have also figm^ed for comparison
the penultimate left upper molar of Swatlierkim giganteum {Jig. 3); both specimens
are of the natural size ; figures of the left upper molars of Bramatherium perimense
will be found in the “ Palaeontological Memoirs ” ( Vol. I, plate XXXIII).

The upper molar teeth of Hydaspitlierium have no waving or crenulation of the
enamel in the central crescent of the crown, by which negative character they
are at once distinguished from the molars of Sivatherium : the central infolds of
enamel in Kydaspitherium are remarkable for them great size and for the depth
of the central portion, so that even in the present much worn state of the specimen,
the two infolds are still connected by a deep channel, and they would not become
completely insulated until the tooth were worn dovm almost to its base ; in Fal-
coner’s figures of the upper molars of Bramatherium, which are in about the same
state of wear as the present specimens, the central enamel folds are completely
insulated. The posterior enamel fold is produced into a peculiar projection at its
postero-internal angle ; the antero-external angle of the inner half of the hind barrel
is produced into the central hollow of the fore barrel ; the fore barrel is the larger
of the two ; the outer sm*faces are placed very obliquely to the antero-posterior
axis of the crown, the anterior costse on the dorsa of the barrels forming the most
prominent points on the outer border of the crown; the median costse are very slight,
which is another point of difference between these teeth and those of Sivatherium.
In the inner valley dividing the two barrels there is a very small tubercle attached
to the hinder barrel only : a similar tubercle occurs in the molars of Bramatherium,
but not in those of Sivatherium: on the anterior surface of the tooth there is a
slight hut well marked cingulum : a similarly placed cingulum occurs in the molars
of Swatherium, hut not in those of Bramatherium. The teeth of Bramatherium
are rather larger than those of the present genus, while those of Sivatherium are
much larger than either. The molars of Kydaspitherium in general form are
very like those of Kelladotherium from the upper miocene of Attica, hut are dis-
tinguished by the presence of the small accessory tubercle in the median internal
valley.

The following table gives the measurements of the specimen : the measure-
ments of the crania of Sivatherium, Bramatherium and Camelopardalis will be
found in the above-quoted paper of Mr. Bettington : —

Length frona foramen magnum to diastema

„ „ „ fractured muzzle

„ „ „ last molar

Length from fractured extremity of nasals to highest point of
„ „ inferior border of orbit to first premolar

„ „ superior „ „

Width at anterior border of orbit
„ posterior „ „

„ constriction above orbits

occipital crest

Inches.

. 16-60
. 17-80
. 9-35

. 18-00
. 6-20
. 10-20
. 10-50
. 12-40
. 8-10

165—78 CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES.

Width across zygomatic arches ....

5, „ tuberosities of maxillae ....

„ in front of first premolar ....

Depth from palate to root of nasals ....

Interval between orbit and fractured summit of frontal (right side)
Whdth of occiput across paroccipital processes
Interval between inferior border of foramen magnum and occipital crest
» i> occipital crest and fractured summit of frontals

Width above occipital crest ....

„ across temporal fossae ....

Interval between external angles of occipital condyles .

Length of right occipital condyle
Vertical diameter of foramen magnum
Transverse „ „ • •

Vertical diameter of left orbit
Transverse „ „ ...

Interval between foramen magnum and posterior basi-occipital tubercles
„ „ posterior basi-occipital tubercles

„ „ foramen magnum and fi-ee border of palate

Width of palate at last molar .

„ „ second premolar

Interval between outer surfaces of last molars
„ „ „ of second premolars

Length of five molars
Length of last molar
Width of hinder barrel of last molar .

„ fore „ „

Length of penultimate molar
Width of fore barrel of penultimate molar
Length of first molar .

IVTdth of fore barrel of first molar
Length of last premolar
Width of „ „ . .

Inches.

13-30

10- 30

5- 00
700

6- 50

11- 40

7- 30
3-80

7- 70
6-30

5- 40

3- 10
2-40

1- 90

2- 60
2-80
2-70
0-60

11-40

4- 90
4-20

8- 00

6- 20
6-35
1-60
1-54
1-70
1-50
1-65
1-42
1-70
1-00
1-50

We maiy now compare the cranium of the new genus with the crania of the
allied genera ; the present cranium is distinguished from that of Bramatherimn by
the absence of the parietal pair of horn-cores, in consequence of which the crest
of the occiput is simple and uninterrupted; instead of the convex profile of
the cranium of Bramatherium the present cranium has a very markedly concave
])rofile. The base of the horn-cores is widely separated from the orbits, and is as wide
as the skull above the orbits ; in Bramatherium, on the other hand, the base of the horn-
cores is approximated to the orbits, and is constricted at its origin from the frontals ;
the occiput of Hydaspitherium is more quadrate and wider in proportion to its
height than the occiput of Bramatherium ; no figure of the basis cranii of the latter
genus is given in Mr. Bettington’s Memoir. The crania of both genera agree in
having the supra-orbital horns arising from a common base — a character which is
found in no other Ruminants. The form of the horn-cores of both the genera is
unknown, though they were probably massive and branched like the posterior horns
of Swatherimn. And here a question arises as to the homology of the horns of

CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES. 79—166

the three allied genera. In his Memoir on Bramatherium, Mr. Bettington considers
the frontal horns of that genus, which arise from a common base, to he homologous
with the -small and conical anterior horns of Sivatherium, wliich arise by a distinct
base close above the orbits ; according to this view the posterior horns of both
genera will also be homologous. The position of the common base of the horns of
Hydaspitherium, which is placed high up on the frontals, and only a short dis-
tance from the occipital crest, seems, however, clearly to show that these horns are
homologous with the posterior horns of Sivatherium ; and I think no one will
doubt but that they are also homologous with the anterior horns of Bramatherium,
which latter will then also be homologous with the posterior horns of Sivatherium.
According to this view, the small anterior horns of Sivatherium have no repre-
sentative in the other genera ; and the posterior horns of Bramatherium are also
unrepresented in the cranium of Sivatherium ; the base of the horns of Hydas-
piotherium is so large that I think it quite probable that higher up it may have
divided into two pairs of horns, which would then be homologous with the two
pairs of horns of Bramatherium ; only in the one case the two pairs take origin
from a common base, and in the other the anterior pair only take origin from
a common base, wliile the posterior pair rise from distinct bases. The supraorbital
horns of Camelopardalis giraffa are probably homologous with the posterior horns
of Sivatherium, while the azygos mesial bony elevation on the frontals of the
former genus is, although not a true horn, perhaps homologous with the anterior
horns of Sivatherium ; the two pairs of horns of Tetraceros quadricornis are res-
pectively homologous with those of Sivatherium ; Helladotherium has no horns.
The accompanying diagram represents these relationships.

Homology of horns of Sivatheridce and Camelopardalis.

Turning now to the skull of Sivatherium, we find very considerable structural
differences between it and the skull of the present genus ; the first striking difference
is the absence of any anterior horns near the orbit in Hydaspithermm ; next we
may notice there is only a very short and shallow contraction above the orbits
in the cranium of Sivatherium, while above this contraction the skull expands

167—80 CRANIA OR RUMINANTS FROM THE INDIAN TERTIARIES.

enormously, the bases of the horn-cores being separated by a long interval from
each other, and closely approximated to the orbits ; in Sydaspitherium, on the other
hand, the skull has a long and deep contraction above the orbit, and is not wider
at the base of the horns, which have a common origin, than at the orbits, while
the base of the horns is separated by a long interval from the orbit ; the larger horns
of SivatJierium were mainly directed outwards, while the direction of those of the
new genus was probably at first nearly vertical ; the nasals of the latter genus were
longer and not arched or pointed like those of Sivatherium. Owing to the position
of the posterior horns, the temporal fossa of Sivatherium forms a comparatively
narrow slit not wider than the orbit, and completely overarched by the base of the
horn ; in the cranium of Sydaspitherium, on the other hand, the temporal fossa
is very wide, and quite open above. On the palatal aspect the cranium of Sivathe-
rium is wider in proportion to its length than is the cranium of Sydaspitherium,
while the space enclosed by the zygomatic arch has its transverse diameter longer
than its antero -posterior diameter, the reverse being the case in Sydaspitherium.
The form of the basi-occipital and sphenoid is nearly the same in both genera,
but the anterior tubercles of the former bone are relatively much larger in the
new genus ; the line of the molars is more curved in Sivatherium ; the occiput
is relatively wider in the latter genus, while the depression for the ligamentum
nuchge approaches nearer to the foramen magnum, and the temporal fossae are
seen as large slits on the occipital surface, while they scarcely show in Sydaspi-
therium.

In the following characters the three alhed genera agree, viz., the presence of
horn-cores on the upper part of the frontal ; the general form of the teeth, which
are coated with a rugose enamel; the great width of the posterior half of the
base of the cranium ; the elongated and wedge-shaped basi-cranial axis ; the pre-
sence of two pairs of tubercles on the basi-occipital, of which the hindmost are
small ; the form and position of the free border of the palatines, and the direction
and position of the orbit ; the approximate parallelism of the basi-cranial axis and
the palate.

In the Memoir quoted above, Dr. Murie lays great stress on the form and
relations of the nasals, as connecting Sivatherium with Saiga ; this, on the face of
it, appears to me a very insignificant character, and is also one which, as stated above,
is not borne out by the present genus, in which the nasals are connected with the
maxiRse ; as showing the slight value of this character, I may refer to a new
Turkestan sheep, Ovis harelini of Severtgoff,’ in which the nasals do not articulate
with either the maxiUse or premaxillse, as in Saiga, and yet no one would think of
classifying the two together: the unimportance of the character in the Bovidce
was also noticed in an earlier portion of the present Memoir. From the form
of its posterior horns Dr. Murie argues with more probability that Sivatherium
was also related to Antilocapra ; he, however, to a great extent ignores its relation-

^ Ann. Mag. Nat. Hist., Ser. 4, vol. xviii. p. 210.

CEANIA OF RUMINANTS FROM THE INDIAN TEETIAEIES. 81—168

ship to Camelopardalis, with which the other genera show very strong points of
affinity. The molar teeth of all the fossil genera agree with those of Camelopard-
alis in general form and in the rugose character of the enamel ; in all the genera the
orbit is situated far below the plane of the frontals, which is a feature peculiar to the
group ; in all there is a vacuity between the lachrymal and the nasals, and there is
no suborbital fossa on the lachrymal : in all the tympanic is compressed and non-
bullate ; while the tympanies of, I believe, all the antelopes are furnished with a
large buUa. In all the four genera the basi-occipital has the same general form,
and the posterior tubercles are similar ; the anterior tubercles are, however, large
in Hydaspitliermm, whereas in Camelopardalis they are nearly obsolete ; this
large size of the anterior tubercles is a character found in most genera of Antelopes,
though not in Tetraceros ; the basi-occipital is, however, more square in the Ante-
lopes, and the anterior tubercles diverge more posteriorly than in HydaspitJierium.
In all the four genera the plane of the basi- cranial axis is single, and is nearly
parallel with the plane of the palate, most nearly so in Camelopardalis and Hydaspi-
therium, and least so in Sivatherium : in all the antelopes the plane of this axis
forms a large angle with that of the palate, and there is generally an angle at the
junction of the basi-occipital and basi-sphenoid : in Tetraceros the plane of the basi-
cranial axis forms a smaller angle with the plane of the palate than in other ante-
lopes ; in the oxen, sheep, and goats, the inclination of. the two planes is the same as
in the antelopes, whereas in the deer the plane of the basi-occipital is nearly parallel
to that of the palate. In Camelopardalis, Bramiatlierium, and Hydaspitherium,
the vertical diameter of the foramen magnum is larger than the transverse; in
Sivatherium the two diameters are nearly equal, while in the antelopes the transverse
diameter is the larger of the two.

The main points by which the fossil genera differ from Camelopardalis are
the following : in the fossil genera the inferior border of the orbit is the most
prominent of the two, while in the living genus the superior border is more promi-
nent than the inferior border ; the nasals articulate with both maxillae and pre-
maxiUae, and are relatively longer in the living genus ; the form of the occiput of
Camelopardalis is also very different, being narrow, with an overhanging crest, with
no distinct fossa for the ligamentum nuchae, and with narrow and distinct paroc-
cipital processes.

In conclusion, I think it may be said that the affinities of the new genus are
much closer to Camelopardalis than to any other living form ; and that, together
with Bramatherium, this new genus forms a link between Camelopardalis and
Sivatherium, the latter having certain points of affinity to the antelopes.

The teeth of Vishnutherium iravadicum^ of which only those of the lower jaw
are known, are of too small a size to have belonged to this genus ; they are
further distinguished by the sculpturing on the enamel being more distinct.

Teeth and Other Remains of Mammalia, p. 37—56, pi. 7, fi>;8 1 2.

169—82 CRANIA OF RUMINANTS FROM THE INDIAN TERTIARIES.

Family^ — CAPBIBjF.

Genus: CAPRA.

Of this division of the Ruminantia, the number of forms hitherto discovered
in the Indian Tertiaries is very small, and individual specimens are also rare. One
species of Capra I have determined from the cast of a cranium, the original of which
is in the British Museum, and which was collected in the Siwaliks of the Markanda
river hy Dr. Falconer ; a second species is founded on a portion of cranium with horn-
cores from Perim Island, now in the Indian Museum ; while a third form, to which
I have not assigned a specific name, is only known to us from several portions of
detached horn-cores, collected hy Mr. Theobald in the Upper Siwaliks of the Potwar
district ; this last form was alhed to, though distinct from, the existing Himalayan
Markhoor, Capra falconeri. The late Mr. Blyth^ announced the discovery of a part
of the cranium and horn-cores of an ibex from the Siwaliks in the following words :
“ I have received a portion of a head and horn-cores of a true ibex, to all appearance
identical with the species {Capra salceen, nobis) which still inhabits the loftiest
Himalayan crags.” I am unable to discover what has become of this specimen, hut
I very much doubt if it is specifically identical with Capra sibirica ((7. sakeen,
Blyth). Mr. Blyth then proceeds to say that the occmTence of this fossil proved
the existence of lofty mountains close to the old Siwalik area ; tliis supposition is
confirmed hy the discovery of the above-mentioned horn-core of a goat allied to
the Markhoor in the Potwar district.

No species of the genus Capra is included hy Dr. Falconer in any of his Lists of
Mammals from the Indian Tertiaries, though the above-mentioned cranium from
the Siwaliks appears to have been referred hy him to this genus. The addition of
this genus to the Fauna of the Indian Pliocene strata, is extremely important, as it
is one of the most modern forms of the modern group of Ruminants ; all the
European fossil species belong to the “ diluvial” or post-tertiary epoch, and its
occurrence in India with Chalicotherium and other extinct forms leading hack to the
Miocene period is only another instance of the strange mixture of genera which we
find in these strata. The species described below from Perim Island seems un-
doubtedly to have been obtained from the same beds in which the jaws and teeth of
Binotherium and Acerotherium were found !

Capra Sivalensis, n. sp. nobis. PL 28, figs. 1 & 2.

Of this species we have several detached horn-cores in the Indian Museum, but
I have figured the cast of a specimen of the cranium now in the British Museum,
which is more complete than any of our own. The specimen figured exhibits tlie
posterior half of the cranium above the orbits, with the horn-cores of both sides.

' Ann. Mag. Nat. Hist., ser. 1, Vol. XI, p. 78.

CRANIA OR RUMINANTS RROM THE INDIAN TERTIARIES. 83—170

The parietals are of considerable length; between the horn-cores and the
occiput they are rounded superiorly and excavated laterally by the temporal fossae ;
the occipital surface forms an obtuse angle with the plane of the parietals, and
its transverse diameter is greater than its vertical diameter ; the basioccipital is
obscured by matrix ; the orbit is situated immediately below the horn-cores, and the
two are separated by a considerable interval. There is a small supra-orbital fora-
men, but no sinus ; the horn- cores are placed very closely together at their base,
and curve upwards and backwards, with an extremely slight outward direction : they
diminish very rapidly in width, and their antero-posterior diameter is very much
greater than their transverse ‘diameter ; their cross-section is somewhat triangular,
presenting sharp ridges at their two internal angles, and a third ridge at the hinder
haR of their outer surfaces. The measurements of the specimen are as follows : —

Length from anterior border of horn-core to occiput
Interval between horn-core and orbit

Inches,

. 4-85
. 2-00

Breadth of cranium at orbits ....

. 420

Width of occiput .....

. 3-25

Height of „

. 2-00'

Length of anterior border of horn-core .

. 7‘20

Transverse diameter of „ . . .

. 1-55

Antero-posterior diameter of „

. 2-75

Interval between bases of horn-cores

. 0-40

The cranium of this species is distinguished from the crania of the small-horned
females of Capra sibirica, and Capra cegagros by the horn-cores being so much
more closely approximated at their bases, by their being very slightly inclined out-
wards, and by their thickness at the base.

Among the living goats the cranium of this species approaches nearest to that
of Hemitragus hylocrinus of the Nilghiris ; the horn-cores of both species have
nearly the same position and direction ; those of the living species, however, are
placed shghtly wider apart at their bases, form a rather more open curve, and have
a nearly flat surface in front : the distance also between the horn-core and the orbit
is somewhat less than in the fossil species, and the parietals are flatter and less
rounded superiorly. It is, however, quite possible that the recent and fossil forms
should be placed in the same genus.

Capua peuimensis, n. sp. nobis. PI. 28, fig. 4.

The specimen on which this species is founded is from the collection of the
Asiatic Society of Bengal ; it was obtained from Perim Island by Captain Eulljames,
and was entered in the Catalogue of the Society’s Collection (No. as the
cranium of a species of antelope by Dr. Ealconer ; to this genus it cannot belong,
as is at once shown by the form of the horn-cores, which have a keel on both
anterior and posterior borders, and a very long antero-posterior diameter, conditions
which are found in no antelopes, and which are characteristic of the goats.

The specimen comprises a portion of the frontals and parietals, with the nearly
complete horn-cores of both sides ; the portions of the frontals remaining are convex

171—84 CRANIA or RUMINANTS EROM THE INDIAN TERTIARIES.

below the horn-cores, and concave between the horn-cores ; the parietals are rounded ;
the orbit is approximated to the base of the horn-core. The bases of the horn-cores
are separated by a considerable interval ; the direction of the horn-cores is upwards,
outwards, and slightly backwards ; they are straight, and decrease in diameter very
gradually from base to summit ; their antero-posterior diameter is much greater than
their transverse diameter ; there is a sharp keel at their anterior and posterior
borders ; the anterior has a very slight twist from within outwards. The dimensions
of the specimen are as follows : —

Interval between base of born-core and orbit
Width of cranium across the orbits
Interval between bases of bom-cores
„ „ summits of „

^ Length of left horn-core
Antero-posterior diameter of base of horn-core
Transverse „ „ „

4-9

1-6

6-5

6-8

2-6

1-8

Erom the cranium of the last species, and from the crania of any of the living
Indian goats, the present specimen is distinguished by the wide interval which
separates the bases of the horn-cores, and by their direction and the marked keel on
their anterior border.

Of the crania of living goats, that which has the nearest approach to the present
specimen is the cranium of the European JEgoceros pyrenaicus ; the horn-cores
of both forms agree in being widely separated at the base and approximated to the
orbit. Those of the European species, however, are distinguished by being longer
and more slender than those of the present specimen. (Eigures of the crania of the
European species will be found in Nouv. Mem. Soc. Hel. 1838.)

Capra, sp. PI. 28, fig. 3.

The specimen in the above figure is a fragment of a horn-core of the left side
of a species of goat ; it was collected by Mr. Theobald in the Siwaliks of the
Potwar district. The specimen has a large antero-posterior diameter, and a small
transverse diameter ; the lateral surfaces are flattened, and the anterior and
posterior borders are sharply keeled ; the horn-core is twisted on its longitudinal
axis, so that the anterior keel curves from within outwards, as in the existing
Markhoor. The dimensions of the specimen are as follows : —

Length of fragment
Anterior-posterior diameter of base
„ „ „ summit

Transverse „ base

Inches.

2-3

1-5

1-4

This horn-core doubtless belonged to a species of goat closely allied to the
Markhoor {Capra falconeri) ; but it is distinguished from the horn-cores of that
species by having a sharp keel on its anterior border, whereas the corresponding
border of the horn-core of the Markhoor is rounded. More complete specimens are
required before we can fully point out the affinities of this specimen.

INDIAN TERTIARY AND POST-TERTIARY YERTEBRATA.

SUPPLEMENT TO CRANIA OF RUMINANTS,

By R. LYDEKKER, b.a.,

GEOLOGICAL SURVEY OF INDIA.

(WITH SUPPLEMENTAL PLATES XXIA AND B, XXIIIA, AND REISSUE OF
PLATES XXI AND XXIV).

The addition of several skulls of Siwalik antilopoid Bovidse to the collection
of the Indian Museum, and the publication of an important memoir on Tertiary
lluminants by Professor Riitimeyer,^ have shown me that in the “ Crania of Rumi-
nants” my identifications of the specimens in the Indian Museum with the genera
Remibos and AmpJiihos of Palconer were generally erroneous, and also that the new
genus Feribos cannot stand. The first part of Professor Riitimeyer’s memoir was
published previously to my own, hut unfortunately had not reached India at the
time of publication of the latter ; the second part was published later than my own
memoir, of which it contains a review, and points out the errors into which I had
fallen. In my memoir I referred to the extreme difficulty of recognizing Pal-
coner’s species of Bovidse merely from the figm?es in the supplemental plates of the
Pauna Antiqua Sivalensis,” without access to the original specimens, and I at one
time thought of discarding Palconer’s names and applying new ones of my own.
The identification of Palconer’s species from the original specimens in the British
Museum by Professor Riitimeyer has now rendered these species distinctly identi-
fiable, and I, therefore, must now alter the names of my wrongly identified species.
I, however, for reasons given below, do not adopt the names assigned by Professor
Riitimeyer. I have reissued descriptions of the plates of the wrongly identified
skulls, with their new names, as well as those under which they were first described.

' “ Die Einder der Tertiar Epoclie, etc.” Abhandlungen der Schweizeriscbea paljintologischen Gesellschaft, Vols. IV
and V. Basel, 1877-78.

173—2

SUPPLEMENT TO CRANIA OE RUMINANTS.

Family — B 0 VTT} /R.

Genus: BOS.

Species : Bos planifrons & Bos acutifrons, Lyd. Plates XII, XIII, XVI,
figs. 2, 4, pp. 109, 112.^

Professor Riitimeyer ^ considers that both these forms are closely related to,
if not varieties of, the European Bos primige^iius, and that they both belong to the
same species, the differences in the form of the skulls and horn-cores being merely in-
dividual or sexual varieties. With regard to the latter hypothesis, I may mention
that we have other imperfect frontlets of these forms in the Indian Museum, showing
that the cranial characters of the figured specimens are constant. Erom the great
differences in the form of the two skulls, already sufficiently pointed out in my
descriptions, I cannot but adhere to my opinion that these two skulls belong to per-
fectly distinct species. I have already pointed out the resemblance of the skull of
B. planifrons to that of B. primigenius ; Bos acutifrons, on the other hand, appears
to me to be so widely different in the form of its occiput, frontals and horn-cores
from B. primigenius^ that I cannot consider them even closely allied, much less as
belonging to the same species. Professor Rutimeyer does not mention that both the
Indian species are certainly not newer than the Pliocene, while the European species
is of Pleistocene age.

Bubalus platyceros, Lyd. Plate XVIII, p. 127.

Synonym : Bubalus sivalensis, Rut.

Professor Rutimeyer® mentions that Bubalus platyceros belongs to the same
species as a cranium in the British Museum, which he has described and figured
under the name of Bubalus sivalensis.^ My own name was first applied to the
species in 1877,^ while Professor Rutimeyer’ s description did not appear till the
following year, after my figm^e and full description were published; my name has,
therefore, the right of priority. Even apart from this, since the name of Bison
sivalensis has been applied to one of the Siwalik Bovidse, it would not be desirable
to have the same specific name for another species which by many naturalists
would be included in the same genus.

* The references here given to the “ Crania of Ruminants” refer to the volume paging.
^ loc. cit., p. 184.

® Ibid, p. 186.

^ Ibid. PI. II, figs. 4, 5, p. 138.

‘ Rec. Geol. Surv. India, Vol. X, p. 31.

SUPPLEMENT TO CRANIA OE RUMINANTS.

3—174

[ Genus: HEMIBOS, Falconer.

(Including Feribos and AmpMbos).

Plates XX to XXIV.

In liis above-quoted memoir on Tertiary Ruminants, Professor Riitimeyer has
pointed out^ that my identifications of Semibos triquetriceros wndL^Amphibos acuti-
cornis of Falconer are wrong, the skull which I have figured as belonging to the former
species in Plates XXII and XXIII really belonging to the latter, and one of the
skulls which I have referred to the latter (Plate XXIV) really belonging to the
former. Professor Riitimeyer has also pointed out that the skull to which the specific
name of occipitalis seems to have been applied by Falconer, and which I have figured
under the name of Feribos occipitalis (Plates XX, XXI fig. 2) is also an aberrant
form of Kemibos triquetriceros of F^alconer. Additional skulls lately acquired by the
Indian Museum, had in a great measure convinced me of my erroneous identifications
previously to seeing Professor Rlitimeyer’s memoir. The new material at hand,
with Professor Rutimeyer’s memoir, appears to me to warrant the union of the genera
Hemibos dinA. Ampliibos (and of course Feribosy. Under the former names I formerly
took the form of the horn-cores as affording grounds for these generic distinctions,
but now find that these characters are insufficient.

Species I : Hemibos occipitalis, Falc. sp. Plates XX, XXI, fig. 2, XXI A, XXIV.

Synonyms : Hemibos triquetriceros. Falconer.

Leptohos trequetricornis. Pale. Mss.

Bos occipitalis, Falc. Mss.

Prohuhalus triquetricornis, Riitimeyer.

Feribos occipitalis, Lydekker (olim).

Ampliibos acuticornis, Lyd. (olim).

The skull figured in Plate XX ® of this volume is so different from the skulls
of Semibos triquetriceros, figured by Falconer, that with the material formerly at
hand I was quite unable to identify it with that species, and accordingly adopted the
specific name of occipitalis apparently proposed for the figured specimen by Falconer,
with the new affix of Feribos. In the supplemental Plate XXIA I have had drawn
another skull lately presented to the Indian Museum by the Roorkee (Rurki) Museum,
which evidently belongs to the same species as the skull drawn in Plate XX. The

' loc. cit, pp. 186-187.

2 From a manuscript of Falconer’s lately shown me by Mr. Davies in the British Museum (referred to at page 136 of
Professor Eiitimeyer’s memoir), it appears that Falconer determined finally to unite the genera Hemibos and Ampliibos
under the common name of Leptobos, as we find in that MSS. the names L. triquetricornis, L. acuticornis,
L. antilopinus, and L. elatus.

^ I have had a reissue of the description of this and other plates issued, in which the specimens bear their new
names, as well as those under which they are described in the text.

175—4

SUPPLEMENT TO CRANIA OE RUMINANTS.

new skull, however, has the horn-cores less pyriform in cross-section than those of the
other s]3ecimen, and with a less forward inclination. The form of the facial part of this
skull also shows that it is identical with Hemihos triquetriceros of Ealconer (E.A.S.
Plate H., figs. 1, 3), and also that the skull which I have figured in Plate XXIV as
AmpMhos belongs to the same sj)ecies. The horn-cores of the Roorkee skull,
and of the specimen figured in Plate XX, are, however, very different from those of S.
triquetriceros as figured hy Ealconer, one of the forms of which (E.A.S. , Plate H., fig. 2)
has straight horn-cores with a most distinctly triangular cross-section, while in
others the section is suh-circular (fig. 1).

In his memoir, Professor Ptlitimeyer has figured the upper part of a skull with
horns (Plate 1, figs. 8, 4) which he names tProhuhalus triquetricornis, being the equi-
valent to Uemihos triquetriceros of Ealconer, hut calls it an abnormal or “trocho-
ceros” ^ form of the species. This skull is similar to the skulls figured in Plates
XX and XXI A of this volume except that the horn-cores are less closely approxi-
mated to the occiput than in the first specimen.

Of what he considers the normal male form of the species. Professor Riitimeyer
has figured a skull (Plate VII, figs. 1, 2), wanting the greater part of the horn-cores ;
in that skull the horn-cores, instead of curving forwards, as in the trochoceros form,
appear, as far as can he judged from the portion remaining, to have been straight,
and to slope backwards from the plane of the forehead : the latter has a high frontal
ridge, and the whole aspect of the skull is extremely huhaline : the cross-section of
the horn-cores is suh-triangular.

Tavo other specimens of the skull of the normal male form are figured in figs. 1
and 3 of Plate H of the “ Eauna Antiqua Sivalensis,” the latter being copied in Pro-
fessor Rlitimeyer’s memoir (Plate IV, fig. 1). The horn-cores in the former of these
specimens aj)pear to present a suh-circular cross-section. The skull represented in
Plate XXIV of this volume appears to belong to the normal form of the male.

In the skull drawn in fig. 2 of Plate H of the “ Eauna Antiqua Sivalensis,’’
the horn-cores have a most distinct triangular cross-section (noticed in the descrip-
tion of the plate as being characteristic), and a marked angle on their frontal aspect.
The horn-cores of this specimen are extremely different from those of the trocho-
ceros-form (Plates XX, XXIA), in which the cross-section is distinctly pyriform,
and the anterior aspect presents no trace of an angle. Erom this great difference,
I had at first, till I saw intermediate specimens, no idea that the skull which I
called JPerihos occipitalis could he Uemihos triquetriceros ; the latter sj)ecific name
being utterly inapplicable to it, except on the lucus a nan lucendo principle. It
indeed appears that the distinctly triangular form of the horn-cores of this species
is an exceptional, in place of a normal character, there being only the one skull of
Ealconer’s with this character well marked : further the horn-cores of AmpMhos
acuticornis of Ealconer are very frequently most distinctly triangular (the cause of
my errors of identification), and I have therefore come to the conclusion that the

' Curved-horned.

SUPPLEMENT TO CRANIA OF RUMINANTS.

5—176

name triquetriceros^ is extremely misleading and totally inapplicable to the species : I
therefore propose to aj^ply to the sjDecies the name of ‘^occipitalis,'' originally
applied by Falconer to the trocboceros form (Plate XX), and wliicli in a wider
sense may apply to all forms of the species.

Professor Rtitimeyer has changed the generic name from Semihos to Prohu-
halus, of which I do not see the advantage ; I propose accordingly that the species
be known as PLemihos occipitalis, my generic term Pcrihos being dropped.

Professor Rtitimeyer has pointed ont the great resemblance of the normal form
of this species to the buffaloes, the species being doubtless a connecting link between
that group and the antelopes. The same author also includes in his genus the
living Anoa depressicornis of Celebes under the name of Prohiibaliis celebensis ; the
resemblance of the normal skulls of Pemihos {Prohuhalus) occipitalis with the skull
of Anoa is certainly very close ; in the earlier part of this work I pointed out the
resemblance of Anoa to Amphihos acuticornis of Falconer, which I now include under
the present genus. If Professor Rtitimeyer is right in this generic identification, it
appears to me that it would have been simpler to have included both the recent
and fossil forms under the genus Anoa, and not to have given the living representa-
tive a new generic and specific name.

In Plate 1, fig. 1, of his memoir. Professor Rtitimeyer has copied from fig. 4
of Plate H of the “Fauna Antiqua Sivalensis ” another skull of Memibos occipitalis,
which he regards as the female form, and which is distinguished by its small cylin-
drical horn-cores.

Professor Riitimeyer has also figured and described^ a hornless form of the
species, from a skull in the British Museum.

In the text the description of the skull figured in Plate XX given on page 141
must be taken as that of the trochoceros male form of H. occipitalis, while that of
the skull figured in Plate XXIV, and described on page 150, must be taken as that
of the normal male of the same species.

Species 2 : Hemibos aceticoenis, Falc. sp. Plates XXI, fig. 1, XXI B, XXII,

XXIII, XXIIIA.

Synonyms: Amphihos acuticornis, Falc. and Eiit.

Leptohos acuticornis, Falc. MSS.

Hemibos triquetriceros, Lyd. (olim).

The cranium figured in Plates XXII and XXIII of this volume and described
under the name of Hemibos triquetriceros, has straight horn-cores, with a most
remarkable triangular cross-section, which are set upon a high frontal ridge. This

iThe word triquetriceros is a barbarism {triqueter and Kepas), and Professor Riitimeyer has accordingly altered
it to triquetricornis. It was written triquetriceras by Falconer, wliicb I altered to ceros, like Rhinoceros ;
Mr. Blanford in the “Manual of Indian Geology” has altered the word to triquetricerve. In the above-quoted
MSS. in the British Museum Falconer has altered the name to triquetricornis.
loc. cit., p. 132, Plate II, figs. I to 3.

177—6

SUPPLEMENT TO CRANIA OE RUMINANTS.

cranium was thus named because of the remarkable resemblance of its horn-cores
to the skull figured under the name of R. triquetriceros in fig. 2 of Plate H of the
“ Eauna Antiqua Sivalensis,” in describing which the notes of Dr. Ealconer/ mention
“ the remarkable triangular form of the horn-core,” and also because in describing
the frontlet of Ampliibos acuticorms,^ it is stated that “the horns are rounded on
their anterior surface and flattened behind.” On these grounds I thought that
the specimen in question must belong to Remihos triquetriceros of Ealconer, and
the specimen figured in Plate XXIV with horns rounded in front, to Ampliibos acuti-
cornis. As, however, the latter specimen is now shown to belong to Remibos occi-
pitalis [triquetriceros), in which the horn-cores are not generally triangular in
section, the skuU drawn in Plate XXII must either belong to Ampliibos acuticornis
of Ealconer or to a new species. None of the skulls figured under the latter
name by Ealconer® have a distinctly triangular cross-section, but in a specimen
figured by Professor Riitimeyer,^ the horn-cores are perfectly triangular in cross-
section, and the skull (of which a side view is given) evidently belongs to the same
species as my specimen (Plate XXII). Another skull recently acquired by the
Indian Museum (Plate XXIIIA) has the anterior angle less distinctly marked
than in my first specimen, and is in this respect intermediate between the latter
and Ealconer’s large male skull the section of the horn-core in the new skull is,
however, most markedly triangular. The horn-cores in the skulls figured by myself
are also directed more upwards than are those of Ealconer’s type specimen, but
these dilferences, with the help of intermediate forms, do not appear to be more
than individual varieties, and I therefore come to the conclusion that the specimens
referred to belong to Ampliibos acuticornis of Ealconer.

Since we have found that both in Remibos occipitalis^ and in Ampliibos acuti-
cornis of Falconer some forms have horn-cores with a triangular cross-section,
and since the two are in other respects most closely allied, I now come to the con-
clusion that they cannot be generically or sub-generically distinguished, and I there-
fore propose to call the latter species Remibos acuticornis.

The triangular-horned variety of Remibos acuticornis (Plates XXII, XXIII,
XXIII A) is distinguished from the triangular-horned variety of R. occipitalis by
its somewhat narrower forehead and less telescopic orbits, as w^ell as by the more
upright direction of the horn-cores : it presents no close resemblance to the
trochoceros form of the latter species.

All the skulls of R. acuticornis hitherto referred to, are evidently those of
males, the specimen figured in Plate XXII being evidently that of an old individual
with the sutures obliterated. Another skull has, however, been figured by Rfiti-
meyeU and Ealconer and Cautley,® which is referred by the former to the female ;

’ Description of Plates of F. A. S.
® F. A, S., Plate I. (lettered series).
5 F. A. S., Plate I, fig. 2.
f loc. cit., Plate III, figs. 4 & 5.

^ F. A. S., Plate I, (lettered series), fig. 2.
^ loc. cit., Plate III, fig. 7.

« F. A. S., Plate H, fig. 2.

8 F. A. S., Plate I, figs. 3 to 6,

SUPPLEMENT TO CRANIA OE RUMINANTS.

7—178

the horn-cores are smaU and suh-cylindrical, the orbits not prominent, and there is
not the high frontal ridge so characteristic of the skull of the male (Plate XXIII).
The cranium, of which a profile view is given in Plate XXI, fig. 1, and a front
view in Plate XXIB of this volume, agrees very closely with the above-men-
tioned skull, and belongs probably to a young female of the same species. The
description of the skull given on pages 145 to 148 of this volume must be taken
as that of the male skull of this species ; the female skull drawn in Plate XXI,
fig. 1, and Plate XXIB is not fully described.

Species III : Hemibos antilopinus. Pale, sp, MSS.

Synonyms : Vrohuhalus antelopinus, Riit.

Amfliibos antilopinus, Pale. MSS.

Leptobos antilopinus, Pale. MSS.

On page 89 of this volume, I mentioned that I had no means of identifying
Ampliihos antilopmus of Palconer, and therefore abandoned the name. Among the
Siwalik fossils in the British Museum, Professor Rlitimeyer has, however, found a
skull of a bovoid ruminant, which he says ^ seems to be that of Ampliihos or {Lepto-
bos) antilopinus ^ of Palconer, and which he has figured in liis memoir ® under the
name of Lrohuhalus antelopinus. Professor Rlitimeyer also mentions that he is not
certain whether the skull be that of a male or female individual. The horn-cores are
subpyriform in cross-section, thereby approaching to those of the trochoceros male of
Remihos occipitalis, and the orbits prominent, which is also a character of the latter
species as opposed to R. acuticornis. It is worthy of note, as indicating the generic
identity of all these animals, that whereas R. 'antilopinus was placed at first by
Palconer in his genus Ampliihos, it is placed by Rlitimeyer in Remihos {Prohu-
halus).

Genus: LEPTOBOS, Pale. MSS.

Species : Leptobos palconebi, Rlitimeyer.

Synonyms ? Ampliihos elatus. Pale. MSS.

„ ? Leptobos elatus. Pale. MSS.

Professor Riitimeyer has described and figured^ several skulls of another
bovoid ruminant in the Siwalik collection of the British Museum, under the name
of Leptobos falconeri, mentioning that it is very possible that these skulls belong to
Ampliihos {Leptobos) elatus of Palconer. Three of these skulls are drawn without
any name, in the unfinished Plate S of the supplement to the “ Pauna Anti qua Siva-

* Loc. cit., p. 136.

* The name is spelt antelopinus by Professor Riitimeyer.

® Plate I, figs. 5, 6.

^ loc. cit., p. 157, Plate I, figs. 7, 8 ; Plate IV, figs. 3, 4, 5, 6 ; Plate VI, fig. 9.

179—8

SUPPLEMENT TO CKANIA OE RUMINANTS.

lensis.” Of this species- there was a horned and a hornless form, the latter, in my
opinion, being probably the female. The horn-cores are sub-cylindrical, and are
placed very far apart at their bases ; the frontals are large and concave, and there is
a distinct triangular parietal on either side. The general shape of the skull, when
viewed from above, is sub-triangular, owing to the great expansion of the temporal
regions, and consequently quite different from that of other Bovidce.

Although it is probable that Falconer intended to class this species finally in
the same genus with the three preceding forms, it appears to me that Professor
Rutimeyer is quite right in placing it in a distinct genus or sub-genus.

Species 2 ; Leptobos (Bibos ?) frazeiii. Rut.

This species is named by Professor Rutimeyer ^ on the evidence of a single
imperfect hornless cranium of a bovoid ruminant, said to have been obtained by the
British Museum from a collection made by Captain Frazer in the Narbada
valley. There are two small crania from the same district in the Indian Museum,
having a very close resemblance to the cranium figured by Professor Rutimeyer ;
these crania have small horn- cores.

Family— 8IV A THEBID^.

Genus: BRAMATHERIUM, Falconer.

Species : Beamatherium perimense. Falconer.

On page 160 of this volume I referred to a cranium of a Sivatheroid, which
I considered to be that of Bramatlierium perimense. During a recent visit to
London I found that the specimen in question is now in the collection of the
Royal College of Surgeons, where it bears the above name.

Genus: HYDASPITHERIUM, Lyd.

Species : Hydaspitherium grande, and H. leptognathus, Lyd.

Since the publication of the “Crania of Ruminants,” some jaws of Eyclaspi-
therium have been noticed by myself ^ ; one of these, BL. grande, is a stout- jawed,
and the other, E. leptognathus a slender-jawed form.

> loc. cit., p. 165, Plate Vll, figs. 11, 12.
2 Eec. Geol. Surv. India, Vol. XI, p. 90.

SUPPLEMENT TO CEANIA OE RUMINANTS.

9-180

’Family^ — C, A VRTT) M.

Genus: BUCAPRA, Riit.

Species : Bucapra daviesii, Riit.

Professor Riitimeyer^ has described under the above name a very peculiar
skull in the Siwalik collection of the British Museum. The skull is hornless and
remarkable for the shortness of the face.

List oe Indian Tertiary Ruminants.

The above corrections and additions render it necessary to re-write the list of
Indian Ruminants given on page 92 of this volume.

PECOEA.

Botid^

ANTILOPIDa;

r 1. Bos namadieus, F. and C.

2. Bos planifrons, Lyd.

3. Bos acutifrons, Lyd.

4. Bos platyrhinus, Lyd.

5. Bubalus platyceros, Lyd. .

( B. sivalensis, Eiit.)

6. Buialus palaindicus, F. and C.

7. Bison sivalensis, Falc. and Lyd. .

8. Semihos occipitalis, Falc., sp,

Hemihos triquetriceros, Falc.
Frohuhalus triquetricornis, Riit.
Leptohos triquetricornis, Falc. MSS.
Bos occipitalis, Falc. MSS.

Beribos occipitalis, Lyd. (olim).
Amphihos acwfocorais, Lyd. (olim).

Narbada.

Siwalik.

Ditto.

Narbada and Upper
Siwalik.

Siwalik.

Ditto.

9. Semihos acuticornis, Falc., sp.

. Ditto.

Amphihos acuticornis, Falc. and Eiit.
Leptohos acuticornis, Falc. MSS.
Semihos triquetriceros, Lyd. (olim).

10. Semihos antilopinus, Falc., sp.

. Ditto.

Amphihos antilopinus, Falc. MSS.
Leptohos antilopinus, Falc. MSS.
Frohuhalus antelopinus, Eiit.

11. Leptohos falconeri, Eiit. . . .

. Ditto.

? Leptohos elatus, Falc. MSS.

? Amphihos elatus, Falc. MSS.

12. Leptohos frazeri, Eiit.

. Narbada.

Bibos (?) frazeri, Eiit.

13. Antilope palceindica,^. OiTodiG.

. Siwalik.

14. Antilope patulicornis, Lyd. . .

. Ditto.

15. Antilope sivalensis, Lyd. .

. Ditto.

16. Antilope porrecticornis, Lyd.

. Ditto.

17. Fortax, sp.,^ Lyd.

. Ditto.

18. Fortax namadieus (Eut.^)

. Narbada.

loc. cit., p. 105, Plate II, figs.
See Preface to volume,

181-10

SUPPLEMENT TO CRANIA OF RUMINANTS,

Sivathkbidj^: .

Camelopaedalid^

C'APRIDiE

OVID^

Cekvid^

Camelid^

f 19. Sivatherium giganteum, P. and C.

. . Siwalik.

1 20. Vishnutherium iravadicmn, Lyd.

. .' Burma.

j 21. Bramatherium perimense, Pale. .

. . Siwalik.

22. Sydaspitherium megacephalum, Lyd.

. Ditto.

23. Sydaspitherium grande, Lyd.

. . Ditto.

1^ 24. Sydaspitherium leptognathus, L3'd.

. . Ditto.

25. Camelopardalis sivalensis. Pale, and Lyd.

. . Ditto.

f 26. Ca^ra sivalensis, Lyd.

. ! 27. Capra ^erimensis, Lj'd.
I 28. Capra, sp., Lyd. .

29. Bucapra daviesii, Eiit.

Ditto.

Perim.

Siwalik.

Ditto.

. 30. ? Ovis, sp., Blyth .

. Ditto.

r 31. Cervustriplidens, Lyd.
j 32. Cei'vii,s simplicidens, Ljd.

^ 33. Cervus, sp.

I 34. Borcatherium minus, Lyd.
35. Borcatherium majus, Lyd. .

. Ditto.

. Narbada.
. Siwalik.

. Ditto.

TYLOPODA.

. 36. Camelus sivalensis, P. and C.

. Siwalik.

For other unnamed species, see Preface to volume.

INDIAN TERTIARY AND POST-TERTIARY YERTEBRATA.

SIWALIK AND NARBADA PROBOSCIDIA,

By R. LTDEKKER, B.A.,

GEOLOGICAL SURVEY OF INDIA.

[WITH PLATES XXIX to XLVI].

^Order-PROBOSCIDIA.

Family— DINO TKFFIFFE.

Proboscidia, in which the whole of the permanent dentition is in nse at the same
time, and in which the second true molar has one ridge less than the preceding
tooth.

Dentition and aeeinities of the Dinotherida:.

The molar dentition of Jjinotherium, the only genus of the family, in its mode
of succession, follows the normal mammalian plan ; in the permanent dentition there
are the full number of these teeth, while the premolars are reduced to two (penul-
timate and last), which have succeeded more complex milk-molars. Both premolars
are two-ridged teeth like those of the Tapir ; the penultimate milk-molar has two
ridges, and the last three : the first true molar has three ridges, and the second and
third have only two. In the equal number of ridges in the last milk-molar, and
the first true molar, the genus agrees with the Mastodons, from which it is distin-
guished by the two last true molars having a smaller, in place of a larger, number of
ridges, than the first.

In the one species in which the complete skull and lower jaw are known (D.
giganteum), there are no incisors or canines in the upper jaw, but a single pair of
large recurved incisors in the mandible. It is probable, from the form of the jaw,
that similar incisors were developed in two Indian species, {D. indicum, D. pen-
tapotamice), but not in a third {D. sindiense).

In the form of its cranium JDinotherium makes a curious approach to that of
some of the Sirenia, especially Salitherium, but has also elephantine characters.
The last milk-molar and the first true molar are like those of the Trilophodons,
while the other true molars are like those of Listriodon (Ungulata) : the premo-
lars are like those of Tapirus. There is also some resemblance between the molars

183—2

SIWALIK AND NARBADA PROBOSCIDIA.

of Dinotherimi and those of TJintatJierium. Marsupial bones are said to have been
developed.^ The genus may he considered as a generalized form connecting the
Ungulata and Sirenia Tpitli the Prohoscidia.

Genus : DINOTHERIUM, Kaup.

Molars with simple ridges and open valleys, and with not more than three
ridges in any one tooth.

Species I : Dinotheeium pentapotami^, Palconer, et noUs.

Pis. XXIX, XXX, and XXXI, fig. 3.

JS-istory. — In a previous fasciculus of the present volume of the “ Palseonto-
logia Indica,” ^ I have already figured and described certain molar teeth of a
species of Indian Dinotlierium under the name of D. penta;potamice^ and in the
same notice I have given the history of the species. At the time of publica-
tion of that notice, the teeth figured and described were the only specimens
belonging to D. 'pentajpotamicB, contained in the collection of the Indian Museum.
Subsequent collections, made by Messrs. W. T. Blanford and Pedden in Sind, and
by Mr. Theobald in the Punjab, have, however, greatly enriched the series of teeth
of that species in the Indian Museum, and have rendered it necessary to add a
second notice to my previous descriptions. The most important of these new speci-
mens, together with two specimens of fragments of the mandible, which were trans-
ferred to the Indian Museum from the collection of the Asiatic Society of Bengal,
are now figured and described in Plates XXIX to XXXI of this memoir. Of the
two last mentioned specimens a short notice by myself has already appeared in the
“ Records of the Geological Survey of India.”^

Mandible. — The specimen represented in fig. 1 of PI. XXIX is one of the two
specimens mentioned above, as having been transferred from the collection of the
Asiatic Society of Bengal. It was found stowed away in an old box with a few other
mammalian fossils, without any trace of label or ticket. That it came from the
Siwaliks is, however, perfectly evident from the circumstance that the specimen is
still covered in many places with a coating of the very characteristic grey Siwalik
sandstone.

The fragment carries two molar teeth, the hindmost and least-worn of which

’ Amer. Journ. Science and Art. Ser. II, vol. XXXVIII, p. 427.

^ “ Molar Teeth and other Remains of Mammalia,” p. 72, (volume paging) PI. IX, — figs. 1 — 5.

* It ma}', perhaps, be not out of place to give the etymology of the ievm pentapotamice, which might not be under-
stood by the non-Indian reader. “ Pentapotamia ” is Falconer’s Greek translation of the Punjab ; the Indian word, as
is well known, means the “ country of the five rivers” {Punj or Panch, five, and ah, water or river) ; Pentapotamia
is formed, like Mesopotamia, from ttIvts (five) and TroVajaof (river).

^ Vol. X, p. 33.

SIWALIK AND NARBADA PROBOSCIDIA.

3—184

is nearly perfect, while the other has been considerably chipped. The worn summits
of the transverse ridges of the hinder molar (right side of figure) are lowest on
the lower border of the figure ; this border will consequently be the outer
side of the specimen, which will, therefore, belong to the left ramus of the
mandible.

Since the hinder molar exhibits no disc of pressure on its posterior surface, and
since both teeth carry two main ridges only, it is evident that these two teeth are
the two last of the permanent molar series, and that consequently the individual to
which they belonged was fully adult at the time of its death. The anterior ridge
of the hinder molar, which, as being the more perfect of the two teeth, is here
selected for description, is concave anteriorly and nearly straight posteriorly ; the
plane of wear of both this and of the second ridge slopes towards the hinder
extremity of the jaw ; the second or hinder ridge is concave on the inner half of its
anterior sm’face, and convex on the outer half ; its posterior surface is nearly straight,
with slight undulations. In the valley between the two main transverse ridges
there is a projecting spur given off from the hinder ridge, passing obliquely
inwards and forwards. In a later state of attrition than the specimen exhibits, as
in the anterior molar (left side of figure), this intermediate longitudinal spur would
display a worn dentine surface, which would ultimately form a narrow bridge of
dentine connecting the dentine surfaces of the two transverse ridges. There is a small
tubercle on the outer side of the main transverse valley. At the posterior extremity
of each tooth there is a large talon-ridge, broader at its outer than at its inner extrem-
ity, and about half the width of the anterior main ridge ; there is a slight trace of a
cingulum on the outer surface of this talon, as is well seen in the anterior tooth ;
eventually the talon-ridge becomes worn down, and forms a part of the grinding
surface of the tooth. Each tooth is implanted in the jaw by three distinct fangs ;
two small ones at the anterior extremity, and a larger and laterally compressed one
at the posterior extremity.

The Indian Museum has obtained recently, through exchange with the Lahore
Museum, two last lower molars of the same species of Dmotheriimi. These two
teeth agree precisely in every detail with the figured specimen. They a23pear to
have been obtained from the Siwaliks of Shaik Budm, on the north-west frontier of
the Punjab. A similar specimen has also been presented to our collection by the
Roorkee (Rurki) Museum, which seems to have been obtained near Roorkee itself —
very possibly low down in the Siwalik series, as it carries fragments of a red matrix
like that of theNahan (lower Siwalik) rocks.

The second specimen which I have figured (Plate XXIX, figs. 2 and 3) appears
to have come from Siwalik beds at Kushalghar, near Attock, where it was obtained
many years ago by the late Lieutenant Garnet ; these beds are probably somewhat
older than the general mammaliferous zone of the Siwaliks and the representative
of the Nahan rocks. As I have already observed in the above-quoted passage in tlie
“ Records,” the figure of this specimen given by Dr. Murchison in the “ Palseonto-

185—4

SIWALIK AND NARBADA PROBOSCIDIA.

logical Memoirs,”^ under the name of Antoletlierium is entirely incorrect. It
appears probable that Dr. Dalconer never saw the original specimen, but made bis
new genus Antoletlierium solely on the evidence of the imperfect drawing from which
the figure was taken ; a copy (or the original) of this drawing was sent by Dr.
Falconer to Professor Owen, whose note on the subject will be found on page 416 of
the first volume of the “ Palseontological Memoirs.”

This sj)ecbnen contains portions of three molar teeth, all more or less damaged,
Imt of which the central one is the most complete; the tooth on the left hand carries
two transverse ridges, connected by a median antero-posterior bridge, and is fairly
well rejoresented in the figure in the “ Palaeontological Memons.” The central
tooth carries three ridges, of which the one on the right side of the figure is the
narrowest, and is separated from the adjacent ridge by a transverse valley, which is
narrower than the valley between the two ridges on the left side of the figure ; antero-
posterior ridges connect the three transverse ridges. In Dr. Falconer’s engraved figure
this central tooth (B) is erroneously represented, as having only two transverse ridges,
its third ridge having been added to the broken tooth on the right side of the figure
(C in Falconer’s figure). The third tooth (on the right side of the figure), has a
higher and larger crown than either of the others ; unfortunately only a portion of
its first ridge remains : the worn surface of this ridge slopes towards the right side
of the figure, and the dentine surface is largest on the lower border of the figure.

From the evidence of the incomplete drawing sent by Dr. Falconer, Professor
Owen remarks “ the tooth B {central tooth) is more worn than A {left tooth in
figure) and A than C {right tooth in figure). B may therefore be molar 1, and C
premolar 4, or the last premolar, A being the second true molar.”

The relative conditions of wear of the three teeth are precisely as stated by
Professor Owen, and the central three-ridged tooth is undoubtedly the first true
molar : the position of the other two teeth in the molar series was, however, wrongly
determined by Professor Owen. If we refer to De Blainville’s or Kaup’s ® figures of
the lower dentition of Dinotherium, we shall find, on looking at the first molar, and
the two adjacent teeth, that the first molar is the most worn, the last premolar the
next most worn, while the second true molar is the least worn of the three. In our
specimen, therefore, the tooth on the left {A of Trofessor Owen) is the last premolar ;
the central tooth, as we have said, is the first true molar, while the broken tooth on
the right {C of Professor Owen), is the second true molar. That this is really the
case, irrespective of the state of wear of the teeth, is proved by the worn smdace
of the ridge of the right-hand tooth, sloping towards the right-hand of the figure,
since in all species of Dinotherium the worn surface slopes towards the hinder
extremity of the jaw. Again, the hindmost ridge and valley of the first lower molar

1 Vol. I, Plate XXXIV, figs. 1-2.

2 Pal. Mem. loc. cit.

Blainville: “ OsteograpLie des Mammiferes,” Atlas, Dinoihernim, PI. Ill : — Kaup. “Ossemeus Fossiles de
Darmstadt,” PI. la.

SIWALIK AND NAEBADA PEOBOSCIDIA.

5— 18G

of Binotherium, are smaller and narrower than the two anterior ridges and the anterior
valley, as is the case in our specimen ; while, as a fourth proof, the right-hand tooth
is the larger of the three, aod must consequently be the last of the series. Prom the
form of the worn dentine surface of the second true molar of our specimen, it is
clear that the fragment belongs to the left ramus of the lower jaw ; the right-hand
tooth of this specimen, therefore, corresponds with the first ridge of the left-hand
tooth (second true molar) of the specimen drawn in fig. 1 of the same plate.

Neither the last premolar nor the first true molar of this specimen carry any
talon-ridges ; the anterior extremity of the second true molar is likewise not fur-
nished with any talon. The longitudinal bridge connecting the transverse ridges of
the last premolar is higher than the similar bridge connecting the two anterior
ridges of the first true molar ; there is a large blunt tubercle placed on the outer-
side of the anterior valley of the last-named tooth.

Lower jaw from Sind. — The fine specimen of the greater portion of the low^er
jaw of a Binotherium represented on Plate XXX, was obtained by Mr. "W. T. Blanford
from the Siwaliks (Alanchars) of Sind in 1877, and has already been briefly noticed
by me in the “ Eecords.”^ The specimen comprises the middle portion of the left
ramus of the mandible, and contains the last premolar and the three true molars, all
in a very fine state of preservation. The two teeth on the left of the figure will corres-
pond in serial position to the two complete teeth represented in fig. 3 of Plate XXIX
(last premolar and first true molar), while the two teeth on the right correspond to the
two teeth represented in fig. 1 of the same plate (second and third true molars).

The mandible is abruptly broken off at both extremities of the series of teeth ;
the middle of the specimen is, however, complete, so that the depth of the jaw can be
ascertained. The enamel of the first ridge of the last true molar is only slightly
perforated by wear ; from which we may conclude that the animal to which the jaw-
belonged had only just attained its maturity at the time of its death.

Last premolar. — The last premolar is a simple two-ridged tooth, distinguished
from the corresponding tooth in the European Binotheria by the absence of
the large anterior talon, which always occurs in the European forms. A low-
longitudinal bridge connects the two transverse ridges of this tooth; owing, however, to
the incapacity of the native artist who drew this specimen, this longitudinal bridge
is not distinctly apparent in the figure : this bridge is not so high as the one in the
homologous tooth represented in fig. 3 of the preceding plate, but if the tooth
were still more worn down, the dentine of this longitudinal bridge w^ould be exposed,
and would then connect by an isthmus the dentine surfaces of the fore-and-aft
transverse ridges, which at present form islands ; this connection would not, however,
take place so soon as in the homologous tooth of the other specimen.

First true molar. — The first true molar is of the same width as the last pre-
molar ; and, as is always the case, this tooth carries three transverse ridges ; the
exposed dentine surfaces of the two anterior transverse ridges are connected by an

187—6

SIWALIK AND NARBADA PROBOSCIDIA.

isthmus of dentine, formed by the abrasion of a mesial longitudinal bridge. We
liave already seen that a similar condition occurs in the homologous tooth of the
jaw represented in fig. 3 of Plate XXIX. The valley between the second and
thii’d transverse ridges is still open, and these two ridges would not be connected by
an isthmus of dentine until the tooth was much more worn down than at present.

Second ^ Third true molars. — The two last molars in the Sind jaw are consider-
ably wider than either of the jDreceding teeth : the penultimate tooth carries two trans-
verse ridges, and has only a very slight trace of a posterior talon ; in the latter
respect it differs from the homologous tooth of the jaw represented in fig. 1 of
Plate XXIX : it also differs from that tooth in having merely a very slight trace
of a longitudinal bridge, and consequently, although the two teeth are in almost
exactly the same state of wear, exhibits no worn dentine surface projecting from
the hinder ridge into the transverse valley.

Third true molar. — The last or third true molar (right side of figure) , also car-
ries two transverse ridges, and has a large three-cornered j)osterior talon, and a low
longitudinal bridge projecting from the second ridge into the transverse valley. In
all its characters, this tooth agrees precisely with the homologous tooth in the jaw
represented in fig. 1 of Plate XXIX, and, therefore, needs no further description.

Dimensions of lower jaws. — In the following table the dimensions of the three
lower jaws of Dinotheriuni described al)ove are compared together. The dimen-
sions in the first column of that table {a) are those of the specimen represented in
Plate XXX ; those in the second column {h) are the dimensions of the specimen
represented in figs. 2 and 3 of Plate XXIX ; while the dimensions given in the
third columir (c) are those of the specimen represented in fig. 1 of the latter
idate.

^ ’ c

Length of four last teeth

Ditto last premolar

Width of ditto

Length of first true molar

Width of ditto

Length of penultimate true molar .

Width of ditto ditto

Length of last true molar

Width of ditto

Depth of jaw at penultimate true molar
'thickness of ditto at ditto

Ditto of enamel of first true molar
Ditto ditto second ditto

9-3

1-8

1-74

205

2-34

2-4

. !■.?

2-4

2-9

2-15

2T5

2-9

3T8

2-25

2-45

5-4

3-4

OTo

0-16

Comparisons. — If we compare the dimensions of the first lower true molar of the
specimen represented ir fig. 3 of Plate XXIX with the dimensions of the homo-
logous upper tooth of Dinotheriuni jpentapotamice, figured and described in a previous
fasciculus of this volume,^ we shall find that the length of the latter tootli is

' “ Molar Teeth and other Ecmains of Mammalia/’ pp. 56-74, Plate IX, fig. 2.

SIWALIK AND NARBADA PllOBOSCIDIA.

7—188

2'75 indies, and its greatest width 2"4 inches. As far, therefore, as the matter of
size goes, the two specimens may certainly he referred to the same species, because
in B. gigantemn the upper molars are considerably larger than the lower. The
ridges of the upper molar are more curved than those of the lower, but that is
always the case in the teeth of this genus. The upper molar has not the connect-
ing longitudinal bridges which are found in the lower molar, but this, as we shall
subsequently see, appears to be a variable character, and I think we may therefore
safely say that the specimen represented in Plate XXIX, fig. 3, belongs to B.
pentapotamicG.

Now, with regard to the other two specimens : if we take the last tooth of tbe
specimen represented in fig. 1 of Plate XXIX and compare it with the homologous
tooth of the specimen represented in Plate XXX, we shall, I think, certainly
arrive at the conclusion that both these teeth belong to one and the same species ;
the only difference between them being that the latter tooth is very slightly the
larger of the two.

The penultimate molars of the two jaws are distinguished, as we have already
seen, by the presence of a hind talon, and an imperfect longitudinal ridge in the
transverse valley of the one, and by their absence in the other. In consequence of
the absence of this talon in the specimen represented in Plate XXX, this tooth is
half-an-inch shorter than the homologous tooth in the other specimen ; both,
however, have precisely the same width. The complete absence of the hind talon in
the one tooth may not improbably be explained by a greater pressure exerted on it
by the succeeding tooth ; the longitudinal bridge, as I have already said, I do not
consider to be a character of rnuch specific value ; and I think that we^ may say that
the lower jaws represented in Plate XXX and in fig. I of Plate XXIX belong to
the same species ; it therefore only remains for us to see whether they belong to
the same species as the specimen represented in fig. 3 of the latter plate, or, in
other words, to B. pe^itapotamics, to which I have referred that specimen.

On referring to the measurements of the three specimens given in the table
above, it will be seen that the last premolar and the first true molar of the speci-
men represented in Plate XXX are slightly smaller than the homologous teeth
of the specimen represented in fig. 3 of Plate XXIX ; it will, moreover, be ob-
served from the figures, that there is a slight difference in the relative develop-
ment of the longitudinal bridges in the teeth of the two specimens. Beyond this
difference in the development of the longitudinal bridge, and the slight difference
in size, I can find no points of distinction between the teeth of the two specimens.
With regard to the development of the longitudinal bridge, we have already seen
that it varies in the two specimens of the two last molars represented in the accom-
panying plates, which certainly belong to the same species, and if we refer to the
figures of the molars of Bmotlierium giganteum given in the third volume of the
atlas of De Blainville’s “ Osteographie,” or in Kaup’s memoirs, we shall find that in
the specimens marked B. cuvieri and B. intermedium (these pseudo-species being

189—8

SIWALIK AND NARBADA PROBOSCIDIA.

now considered merely as varieties of D. giganteum) the longitudinal bridge is inucli
more developed than in tlie otlier specimens. In a large collection of the molars of
D. giganteum from Eppelsheim, in the Indian Museum, I also find that there is a
very great variety in the relative development of this bridge. It seems therefore
that, as far as regards this longitudinal bridge, no valid grounds for specific distinc-
tion can be adduced.

There now remains, therefore, merely the question of the difference of size, and
this is so slight that it may well be due to sexual difference only : I have, there-
fore, no hesitation in referring all the three specimens in question to D. penta-
potamice.

Comparison with D. indicum and D. giganteum. — It now remains to compare
these lower jaws with those of the other two well-established sj)ecies of Dmotherium,
namely, the Indian D. indicum and the Euro]3ean D. giganteum. The remains of
the first of these species were originally obtained from Perim Island, and consisted
of a broken molar and a portion of the mandible, which are described in the
“ Palseontological Memoirs.^ Specimens of the teeth of this second Indian species,
which was slightly larger than D. giganteum, are described in the sequel of the
present fasciculus, where the characters which distinguish them from the correspond-
ing teeth of D. pentapotamice will be found pointed at. Here it will be only neces-
sary to compare together the dimensions of the three ja'vrs of the last named
species with the corresponding dimensions of the jaws of D. indicum and D. gigan-
teum ; the dimensions of the two latter species are nearly all taken from the table

given by Dr. Ealconer in

his above- quoted

notice.'

^ The three

jaws of

D. penta^

potayyiicB

are Respectively

indicated by

the

same

letters as were used in the table

given on page 187 of this

memoir.

D. pentapotamice.

rt J\ ^

D. indicum.

H. giganteum.

Length of last premolar

1-8

2-9

3-3

Width of

ditto

1-74

2-05

2-6

2-7

Length of first true molar’

2-34

2-4

3-6

Width of

ditto

1-8

2-8

Length of second true molar

2-4

2-9

Widfli of

ditto

215

2-15

3-5

2-9

liength of third true molar

2-9

3-18

...

3-8

Width of

ditto

2-25

2 15

3’5

Depth of jaw at second premolar

9-2

6-9

Ditto

at first true molar

5-2

... ...

8-7

6-2

Ditto

at second true molar

5-4

6-6

Ditto

at third true molar

6-4

Width of jaw at second premolar

5-1

4-0

Ditto

at first true molar

3-4

6-4

5-0

Ditto

at second true molar

3'4

Ditto

at third true molar

4-1

fil

* Vol. I, pp. 39fi, 404-
" Ihid., p. 407.

SIWALIK AND NARBADA PROBOSCIDIA.

9—190

Prom a perusal of this table, it will be at once evident that the great difference
in the matter of size alone, would be a sufficient distinction between DinotJierium
pentapotami<E and D. itidicum., even if there were not differences in the form of the
teeth themselves, which I shall point out when I come to the description of those
of the latter species. The jaw of is less compressed than that of

D. indicum ; if the jaw of the latter at the second premolar had the proportions of the
former at the first true molar, it would have a width of 6T inches in place of only 5T
inches, which is its real width. The less elliptical cross-section of the jaw of J).
pentapotamicB is a character in which it approaches typical Mastodons. The specimens
are too imcomplete to admit of any more elaborate comparison of the form of the jaw.
With regard to D. giganteum, there is a closer resemblance in the matter of size
between it and D. pentapotamics ; indeed, if we compare together the larger teeth
of the latter, and the smaller teeth of the former, we might possibly find specimens
which showed extremely little difference in the matter of size. This, however,
would obviously be an unfair method of comparison, since we ought to compare
together the large teeth of one species with the large teeth of the other, and mce versa.
If such a comparison be made, it will always be found that the teeth of D. gigantetim
are considerably larger than those of D. pentapotamice. Burther, the last lower true
molar of the latter species has a much larger hind talon than the corresponding
tooth of the larger European species, and there seems to be generally a greater
teirdency to a development of the longitudinal bridge in the lovv^er molars of the former
than in those of the latter sj)ecies. In my previous notice of the molars of the same
species, I have pointed out certain distinctive characters of the upper molars, among
the most important of which, is the complete blocking of the lower portion of the
outer extremity of the transverse valley in the second upper true molar, a character
which causes the structure of that tooth to approach to that of a premolar. These
combined differences, together with those which I shall immediately show to occur
in the form of the last upper premolar, confirm the conclusions previously arrived
at, as to the specific distinctness of Dinotherium pentapotamice.

It may be not out of place to notice that in teeth which are constructed on such
an exceedingly simple plan as are those of Dinotherium, the variations in form
which they are capable of undergoing must of necessity be extremely small, and
there is consequently in most cases very great difficulty in deciding the specific
identity or distinctness of isolated teeth, as has been remarkably illustrated in the
case of the European forms of the genus. Any constant point of difference in the
teeth, however small, must, I think, in consequence of this small amount of possible
variation, always be regarded as of specific value.

Last upper premolar. — The last tooth of the present species, which calls for
notice here, is the specimen represented in fig. 3 of Plate XXXI, which is the last
upper premolar of the left side, and which has not previously been described. The
preceding or penultimate upper premolar of the right side has already been figured
in Plate IX, fig. 1, of an earlier fasciculus of this volume.^

* “Molar Tc-etli and Other Ecinains of Mammalia."

191—10

SIWALIK AND NAEBADA PliOBOSCIDIA.

The tooth under discussion was obtained by Mr. W. T. Blanford from the
Sivvalik (Manchar) rocks of Sind ; it is in an intermediate condition of wear, and
has unfortunately lost a portion of the centre of its hinder ridge, and the outer
extremity of the anterior ridge. The crown is approximately square, and carries
two transverse ridges, of which the anterior one (left side of figure) extends com-
pletely across the crown, with only a slight depression near the middle, while the
posterior is interrupted by a wide and deep gap. These gaps or depressions in the
transverse ridges show that the tooth does not belong to the true molar series, in
which the ridges are uninterrupted. At the outer side of the crown (top of figure)
the two ridges are connected by a low wall with a V-shaped gap in the middle ; on
the inner side (lower border of figure) the transverse valley slopes down between
the extremities of the ridges, without any tubercle at its entrance. The inner
extremities of the ridges are wider and more worn than the outer. On the dorsal, or
outer surface of the tooth, there is a deep vertical groove, occupying the space
between the extremities of the two ridges. The tooth is distinguished at once from
the penultimate premolar (Plate IX, fig. 1) by the greater continuity of the
transverse ridges, and by the absence of the high and externally flat wall which
forms the outer side of that tooth. The dimensions of this tooth are compared
below with those of a corresponding tooth of D. giganiemn from Eppelsheim, in
the collection of the Indian Museum : —

D ‘pentapotamia. D. giganteum

Length of outer border 2’1 3‘0

„ „ inner 2‘0 2-1

Width of last ridge 2'5 3 2

The tooth of D. giganteum is rather a small specimen. On comparing together
the two teeth, it will be found that the crown of the Indian tooth is more nearly
square than that of the European ; the groove on the outer surface is considerably
deeper in the former than in the latter ; the anterior ridge is less complete in the
European tooth, and the two outer extremities of the ridges are more closely
connected than in the Indian tooth. In the latter there is no trace of the tubercle
which occurs at the entrance of the transverse valley in the former.

I have lately made a comparison of the penultimate upper premolar of D,
pentajpotamice (Plate IX, fig. 1), with a corresponding tooth of D. giganteum from
Eppelsheim, and I find that the European tooth is more square (this is especially
exemplified in some of Kaup’s figures)' than the Indian; that the hind column
(representing the inner extremity of the second ridge) extends further across the
tooth in the former. The entrance to the transverse valley is blocked by a
conical tubercle in the European tooth, while in the Indian the valley is extended
outwards by a kind of spout-shaped j)i’ojection ; there are also differences in the
shape of the cingulum, which are not easy to describe. The differences between
the premolars confirm the conclusions previously arrived at, as to the specific
distinctness of the Indian form.

‘ Kaup, loc. cif., Plate II, fig. 6,

SIWALIK AND NAEBADA PROBOSCIDIA.

11-192

Correction. — The perfect state of preservation of the specimens represented in
the accompanying plates of this memoir, has shown me that I was misled as to the
structure of the imperfect lower molar from Sind represented in fig. 5 of Plate IX
of an earlier fasciculus of this volume.^ In that specimen I considered the promi-
nence on the right side of the figure as being a true transverse ridge, and conse-
quently that the tooth was a three-ridged one, of which the first ridge was broken
away : the specimen was accordingly considered as the first lower true molar. A
comparison of that tooth with the last lower molar of the specimen represented on
Plate XXX will , however, at once show that the prominence on the right side of
the broken tooth is in reality merely a talon-ridge, and consequently that the tooth
is, in truth, the last, and not the first, lower true molar.

Dentition of D. pentapota7ni(B. — Taking the specimens of the teeth of D.penta-
potamim described here, together with those figured and described in the earlier
fasciculus of this volume (Plate IX), we are now acquainted with the greater part
of the permanent dentition of the species. The known teeth are, in the upper jaw,
the two premolars, and the first and second true molars, and in the lower jaw the
last premolar, and the three true molars. The following table gives the dimensions
of all these teeth, together with those of D. gigantemn : — •

r Length of penultimate premolar
I Width of ditto ditto
Length of last premolar

I Width of ditto

’ Length of first true molar

Width of ditto

- Length of second true molar
I^Width of ditto

Length of last premolar
Width of ditto

^ Length of first true molar

S Width of ditto

w i Length of second true molar .

o Width of ditto

1-^

Length of third true molar .
j^Width of ditto

D. peniapotamia. D. gxganieum.

2-3 3-4

2-1 31

21 3-4

2-5 3-7

275 41

2-4 3-2

2-7 3-9

2-4 3'6

1-8 3-3

1- 74 2-7

2- 34 3-6

1- 8 3 0

2- 4 3‘3

215 2-9

2-9 3-8

2-25 2-45

The proportionate relations of the upper molars of the two species show that
those of D. pentapotami(B, which have all been found as detaclied specimens, are
rightly referred to one and the same species.

Species 2 : Dinotheiuum indictjm, Palconer. PL XXXI, figs. 1 & 2.

Distory. — We have now to consider two other molars of an Indian Dinotherimn,
which are of larger size than any of the previously described specimens, and which
also difiler from them, to a considerable extent, as regards form. Prom considera-

’ loc. cit.

193—12

SIWALIK AND NAEBADA PROBOSCIDIA.

tions wliich will be detailed below, these teeth have been referred to Dinotherium
indicum of Falconer, the history of which species has already been referred to when
treating of the preceding species of the genus.

Second upper true molar. — The first of the two specimens in question is the
half of an upper molar represented in fig. 1 of Plate XXXI. This specimen was
sent to the Indian Museum among a small collection of Punjab, and other North
Indian fossils made by the late Dr. Verchere, and appears to have been obtained
from the Siwaliks of Dehra Ghazi Khan, on the north-west frontier. This speci-
men, together with another tooth described below, has been already shortly noticed
by me in the “ Records. The fragment consists of the hinder ridge of the
second left upper true molar. The position of the tooth is determined by there
being a disc of pressure on the remaining ridge, indicating the presence of another
tooth on this side, and by the width of the ridge being equal to that of the valley,
in place of being narrower, which it would be were the tooth the third true molar.
The concave side of the ridge shows that this is the hinder side of the tooth, and
the broken tubercle on the lower border of the figure shows that this border is the
inner side of the tooth. The great general size and width of the ridge shows
that the tooth cannot be the first true molar. The one remaining ridge is but
slightly abraded by detrition, the enamel not having been perforated ; the plane of
wear slopes towards the anterior side of the specimen.

Since the corresponding molar of the smaller D. pentapotamice has already been
figured on Plate IX, fig. 3, and described on page 73 of an earlier part of the present
volume,^ it will be simpler to commence by comparing these two teeth together at
once, instead of describing the larger tooth separately.

Placing the two figures side by side, it wdll at once be seen that the tooth of
D. indicum is widely distinguished from that of D . pentapotamice- by its greatly supe-
rior size. In addition to this distinction, however, tliere are points of difference in
the form of the two teeth, which afford more certain grounds for the specific dis-
tinctness of the two animals to which they belonged. It wdU be observed that the
single ridge of the larger tooth is considerably straighter, and has a much slighter
cavity on the posterior side, than has the corresponding hinder ridge of the smaller
tooth. Again, the comparatively flat hinder surface of the larger tooth is entirely un-
interrupted by ridges or projections, whereas the corresponding surface of the smaller
tooth has a very distinct curved ledge, or cingulum near its base, and an imperfect
ridge somewhat higher up. In the transverse valleys of the two teeth, another im-
portant distinetion presents itseff ; in the smaller tooth the outer extremity of that
valley is blocked by a low wall, which is the highest part in the whole valley ; in
the larger tooth, on the other hand, the valley is perfectly open externally, the
middle point of the valley being the highest, from which point it slopes away rapidly
on either side. In both the larger and smaller teeth there is a very large rounded

’ Eec. Geol. Surv. of India, Vol. XII, p. 42.

? “ Molar Teeth and other Remains of Mammalia” (volume paging ).

SIWALIK AND NAEBADA PEOBOSCIDIA.

13—194

tubercle placed at tlie inner entrance of the transverse valley. Prom tlie height
of the centre of the transverse valley in the larger tooth, it appears that in the
complete tooth there must have been a slight longitudinal bridge, connecting the
transverse ridges. The larger tooth is further distinguished from the smaller by
the much greater thickness of the enamel. In the following table I have compared
together the dimensions of the two teeth, together with those of the corresponding
tooth of the European Binotherium giganteum :■ —

Z). indicium, D. giganteum. D. pentapotamim.

Width of last ridge SV 3’4 2-3

Thickness of base of ridge 1'8 1-6 I’l

Thickness of enamel 0-25 0‘i8 0-09

Although, from its incomplete state, we are unable to compare the whole of the
larger tooth with the corresponding tooth of Binotherium pentapotamice, yet the
comparisons we have made are, I submit, sufficient to show the specific distinctness
of the two. It now remains to see to what species the larger tooth should be
referred. If we turn to the “ Palseontological Memoms ” of Dr. Ealconer,^ we shall
find that the fragmentary molars and lower jaw of Binotherium indicum are of
somewhat larger size than those of B. giganteum, while the former are further dis-
tinguished by the great thickness of their enamel. In both these characters the
imperfect tooth described above agrees with Dr. Falconer’s specimen, and I have
therefore referred it to the same species.

Dr. Falconer, in the notice mentioned above, has shown in what respects the
lower jaw of Binotherium indicum differs from the lower jaw of B. giganteum. The
fragmentary upper molar of the former species is distinguished from the correspond-
ing tooth of the latter by its larger size and thicker enamel, by the great size of the
tubercle at the inner extremity of the transverse valley (which is very small or
wanting in B. giganteum), and by the elevation of the middle of the transverse
valley.

First lower true molar. — The only other tooth of Binotherium in the Indian
Museum which I can refer to the same species, is the specimen represented in fig. 2
of Plate XXXI, which was collected by Mr. W. T. Blanford in the Laki hills
of Sind. The specimen is a complete three-ridged tooth, with all three ridges well
worn. From the number of ridges it might be one of the “intermediate” molars
of a trilophodont Mastodon, but the absence of any trace of a median longitudinal
cleft across the ridges, together with other characters which I shall subsequently
notice, shows that it cannot belong to a Mastodon, and that it must therefore be
referred to a species of Binotherium. In the latter genus a three-ridged tooth can
only be either a last milk-molar or a first true molar ; the great size of our speci-
men shows that it cannot be a milk-molar, and it must consequently be a first true
molar. The large size of the antero-posterior diameter in relation to the transverse,
together with the convexity of the ridges being on the hinder side, shows that the

* Vol. I, pp. 396, 404.

195—14

SIWALIK AND NARBADA PROBOSCIDIA.

tooth is the first lower true molar, while the greater height of the inner extremities
of the transverse ridges (top border of figure) shows that the tooth belonged to the
left ramus of the mandible. The tooth, therefore, corresponds to the second tooth
(counting from the left) in the left ramus of the mandible of Binotherium penta-
potamicB represented on Plate XXX ; the two teeth are almost in the same relative
condition of wear, which will greatly facilitate our comparison.

It will firstly be noticed that there is a very great difference in the matter of
size alone in the two teeth, which, as far as it goes, would lead us to refer the two
to distinct species. There are, however, more important points of differences in the
form of the two teeth which I now proceed to consider. The larger tooth is dis-
tinguished by having on its outer and hinder surfaces a very stout and thick
cingulum, extending from the middle of the anterior ridge (left side of figure)
along the outer and posterior surfaces ; (in the specimen this cingulum has been
broken away on the posterior surface, hut the fractured surface shows its former
extent) ; no trace of this cingulum is to he seen on the smaller tooth. Again
in the smaller tooth, the worn surfaces of the first and second ridges are united by
a longitudinal bridge of dentine, while in the second and third ridges the worn
surfaces are still separated by the transverse valley. In the larger tooth, all three
of the worn surfaces of the ridges are still separate, but it will he seen that the
second transverse valley is as nearly obliterated as the first, and that consequently
the worn surface of the third ridge would he united to that of the second, as soon
as the latter would to that of the first. Pinally, the enamel of the larger tooth is
very much thicker than that of the smaller. In the following table the dimensions
of i\\e) ivViQ oi Binotherium pentapot amice and B. giganteum are

compared vdth those of the corresponding new tooth : —

New tooth. D. giganteum. D, pentapotami<B,

Length of tooth . . . .

3'9

3-5

2-35

Width of first ridge . . . ,

2-5

2-6

1-8

Ditto of second ditto

2-6

1-8

Ditto of third ditto

2-4

2'2

1-7

Thickness of enamel ....

0-25

0-19

OTl

These differences in size and

form make it

clear that the

new

tooth cannot

belong to B. pentapotamicB. In

the lower jaw

of Binotherium indicum described

by Dr. Palconer from Perim Island,^ the crowns of the molars have unfortunately
all been broken off, and we eannot therefore compare the form of the first lower
molar of that specimen with the Sind tooth. The bases of the molars, however,
remain in the Perim Island jaw, and the dimensions of the first true molar, as given
by Dr. Palconer,^ are as follows : Length 4 inches, width behind 2’8 inches. Now,
we have already seen that the Sind tooth is slightly broken, and that therefore its
length, where complete, must have been almost exactly equal to the Perim Island
tooth ; the latter, however, is rather the wider of the two. As the Sind tooth

* Palseontolo^ical Memoirs, loc, cit.

- Ihid, p. 407.

SIWALIK AND NARBADA PROBOSCIDIA.

15—196

agrees so closely in size with the Perim tooth, and also with the upper molars of
the same species in the thickness of the enamel, I have referred it to the same
species.

The first lower molar of D. indicwm agrees with the other remains of the same
species in being slightly larger than the correspondmg tooth of D. giganteum.
The presence of the large cingulum on the former distinguishes it from the latter.

Distribution,— The, distribution of Dinotherium indicum now extends from
Perim Island through Sind to the north-western Punjab.

Characters. — The species is still very imperfectly known. It appears to he the
largest known species of the genus, and is mainly characterized by its thick enamel.
Dr. Palconer thought from the study of the jaw, that he could trace indications of
a closer affinity with the Mastodons, than exists in other species of the genus.
It is much to be hoped that we may eventually discover the skull of tliis and the
other Indian species.

Species 3 : Dinotherium sindiense, n. sp. nobis. PI. XXXI, fig. 4.

Besides the two foregoing species of Dinotherium, we have one specimen of a
lower jaw in the Indian Museum, which indicates the existence of a thud Indian
species of the genus, which must have been of very small size, and which I propose
to call Dinotherium sindiense, as the specimen in question was obtained in Sind by
Mr. Pedden. A notice of the discovery of this specimen has been previously given
in the “ Records.”^ The specimen is represented of the natural size in fig. 4 of
Plate XXXI, and will he seen to consist of a fragment of the hinder part of the right
ramus of the mandible, carrying two broken molars. Although these molars are
considerably damaged, enough of them remains to determine their serial position.
The tooth on the left side of the figure will be clearly seen to be a molar, which
originally carried two ridges and a large hind talon, and in fact corresponded in
form to the molar on the right hand of the jaw represented in Plate XXX, the
two specimens belonging to opposite sides of the mandible. This tooth therefore
is the last lower true molar, and the preceding more broken tooth (on the right side
of the figure) will consequently he the second true molar. The position of these
teeth being determined, it follows that the jaw belonged to a fully adult animal, and
that it cannot be a young specimen with the milk-dentition, in which case the
last tooth would have had three ridges.

Although the teeth of this specimen are considerably smaller than the corre-
sponding teeth of Dinotherium pentajpotamice, I should not, probably, on these
grounds alone, have referred the Sind mandible to a different species. The form of
this new lower jaw is, however, so totally different from the lower jaw of D. penta-
potamice, of which the teeth are represented in Plate XXX, that there can be no

' Eec. Geol. Surv. of India, Vol. XII, p. 43.

197—16

SIWALIK AND NAEBADA PEOBOSCIDIA.

question but that they belong to distinct species. In the following table the
corresponding dimensions of the two specimens are compared together : —

Z). pentapotamiee. B. sindiense.

Length of second molar 2'4 2 3

Width of ditto 2’15 1’55

Length of third molar 2’9 2-5

Width of ditto 2'25 1’9

Depth of jaw at second molar 5 3 3'1

Ditto at third molar S'S 3‘3

Width of jaw at second molar 3'4 2'9

Ditto at third molar 41 3’5

The great difference in the shape of the two jaws will be gathered from a
perusal of the foregoing table. The lower jaw of D. jpentapotamice is laterally
compressed, though, as before said, to a somewhat less extent than that of some other
species, and its vertical diameter is consequently very much greater than its trans-
verse diameter. In D. sindiense, on the other hand, the lower jaw is nearly cylin-
drical, and at the last true molar the transverse diameter is greater than the vertical
diameter. Whereas the lower jaw of penta'potamicB on its outer side descends
vertically down from the teeth, that of D. sindiense is expanded into a distinct ledge
below the roots of the molars.

It is unfortunate that the molars of J). sindiense are in such a battered con-
dition, so as to preclude any exact comparison with those of D. pentapotamiee, but
the differences in the size and form of the two jaws are amply sufficient to afford
grounds of specific distinction. In its cylindrical form, indeed, the lower jaw of
Dinotherium sindiense seems to differ from the jaws of all other known species
of the genus : from its sub-circular cross-section and generally slender form, I think
it not improbable that this species of Dinotherium was unprovided with a long
deflected symphysis and mandibular incisors. In all other species of Dinotherium
the mandible is laterally compressed to offer more resistance to the enormous
weight of the symphysis and tusks, and in all Mastodons with a long heavy symphysis
and mandibular tusks {M, angustidens, M. longirostris, M. pandionis), the rami of
the mandible are similarly compressed. If this suggestion should turn out to be
true, it may be necessary to form a new genus for D. sindiense. In its sub-cylin-
drical mandible this species shows affinity to the Mastodons with short mandibular
symphyses.

Family II—FLFFMAFTIDM.

Prohosciuia in which there are never more than three teeth in use at the same
time, and in which the number of ridges in the second true molar is either equal to,
or greater than, the number in the preceding tooth,

SIWALIK AND NAEBADA PEOBOSCIDIA.

17—198

Dentition oe the Elephantidje.

The dentition of the Elephantidce is perhaps more remarkable than that of any
other group of the Mammalia, in regard both to the structure of the teeth and in
their mode of succession, and is evidently the product of a long course of special-
ization. With regard to the structm’c of the teeth, I shall not enter into details here,
since this has been most amply illustrated by Ealconer,’^ Owen^ and other writers.
With regard to the number and succession of the teeth, a short notice is necessary,
in order that the reader who has not specially studied the dentition of this interest-
ing group may have no difficulty in following the descriptions given in the sequel.

It will not he necessary to notice the systems of classification proposed for the
molars of the Elephantidce, before their homology was thoroughly understood, which
has been already lucidly done by Dr. Falconer,® and I therefore proceed at once
to the classification of those teeth as given by the last-named writer.

In the living Indian elephant, it has been demonstrated that six grinders are
developed in a horizontally succeeding series, on either side of both upper and lower
jaws, and that the hinder teeth are more complex than the anterior. The whole of
these teeth are never in use at any one time, the anterior ones falling out as they are
worn down, and the posterior ones gradually coming into use and replacing them ;
the animal when quite adult has only the last molar in use, the length of wear of
which of course determines the life of its owner. There are never more than three
teeth on one side of either jaw in use at any one time.

The Mastodons present a like number of horizontally succeeding teeth. Of
these six teeth in that group, the thEd, fourth and fifth usually carry the same
number of ridges, and have hence been termed by Dr. Falconer the “ intermediate”
molars. In the Trilophodont section of the group the number of ridges on each of
these three teeth is always three, and in the Tetralophodont, normally four ; the
tooth (of the horizontal series) immediately in front of the first “ intermediate”
molar, has always one ridge less than the latter, and tlie tooth behind the third
“ intermediate” molar, one ridge more than that tooth. The serial position of these
teeth is, therefore, generally pretty easy to determine, since the three isomerous
“ intermediate” molars increase regularly in size from the first to the third.

Of these six teeth, it will subsequently be shown that in many species of
Mastodons, and in one elephant, the second and third (and sometimes the first) are
replaced by vertically succeeding teeth, which consequently correspond to the pre-
molars {dents des remplacement) of the normal mammalian dentition, as exemplified
in the Pig. The three first teeth in the horizontal Series must therefore correspond
to the last three teeth of the milk dentition of the Pig, and may therefore be called

* Pal. Mem., Vol. I, p. 43, et seq. Fauna Antiqua Sivalensis, Pis. I & II.

- Odontography, p. 613 et seq.

3 loc. cit.

399—18

SIWALIK AND NARBADA PROBOSCIDIA.

the antepenultimate, penultimate, and last milk molars, respectively^ The three last
teeth in the same series, as being developed behind and after the milk-molars, must
consequently correspond to the true molars of the Pig ; and are, therefore, called
respectively the first, second and third true molars.

Now we come to the consideration of the dents des r emplacement or premolars,
the succession of which requires a little more discussion. In a memoir on elephant
and Mastodon, by Dr. Palconer,^ puhhshed in part in 1816, the writer seems to have
been of opinion that apparently in all Mastodons only the penultimate milk-molar
was succeeded by a premolar [dent de r emplacement'). This conclusion seems to
have been previously arrived at by Professor Owen, who retains it up to the late
date of 1861.® Dr. Palconer, however, in a subsequent paper, ‘‘ announced that
in many species, at least, two premolars were developed, which I infer, though
this is not clearly expressed, replace the second and third milk-molars. That the
last upper milk-molar was succeeded by a ‘ dent de r emplacement' is proved by the
specimens drawn in figs. 6 and 8 of Plate XXXVII of this memoir, as well as by
many other foreign specimens, from which we arrive at the conclusion that the
two last milk molars were respectively rejDlaced by vertically succeeding premolars
in the upper jaw : other specimens (especially a fine lower jaw of M. angustidens)
show that the two last lower milk-molars were similarly replaced.

Although it is frequently difficult, as will be noticed in the sequel, to refer
detached milk-molars and premolars to their proper serial position, there is never any
difficulty when the teeth occur in the jaw. When all the teeth in a Mastodon jaw belong
to the milk and true molar series, any one tooth is more worn than the tooth imme-
diately behind it : when premolars are present, on the other hand, the last premolar
is less worn than the tooth immediately behind it, whether such tooth be a milk-
molar or the first true molar.

Besides the penultimate and last milk-molars, there is good evidence that in
one species of Mastodon, at least, an antepenultimate premolar was also developed.
In a young upper jaw of the American trilophodont M. productus, figured and de-
scribed by Professor Cope,® in front of two three-ridged teeth, which from their
isomerous ridge-formula, must be the first and second “intermediate” molars, or
respectively the last milk-molar and the first true molar, there are developed two small
teeth, which are less worn than the third milk-molar, and which consequently are
true “ dents des r emplacement," or respectively the antepenultimate and penultimate

’ For the sake of brevity in the text, I have usually termed these teeth, respectively, the first, second and third milk,
molars, though, of course, they really correspond to the second, third and fourth of the milk-molars of the typical series.
According to Falconer the pre-antepenultimate milk-molar is occasionally developed as an abnormality in the African
elephant.

^ loc, cit., p. 95.

® “Odontography,” p. 615, ed. 1840.45. “Palaeontology,” p. 388, ed. 1861.

* loc. cit., Vol. II, p. II (published in 1857).

* “ Extinct Vertebrata of New Mexico.” U. S. Geographical Survey, W. of 100th merid., Vol. IV, Pt. II, p. 309,
PI. LXXI, fig. I.

SIWALIK AND NARBADA PROBOSCIDIA.

19—200

premolars.. We have tlierefore now proof that all the milk-molars of Mastodons may
be replaced by premolars.

I shall, I hope, be pardoned for pointing out what appears to me to be a want
of precision in Professor Cope’s description of the above-mentioned teeth. In that
description the learned author, speaks of all the three anterior teeth as “ premolars,”
apparently not recognizing their belonging to totally different series, though he men-
tions that the last “premolar,” according to Owen, belongs to the milk-series. This
fusion under one serial name, of teeth belonging to two totally different series, is a
retrograde step from Palconer’s clear definitions. It does not appear that Professor
Cope recognized the importance of his specimen in further elucidating the dentition
of the Mastodons.

Although we have thus seen that the antepenultimate and penultimate milk-
molars may be replaced by premolars it appears to me that certain pre-molars of
M. angustidens classed by the late Professor Herman von Meyer as the antepenul-
timate and penultimate, are really the penultimate and last. In a left maxilla of a
young M. angustidens figured by that distinguished palseontologisP, there are seen
three teeth, the two first of which are less worn than the third, and which are
consequently premolars ; the third tooth carries three ridges, and, the species being
trilophodont, might be either the last milk-molar or the first true molar. By
von Meyer the three* ridged tooth is classed as the third milk-molar {dritter milch
backenzahn), while the two premolars are respectively classed as first and second
(erster und zweiter erzatzhackenzahn) . This first and second, as being in front
of the third milk-molar, must be equal to penultimate and antepenultimate. In
fig. 12 of Plate V of the same memoir, von Meyer figures a two- ridged premolar,
which he calls the second or penultimate {zweiter ersatzhackenzahn) . Now,
in a cast of a young lower jaw of M. angustidens in the Indian Museum, below
a three-ridged tooth, there is seen in its alveolus a replacing premolar, which is
similar to, though smaller than, von Meyer’s lower premolar ; the former, as replac-
ing the last milk-molar must be the last or third (fourth of the typical series) pre-
molar, from which I infer that von Meyer’s is the same.^ Again, in the upper jaw
the two premolars are far larger and more complex than the antepenultimate and
penultimate premolars of M. 'productus., and I therefore think it is pretty clear
that the former teeth really are the two last of the premolar series. The third tooth
in von Meyer’s specimen will consequently be, if I am right, the first true molar.
I shall have to refer to this question again, as it has an important bearing on the
determination of certain teeth figured in this memoir. I may also mention that in
fig. 16 of Plate V of the same memoir von Meyer has figured a detached three-
ridged lower molar of M. angustidens, as the second lower milk-molai {zweiter milch
hackenzahn) ; this tooth, I think, is certainly the thii*d milk-molar. The second
lower milk-molar, as seen in a specimen from Eppelsheim in the Indian Museum
(preceding a three-ridged tooth, and with the pressure-mark of a preceding tooth
anteriorly) is a two-ridged tooth; tlie third milk-molar, succeeding the s(;X?ond, in the

' Palffiontographica Vol. XVII, PI. Ill, fig. 1.

^ See “ Les Enchainements du Monde Animal,” fig. 242.

201—20

SIWALIK AND NARBADA PROBOSCIDIA.

above specimen, is quite like von Meyer’s specimen, but rather smaller. If von
Meyer’s determination be not erroneous, Dalconer’s ridge-formula for this Trilopho-
tlon will not hold.

It thus appears that among the Mastodons, all together nine molar teeth may
be developed, which may be tabulated as follows : —

Horizontal series

Vertical series .

. 1

Intermediate molars.

2 3 4 5 6

Milk-molars. True molars.

(2 3 4) {12 3)

(of type series.) (of type series. 1
. 1 2 3

Premolars.

(2 3 4 of
type series.)

Milk-molars |5| Premolars True molars

In all species the whole of the horizontal series is always developed ; of the
preniolar series in one species, the two first (and probably also the last) are developed
{M. productus) ; in others only the two last (Jf. angustidens, according toPalconer’s
and my own view, and Jf. latidens), and in others none {M. giganteus, M. sivalensis).

In the elephants, premolars are only known to be developed in one species
[E. planifrons): all the horizontal series are developed, but the isomerous ridge-
formula of the intermediate molars is not preserved in tlie more specialized forms.

We now come to the consideration of the rest of the dentition of the Elephantidce,
which need not detain us long. In the upper jaws of all species there are
normally developed a single pair of tusk-like incisors, wdiich apparently correspond
to the outer pair of the three typical mammalian incisors. In the females these
tusks are usually smaller than the male ; in some species, however, {E. africanus),
the tusks are frequently large in both sexes. In one variety of the Indian elephant
the Miikna (Mooknah), the tusks of the male are small, like those of the female.
In the living species, and certainly in some fossil species, of ElepliantidcB, the per-
manent tusks were preceded by deciduous tusks : it is probable that this replace-
ment occurred in all members of the family.

In many Mastodons {M. longirostris,M. angustidens) a pair of tusk-like incisors
were also developed in the lower jaw, but it is not known whether these were pre-
ceded by deciduous incisors. In, at aU events, one species [M. angustidens) inferior
tusks were developed in both sexes in others {M. longirostris, M. perimensis)
in only one sex, which was doubtless the male : in other species again ( Jf. sivalensis
M. latidens, M. falconeri), there appear to be no lower incisors ever developed. In
no elephant is there ever any trace of lower incisors.

It will be apparent from the above summary that no generic determinations
of the Elephantidce can be made from the presence or absence of certain teeth.

* Prof. Owen, “ Odontography ” p. 615, gives no lower tusks to the females of all species, hut to all males of
the genus.

SIWALIK AND NAEBADA PROBOSCIDIA.

21-202

Genus 1 : MASTODON, Cuvier.

Elephants in which the number of ridges in the intermediate molars is never
greater than five, and in which cement is either entirely absent, or present in
comparatively small quantities.

Section A. — Teilophodon, Ealconer.

Intermediate molars with three ridges only.

Species 1 ; Mastodon ealconeei, n. sp. nobis. Pis. XXXII & XXXIII.

Kistory. — This species of Siwalik trilophodont Mastodon has been hitherto
known chiefly by name only, since in the previously published notices of the molars
upon which I founded the species, I did not give any detailed descriptions, hut
merely stated that I thought that these teeth were distinct from those of Mastodon
pandionis (the only other Indian species of the genus with a ternary ridge-formula),
and also from those of all other species of the genus. The first notice of this
species will he found in the “ Records,”^ where two lower molars are noticed : a
second notice of the imperfect cranium of a young animal and of some detached
molars was given subsequently.^ Since these two preliminary notices were pub-
lished, several additional molars of the species have been obtained by different
Officers of the Survey, and the more perfect specimens from this collection are
now for the first time described and figured under the name of Mastodon falco-
neri (Lyd.), which was originally proposed by me for the teeth noticed in the
“ Records.”

Specimens figured. — The molars figured in the present memoir comprehend the
second and third upper milk-molars, the first and second upper true molars, the
second lower milk-molar, and the first and second lower true molars. This series
of teeth affords us ample material for showing the distinctness and affinities of the
species, though it is to be hoped that subsequent researches may bring to light a
complete cranium, in order that we may compare this with the cranium of such of the
Siwalik species in which it is known.

Second upper milk-molar. — The two upper milk-molars represented in figs. 2
and 3 of Plate XXXII belong to the imperfect and much crushed cranium of a young
individual noticed above, which was obtained by Mr. Theobald in 1877 from the
Siwaliks of the Punjab. The smaller of these two teeth (fig. 2) is evidently
from its size and shape the second upper milk-molar of the left side.® The specimen
was implanted in the jaw by two fangs, the hinder one of which is considerably the
larger of the two. The form of the crown-sinface is rudely oblong, and is narrower

* Eec. Geol. Surv. India, Vol. X, p. 83.

5 Ihid., Vol. XI, p. 70.

* This tooth, being more worn than the succeeding tooth (fig. 3), must be a milk-molar, and not a premolar.

203—22

SIWALIK AND NAEBADA PEOBOSOIDIA.

in front than behind ; tlie triturating surface carries two main transverse ridges
separated on the outer side (top of figure) by a wide and open valley, but on the
inner side in contact. Each of tliese transverse ridges is mesially divided by
an antero-posterior cleft into a distinct inner and outer column, of which the latter
is the higher. In both columns of the anterior ridge the enamel has been per-
forated by trituration; the inner column of the second ridge also exhibits a
considerable islet of dentine, while tlie outer column of the same ridge has
merely its summit obliquely abraded, but the enamel not perforated : the den-
tine islets are in the form of irregular disks. In advance of the first transverse
ridge there is a small and narrow talon, somewhat broken in the specimen, and
on which the dentine is exposed by wear; behind the second transverse ridge
tliere is also another talon, of unusually large size, and carrying a distinct tubercle
at its inner extremity. There is a small tubercle at the outer extremity of the
main transverse valley. The length of this tooth is 1“8 inches and its greatest
width 1'6 inches.

Third upper milk-molar. — The larger tooth (fig. 3) immediately succeeded the
lu’eceding specimen in the young cranium, and is consequently the third, or last,
upper milk-molar of the left side. This tooth is implanted in the maxilla by four
distinct fangs ; the crown is oblong in shape and carries three transverse ridges and
fore-and-aft talons. The presence of these three transverse ridges on this tooth,
which is the first of the ‘‘ intermediate ” molars, proves that the cranium in question
belonged to a trilophodont Mastodon ; the corresponding and two succeeding teeth
of a Tetralophodon would of course carry four ridges each. Each ridge, as in the
preceding tooth, is mesially divided by an antero-posterior cleft into an inner and
an outer column, and each ridge carries five cusps or denticles. The inner portion
of both the second and third ridges is placed somewhat obliquely to the long axis
of the tooth ; the postero-internal angle of the last ridge joins the hind talon ; there
is a small but distinct tubercle on the hinder side of both the first and the second
ridges ; these tubercles at their junction vdth the centre of the next ridge very
slightly block the bases of the transverse valleys ; the latter are, however, mainly
characterized by being fairly free and open. The anterior talon is small, and its
cusps indistinct ; the posterior talon is larger, and its cusps, especially the three on
the inner side, are of large size and well defined. There is a faint trace of
abrasion on the anterior ridge of this tooth, otherwise it is untouched by Avear.
The length of the specimen is 3’04i inches and its width 2 inches.

Young cranium. — The imperfect cranium of a young individual of Mastodon
falconeri, to which the two last described milk-molars belong, is in much too
damaged a condition to afford a figure, or indeed any idea of the form of the perfect
cranium. Behind the tooth, which lias been described above, there occurs in that
cranium the germ of a third tooth which, like its predecessor, carries three transverse
ridges, and which from its position must be the first true molar. This tooth agrees
in general characters exactly with the last milk-molar described above ; as it cannot

SIWALIK AND NARBADA PROBOSCIDIA.

23—204

be tliorouglily cleaned from matrin, I have not given a figure of it. Its length is
3-5 inches, and its greatest width 2’ 2 inches.

First upper true molar. — The much worn tooth represented in fig. 4 of
Plate XXXII has a very convex masticating surface, and therefore belongs to the
upper molar series, and to the left side of the cranium, as is shown by the curvature
of the ridges ; as it agrees so closely in size with the last-mentioned specimen, and
carries three transverse ridges, it must be the first true molar. The first and the
tliird ridges are the most worn, all traces of the median longitudinal cleft having
been obliterated in these ridges ; in the second ridge a portion of this cleft still
remains, the masticating surface being consequently divided into two incomplete
dentine islands, of which the one on the lower side of the figure is the larger ; this
side is consequently the inner side of the tooth. The length of tliis specimen is
3 '6 inches, and its greatest width 2’3 inches.

This tooth is of great value, because it enables us to show that the first
true molar of FinotJierium indicum represented in fig. 2 of Plate XXXI cannot
belong to this species of trilophodont Mastodon, to which on first sight it might be
assigned. It will be seen, on comparing together the two teeth, that though the
tooth of the Mastodon is the most worn of the two, as is shown by its lower ridges
and wider dentine surfaces, yet still a trace of a median longitudinal cleft remains
in the second ridge, while in the much less worn 'Dinotherium tooth no such trace
is visible. The latter tooth, as a minor character, is further distinguished by the
much greater thickness of its enamel.

Second upper true molar. — The large tooth represented in fig. 1 of the same
plate was collected by Mr. Theobald in the Punjab, and is an upper molar of the
left side, and from its general characters evidently belongs to the same species
as the last. Since this tooth carries three transverse ridges, and is larger than any
of the preceding specimens, it must evidently be the last of the ‘intermediate’
molars, or, in other words, the second true molar. The greater portion of the
anterior ridge and talon of this tooth has been broken away, though the other two
ridges are complete ; the median ridge, and especially its inner column, has been
considerably abraded by detrition ; the enamel of the last ridge is not perforated.
A distinct median and longitudinal cleft divides each ridge into an inner and an
outer column ; in each of the two transverse valleys there is a tubercle on the
inner side of the median cleft closely applied to each side of the ridges. There is
in consequence a slight obstruction at the bottom of each of the valleys which are,
however, mainly characterized by being free and open; a large blunt tubercle
occupies the mouth of the inner side of each valley. The ridges are nearly straight,
which seems to indicate that the curvature which we observed in the last milk-
molar gradually dies out in the succeeding teeth ; the ridges when unworn seem to
have carried cusps like those of the preceding teeth. The posterior talon carries
two distinct cones, the innermost of which is the larger of the two. The enamel
is smooth and of considerable thickness ; the dentine islands of the inner columns

205 — 24

SIWALIK AND NARBADA PROBOSCIDIA.

are distinctly trefoil-shaped, caused by the union of the lateral tubercles with the
main column. The length of this tooth is 5-2 inches, and its greatest width
3 ’5 inches.

Second lower milk-molar. — Turning now to the lower molar series of this
species, of which three specimens are figured on Plate XXXIII, we may first con-
sider the smallest of the three. This specimen (fig. 2) was found by Mr. Theobald
with the young cranium noticed above, and belonged to the same individual.
The tooth is implanted in a small fragment of the mandible, the remainder of
which seems to have been unavoidably destroyed during the process of extricating
the specimen, which was in juxtaposition with other valuable fossils : the tooth is
implanted in the fragment of the jaw by two fangs. The outline of its crown in
plan roughly approximate to an isosceles triangle, of which the apex is placed
anteriorly ; the base of the triangle is somewhat damaged. The crown may be said
to consist of two ridges, separated by a deep transverse valley, and of a hind talon.
The anterior ridge has its longer diameter coincident with the longer diameter of
the tooth, and consists of five agglomerated columns ; the three anterior columns of
this ridge exhibit small surfaces of dentine exposed by detrition. The dentine
surface on the lower border of the figure is the most worn ; this border, therefore, is
the outer side of the tooth, which must consequently belong to the right ramus of
the jaw. The second ridge has its transverse diameter longer than the antero-
posterior, and runs somewhat obliquely to the long axis of the crown ; it is divided
by an indistinct median cleft. The external column of this ridge exhibits a plane
of detrition on its anterior surface, but the enamel has not been perforated. The
enamel, as in the upper milk-molars, is much corrugated. The fragment of the
mandible in which the tooth is implanted is laterally compressed to a slight extent.
The dimensions of the specimen are as follows : —

Length of tooth ........... 212

Width of ditto ........... 1-45

Height of crown of ditto . . . . . . . . . I’l

Depth of jaw ........... 2'6

Width of ditto ........... 2'0

We shall subsequently see that this slight lateral compression of the mandible
is only a character of the young animal, or possibly of the anterior portion of the
mandible in the adult.

With regard to the serial position of this tooth, it is evident that the specimen
belongs to the milk-molar series, and also that it is of too large a size to be the
first of the series, while it cannot be the third, as in that case it should carry three
transverse ridges ; it must, therefore, be the second of the milk-molar series, and
must consequently correspond in position with the upper tooth represented in
fig. 2 of Plate XXXII. The relative size of the two teeth agrees well enough
together ; the upper tooth, however, is somewhat more worn than the lower, which

SIWALIK AND NARBADA PROBOSCIDIA.

25—206

appears to be the reverse of what occurs in other Mastodons. If, for instance, we
refer to the figures of the upper and lower milk dentition of Mastodon pentelici and
M. tapir oides {turicensis) given by Professor Gaudry,^ we shall find that the lower
teeth are somewhat more worn than the upper. As I have not the least doubt but that
the two Indian teeth belong to the same cranium, we must assume that the relative
position of the upper and lower milk-molars of M. falconeri was slightly difi'erent
from that which occurs in M. pentelici and M. tapiroides : that is, that the second
upper milk-molar M^as partially abraded by the first lower milk-molar in the former.

First lower true molar. — The next tooth we have to consider is the specimen
represented in fig. 3 of Plate XXXIII. This specimen is an almost unworn germ
of the crown, with the fangs broken off, collected by Mr. Pedden in the Laki hills
of Sind. This tooth carries three transverse ridges and fore-and-aft talons ; it is
slightly larger than the first upper true molar in the young cranium noticed above,
but is evidently the corresponding lower tooth. The column in the lower right-hand
corner of the figure is very slightly abraded by wear, which shows that this side
is the front of the tooth ; from the flatness of the crown the tooth is evidently
from the lower jaw, and from the accessory columns being placed externally to the
median cleft, the tooth is as evidently of the right side, the arrangement of these
columns being exactly the reverse of that which occurs in the upper molars.

The general form of the crown, as viewed in plan, is oblong, with an average,
even width ; the ridges are blunt and short, with fairly open valleys ; each ridge
is divided by a deep median longitudinal cleft into an outer and an inner column ;
the three inner columns carry two cusps each ; the first outer column has a large
accessory column or tubercle, placed at its antero-internal angle, on the outer
side of the median cleft ; a faint rudiment of another accessory column occurs on
the postero-internal angle of the same main column. The outer column of the
second ridge has a large accessory column on either side adjoining the median
cleft ; the outer column of the third ridge is divided into two cones by a cleft. The
anterior talon (right side of figure) forms a broad projecting ledge on this side of
the tooth, divided into numerous cusps, of which the one at the inner extremity is
considerably larger than any of the others. The posterior talon consists of a low
double ridge, the first of these minor ridges being the larger of the two, and bearing
two large outer cusps, between which are three much smaller cusps ; the second ridge
of this talon consists of three or four ill-defined cusps. There is a slight rudiment
of a tubercle at the outer extremity of each of the transverse valleys. In tlie
centre of each of these valleys there is also a slight obstruction at the very base,
caused by the approximation of the bases of the contiguous accessory columns.
When worn down, the outer columns would evidently present distinct trefoil-shaped
islets of dentine. The length of this tooth is 4T inches ; its width 2’4 inches, and
the height of its ridges 1*4 inches.

Second lower true molar. — Continuing our serial survey of the lower molars

‘ “ Animaux Fossiles el Geologie de I’Attique,” Atlas, Pis. XXII and XXIY.

207—26

-SIWALIK AND NARBADA PROBOSCIDIA.

we come to the large specimen represented in fig. 4 of Plate XXXTII. This tooth
is implanted in a fragment of the right ramus of the mandible, represented
on a smaller scale in fig. 1 of the same plate ; in front of the figured tooth there
is in the jaw a fragment of another much worn tooth, which is too imperfect for
description ; the jaw was collected by Mr. Theobald in the Punjab. The figured
specimen carries three stout transverse ridges and a very large hind-talon, and is
slightly narrower in front than behind. Prom its great size, and from the number
of its ridges, it is quite evident that this tooth must be the second lower true molar,
and that it consequently corresponds in serial position to the upper molar repre-
sented in fig. 1 of Plate XXXII ; the two teeth, however, belong to opposite sides.
The lower molar is considerably more worn down than the upper, since in the
former large dentine islets are exposed on the third ridge, whereas in the latter, the
enamel of that ridge is not perforated.

The lower molar is considerably the larger of the two, and probably belonged to a
male animal ; the two teeth, however, have essentially the same form and structure,
and it will be needless to fully describe the lower tooth, and I will, therefore, content
myself witli pointing out how this tooth agrees with the upper tooth.

Both teeth, it will be observed, carry a very large hind-talon, which is divided
into two columns, but less distinctly in the lower than in the upper molar. In the
lower molar the accessory columns in the transverse valleys, on the outer side of the
median longitudinal cleft, are somewhat more worn down than the corresponding
columns, on the inner side of the same cleft, in the upper toot h, and the valleys
are apparently blocked to a somewhat greater extent than in that tooth. The
lower molar agrees with the corresponding tooth in the upper jaw (Plate XXXII
hg. 1) in presenting trefoil-shaped islets of dentine exposed by detrition, in one
of the columns of each ridge, when well worn down. These trefoils are situated
on the inner columns of the upper, and on the outer columns of the lower molar.
At the outer extremity of each transverse ' valley in the lower molar, there occur
low blunt tubercles, the homologues of those situated at the opposite extremities of
the valleys of the upper molars. A small anterior talon occurs in the lower molar,
its exposed dentine islet being in direct connection with that of the first ridge ; it
is probable that a similar talon, which has been broken off, originally occurred in
the upper molar.

Mandible. — The fragment of the mandible in which the last described molar
is implanted (Plate XXXIII, fig. 1) comprizes the middle portion of the horizontal
ramus of the right side. This fragment is of great thickness and presents a nearly
circular transverse section below the last molar, the transverse diameter being
slightly larger than the vertical. Its inner surface is nearly flat, but its outer sur-
face is convex, both from above downwards and from before backwards, and bulges
out suddenly below the crown of the molars. The lower border is highly convex* ;
where broken off at the commencement of the symphysis it has only a depth of

This border iu the figure is too straight.

SIWALIK AND NARBADA PROBOSCIDIA. 27—208

about 4 inches, and is sub-circular in section. The dimensions of the ramus and
molar are as folloArs : —

Length of last molar

Width of ditto . . , ,

Height of hind-talon of ditto .

Depth of iaw at last molar

Ditto at commencement of symphysis

Width of ditto (greatest)

6-4

3-7

2-2

6'4

3-9

6-9

From the arcuated form of the inferior border of the mandible, we infer that the
mandible had a short symphysis^ because in Ilasfodons with a long symphysis
(AT. migustidens,^ M. longirostris,^ M. pandionis^), the portion of the mandible
beloAV the molars has a straight, or slightly sinuous, inferior border, but never an
arcuated one. The same inference is draA^n from the sub-circular section of the
mandible ; the Proboscidia with long mandibular symphyses {Dinotherium [except
D. sindieme~\, M. angustidem, M. longirostris^ M. pandionis) having always a
laterally compressed mandible. This lateral compression of the mandible is necessary
to bear the strain of the great symphysis, it being a well known principle in engi-
neering to have the larger diameter of your beams at right angles to the strain.
Again, if M. falconeri had a long symphysis, the cross section of the ramus of the
mandible at the symphysis, in place of being sub-circular, would have a thick base
and a thin outer wall superiorly.

General characters. — The characteristic points of the molars of M. falconeri may
be summarised as follows : “ intermediate molars with three transverse ridges and
large talons ; ridges running directly across the crown of the tooth, and divided into
inner and outer columns by an approximately mesial cleft; accessory columns,
situated on the inner side of this cleft in the upper, and on the outer side in the
lower molars, slightly blocking the middle of the transverse valleys; dentine islets
of one column in each ridge trefoil-shaped® ; blunt tubercles at the inner extremities
of the valleys in the upper, and at the outer in the lower molars ; no cement.” Tiie
ridge formula of this species, as far as it is known, follows the normal trilophodont
law, and may be written as follows ; —

Milk-molars. Molars.

l(P)4-2-l-3 3-k3-k4(?)

l(.P)-k2-|-3(?) 3-|-3-h4(.?)

From the absence of any trace of premolars in the young cranium noticed al)ove,
(in which the alveoli of the milk-molars have been cut into) it is almost certain that
these teeth were not developed in this species. We know nothing regarding the

‘ See “ Fauna Antiqua Sivalensis,” PI. XLIV, figs. 4, 16, 23 A.

2 Ihid., pi XLIV, fig. 5. “ Abhand. der k. k. Geol. Reich.,” Wien., vol. vii, pt. 4, 1877, pi. I.

3 F. A. S. pi. XLV, fig. 10.

See below.

^ In the tooth drawn in Plate XXXII, fig. 4, it will he observed that the exposed dentine surfaces are of an
irregular, and not of a trefoil-shape ; this is owing to the great extent to which the tooth has been worn away.

209—28

SIWALIK AND NARBADA PROBOSCIDIA.

form of the cranium. The mandible was sub-circular in section, and had a short
symphysis.

In my first notice of the molars of this species in the “ Records,” it was stated
that cement was present in these teeth ; this I now find to be an error, into which
I was led by the exfoliation of the enamel in those specimens.

Specific distinctness. — Our next point is to institute a comparison between the
molars of Mastodon falconeri and those of other species of the genus, in order
that we may he assured of its distinctness as a species. In this comparison, we
may at once dismiss all species belonging to the section Tetralophodon, in which
the “ intermediate” molars never carry less than four transverse ridges, and wq
may accordingly confine ourselves to the trilophodont section of the genus.
Of that section the following list gives the names of all the described species
(exclusive of the present) with which I am acquainted, > ; —

1. M. andium (Cuvier) — N. America.

2. 31. angustidens (Cuvier) — Europe.

3. M. horsoni (J. Hays) — Europe.

4. 31. humioldtii (Cuvier) — S. America.

5. 31. maximus (Cuvier) — America.

6. M. ohscurus (Leidy) — N. America. ^

7. 3/1. pandionis (Falconer) — India.

8. M. pentelici (Lartet and Gaudry) — Europe.

9. 31. productus (Cope) — N. America.

10. M. pgrenaicus (Lartet) — Europe.

11. M. tapiroides (Cuvier) — Europe.

(M. turicensis, Schinz).

12. 31. virgatidens (Meyer) — Europe.

Commencing with Mastodon angustidens^ we shaU find in the Fauna Antiqua
Sivalensis,”^ figures of the third upper milk-molar and of the second upper true
molar. These teeth correspond with the two molars of Jf. falconeri represented in
figs. 3 and 1 of Plate XXXII of the present memoir. In the following table the
dimensions of these four teeth are arranged side by side, viz. : —

Length of third milk-molar
Width of ditto ditto
Length of second true molar
Width of ditto ditto

M. angustidem. M. faXconeri.

. 2-8 3-04

. 1'6 20

. 4-5 5-2

. 2-6 3-5

As far, therefore, as mere size goes, there is a very considerable difference be-
tween the molars of the two species, which appears to be pretty constant. The
molars of both species agree in having trefoil-shaped discs of wear on one column

^ Figures (of the natural size) of the molars of this species are given in Herr Vacek's Memoir on Mastodon.
“ Abhand der k. k. Geol. Reich Wien., vol. vii,pt. 4, 1877, plates IV, V.
s Plate XL, figs. 7—9.

SIWALIK AND NARBADA PROBOSCIDIA.

29-^-210

of each ridge, but these trefoils are nearly always less distinct in the European teeth,
and in some are very imperfect. In most of the teeth of M. angusticlens, the accessory
columns are more developed than in M. falconeri, and the valleys are consequently
blocked to a greater extent. The second true molar of the former species carries
a greater number of cusps on its ridges, and its talons are more complex than in
the corresponding tooth of the latter. Again, the true molars of the European
species generally diminish in width anteriorly to a greater extent than do those of
the Indian species.

Comparing the tooth represented in fig. 8 of Plate XL of the Eauna Anti-
qua Sivalensis,” described as being the first upper true molar of Mastodon angus-
tidens, but which is of the same size as the second true molar represented in the
next figure, with the second upper true molar of Jf. falconeri described above, we
find that the former tooth is somewhat longer than the latter, and is wider posteri-
orly and narrower anteriorly. The valleys in the European tooth are more blocked,
owing to the greater development of the accessory columns ; the latter being placed
on the line of the median cleft, instead of on one side of it, as in M. falconeri.
Again, the ridges of the molar of M. angustidens are considerably higher than are
those of M. falconeri : thus, the height of the last ridge in the so-called first lower
molar of the former species is 2-2 inches, whereas in the nearly equal-sized tooth
of the latter it is only 1*4 inches. In an undoubted first lower true molar of M'.
angustidens in the collection of the Indian Museum, which is less than 3 inches
in length, the height of the ridges is equal to the height of those of the corre-
sponding tooth of M. falconeri with a length of 4’1 inches.

A second upper true molar of Mastodon angustidens represented in fig. 3 of
Plate III of the second volume of the “ Palaeontological Memoirs,” makes a nearer
approach to the corresponding tooth of M. falconeri than does any other tooth that
I have seen. The ridges on that tooth, however, are higher, the talons less develop-
ed, the accessory columns larger, and placed nearer the median cleft, than in M.
falconeri. An excellent figure of what seems to be the first upper true molar of M.
angustidens is given by H. von Meyer L that specimen is a regularly oblong tooth,
and has the columns of the ridges more distinct, and a deeper mesial cleft than in
the last milk-molar of M. falconeri (Plate XXXII, fig. 3) : the European tooth
has a lateral cingulum continuous with the fore-and-aft talons, which is entirely
wanting in the Indian tooth ; in the latter the valleys are less open and more
sinuous than in the former.

The presence of a cingulum in the upper molars of M. angustidens seems from
von Meyer’s figures to be very constant. The specimen of an apparently second
upper true molar figured in the above quoted plate (fig. 7) of von Meyer’s, has an
almost complete cingulum, which is entirely wanting in the corresponding tooth of
M. falconeri (Plate XXII, fig. 1) : in the former tooth the valleys are less entirely
open than in the latter : other teeth figured by von Meyer exhibit corresponding
differences from those of M. falconeri, which the reader can examine for himself,

’ Palseontographica, Vol, XVII, Piate III, fig. 1, (the lowest tooth in the figure).

211—30

SIWALIK AND NARBADA PROBOSCIDIA.

Pine specimens of the associated penultimate and last lower true-molars of
M. angmtidens are figured by Herr. Vacek the penultimate molar there figured is
a much smaller, narrower, and simpler tooth than the corresponding tooth of M.
falconeri (Plate XXXIII, fig. 4) ; the valleys of the former are much more open,
the accessory columns and the dentine trefoils less distinct than in the Indian tooth ;
the Euroj)ean tooth has a very small hind-talon. The mandible of M. falconeri has a
sub- circular cross section, and a short symphysis, while that of M. angustidens ^ has
the symphysis produced into a long rostrum ; whether the former had tusks is not
known ; it seems almost certain, from the absence of any trace of premolars in the
young cranium of the former species described above, that those teeth are not
developed in 31. falconeri, as they are in M. angustidens.

M. horsoni. — After 31. angustidens no species of 3Iastodon comes very close to
31. falconeri in the form of its molars. The molars of 31. horsoni are at once dis-
tinguished from those of the Indian species by the almost complete absence of the
median longitudinal cleft, which in general is only visible in germ or quite unworn
specimens.^ Pigures of the teeth of this species are given in Messrs. Lortet and
Chantre’s monograph on the Mastodons of the Rhone basin.^

31. andium. — In 3Iastodon andium ® the dentine discs of wear of the columns
of the ridges are of nearly equal size, and do not form such distinct trefoils as in 31.
falconeri. The second lower molar in the former has no hind-talon, while the last
lower molar has five ridges, which we have no reason to suppose ^was the case with
31. falconeri.

31. humholdtii.- — Pine specimens of the associated second and third lower
true-molars of this species are figured by the late H. von Meyer.® The last tooth has
four ridges and a very large hind talon, which might almost be classed as a fifth
ridge. The penultimate lower molar has three ridges and a small hind talon : if the
figure of this tooth be compared with that of the corresponding tooth of M. falconeri
(Plate XXXIII, fig. 4), it wiU be seen that the European tooth irrespective of its
smaller size, differs from the Indian by its much simpler structure, and smaller hind-
talon ; the columns in the former are not sub-divided into smaller cones, the acces-
sory tubercles are small, and the valleys much more open than in the latter.

31. maximus. — In 3Iastodon maximus the ridges of the molars are relatively
higher and narrower than in Jf. falconeri ; the transverse valleys are completely
open, and the hind-talon is very small.

* “ Abhandl. der. k. k. Geol. Eeich.” Wien 18, Vol. VII, pt. 4, Plate IV, fig. 2.

^ “ Fauna Antiqua Sivalensis”, Plate XLIV, fig. 5. Abhandl. der. scbweiz, Pal. Gesell., Vol. Ill, Plate II, fig. 4.

® Falconer : “ Palaeontological Memoirs”, Vol. II, p. 18.

“ Mastodontes du Bassin du Ehone.” Arcbiv. du Mus. d’Hist. Nat. de Lyons, Vol. II, Plates XI, XII, XVI,
XVI bis. Lyons, 1878.

® M. andium was originally classed by Falconer as a Trilophodon, though subsequentlj as a Tetralophodon ; it
appears in the latter group on page 14 of the second volume of the “ Palaeontological Memoirs,” but a note is there
appended, saying that Dr. Falconer thought of once again placing it among the Trilojohodons, The species seems in
reality to have been one forming a link between the two sections of the genus.

® “ Palacontographica,” Vol. XVII, Plate VI.

SIWALIK AND NADBADA PEOBOSCIDIA.

31—212

M. obscurus. — In Mastodon ohscurtis ^ tlie lower molars are relatively much
narrower than are those of M. falconeri.

M. pandionis. Some of the molars of Mastodon pandionis are described in
the succeeding pages, where their distinctness from those of M. falconeri will be
pointed out.

Of M. pentelici ^ only the milk-molars appear to be known. The second
and third upper milk-molars of that species are considerably larger than the cor-
responding teeth of M. falconeri, and there seems to be a greater number of accessory
tubercles in the transverse valleys of the former teeth than in the latter. The second
lower milk-molar in the Athenian species has a relatively larger talon than the
corresponding Indian tooth ; the two columns of the hinder ridge of the former
tooth are moreover less closely connected than are those of the latter. Both these
species seem to have had the common character of being unprovided with pre-
molars.

M. productus. — The mandible of M. productus ® is laterally compressed, and has
a long tusked symphysis, and is, therefore, different from that of M. falconeri.
The molars are also of simpler structure, and the columns of the ridges wear into
irregular disks, and not into distinct trefoils of dentine.

M. pyrenaicus. — I cannot find any description of this species wliich seems to
be only known from a manuscript note of the late E. Lartet.

M. tapiroides ^ is sufficiently distinguished from the present species by the
fact that the transverse ridges of the molars are not distinctly bisected by a
median longitudinal cleft, as well as by the transverse valleys being entirely unin-
terrupted by accessory tubercles. The early teeth of the species are figured by
M. Gaudry® under the name of M. turicensis ; M. Gaudry also seems to include
M. horsoni under the same specific name, and there does not indeed apj)ear to be any
difference of specific value between the molars of the two. A fine series of the
molars of this species are figured in Plates VIII and IX of the above-quoted memoir
of Lortet and Chantre.

M. mrgatidens. — Figures of three of the molars of this species are given
in figures 1, 2, 4 of plate IV of the above quoted memoir of von Meyer. The teeth
are quite unlike those of M. falconeri, as the ridges and valleys extend without
interruption across the crown.

Since none of the Trilophodons with which I am acquainted agree in the
characters of the teeth with the above described Indian trilophodont teeth, there
seems to be no doubt as to the specific distinctness of the animal to which the latter
belonged.

1 Leidy : “Extinct Vertebrate Fauna of the Western Territories,” pi. 22.

2 Gaudry: “ Animaux Fossiles et Geologie de I’Attique,” Pis. XTII, XXIII.

2 Cope; “ Extinct Vertebrata of New Mexico,” United States Geographical Survey, west of 100th naeridian,
VoL IV, pt.II, P, 306, Pis. LXX— -LXXII.

Falconer : “ Palaeontological Memoirs” Vol I, p. 103

® Gaudry : loc. cit. Plate XXIV.

213—32

SIWALIK AND NAEBADA PEOBOSCIDIA.

Distribution. — Eemains of this species have hitherto only been procured from
the Siwaliks of the Punjab to the westward of the Jheluin river, and from the
Manchars (Siwaliks) of Sind. No species of Trilopliodon was known to Palconer
and Cautley at the time of publication of the “Pauna Antiqua Sivalensis,”
whose main collections were obtained far to the eastward of the Jhelum (Sutledge
and Ganges valleys), and from the vast series of mammalian remains collected
by those writers it may he fairly inferred that the present species does not occur
in the more easterly Siwaliks.

Species 2 : Mastodon pandionis, Palconer. Plates XXXIV-V-VI-VII, fig. 3.

Distory. — This species was named by Dr. Palconer on the evidence of two
molar teeth in the India House collection said to have been obtained from the Deccan.
No description of the species was apparently ever published by Dr. Palconer, but in
his synoptical table of Mastodons given on pages 14 and 15 of the second volume of
the “Palaeontological Memoirs,” the species is classed by Dr. Palconer among the
Trilopliodons, in the division characterized by having molars with “ colliculi obtusi
alternatim mammillati, valliculce interruptce.'^

Although no description of the molars of this species was ever published by
Dr. Palconer, it appears that that writer made a comparison of these molars with
certain of those of Mastodon angustidens in the collection of the late M. Edouard
Lartet, the results of which comparison were embodied in a manuscript note. This note
has been published on page 124 [et seq.) of the first volume of the “ Pala3ontological
Memoirs,” but unfortunately in such a manner as to give the impression that the
sjiecimens in M. Lartet’s collection belonged to the same species as the Indian
teeth, though they really belonged to M. angustidens.^ I have copied the note
below, and have interpolated in brackets the name of the latter species where it is
necessary to the proper comprehension of the text.

“No. 1. — The principal piece is a penultimate molar, upper jaw, left side; so
determined from comparison with a germ specimen from M. Lartet of an antepenul-
timate (molar of M. angustidens) . The crown of the tooth is perfectly entire, the
front ridge alone being a little touched by wear on the inner side ; but the fangs and
base of the tooth are broken off right across on a line with the termination of the
enamel shell. It exhibits three well defined ridges, with a thick strong front talon
and a hind-talon confluent with the last ridge. It is a true and unmistakeahle
Trilopliodon, the only one yet yielded by India, and very different in its crown
characters from all the Siwalik, Ava, or other fossil Mastodons of the East.

“ The general form of the crown resembles very strongly that of Trilopliodon
angustidens, the principal difference being that the ‘ col ’ of vallecular flanking
mammillae is stiU more developed than in that species. The crown is traversed, as

^ The footnote on page 124 of the “ Palseontological Memoirs ” by Dr. Murchison, only adds to the confusion.

SIWALIK AND NAEBADA PEOBOSCIDIA.

33—214

usual, by an indistinct longitudinal cleft along the axis, marking off an outer and
inner division. Each of the three ridges has the outer division simple, and composed
of a thick conical transverse mass, the summit of which is somewhat compressed
and indistinctly hi- or tri-lobed by corresponding longitudinal furrows. The inner
division is more massive and complex, each ridge throwing out from its anterior and
posterior surface — the former diagonally forwards, tlie latter diagonally backwards —
a ‘ col ’ of robust tubercles, which meet in a chevron form in the middle of the valley,
so that, when the inner division of the crown is regarded in plan apart from the
outer, it presents, in conjunction with the outlying tubercles, a series of zig-zags
closely resembling the letter W. The complexity of pattern is further increased by
the salient apex of the connecting ‘ col ’ being continued outwards towards the margin
in a single line of cylindrical mammillae, which completely obliterate the bottom of
the outer half of each valley ; while the inner half, corresponding with the re-enter-
ing angle between the large inner cones, forms a gorge which is entirely free from
tubercles. The anterior talon forms a subordinate ridgelet, which is thrown off in
the usual manner from the anterior portion of the inner cone, and is continued out-
wards towards the margin, with less inclination downwards than is usually the case
in the other allied species {M. angustklens.) It is composed of about four robust com-
pressed tubercles, which are separated from the anterior ridge by a well-marked chasm.
The posterior talon consists of a cluster of indistinct confluent tubercles thrown off
from the posterior part of the inner tubercle of the last ridge, and so adpressed to the
ridge that it does not yield the defined and separate appearance seen in the talon of
Trilophodon angustidens. After a little wear the posterior talon would be involved in
the disc of detrition of the last ridge.”

Dr. Ealconer then proceeds to compare a small two-ridged tooth of 31. pandi-
onis with a corresponding tooth of Jf. angustidens. I shall have occasion to refer
to this tooth again in the sequel.

Besides the two above-mentioned molars of 31. pandionis, there is a left an-
tepenultimate upper molar of the same species in the British Museum, repre-
sented on Plate XXXIV (figs. 6 and 7) of the first volume of the “ Palseontological
Memoirs,” which was found in Dr. Ealconer’s private collection, the locality whence
it was obtained being apparently unknown. Of this specimen I have obtained
a cast for the Indian Museum through the courtesy of the Trustees of the British
Museum.

A short note on some lower jaws of this species was joublished by myself in the
“Eecords.” ^

Lower jaws and molars. — Having thus sketched the history of the known molars
of 3Lastodon pandionis, I proceed to describe some lower jaws and molars of a tri-
lophodont 3iastodon collected by Mr. Theobald in the Siwaliks of the Punjab, and
which I have referred to that species from the general resemblance of these molars
to Ealconer’s original specimens.

' Eec. Geol. Surv, ludia, Vol. XII, p. 43.

215—84

SIWALIK AND NAEBADA PEOBOSCIDIA.

The first specimens we have to consider are the two lower molars represented
in Plate XXXIV. These two specimens are associated molars from the right ramus
of an incomplete mandible collected hy Mr. Theobald at Halwar ih the Punjab.
Since the smaller of the two teeth (fig. 1) carries three ridges and a hind-talon,
wliile the larger and succeeding tooth (fig. 2) carries four ridges and a double hind-
talon, or perhaps five ridges, it is evident that the two teeth are respectively the
second and third true molars of a trilophodont Mastodon. The smaller tooth is in
an intermediate state of wear, while the larger tooth has only its first and second
ridges slightly touched hy wear, the remaining three ridges having been still con-
cealed hy the gum at the death of the animal to which the jaw belonged. As I
have said, the smaller tooth consists of three complete transverse ridges and a large
hind-talon ; each of these ridges when not too much worn away, as is shown in the
third ridge, is divided into an inner and an outer portion by a median longitudinal
cleft. The summit of each of these two portions of the transverse ridges, when
worn down, exhibits irregularly shaped discs of dentine, which never assume a
trefoil figure, and by which character the molars of this species are at once distin-
guished from those of Mastodon falconeri. In the transverse valley between each
pair of ridges, there are two or more accessory tubercles or columns, wedged in
between the main ridges, completely blocking the valleys, almost as high up as
the summits of the ridges, as is shown in the larger and less worn tooth. The inner
and outer columns of each transverse ridge are placed somewhat alternately to one
another, the centre of the inner column of each ridge being slightly in advance of
the centre of the outer column. The hind-talon of the smaller tooth is of unusually
large size, and, like the true ridges, consists of an inner and an outer column,
separated by a median longitudinal cleft, and having two accessory columns in
the preceding transverse valley. The talon is distinguished from a true ridge by
being much narrower ; it is probable, however, that we have here (as in the suc-
ceeding tooth) a molar indicating the transition from a Trilophodon to a Tetra-
lophodon. In the two anterior ridges of the penultimate tooth the exposed dentine
discs of wear have united, hut there is still in both a large trace of the median
cleft.

The larger and nearly unworn tooth (fig. 2) shows the terminations of each
column more distinctly ; but these columns have precisely the same arrangement
as in the smaller tooth. There is an anterior talon consisting of four agglomerated
columns, and a large double hind-talon.' The columns of the transverse ridges are
of great vertical height, and at their summits curve towards the anterior extremity
of the tooth. The valleys contain a considerable quantity of cement, which extends
to within half an inch of the summits of the columns ; this cement appears to be
loosely attached, and to fall off as the teeth are worn down. It, however, exists in all
the specimens of unworn molars, which are contained in the collection of the

* Tlie anterior part of this talon might almost be countec as fifth true ridge, but it will be seen to correspond with
the hinder part of the smaller tooth, which is evidently a talon.

SIWALIK AND NADBADA PEOBOSCIDIA. 35—216

Indian Museum, and may therefore he considered as characteristic of the species.
The dimensions of the two figured teeth are as follows : —

Length of second true molar . 4-1

Width of ditto ditto 2-6

Length of third ditto . .7-4

Width of ditto

Height of ridges in ditto 2-9

Comparison. — If the description of these two molars he compared with fal-
coner’s description of the upper molar of Mastodon pandionis, and with the figure
of the unworn upper molar, represented in figs. 6 & 7 of Plate XXXIV, of the first
volume of the “ Palaeontological Memoirs,” it will he evident that aU these three
teeth have the same general characters, viz., very high and alternately arranged
columns in the transverse ridges, and the transverse valleys completely blocked hy
outlying columns from these ridges. There appears, indeed, from falconer’s descrip-
tion, to he a somewhat greater complexity in the arrangement of the outlying
columns in the upper molars than there is in the lower molars. AU the other charac-
ters of the teeth are, however, so exactly similar, that there is every prohahility of
theu having belonged to the same species, and I have accordingly referred them aU
to Mastodon pandionis. The molar figured in the “ Palseontological Memoirs,”
considered by Dr. falconer to he the first or antepenultimate true molar, has a
length of 4 and an extreme width of 2’5 inches ; these dimensions are very nearly
the same as those of the second or penultimate lower true molar from the Punjab ;
as wiU he seen subsequently, however, there is some considerable variation in the size
of the molars of this species.

Form of mandible. — The portion of the mandible to which the two figured
molars belong, comprehends the greater part of the horizontal ramus of either side ;
posteriorly each ramus is broken off at the hinder end of the alveolus of the last
molar, and anteriorly a short distance in advance of the penultimate molar. The
inner surface of the ramus is convex from above downwards, and the outer concave,
the inferior border being slightly sinuous : the angle included between the inferior
and alveolar borders is an acute one, the bounding lines receding from one another an-
teriorly, and the jaws consequently becoming deeper towards that extremity. The
whole ramus is much compressed from side to side, the vertical diameter being much
greater than the horizontal. The ramus is continued in the same line as the alveo-
lar border up to the point where it has been broken off in front ; at the fracture
the depth of the sides of the mandible is close on 8 inches ; the jaw presents in
section at this point a U-shape, the base of the JJ being thick, and the sides very
thin, and tapering to a sharp edge at their summits. The form of this transverse
section, together with the direction of the upper and lower borders of the ramus,
shows that the mandible when complete would have been produced into a long
trough-hke symphysis, and could not ha’ve had a short symphysis like that of the
living Indian elephant. We shall subsequently describe a specimen showing this

217—36

SIWALIK AND NALiBADA PEOBOSCIDIA.

trougli-like symphysis. Where there is a short symphysis in Proboscidians, the
alveolar border of the ramus descends suddenly in front of the anterior molar, as is
well shown in Plates XLIV — V of the “Fauna Antiqna Sivalensis,” and the
transverse section near the commencement of the symphysis is suh-circnlar in
each ramus, instead of having a thick base and a tall and slender side-wall. The
mandible of 31. pandionis presents two nervous foramina of unusually large size,
the one situated immediately below the anterior border of the penultimate molar,
and the other some 4 inches in advance, nearer to the inferior border of the ramus.
The great size of these two foramina indicates a very large inferior branch of the
fifth nerve, which would be necessary to supply the large amount of tissues in-
vesting a prolonged symphysial rostrum. The mandible has a vertical depth of
7’0 inches at the middle of the last molar, and of 7 ’8 inches at the commencement,
of the symphysis, its thickness at the same point being 4 inches. If these dimen-
sions are compared with those of the mandible of 31. falconeri given above, it will
be seen that the form of the ramus of the mandible is alone an amply sufficient
distinction between the two species.

Last lower molar. — The next specimens I wish to bring to notice, are the right
lower molar, and the symphysis of a mandible, represented in fig. 4 of Plate XXXV
and in fig. 1 of Plate XXXVI. The molar is implanted in a fragment of the right
ramus of a mandible, which is a part of the same specimen as the symphysis. When
first obtained in the Punjab, Mr. Theobald tells me that the whole mandible was
complete ; the specimen was, however, in an extremely friable condition, and when it
arrived in Caleutta, the only j)ortions which were not reduced to fragments were a
part of the right ramus and the figured symphysis.

The figured tooth, which is about one-third worn down, carries four ridges and a
large hind-talon. If this tooth be compared with the last lower molar represented
in fig. 2 of Plate XXXIV, it will be seen (making allowance, of course, for the one
tooth being well worn and the other a germ) that the two molars are similar in
form, and evidently belong to the same species, both being in fact last right lower
molars. It will not be necessary to describe the well-worn tooth, since I have mainly
figured it in order to prove that the symphysis of the mandible associated with it really
belonged to 31astodon pandionis.

The worn last molar is slightly larger than the corresponding unworn tooth,
and the jaw in which the former is implanted is also correspondingly larger than the
jaw which carries the unworn tooth ; both jaws have, however, preeisely the same
form. In front of the worn last true molar, there remains in the jaw the fragment-
ary base of a smaller tooth, which from its size seems to have carried three ridges,
and which would therefore correspond to the penultimate molar, represented in
fig. 1 of Plate XXXIV. The length of the last molar is 8’4 inches, and its width
3’3 inches.

Symphysis of mandible. — The symphysis of the mandible, Plate XXXVI, fig. 1,
consists of a long spout-like rostrum, the figured fragment being some 20 inches in

SIWALIK AND NARBADA PROBOSCIDIA.

37—218

length. The medial line of this rostrum is deeply hollowed out posteriorly, the hollow
gradually diminishing anteriorly ; its depth posteriorly, where the specimen is broken
off, is upwards of 7 inches, the lateral walls are thin, and nearly vertical externally.
There are no traces of tusks in this specimen. There must have been an interval of
at least a foot between the proximal extremity of this symphysis and the fragment
of the ramus of the mandible where broken off in front of the penultimate
molar.

Tushed symphysis. — We have now shown that Mastodon pandionis had a pro-
duced symphysial rostrum, which in some individuals was tuskkss ; this leads us to
the consideration of another mandibular symphysis (Plate XXXVI, fig. 2), which
from its form evidently belonged to a Mastodon^ and I think most probably to M.
pandionis. This specimen was, like the other, obtained by Mr. Theobald in the Punjab.
It consists of the distal extremity of the symphysis, the fragment being about one
foot in length. It is furnished with a pair of large and laterally compressed tusks,
of which the middle portions alone remain. These tusks are strongly arcuated, their
superior border being concave and their inferior convex : they consist entirely
of ivory, and present no trace of any band of enamel, as seems to be uni.
versally the case in the genus. Their transverse diameter is 2 inches, and the
vertical 3'1 inches. The rostrum itself is concave superiorly and convex inferiorly
in the antero-posterior direction. The anterior portion of the superior surface
presents a lozenge-shaped hollow, externally to which this surface slopes away to
join the lateral surfaces. In this specimen there is no trace of the deep longitudinal
groove which occurs in the tuskless symphysis represented in fig. 1 of the same
plate. Such a groove could not, however, exist between the large tusks of the
second symphysis, and if, as I think is most probably the case, the latter belonged
to the same species as the former, it would seem that the presence of the large tusks,
in the one and their absence in the other are the cause of the differences in form of
the two specimens. The symphysis of the mandible is known in Mastodon latidens,
M. perimensis, and M. sicalejisis, as we shall see when we come to the description of
those species, and the tusked symphysis cannot, therefore, belong to either of them,
all of which have much shorter symphyses. The complete mandible of 31. falconer i
is unknown, but the specimen already referred to (Plate XXXIII, fig. 1) proves
that this species had a short and probably edentulous symphysis. If, therefore,
the tusked symphysis under consideration does not belong to M. pandionis, it must
belong to a new species of Siwalik Mastodon, which is somewhat improbable. On
the supposition that the symphysis in question belongs to M. pandionis, it must
have belonged in all probability to a male individual, while the tuskless specimen
represented in the other figure of the same plate must have probably belonged to a
female.

Last upper true molar. — A last left upper true molar of a trilophodont Mastodon
in the Indian Museum seems to belong to the present speeies. The specimen was
collected by Mr, Theobald in the Punjab ; but is not in a good condition for fi guring

219—38

SIWALIK AND NARBADA PROBOSCIDIA.

owing to tlie sandy matrix having become mingled with the enamel. The tooth is in
germ, and has a convex crown surface, carrying four transverse ridges, and fore-and-
aft talons. The anterior talon consists of a semi-circular ledge attached to the first
ridge. The true ridges and the hind- talon correspond exactly in form with those
of the corresponding lower molar represented in fig. 2 of Plate XXXIV : they
are tall, placed somewhat alternately, and divided by a deep longitudinal cleft. The
accessory tubercles appear to he arransred as in Falconer’s specimens of the earlier
upper molars ; the valleys are completely blocked by outlying tubercles, and also
contain a comparatively large quantity of cement. The length of the specimen is
7*7 inches, its greatest width 3*8 inches, and the height of the second ridge 3*3
inches.

First loioer true molar. — The imperfect small tooth represented in fig. 1 of Plate
XXXV is implanted in a fragment of the right ramus of a mandible, and, like
the preceding specimens, was collected by Mr. Theobald in the Siwaliks of the
Punjab. Prom its small size, this tooth must evidently he either the last milk-molar,
or the first true molar, and from carrying only three transverse ridges, it must
evidently belong to a trilophodont Mastodon. As the tooth is only slightly smaller
than the second lower true molar represented in fig. 1 of Plate XXXIV, it is
probable that it is the first true molar. The tooth is entirely unworn, and has lost
the inner column of the first ridge. The posterior extremity of the specimen is
slightly narrower than the anterior ; and there seems to have been a very small
mterior talon, judging from the fragment remaining on the outer column, and there
is a very distinct hind-talon, of a much lower elevation than the third ridge. The
ridges are all of great relative height, and are divided mesially by a cleft along the
antero-posterior axis of the tooth ; each ridge is slightly convex anteriorly, and con-
cave posteriorly. Prom either side of the outer column of each ridge accessory
columns or tubercles project into the transverse valleys, completely blocking them^
nearly up to the summits of the ridges. Cement occurs in some quantity in the
valleys. The length of this tooth is 4*2 inches, its greatest width 2*3 inches, and
the height of the outer column of the first ridge 2 inches.

In being furnished with cement, as well as in the great relative height of the
transverse ridges, this tooth agrees with the penultimate lower molar of M. pandionis
represented in fig. 1 of Plate XXXIV. The accessory tubercles or columns are,
however, less complex, and the ridges are placed less alternately than in the last
lower molar represented in fig. 2 of the last-named plate. Now, we have already
seen that the accessory columns are less developed in the second lower molar than in
the upper molars, and it appears from the present specimen that these tubercles
are still less developed in the earlier lower molars.

The present specimen is readily distinguished from the corresponding tooth of
the other Indian Trilopliodon [M. falconeri, Plate XXXIII, fig. 3) by its general
shape, by the ridges being much higher (2 inches in place of 1*4 inches), by the
valleys being blocked to a greater extent, and by the presence of cement.

SIWALIK AND NARBADA PROBOSCIDIA.

39—220

Third lower milk-molar. — Of the third lower milk-molar there is a damaged
specimen in the Indian Museum from Perim Island. This tooth carries three trans-
verse ridges, and small fore-and-aft talons ; it is narrower in front than behind, and
in aU essential characters it agrees with the preceding specimen. Its length is 3'4
inches, and its greatest width 2-3 inches.

Second tipper milk-molar. — The specimen represented in fig. 3 of Plate
XXXV is evidently from its size a milk-molar or a premolar, while from its shape
it as evidently belongs to the upper jaw. It carries two transverse ridges, a very
small anterior talon, and a large posterior talon. Prom its carrying only two com-
plete transverse ridges, it might he either the first milk-molar of a tetralophodont,
or the second milk-molar of a trilophodont Mastodon. Now, from its size, the
specimen is too large to he a first milk-molar, and must consequently he the
second milk-molar of a Trilophodon, if it be not a premolar. Of the trilophodont
Mastodon falconeri, the second upper milk-molar is known (Plate XXXII,
fig. 2), and there, therefore, only remains M. pandionis to which the specimen can
belong ; as we shall see immediately, its characters agree sufficiently well wdth those
of the other teeth of that species to make it probable that the specimen is rightly
referred to it.

The specimen was collected by Mr. Theobald in the Siwaliks of the Punjab,
and is entirely untouched by wear, never having been protruded from the gum : we
may, therefore, infer that the animal to which it belonged must have died in
early infancy. The crown is considerably narrower in front than behind, and the
ridges borne on it are of unusual relative height, the second being rather lower
than the first. The latter, which is divided like the former, into an outer and an
inner portion by the median longitudinal cleft which traverses the crown, has two
large tubercles at either extremity : between these are two smaller ones — one on
either side of the median cleft. The second ridge has tw^'o tubercles externally, and
a mass of four small agglomerated tubercles internally. The larger size of this
second mass of tubercles shows that this is the inner side of the tooth, which will
consequently belong to the left side of the jaw. The hind-talon has two distinct
tubercles externally, while internally it merges into an indistinct ledge on the second
ridge. The main transverse valley is completely blocked up by an accessory column
or tubercle, rising from the hinder side of the first ridge, and uniting at its base
with one of the tubercles on the opposite side of the second ridge. A considerable
quantity of cement occurs in this valley, at once distinguishing this tooth from
those of all other Siwalik Mastodons except M. perimensis. The length of the tooth
is 2’ 2 inches, its greatest width 1*8 inches, and the height of the first ridge
1*3 inches.

In the height of its ridges, their general form, as well as in the presence of
cement, and in the blocking of the transverse valley, this tooth agrees with the first
lower molar of Mastodon pandionis, and I accordingly provisionally refer it to the
same species.

221—40

SIWALIK AND NARBADA PROBOSCIDIA.

In fig. 1 of the fortieth plate of the “ Eanna Antiqna Sivalensis,” there is
figured a tooth-germ which, except in being shghtly smaller, agrees in all respects
with the present specimen, and which is probably the corresponding tooth of the
opposite side of the jaw. The description of the specimen as given by Ealconer is
as follows: “Eigs. 1 and la. — Mastodon latidens. Second (?) upper milk-molar
with two ridges, Length 1‘9 inches, width 1*4 inches.” Now, it is quite clear that
the specimen cannot belong to Mastodon latidens, as the corresponding tooth of that
species (Plate XXXVII, fig. 4) has three transverse ridges, which are very much
lower, and the valleys are open and without cement. I shall again have occasion to
refer to this question when I come to consider the second upper milk-molar of the
latter species.

The present tooth has a considerable general resemblance to the corresponding
tooth of Mastodon angustidens ^ (with which species M. pandionis seems to be closely
allied), but is distinguished by the greater extent to which the transverse valley is
blocked, and by the smaller size of the hind talon as well as by the generally larger
size of the tooth.

Last lower premolar and second loioer milk-molar. — The small specimen re-
presented in fig. 2 of Plate XXXY is evidently either the second (penultimate)
lower milk-molar, or the last lower premolar of a trilophodont Mastodon ; it is
similar to a tooth formerly in the j)ossession of Ealconer, and by him classed as the
last lower premolar of M. pamdionis, and is also very similar to the tooth generally
classed as the last lower premolar of M. angustidens, from, which I think thnt
Ealconer was probably right in referring his specimen to 31. pandionis. As both the
second milk-molar and the last premolar of Trilophodons have only two ridges it is
extremely difficult to assign either of these teeth, when found isolated, to its proper
serial position: I shall, however, show why I class the figured tooth as the latter. This
tooth belongs to the left ramus of the mandible ; its crown is elongated antero-
posteriorly and is somewhat narrower in front than behind. This specimen, together
with a slightly smaller but similarly shaped tooth, was obtained by Mr. W. T. Blanford
in Sind. The smaller specimen shows very distinctly small discs of pressure on both
the anterior and posterior surfaces, proving that the tooth when in position had
another tooth on either side of it in close apposition : the figured tooth has only a
disc of pressure posteriorly, showing that the preceding tooth must have dropped
out at an early period. As the two teeth are precisely similar there cannot be the
least doubt but that they are homologous.

This tooth, as we have seen, carries two complete transverse ridges, and small
fore-and-aft talons, a wide and nearly open valley separating the two ridges. Both
ridges are considerably worn, the plane of wear of the first being nearly horizontal,
while that of the second is directed obhquely forwards. In the centre of the valley,
between the two ridges, there is a small circular disc of dentine exposed, which is
formed by the abrasion of one of the ridges of the corresponding upjoer tooth

' Gaudrj’ : “ Les Euchainements du Monde Animal,” Mammiferes Tertiaires, p. 180, fig. 239.

SIWALIK AND NARBADA PBOBOSCIDIA.

41—222

against the summit of a tubercle situated in this valleyd The first ridge is short,
and has only a very faint trace of a median division ; the second ridge is longer,
and is distinctly divided into an outer and an inner column by a median cleft ; each
column of this ridge appears to have been capped by tubercles. The inner one of
these two columns is placed somewhat in advance of the outer one. The worn
tubercle in the transverse valley shows that the latter was somewhat blocked at its
base ; the transverse ridges are of considerable height.

The length of the tooth is 2’05 inches ; its greatest width 1’5 inches ; and the
height of tlie worn front ridge 1 inch.

This tooth is of simpler construction than any of the other molars of Mastodon
pandionis ; we have, however, seen that in that species the lower molars are less
complex than the upper, and the anterior lower teetli less complex than the succeed-
ing ones. In describing the known teeth of Mastodon pandionis in the above quoted
note in the “ Palseontological Memoirs,”^ Dr. Palconer describes a small tooth of
that species in the following words : — “ Among the teeth presented (to the India
House) by Colonel Sykes, there is also a small two-ridged Mastodon tooth, with very
smooth enamel, which in form, through every detail, agrees so exactly with a speci-
men (of a lower premolar of Mastodon angustidens) of Lartet’s, that I unhesitatingly
consider them to be homologous teeth of the same side and of nearly the same age.

Dimensions of premolar. Lartet’s. Indian.

Extreme length I'/S 1’9

Width of front ridge . . I'l 1’15

Ditto of back ridge 1-4 1’4

“ Lartet’s specimen is a detached tooth, labelled in his list, ‘Last lower
molar, left side.’ It consists of two ridges, both of them worn ; but the outer and
inner discs not continuous, and the middle of the valley occupied by a tubercle,
which is worn low down, leaving a circular disc. There, is a well-marked anterior
talon of two worn tubercles, but deeply impressed by an anterior disc of pressure
against a preceding tooth in position. Behind there is also a talon, but very
strongly impressed by a disc of pressure, so that the posterior talon only exists on
the inner side. The crown presents the cucumber shape, so remarked on by De
Blainville, «.<?., the first ridge is narrow (short), the second broad (long).

“ The Indian fossil shows precisely the same characters, i.e., two ridges worn,
and two talons, with a connecting tubercle between the ridges. The back talon is
marked with a disc of pressure, but the anterior talon consists of two confluent
prominent tubercles, free from any mark, showing either that there was no penulti-
mate premolar, or that it was very caducous and dropped out without pressure from
liehind. The anterior ridge is narrow (short) ; the posterior broad (long), as in
Lartet’s (specimen) ; but the ridges are more worn, and the discs confluent. The
crown slopes from the inside, which is higher, to the outside, which is lower, but less

* This abraded spot is scarcely perceptible in tbe figure,

2 Vol. I, p. 125.

223— 42

SIWALIK AND NAEBADA PBOBOSCIDIA.

so than in Lartet’s (specimen). The intermediate tubercle is worn down as in his
(specimen), and the posterior talon is only exhibited free on the inner side.

“ The specimens are so exactly alike that they might have been taken for the
same species, hut that the Indian is a little larger.”^

Now, a perusal of the foregoing note, will, I think, render it perfectly clear to
the reader that the Indian tooth described therein is the corresponding tooth to our
Sind specimen, and that consequently both teeth belong to the same species, which
is probably M. pcmdionis of Balconer. With regard to the serial j)osition of this
tooth we find that according to Balconer it agrees exactly with a tooth in the pos-
session of Ed. Lartet, and considered as the last lower premolar of M. angustidens.
It also agrees with- a similar tooth figured by von Meyer, in figures 12, 13 of plate
V of Vol. XVII of the “ Palaeontographica,” as the second lower premolar of that
species ; normally the number of premolars in Mastodons is two, and the last might
therefore well he called the second. Von Meyer has, however, figured a specimen
of a young u2Dper jaw of IL angustidens (plate III, fig. 1), which he describes as
containing two premolars, which he calls first and second, and a third tooth which
he calls the third milk-molar. There can he no question hut that the two first
teeth in that sjDecimen are premolars, since they are less worn than the thuxl tooth.
The posterior of the two premolars is a complex two-ridged tooth very like the
second upper milk-molar of M. falconeri ; according to von Meyer’s classification
this tooth must have succeeded a tooth (the second milk-molar) no more complex
than itself, and a third premolar would have appeared above the third tooth in the
jaw. As the second premolar in that jaw is as complex as the last premolar in the
tetralophodont M. latidens, I cannot hut think that the third tooth in von Meyer’s
jaw is really the first true molar, the jaw belonging to a very small animal. The
second premolar will then he the last, and not the penultimate. If this view
he admitted, the lower premolar, which he figures as the second or penultimate,
will he the last, and indeed this must he so, because in a cast of a lower jaw of a
young 31. angustidens in the Indian Museum, we find that the premolar succeeding
the last milk-molar is a two-ridged tooth, much like Meyer’s specimen, while the
preceding premolar has only a single column, and corresponds to the first (penul-
timate) premolar in Meyer’s upper jaw. In this lower jaw the first true molar is
no larger than the tooth which I consider the same in the uj)per jaw figured by
Meyer.

The tooth of M. angustidens, which is so like our specimen, is therefore the
last lower premolar, and I accordingly class the figured specimen as the correspond-
ing tooth of 31. pandionis.

I have, however, another reason why this tooth should he classed as the last
premolar, and not as the second milk-molar. The small s^oecimen represented in
figure 3 of plate XXXVII, obtained by Mr. Eedden in Sind, from its narrow and
elongated form, evidently belongs to the lower jaw, and seems to belong to a Trilo-^

* The italics and hrachets in the above note are my own.

SIWALIK AND NARBADA PROBOSCIDIA.

43—224

pJiodon, being either the second milk-molar or the last premolar. The tooth belongs
to the left ramus of the mandible, as is shown l)y the inner extremity of each of
the two transverse ridges (top border of figure) being considerably higher than the
outer ; the first ridge is slightly smaller than the second, and there are small fore-
and-aft-talons. The plane of detrition of the first ridge is directed obliquely back-
wards, and that of the second ridge obhquely forwards ; each ridge is mesially
bisected by a distinct antero-posterior cleft into an outer and an inner column.
The length of this tooth is 1’6 inches and its greatest width I'l inches.

With regard to the position of the tooth and the species to which it belongs, I
have but little doubt that the specimen belongs to a TrilopJwclon of the two
Indian Trilopliodons, the second lower milk-molar of M. falconeri is known, and
it is probable that there were no premolars developed in that species ; there only
remains, therefore, 31, pcmdionis.

Of two specimens of the lower young dentition of M. angustidens in the
Indian Museum, the one shows the last milk-molar, with the last premolar in
alveolo below it, and the other the second and third milk-molars. In both these
specimens the last milk-molar has a length of about 2^ inches ; the second milk-
molar is 1’3 inches in length and the last premolar 1‘8 inches. It is, therefore,
apparent that in jaws having the last milk-molar of the same size, the last premolar
is larger than the second milk-molar. On these grounds I have provisionally
classed the tooth represented in figure 2 of plate XXXV (closely resembling the
last lower premolar of M. angustidens) as the last lower premolar of 3£. pandio-
nis, and the smaller tooth represented in figure 3 of plate XXXVII (closely
resembling the second lower milk-molar of 31. angustidens) as the second lower
milk-molar of 31. pandionis}

The tooth classed as the second lower milk-molar of 31. pandionis is small as
compared with the second upper milk-molar of the same species figured in this
memoir (PL XXXV, fig. 3). Palconer’s specimen of the latter tooth referred to
above, however, is smaller and agrees well in this respect with the lower tooth.
The difference in size of the last lower premolars, shows the variations in size which
the teeth of this species may undergo.

General characters. — Having now described such of the remains of 31. pan-
dionis accessible to me, as are of the most importance in illustrating its dentition
and osteology, I proceed to summarize the general and distinctive characters of the
species, as far as they are at present known. We are unfortunately totally unac-
quainted with the cranium, and nothing, therefore, can be said regarding this
important part of the skeleton. The mandible is charaeterized by its compressed
rami, and by being produced into an enormous symphysial rostrum, which in some
individuals was tuskless, and in others (unless the symphysis figured in Plate
XXXVI, fig. 2, belongs to a new species) carried large and much curved incisors.

1 The specimen figured in figure 6 of plate V of the above-quoted memoir of Meyer as the second milk -molar of
Jlf. angustidens is a three-ridged tooth, and would therefore seem to he the third milk-molar.

225—44

SIWALIK AND NARBADA PROBOSCIDIA.

With regard to the dentition, the molars, in regard to the number of ridges borne
on their crowns, followed the normal trilophodont rule. The pattern of the molars
approaches nearest to that of the molars of M. angustidens, hut presents important
differences which will he pointed out below. If the above identifications are correct,
t he species was provided with premolars.

Distinctions and differences. — From Mastodon falconeri (the only other Indian
species with a trilophodont ridge formula to the molars), the present species is at once
distinguished by its compressed mandible with its long symphysial rostrum. It is
further distinguished by the more complex pattern of the crowns of the molars,
and by the irregularly shaped discs, in place of the perfect trefoils of dentine, pro-
duced by the detrition of the ridges. The molars of 31. pandionis are further,
distinguished by the greater absolute height of the transverse ridges, and by the
greater extent to which the intervening valleys are blocked by tubercles ; there is
also a greater tendency to an alternate arrangement of the columns of the ridges.
Cement is present in 31. pandionis and absent in M. falconeri.

3Iastodon angustidens. — It is imfortunate that in the mandible of M. pandionis
the jiortion connecting the alveolar part of the ramus with the sym23hysis is wanting,
so that we are unable to institute a comjDarison between the comjolete mandible of
31. angustidens ^ and M. pandionis. A very beautiful specimen of the extremity
of the mandible of M. angustidens, is figured by Dr. Biederman '3 that specimen
carries two small sub-cylindrical tusks, and is greatly contracted laterally near
the 2>roxima,l end of the symphysis, jn’oximately the rostrum has a deep trough,
which expands into a flattened plate anteriorly. Tliis symphysis is quite different
in form from that of the tuskless symjfliysis of M. pandionis flgured in figure 1 of
Plate XXXVI : it approaches nearer to the tusked specimen in figure 2 of the
same plate. There are, however, very considerable differences in the form of the
trough and tusks, which are somewhat difficult to point out, but which will be
very apparent if the two figures are compared together : the tusks in M. pandionis
are much curved, the concavity looking upwards, and very highly compressed, while
those of 31. angustidens are quite straight,^ and in the young individual, according to
Falconer,^ had a pyriform cross section. To the tuskless variety of M. pandionis,
there is no corresponding form in 31. angustidens, since in that species both sexes
were alike furnished with inferior tusks ; ® the peculiar trough-hke symphysis of the
tuskless variety of M. pandionis has, therefore, no representative in 31. angus-
tidens. The two species resemble one another in the compressed rami and long
symphysis of the mandible.

The molars of the Indian seem to be distinguished from those of the European
species by their somewhat more complex pattern, and by the surfaces of dentine

^ “ Fauna Antiqua Sivalensis,” PI. XLIV, fig. 5.

2 Abbandl. der. Schweiz. Pal. Gesell, Vol. Ill, pi. II, fig. 4 — (Biederman’s memoir),

^ Von. Meyer : “ Palaeontographica,” Vol. XVII, pi. VJII.

“ Palaeontological Memoirs,” Vol. II, p. 41.

6 Ihid., p. 13.

SIWALIK AND NARBADA PROBOSCIDIA.

45—226

on the worn crown surface presenting irregularly, or somewhat triangular, in place
of trefoil-shaped islets, as well as by the presence of large quantities of cement in
the valleys. The second and third lower true molars of Mastodon angustidens ^ are
much simpler teeth than the corresponding teeth of M. pandionis (Plates XXXIV,
XXXV, fig. 4). In the latter the valleys are completely blocked by outlying
tubercles whereas in the former they are tolerably open. There is a very small hind-
talon in the penultimate molar of M. angustidens, and a very large one in that of
M. pandlonis : in the last molar of the latter species the hind talon is developed into
a fifth ridge, whereas in M. angustidens there are generally only four ridges in the
last molar, with a very small hind-talon : M. pandionis in these respects is evidently
transitional between the Trilophodons and Tetralophodons. A penultimate lower
true molar of M. angustidens figured by von Meyer ^ has a larger hind-talon than
in the last mentioned specimen ; this talon, however, is in the form of an elongated
cingulum, and quite different from the two-coned talon of M. pandionis ; the whole
tooth of M. angustidens is of less complex structure than that of M. pandionis. That
the two species are closely allied is, however, evident from the general structure
of the molars and the mandible.

M- productus. — The mandible of the trilophodont M. productus of the Loup-
fork group of New Mexico^ has a produced tusked symphysis which has more
resemblance to the tusked symphysis drawn in fig. 2, of Plate XXXVI of this
memoir, and provisionally referred to M. pandionis. In M. productus (Plate LXX
of Cope’s memoir) the tusks are straight, and have their largest diameter in the
transverse, and not in the vertical direction, in which respect they differ from the
tusked mandible referred to M. pandionis. The molars of M. productus are smaller
and of simpler structure than those of M. pandionis, and the last molar has only
four ridges, without a hind-talon.

Mastodon pandionis does not approximate closely to any other described species
of Trilophodon in its dental and mandibular characters.

Mastodon siralensis. — Curiously enough the last lower molar of this species
has a most remarkable resemblance to the corresponding tooth of the tetralopho-
dont Mastodon sivalensis, as may he seen by comparing this tooth of M. pandionis,
represented in fig. 4 of Plate XXXV, with the corresponding tooth of the same
side of the jaw of M. sivalensis represented in fig. 3 of Plate XLIV. Both teeth
appear to have the same number of ridges, though in the former the last ridge
(from the characters afforded by the penultimate molar) is described as a talon, and
in the latter (from the same considerations) as a true ridge. Both teeth agree
in the general form of the ridges, of the columns, and of the accessory columns,
and also in the general shape of the dentine islets : the main difference between
the two teeth being that the molar of M. sivalensis is somewhat narrower than

’ “ Abhandl. der. k. k., Geol. Reick.” Wien 1877, Vol. VII, Pt. 4, Plate IV, fig. 2.

- “ Palseontographica”, Vol. XVII, pi. IV, fig. 6,

*Cope: “Extinct Yertebrata of New Mexico,” United States Geographical Survey, west of 100th meridian,
Vol. IV, pt. II, p. 306, Pis. LXX— LXXII.

227—46

SIWALIK AND NATLBADA PROBOSCIDIA.

the other, and that the columns are rather more perpendicular in M. pandionis.
The two teeth are, however, so alike that it appears to me to he extremely
doubtful if they could he distinguished as belonging to separate species, if they had
been found detached from the jaws. The tooth of M. pcmdionis, it will he remem-
bered, is implanted in the ramus of the mandible with the long spout-like rostrum
figured in Plate XXXVI, fig. 1, and as we find from other specimens, was preceded
by a three-ridged tooth. The tooth of M. sivalensis, on the other hand, belongs
to a nearly complete mandible with a short symphysis (described below), like that
of the living Indian elephant, and was preceded by a four- (or occasionally five-) ridged
tooth. There can, therefore, he no question as to the specific distinctness of the
animals to which the teeth belonged. Prom this we learn that certain of the teeth
of some species of Prohoscidia may be almost indistinguishable from those of
a totally distinct species, and from this it is not a very long step to the whole of the
teeth in two species being of the same character, though the crania and mandibles
are very distinct : this has an important bearing on the case of §tegodon insignis
and 8. ganesa, to be noticed below. I ought to add that the molar of M. siva-
lensis has no traces of the cement which occurs in that of M. pandionis.

Distribution. — Bemains of Mastodon pcmdionis in the collection of the Indian
Museum, have been obtained from the Siwahks of the Western Punjab, Sind, and
Perim Island : according to Dr. Falconer they have also been obtained from the
Deccan, whence molars were sent by the late Colonel Sykes. As no remains of this
species were obtained by Cautley and Falconer among their immense collections of
fossils from the more easterly Siwaliks, it seems probable, as in the case of the last
described species, that the geographical range of M. pandionis did not extend
much to the eastward of the Jhelum river. The second milk-molar figured in
the F. A. S. and referred above provisionally to this species, is an apparent excepi-
tion ; there is, however, no record of the locality whence that specimen came.

Section B. — Tetralophodon, Falconer.

Intermediate molars normally with four, but in some species occasionally with
five , transverse ridges.

Species 3 : Mastodon latidens. Clift, {in parte.) Plates XXXVII, XXXVIII,

& XXXIX.

History. — This species of tetralophodont Mastodon was originally named and
described by the late Mr. Clift upon the evidence of certain molars brought from
the valley of the upper Irawadi by the late Mr. Crawfurd early in the century.'
These fossils appear to have been the first winch were obtained by Europeans

* Trans. Geol. Soc. Lon., Ser. 2, Vol. II, p. 369,

SIWALIK AND NAEBADA PROBOSCIDIA. 47—228

from the mamnialiferous strata of the Indian region. With the teeth which rightly
belong to Mastodon, Air. Clift appears to have included, under the same specific
name, certain other molars with a higher ridge-formula, which really belonged to the
suh-genus Stegodon, subsequently formed by Dr. Dalconer. Tliis error was detected
and corrected by the late Dr. Palconer,^ who described and figured other similar
molars from the Indian Siwaliks ^ under the name of Mastodon latidens, at the
same time clearly defining the characters of the species and the genus. ^ The
distinctive characters of the molars of this species, as given by Dr. Dalconer, are that
the “ intermediate molars” each carry four transverse ridges, which are low and
blunt, and extend straight across the crown of the tooth. These ridges are sepa-
rated by transverse valleys, which are quite open, and uninterrupted by outlying
tubercles. Cement is virtually absent, or, at all events, is present in such small
quantities that it may be practically disregarded.

In Plate XXXI of the “ Pauna Antiqua Sivalensis” figures of more or less
perfect specimens of the following teeth of this species have been given, name-
ly, the third (?) upper milk-molar (fig. 2), the third upper milk and first true
molar (fig. 3), the incomplete first and second upper true molars (fig. 4), and
the last upper true molar (fig. 6). In Plate XL of the same work are given
figures of the following teeth, namely, a tooth doubtfully considered as the. second
upper milk-molar (fig. 1), which, as we shall see below, does not belong to this
species at all, the third upper milk-molar (fig. 2), and the first upper true molar
(fig. 3). Excluding, therefore, the misnamed tooth, the upper molars of Mas-
todon latidens, known to Ealconer and Cautley, were the third milk-molar and
the three true molars ; no complete specimen of the second true molar was, how-
ever, figured by them.

In the accompanying Plates XXXVII to XXXIX I have given figiwes of
the three milk-molars,^ of the second and third true molars, and of the penultimate
and last upper premolars, so that, with a few exceptions, the whole of the dentition
of this species is now known.

Distribution. — Before proceeding to describe the figured molars, I wish to say
a few words relating to the distribution of this species. In the synoptical view of
the species of Mastodon given by Dr. Ealconer,® the distribution of Mastodon latidens
is given as “ Ava and Southern India this statement appears clearly to be erro-
neous. The real distribution of the species is as follows ; — “ Ava,® Siwaliks from
Dehra-Dun to Punjab,’’ Sind,® and doubtfully from Perim Island.® T he species?

1 Pal. Mem., Vol. I, p. 59.

^ Fauna Antiqua Sivalensis, Pis. XXXI and XL.

® Pal. Mem., Vol. II, pp. 11, 14.

^ Except the first lower.

® Pal. Mem. Vol, II, p. 14.

® Clift : sujp. cit.

^ Specimens in Indian Museum.

® Falconer : Cat. of Vert. Foss. Mus. As. Soc. Bengal, p. 256. Ind. Mus.

^ Falconer : sup. cit., p. 206.

229—48 SIAYALIK AND NARBADA PROBOSCIDIA.

therefore, has a wide range in space, though it is extremely doubtful if it was ever
obtained from Southern India.

First upper milk-molar.- — The small tooth represented in fig. 5 of Plate
XXXVII is the first upper milk-molar of a tetralophodont Mastodon, as may at
once be seen by a comparison of the specimen with the palate of a young Mastodon
figured by the late Professor Kaup.^ The first milk-molars of Masto-
dons are so alike, that it is a matter of extreme difficulty to refer them to their res-
pective species. In the present instance there are, however, two points from the
evidence of which the tooth may with a tolerable degree of certainty be referred to
the present species.

The tooth belongs to the left side of the jaw, and is of a triangularly oval
shape, narrower in front than behind ; the crown bears two transverse ridges and
an anterior talon or cingulum. The hinder ridge is divided into two distinct
columns by a median cleft ; the anterior ridge consists of a single thick and blunt
cone. The enamel is thrown into a series of corrugations. The length of this tooth
is I' 2 inches, its greatest width I inch, and the height of the hinder ridge *6 inch.

In a note to Plate XL, fig. 1, of the “ Pauna Antiqua Sivalensis,” Dr. Falconer
notices the left upper jaw of a young M. latidens, containing the first and second
milk-molars. The first milk-molar has two ridges and a heel, and is oval-shaped. In
having a heel or talon, the tooth of tjiat specimen agrees with our figured specimen ;
the absence of a talon in another first upper milk-molar from the Siwaliks, which
I have referred to M. perimensis, shows that that tooth cannot belong to M. latidens.
The corrugated enamel is a character common to some of the later milk-molars of
M. latidens, and does not occur in either Ml. perimensis or M. sivalensis. On these
grounds I have referred the figured tooth to M. latidens.

In the Indian Museum there is another tooth ^similar to the figured specimen ;
both were obtained^from the Punjab.

First lower milk-molar. — I cannot find any tooth in the Indian Museum which
I can with any certainty consider as the first lower milk-molar of this species,
neither is there any specimen figured in the “ Fauna Antiqua Sivalensis.”

Second upper milk-molar. — The next tooth for consideration is the specimen
represented in fig. 4 of the last quoted plate, which was collected by Mr. Theobald in
the Western Punjab. This tooth has its crown narrower in front than behind, and
rudely oblong in form ; it carries three transverse ridges, and fore-and-aft talons,
the hinder one being somewhat broken. It presents a small disc of pressure ante-
riorly and a larger posteriorly, indicating the apposition of other contiguous teeth.
The enamel is thrown into vertical corrugations. The ridges and valleys are perfectly
straight and simple, the latter being entirely free from outlying tubercles ; the ridges
are very blunt and low, and are divided into outer and inner columns by a longitu-
dinal cleft ; the hindmost ridge is scarcely touched by wear. The middle ridge
forms the highest point of the crown, showing that the tooth belongs to the

' “ Ossements Fossiles du Museum de Darmstadt,” PI. xvi, figs. 1, la.

SIWALIK AND NA.EBADA PROBOSCIDIA.

49—230

upper jaw, while the state of attrition j)roves that it belongs to the right side. The
characters of this tooth are so essentially those of IKastodon latidens, that there can
he no hesitation in referring it to that species ; from the number of ridges carried
on the crown, the tooth must be the second or penultimate upper milk-molar. The
length of this tooth is 2-05 inches ; its greatest width 1-5 inches ; and the height
of the unworn ridge O' 7 inch.

This tooth being the second upper milk-molar of Jf. latidens, it is perfectly
evident that the tooth, represented in fig. 1 of Plate XL of the ‘‘Pauna Antiqua
Sivalensis, ” cannot belong to that species ; that tooth has a semi-oval crown, and
only carries two main transverse ridges, which are higher than in our specimen ; it
corresponds in form with the second upper milk-molar represented in Plate XXXY
fig. 3, of this memoir, which I have provisionally ref erred to M. pmidionis, to which
species Palconer’s specimen must likewise be referred. On page 51 of the Index
to the Plates of the “ Pauna Antiqna Sivalensis,” Dr. Palconer makes a note to the
effect that in the Museum of the Boyal College of Surgeons, there is “ the left
side of the upper jaw' of a young Mastodon latidens containing the first and
second milk-molars. The anterior tooth is about 1 inch long and *8 inch wide,
and has two ridges with a heel. The main ridge is transverse ; the anterior one is
an obtuse cusp. The tooth is oval, the sharp end being in front. The second
milk-molar is 2 inches long by about 1*5 inches wide. It has three main trans-
verse ridges and a small bourrelet ridge in front, and a heel ridge behind. It
expands very widely in the direction of the orbit. A vertical section shows some-
thing like the enamel of another small tooth half an inch long.” Prom this note
it will be clear that the specimen figured in the “ Pauna Antiqna Sivalensis ” as
the second upper milk-molar of Mastodon latidens could not be that tooth if the
specimen in the College of Surgeons were, because the one tooth had two ridges and
the other three. The latter tooth, judging from Palconer’s description, seems to
agree exactly with the second upper milk-molar of Mastodon latidens figured in this
memoir, and I have, therefore, no doubt but that the jaw in the College of Surgeons
is rightly referred to that species. The first milk-molar in that jaw seems, as I
have already said, to agree with the tooth referred above to the same serial position.

Second loioer milk-molar. — In fig. 2 of plate XXXVII we have the corre-
sponding lower tooth to the last specimen. This tooth was likewise obtained by
Mr. Theobald in the Punjab. Its crown surface is concave superiorly, and as the
ridges are most worn on the lower border of the figure, the tooth must have
belonged to the right ramus of the mandible, while the presence of three transverse
ridges and the small size of the crown prove the tooth to be the second, or penulti-
mate milk-molar. The anterior part of the tooth has been somewhat damaged,
the fore-talon being in consequence broken away. The posterior talon (left side of
figure) is complete, and the two columns of which it is composed have their sum-
mits slightly touched by attrition ; the three transverse ridges are all well worn.
The tooth is narrower in front than behind ; the ridges and vaUeys are low and

231—50

SIWALIK AND NAEBADA PEOEOSOIDIA.

shallow I each ridge is slightly convex anteriorly, and there is also a slight ten-
dency to a blocking of the transverse valleys by incipient accessory tubercles
bulging out at the centres of the transverse ridges. This curvature of the trans-
verse ridges and consequent partial obstruction of the valleys seems to be a charac-
ter which distinguishes the lower from the upper molars of Mastodon latidens. This
character is well exhibited in the specimen under consideration, in the succeeding
lower milk-molar represented in fig. 1 of the same plate, in the lower true molar
represented in fig. 8 of Plate XXXI of the “Pauna Antiqua Sivalensis,” and
in all the lower molars of this species contained in the Indian Museum. The
enamel is quite smooth. The length of the second lower milk-molar (allowing a
quarter of an inch for the broken talon) is 2-3 inches ; and its greatest width 1*7
inches. The tooth is, therefore, slightly larger than the second upper milk-molar :
a slight increase of size in the lower over the ujiper molars seems to be of very
general occurrence in the Mastodons.

Thwd upper milk-molar. — Continuing our serial description of the molars, the
next tooth we have to consider is the third or last upper milk-molar, of which a
specimen is represented in fig. 8 of Plate XXXVII. This tooth is the specimen
referred to on page 72 of the eleventh volume of the “Eecords of the Geological
Survey of India,” in connection with the presence of premolars in this species.
The tooth was obtained by Mr. Theobald in the North-Western Punjab, and is im-
planted in a fragment of the left maxilla ; above this tooth there was found in the
l^one the germ of a smaller premolar tooth (represented in fig. 6 of the same
plate), which would subsequently have displaced the larger tooth vertically, had
the animal lived long enough. The presence of this premolar above the larger and
more complex tooth proves that the latter belongs to the milk-molar series, and
the presence of four ridges on that tooth shows that it must be the last of that
series, or the first of the “ intermediate” molars of Dr. Palconer’s arrangement.
The tooth has an oblong crown, slightly narrower in front than behind : the three
first ridges are well worn, and their dentine disks have become partly united : the
last ridge is somewhat worn on the inner side, and very slightly on the outer,
while the posterior talon is almost intact. In profile the crown is strongly convex.
The anterior talon, judging from the width of the first dentine surface, must have
been small, while the posterior talon is of relatively large size and is double. The
transverse ridges are very low and blunt, and extend almost straight across the
tooth ; the valleys are open and entirely unobstructed by accessory tubercles. As
is seen in the slightly worn last ridge, there is a longitudinal cleft, placed slightly
on the outer side of the median line of the tooth, and dividing each ridge into an
inner and an outer column, of which the inner is considerably the larger. In respect
of this longitudinal cleft, and indeed in all characters (except of course size and the
number of ridges), the tooth agrees precisely with the smaller second upper milk-
molar represented in fig. 4 of the same plate. The enamel is corrugated. The
length of the last milk-molar is 3*7 inches, and its greatest width 2’2 inches.

SIWALIK AND NARBADA PROBOSCIDIA.

51—232

In fig. 2 of Plate XL of the “ Pauna Antiqua Sivalensis,” there is repre-
sented a small upper tetralophodont tooth of Mastodon latidens, which is doubtfully
described as the third milk-molar. That tooth, though unfortunately represented on
a very small scale, seems to correspond in every particular with the tliird upper
milk-molar described above (in comparing these teeth, allowance must of course be
made for the fact that our specimen is much worn, while Palconer’s is an unworn
germ), and I have no doubt but that they are the corresponding teeth. Ealconer’s
specimen is, however, somewhat smaller than Mr. Theobald’s, which may very pro-
bably be accounted for by the one tooth having belonged to a female and the other
to a male individual. The length of Ealconer’s specimen is 3 inches, and its width
1-8 inches.

In figs. 3 and 3a of Plate XXXI of the “ Eauna Antiqua Sivalensis,” there
is represented a fragment of the right maxilla of a young individual of M. latidens
containing two teeth, which were doubtfully considered by Dr. Ealconer as being the
third milk-molar and the first true molar. The most anterior of these teeth (remem-
bering that the two specimens belong to opposite sides of tlie jaw) will be found to
agree very closely in all respects with the third upper milk-molar from the Punjab,
described above ; the only difference between the two being that the hind-talon is
rather less developed in Ealconer’s specimen. The length of the latter specimen is
3‘7 inches, and its width 2-6 inches. This length agrees precisely with that of our
Punjab specimen, but Ealconer’s specimen is rather the wider of the two. There
seems, however, to be no doubt but that the two teeth correspond in relative position.

Third loiver milk-molar. — Of the third lower milk-molar, a nearly perfect speci-
men is represented in fig. I of Plate XXXVII. This specimen was formerly in
the collection of the Asiatic Society of Bengal, and was from thence transferred to
the collection of the Indian Museum ; it is described by Dr. Ealconer at page 206 of
the Catalogue of the Eossil Vertebrata in the Museum of the Asiatic Society. It
is uncertain whence this tooth came, but Ava seems to be the most probable locality.
The tooth is implanted in a fragment of the right ramus of the mandible ; in front
of this tooth there is the remnant of a smaller tooth ; the jaw is very small and
narrow. In describing this specimen. Dr. Ealconer considered it as being probably
the first true molar, but the jaw is much too small to have carried three teeth in
front of this tooth, while the proportion of this tooth to the last upper milk-molar
represented in fig, 8 of Plate XXXVII is very much the same as exists between
the lower and upper second milk-molars represented in figs. 2 and 4. I, therefore,
think it probable that the specimen represented in fig. I is really the third
lower milk-molar of Mastodon latidens, and I shall accordingly so consider it. The
anterior ridge and talon are a good deal damaged, but the tooth when complete
carried four transverse ridges and a large hind-talon. The charaeters of the tooth
are so essentially those of the second lower milk-molar represented in fig. 2 of
the same plate that no further description is necessary. The length of the tooth is
4-1 inches, and its greatest width 2*2 inehes.

233—52

&IWALIK AND NARBADA PROBOSCIDIA.

First upper true molar.— -Oi the first upper true molar, we have no complete speci-
men in the collection of the Indian Museum. A tooth is, however, represented
in figs. Sand 3a of Plate XL of the ‘‘Pauna Antiqua Sivalensis,” considered by
Palconer to he the first true molar, and which I also think is so, as it is consi-
derably larger than the third milk-molar represented in figs. 2 and 2a of the
same plate of that work. The length of the first upper true molar is 4 inches, and
its width 2*3 inches. The length of this tooth, it will be observed, is not much in
excess of that of the third upper milk-molar from the Punjab described above, and
it is probable that Palconer’ s first true molar, like the last milk-molar figured on
the same plate, belonged to a female individual. The first true molar carries four
ridges and two talons, and agrees in all essential characters with the upper milk-
molars described above. We have already referred to the fragment of the maxilla
of M. latidens which is represented in figs. 3 and 3a of Plate XXXI of the
“Pauna Antiqua Sivalensis,” and have shown that the first of the two teeth con-
tained in that specimen is in all probability the last milk-molar. The succeeding
tooth in that sj)ecimen must consequently be the first true molar. This tooth is
quite different from the last-mentioned first true molar; it differs from that
tooth in being larger, and, in place of carrying only fom* transverse ridges, carries
five distinct transverse ridges and two talons ; the tooth narrows behind, and has
the general form of a last, in place of a first, true molar. The length of this tooth
is 5’6 inches, and its greatest width 2’9 inches. In regard to width, therefore, this
tooth has a relative size corresponding to that of the last upper milk-molar represented
in fig. 8 of Plate XXXVII of this memoir, and probably belonged to a male animal.
The additional ridge, and the great size of the hind-talon of the first true molar
represented in the “ Pauna Antiqua Sivalensis,” can only be looked upon as abnor-
malities. We shall see presently that a corresponding abnormality occasionally
shows itself in the succeeding molars of this species.

First loioer true molar. — Of the first lower true molar of Mastodon latidens
there is no complete specimen either in the collection of the Indian Museum or
among the teeth figured in the “ Pauna Antiqua Sivalensis.”

Second upper true molar. — The large molar represented in fig. 2 of
Plate XXXVIII belongs to a splendid specimen of the palate of Mastodon latidens
obtained from a Burmese Pagoda by Sir Arthur Phayre, and by him presented to the
Museum of the Indian Geological Survey (now Indian Museum) . This palate con-
tains the second true molar of either side (of which the right is figured), and the
first ridge of each of the last true molars. The molars in this palate, as well as in
another splendid specimen, with the corresponding teeth, in the Indian Museum,
collected by Mr. Theobald in the Punjab, converge anteriorly to a greater extent than
occurs in any other species of the genus, and present in this respect a similar relation
to that which v^e find prevailing in the Stegodons and true elephants. The figured
tooth is in a medium state of wear, the hind-talon being still scarcely touched.
There is a small anterior talon and a very large posterior talon,' the latter indeed

SIWALIK AND NARBADA PROBOSCIDIA.

63—234

almost assuming the proportions of a fifth ridge. Each ridge is very low and blunt
and closely approximated to the adjoining ridges. A longitudinal cleft, placed
slightly on the outer side of the median line of the tooth, divides each transverse
ridge into an inner and an outer column, each of which is again suh-divided into
smaller mammillae or cusps, of which there are generally two on the outer, and four
on the inner column. The transverse valleys are perfectly uninterrupted, except
at the inner extremities of the third and fourth, where there are low and blunt
tubercles. There are semi-distinct tubercles on the hinder side of the latter ridges.
The tooth agrees in all essential characters with the milk-molars already described.

In the first column of the following table are given the dimensions of the
palate to which the figured tooth belongs, while in the second column are given
those of the somewhat larger but very similar palate referred to above : —

Interval between outer surfaces of posterior extremeties of second molars .

10-0

ll-O

Ditto ditto of anterior

ditto

8-0

9-0

Ditto inner surfaces of posterior extremeties of second molars .

3-4

4-0

Ditto ditto of anterior

ditto

1-5

2‘4

Length of second molar

6-6

6-3

Greatest width of second molar

3-6

4-0

Width of first ridge of third molar .

4-1

4-3

The second upper molar in the larger palate differs f]’om the figured specimen
of the same tooth, in that its posterior talon is of very small size, and makes
no approach to a fifth ridge. In fig. I of Plate XXXVIII an imperfect second
upper molar of M. latidens is represented, which in the form of its ridges and valleys
wull he seen to correspond precisely with the second upper molar represented in
fig. 2 of the same plate. Both teeth are from the right side of the upper jaw,
and both are very evidently in the same state of detrition. The imperfect tooth,

wdiich was collected by Mr. Theobald in the Punjab, has lost a part of its first

ridge and anterior talon, and also a portion of the inner extremeties of the second
and third ridges. This tooth is of slightly smaller size than the complete specimen,
but is doubtless the corresponding tooth. In describing the perfect tooth repre-
sented in fig. 2, it was noticed that there might he some question in deciding whether
that tooth had five ridges, or four ridges and a large hind-talon. The imperfect
footh has, however, most distinctly five ridges, though the fifth is somewhat smaller
and lower than the others : at the hack of this fifth ridge there is a very small

hind-talon. Now, there can he no doubt but that this fifth ridge is the represent-

ative of the talon of the more perfect tooth (fig. 2), and we have, therefore,
established the important fact that in the second upper molar of this species there
is a complete transition from a foim-ridged to a five-ridged tooth.

Second lower true molar. — Of the second lower true molar there are several
specimens in the collection of the Indian Museum. In their general characters
these teeth are merely enlarged repetitions of the lower milk-molars represented in
figs. 1 and 2 of Plate XXXVII, and I have not, therefore, thought it necessary to
give figures of any of them. Some of these teeth carry four -transverse ridges

235—64

SIWALIK AND NAUBADA PEOBOSCIDIA.

with a large hind-talon, others five ridges, and others again five ridges and a large
hind-talon, so that these teeth present modifications, in this respect very similar
to those which we have found to prevail in the corresponding upper teeth. The
length of a tetralophodont second lower true molar in the Indian Museum is
6’ 8 inches, and its width 3’ 6 inches.

Third upper true molar. — The large tooth represented in Plate XXXIX is
a specimen of the last upper molar of the right side of M. latidens, collected
by Mr. A. B. Wynne near the village of Lehri, in the Punjab. This speci-
men is implanted in a fragment of the maxilla, which also contains the two
last ridges of the preceding or penultimate tooth. The figured tooth is entirely
unworn, and was still covered by the gum at the death of the animal. The crown
carries six transverse ridges, the hindmost of which is considerably smaller than
the others, and probably represents an ultra- developed talon, as we saw to be the
case in the penultimate molar represented in fig. 1 of Plate XXXVIII. The
tooth consequently belongs to a “ pentalophodont” type of dentition. The ridges
are low and simple, and with the exception of the first, are slightly convex an-
teriorly, and as concave posteriorly. Each ridge is divided by a longitudinal cleft
placed somewhat externally to the mesial antero-posterior axis of the tooth.
The internal moiety of each ridge, with the exception of the second and sixth,
bears three mammillEe or cusps, while the external moiety bears only two on each
ridge. The valleys are quite simple and uninterrupted. In this tooth, as also in
all the previously described specimens, there is no trace of cement. The length of
the specimen is 8’6 inches, and its greatest width 4-2 inches. The last upper molar
of this species represented in fig. 6 of Plate XXXI of the “ Eauna Antiqua
Sivalensis” has the normal five ridges of the tetralophodont type, and there is a
similar specimen in the Indian Museum. This tooth may, therefore, be either
penta or hexalophodont.

Third lower true molar. — I have not yet seen any complete specimen of the
third lower true molar of this species. An imperfect specimen, in the collection of
the Indian Museum, from the Punjab, has four ridges remaining, the first of which
has a width of 4T inches.

Tremolars. — In figs. 6, and 7 of Plate XXXVII are represented two small
teeth which are probably both premolars, and one of which (fig. 6) certainly belongs
to M. latidens. I cannot be sure whether the other specimen is rightly referred to
the present species, and I may add that it is often a matter of extreme difficulty
to refer these small isolated teeth to their respective species and serial position,
when we have a large number of species to deal with.

Last upper premolar. — With regard to the tooth represented in fig. 6, no
difficulty presents itself, since this tooth is a germ which was embedded in the
bone above the base of the last left upper milk-molar represented in fig. 8 of
the same plate, and which must consequently be the last upper premolar of the
same side. The crown of this germ has its angles rounded off, and carries two

SIWALIK AND NARBADA PROBOSCIDIA.

55--236

equal-sized transverse ridges, and a smaller row of cusps representing a third
ridge posteriorly. The two main ridges are divided by a median cleft, and are
low and straight, and the valleys are open and uninterrupted. The length of
this tooth is 2-2 inches, and its greatest width 1-8 inches. Two very similar,
but worn specimens, of the same tooth are contained in the collection of the
Indian Museum. This tooth is of great importance as proving the existence
of premolars in the species. Since this tooth is so much smaller than the tooth
which it succeeds (Plate XXXVII, fig. 8), it is quite clear that when protruded,
it could not have touched the teeth on either side of it, and probably stood quite
isolated, as the corresponding tooth of M. pei'imensis represented in Plate XL.

Penultimate upper premolar.— Tho, next specimen (Plate XXXVII, fig. 7) was
also obtained from the Punjab. This tooth has no disc of pressure on either side •
from its form it might be either the first upper milk-molar or the penultimate pre-
molar of a Tetralopliodon, or the second upper milk-molar or last premolar of a Tri-
lopJiodon. The absence of any disc of pressure in front shows that it cannot be the
second milk-molar of a Triloplwdon, while its large size and absence of posterior disc
of pressure shows that it cannot be the first milk-molar of a Tetralopliodon.
Among the Tetralophodons, M. latidens and 31. perimensis are known to have been
furnished with premolars, and it is, therefore, very probable that the tooth is the
penultimate upper premolar of one of those species. As the tooth has low ridges, I
have thought it not improbable that it belongs to the former species, though I cannot
be at all positive in this determination. The tooth is considerably worn down, and
has an irregularly oval-shaped crown, somewhat narrower in front than behind. It
carries two low and broad ridges, closely approximated externally, and a small hind-
talon. The side of the ridges on the lower border of the figure is the most worn
showing that this border is the inner side of the tooth, which must consequently
belong to the right side of the maxilla. The transverse valley is not very distinct,
which might lead to the belief that the tooth does not belong to 3T. latidens. Pre-
molars, however, not unfrequently vary somewhat from the type of the other teeth.
The length of the specimen is 1-6 inches, and its greatest width 1-4. Another very
similar specimen, also from the Punjab, is in the collection of the Indian Museum.

General characters, — Having now passed in review the molar series of Jf, lati-
dens, we may sum up what is known regarding the species. The adult cranium is
unfortunately quite unknown ; the palate is noticeable from the extent to which the
molars converge anteriorly. The mandible is known by a specimen of the greater
part of the right ramus, containing the two last molars, in the collection of the
Indian Museum. This mandible is very long and slender, and sub-circular in cross-
section, in the middle its vertical diameter at the penultimate molar being 5‘8 inches
and its transverse diameter 5'6. inches. The lower border is nearly straight up to the
symphysis, with a slight convexity in the middle. Prom the extremely small size and
circular section of the ramus where broken off at the commencement of the sym-
physis, the latter must have been short, and was probably unprovided with incisors.

237—56

SIWALIK AND NAEBADA PEOBOSCIDIA.

The ridge formula of the milk-molars is constant, and conforms to the normal
tetralophodont order ; in the true molars, however, there is not an unfrequent ten-
dency to assume a pentalophodont type, the hind- talon of many of these teeth not
unfrequently, partly or entirely, taking the form of a fifth ridge. In the following
ridge-formula these varieties are indicated : —

Milk-molars. True molars.

2 -f- 3 + 4 (4-5) + (4-5) + (5-6)

2 + 3 H- 4 4 -f (4-5) + 5

This tendency to the production of an additional ridge in the true molars of
M. latidens will he subsequently shown to he a character which it possesses in
common with M. sivalensis. This tendency to variation enables us easily to com-
prehend how the passage from the tetralophodont Mastodons to the hexalophodont
Stegodons {S. cliftii) was effected. A similar tendency to variation in the ridge
formula of various species of European Mastodons is noticed by M. Gaudry.' With
these very important variations in the molars, which, as in M. latidens, can be traced
gradually from an incipient to a perfect ridge, it seems impossible that any naturalist
can continue to maintain that there is at present no evidence of the passage of one
species of animal into another. If the very small hind-talon of the five-ridged
second upper molar of Mastodon latidens described above (Plate XXXVIII, fig. 1)
were to develope into a true ridge, as we have seen to be the case with the talon
in the four-ridged tooth {ihid. fig. 2), and were the median cleft to disappear
and a little cement to appear in the valleys, the hypothetical six-ridged tooth would
then be almost, if not quite, indistinguishable from the corresponding tooth of
Stegodon cliftii. Such considerations appear to me to make it very doubtful whether
the genus Stegodon should not be abolished and united with Mastodon. But here
a new difficulty presents itself, since the genus Stegodon at one end presents
elephantine characters in its molars, and at the other mastodontine. Thus S. cliftii
(including S. sinensis) has the inner side of the upper molars the lowest, as in the
Mastodons, while all the other Stegodons have the centre of each ridge the highest, as in
the Loxodons and true Elephants. On the other hand, all the Stegodons have the
elephantine character of having generally no median cleft to the molars, but this is
a character which is equally shared with Mastodon horsoni and some other species of
that genus. On the other hand again, the molars of Stegodon insignis are very close
in general structure to those of Loxodon planifrons, and are certainly nearer to those
of the true Elephants, than to those of the Mastodmis. Stegodon insignis, however,
cannot be generically separated from S. cliftii, and we are therefore led to the con-
clusion that if the latter species were united to the Mastodons, then the latter would
have to be classed in the same genus as the true Elephants. There is, indeed, no real
distinction between Elephants and Mastodons, although the species at either of the
chain {Mastodon maximus and Euel&plias primigenius) are widely separated from

Les Enctainements du Monde Animal.” Mammiferes Tertiaires, p. 181, Paris, 1878.

SIWALIK AND NAUBADA PDOBOSCIDIA.

57—238

each other. Premolars are common to some Mastodons and one Loxodon {L, platii-
frons) ; cement is found in the valleys of the molars of M. pandionis, M. humboldtii,
and M. perimensis, and we have already referred to the mesial cleft and plane of
wear of the molars in Mastodon and Stegodon. At present, it is true, there is no
Mastodon known which carries more than five ridges in the intermediate molars,
but Stegodon cUftii has but six, and is therefore merely the next step in the ascend-
ing scale towards the true Elephants. The only other point in the dentition of
Mastodons in which they differ entirely from Elephants, is in the presence of lower
incisors, but this is only a character possessed by some species of the genus. It
appears, therefore, that there is no definition by which all Mastodons can be dis-
tinguished from all Stegodons and the other Elephants ; and although it may be
desirable to retain these generic (or sub-generic) terms as convenient landmarks in
the evolutionary series from Mastodon to Elephant, I cannot but think that they
cannot be philosophically justified, and that all the known Proboscidians, with the
exception of DinotJierium, ought to be classed in one large genus. ^

The line of descent of the true elephants from the Dinotheria has evidently
gone through some form allied to a Trilophodon of the type of M. tnaximus, then
through a tetralophodont form like M, latidens, then through the MastodonMke
Stegodons, like S. cliftii and homhifrons, then the higher-ridged Stegodons, like S.
insignis, and finally, through the Loxodons. The trilophodont and tetralophodont
Mastodons, with alternately arranged columns, like M. angustidens and 31. sivalensis,
belong to a group which seems never to have been further modified in the direction
of a more complex ridge formula than a pentalophodont type in the true molars.

It will be noticed in the above descriptions that all the upper milk-molars of
M. latidens have corrugated enamel, while the lower have the enamel smooth and
polished. As I have at present only seen one specimen of each tooth (excepting the
first upper milk-molar), I am unable to say whether this is a constant peculiarity,
or whether, as is perhaps more probable, there be two varieties of the species, in one
of which the enamel is smooth and in the other polished.

Distinctions and differences. — Mastodon latidens does not seem to come close
to any other species of the genus, the nearest approach being made, in regard to the
characters of the molars by the European 31. longirostris. The molars of the latter
species, however, have much taller ridges, with a deeper median cleft, and with acces-
sory columns partly blocking the bases of the valleys : the series of molars on either
side of the palate in 31. longirostris do not iucline together anteriorly to tlie extent
they do in M. latidens. The two species are further widely distinguished by the
form of the mandible, that of the Indian species being nearly straight interiorly, sub-
circular in section, with a short and probably tuskless symphysis, v hile that of the

' On page 12 of a paper on the “ Extinct Animals of North America,” read before the Eoyal Institution on
March 10th, 1876, by Professor Plower, the author remarks that “ Mastodons, however, were, after all, very like
Elephants, only being distinguished by some peculiarities of the teeth ; and by means of intermediate species the two
forms pass so gradually into one another, that it is difficult to say, in the case of some species, with which they ought
most properly to be classed.” •

239—58

SIWALIK AND NARBADA PROBOSCIDIA.

European species is laterally compressed, has a long tusked symphysis, and its in-
ferior border bends suddenly upwards below the first molar. (E. A. S., PL XLV,
fig. 10).

Species 4 : Mastodon perimensis, Ealconer and Cautley. Pis. XL, XLI, figs. 1,

3, 4, XLII, XLIII.

History and specific characters. — The remains of a species of Mastodon obtained
from Perim Island were referred by the authors of the “ Eauna Antiqua Sivalensis ”
to a new species, under the name of Mastodon perimensis^ and figures of the cranium
and of some of the molars are given in Plates XXXI, XXXVIII, XXXIX, and XL
of that work, though no detailed description of the species was ever published.
Various notes, however, on the teeth of this species will be found scattered through
Dr. Ealconer’s papers, as collected in the “ Palaeontological Memoirs,” and figures
of two molars are given on Plate IX of the first volume of that work. In the
synopsis of the species of Mastodon and Elephant given on pages 14 and 15 of the
second volume of the “ Palaeontological Memoirs,” Mastodon perimensis is classed
among the Tetralophodons which have the columns of the transverse ridges arranged
alternately, and the intervening valleys blocked; while on page 12 of the same
volume, we find it stated by Dr. Ealconer that the molars of this species have a
considerable quantity of cement in the valleys. The figures of the molars in the
“ Eauna Antiqua Sivalensis ” show that they are characterized by the valleys being
mainly transverse, but blocked in tlie middle by accessory columns, while one of the
columns of each of the transverse ridges, when worn, presents a rudely trefoil-shaped
dentine surface. In the “ Records of the Geological Survey of India,”^ I have
mentioned the discovery of a mandible of this species in the Punjab, and have also
shown that the species was provided with upper premolars, and that the male was
provided with small cylindrical mandibular incisors. The teeth figured in the
“ Eauna Antiqua Sivalensis” are in most instances more or less imperfect specimens,
as is so very frequently the case with fossils from Perim Island, which have usually
been subjected to a long course of washing and rolling on the sea-beach. The teeth
of M. perimensis collected by Mr. Theobald in the Siwaliks of the Punjab, which
form the subject of the present notice, are in far more perfect condition. The milk-
molars of this species are unfortunately at present, with one exception, unknown,
but specimens have been obtained of the whole of the series of true molars, and
one specimen of the last upper premolar is known. In describing these teeth I shall
take them in the order of their serial succession,

T'irst upper milk-molar. — The small tooth represented in fig. 3 of Plate XLI
was obtained by Mr. Theobald in the Siwaliks of the Punjab. Erom its size and
shape, it is evidently the first upper milk-molar of a Mastodon, and from the dis-
tinctness of the two ridges, in all probability of a tetralophodont species ; from the

' Vol. XI, p. 71 ; Vol. XII, p. 45.

SIWALIK AND NARBADA PROBOSCIDIA.

59—240

(‘xtreme smoothness and j)olished state of the enamel, I have thought that the
specimen should very probably he referred to Mastodon perimensis, the teeth of
which present that character to a very noticeable extent. The tooth belongs to the
right side of the jaw, and is rudely oval in outline, the posterior border, however,
being nearly straight. The crown bears two ridges, the hindmost of which is acute
and wedge-shaped, and is partially divided by a slight cleft into two columns. The
anterior ridge forms a large blunt cone, the summit of which has been largely worn
away. A longitudinal ridge on the inner side of the tooth connects the fore-and-aft
ridges ; there is no trace of any talon. The absence of any talon, as well as the
shape of the ridges, at once distinguishes this tooth from the corresjionding milk-
molar of Mastodon latidens, which has already been described. The length of the
specimen is 1'3 inches, and its greatest width I'l inches.

Last upper premolar and first true molar. — The fragment of the left maxilla
of a Mastodon containing two teeth, which is represented in Plate XL, was also
obtained during the winter of IS'ZS by Mr. Theobald in the Siwaliks of the Punjab,
and has already been shortly alluded to at page 45 of the twelfth volume of the
“ Records.” The larger tooth carries four transverse ridges on the crown, and very
small fore-and-aft talons : the fourth ridge is smaller than either of the others, and
is considerably curved. The tooth has not been much used, and the rate of wear is
very unequal on the two sides of each ridge, the inner column showing a large
surface of dentine, while the enamel of the outer column is still imperforate. The
ridges are of considerable vertical height, and each is divided into two portions by
the cleft wdiich traverses the longitudinal axis of the tooth. The portion of each
ridge on the outer side of this cleft consists of two agglomerated tubercles ; while
the portion on the inner side consists of one large column or tubercle in the line of
the ridge, to the external extremity of which two smaller tubercles are closely
attached, projecting on either side into the valleys, and uniting with the corre-
sponding tubercles of the other ridges, thus obstructing the base of each valley.
The arrangement of the accessory tubercles causes the worn dentine surface of each
main column to form a trefoil-shaped islet, as is shown in the first ridges. Very
slight tubercles occur at the inner extremities of the transverse valleys. The
enamel is somewhat corrugated, and there are some slight traces of cement in the
valleys.

The anterior surface of the tooth presents a large flattened and smooth disc,
produced by the pressure of a preceding tooth, which must have been in apposition
with it.

The smaller tooth, which has a sub-circular crown, is entirely unworn, and had
not perforated the gum at the death of the animal. The summits of the crown of
this tooth are nearly on a level with the base of the crown of the larger tooth. The
smaller tooth is placed on the anterior side of the larger. The position and condition
of wear of the smaller tooth, show that it must be a premolar, and that it has vertically
succeeded a worn-out milk-molar, which produced the facet on the anterior surface

241—60

SIWALIK AND NilRBADA PROBOSCIDIA.

of the larger tooth. The relatively large size of the premolar shows that that tooth
must be the last of the premolar series, and consequently that the larger tooth must
he the first true molar.

The premolar has two ridges, which are considerably higher than those of the
corresponding tooth of Mastodon latidens described above (Plate XXXVII, fig. 6) ;
the ridges are crowned by distinct cusps. There are weU developed fore-and-aft
talons. The dimensions of the two teeth are as follows : —

Length of premolar . . . . ... . • . . 2-15

Width of ditto .......... 1’8

Height of first ridge of ditto ......... 1'3

Length of true molar ......... 4*2

Width of ditto .......... 2'4

Height of second ridge of ditto ........ 1‘7

Comparisons. — The first true molar in the specimen described above • differs
entirely from the molars of Mastodon latidens, which is the only Indian TetralopJio-
don which we have yet described, and it will also he shown to differ from the molars of
M. sivalensis which are described below. There, therefore, only remains M.perimen-
sis to which the specimen can belong, and we find that the characters of the tooth
agree precisely with those of the molars of that species noticed above, and I have
consequently referred the specimen to that species. In figs. 3 and 4 of Plate IX
of the first volume of the “ Palaeontological Memoirs,” a specimen of the same tooth
of M. perimensis from Perim Island is figured on a small scale, in profile and plan.
That tooth was named by Dr. Palconer, and is now in the collection of the Indian
Museum. It is rather less worn than the specimen figured here, but agrees exactly,
in all its characters, with the exception of being somewhat larger. The length
of the Perim Island specimen is 4‘5 inches, and its width 2-7. It is probable that
the latter belonged to a male, while tte Punjab specimen belonged to a female
individual.

Comparison with M. longirostris. — The molars of M. perimensis present, at first
sight, a very striking resemblance to the molars of the European M. longirostris,^
though a closer examination shows certain differences between the two. Wq have
already seen that in the upper molars of M. perimensis, only the inner columns of
the ridges carry accessory tubercles next the longitudinal cleft, and that conse-
quently these columns alone present during detrition a trefoil-shaped surface of
dentine, while the outer columns present an oblong surface. In the molars of Masto-
don longirostris, on the other hand, both the inner and outer columns of the ridges de-
velope accessory tubercles next the cleft, and both in consequence wear into trefoils, as
is shown in Kaup’s figures. The trefoil of the outer column, however, does not appear
till the inner is partly disappearing, owing to the lesser height of the accessory tuber-
cles of the former. In consequence of this a molar of M. longirostris, when slightly
worn, carries a trefoil dentine-surface on the inner ridge, and an oblong one on the

* See Kaup : “ Ossements Fossiles du Museum de Darmstadt,” PI. XVI.

SIWALIK AND NARBADA PROBOSCIDIA.

61—242

outer, and tliese at first sight very closely resemble those of the molars of M. peri-
mensis. The molars of the two species are further distinguished by the presence of
cement in those of the Indian, and its absence in those of the European species. The
general similarity, however, in the plan of the molars of the two species shows that
they must he closely related, though the differences noticed above, as well as the
dissimilarity in the mandible of the two species, which will he noticed below, affords
marked points of distinction between them.

Second upper true molar. — Of the second upper true molar of M. perimensis, a
specimen from the right side, is represented in profile and plan in figs. 5 and 6 of the
plate of the “ Palseontological Memoirs ” already referred to above ; that specimen
was obtained from Perim Island, and is now in the collection of the Indian Museum.
The figure in plan (6) is somewhat indistinct, and on a small scale, but it suffices
to show that the tooth is but little worn, and that it carries four ridges and a
large hind- talon (broken in the specimen), and that the inner columns of the ridges
wear into trefoils, and the outer into oblongs, while the valleys contain cement.
The length of the specimen, allowing slightly for the broken talon, is 5-0 inches,
and its width 3' 5 inches. The tooth, which is represented in fig. 4 of Plate XLI
of this memoir, was obtained by Mr, Theobald in the Punjab, and is from the right
side of the upper jaw. It carries four ridges, and a very large hind-talon, and,
except in the matter of size, agrees so exactly with the second upper true molar
of Mastodon perimensis from Perim Island, that I have no doubt that it is the homo-
logous tooth. The presence of the large hind-talon shows that the specimen cannot
be the first true molar, in which, as we have already seen, that talon is but very
slightly developed. The Punjab second molar is rather smaller than the Perim
tooth, as we found to be the case with the first true molars from the same localities.
The smaller tooth is probably that of a female.

The figured tooth is well-worn, the dentine discs having become united in the
first and second ridges. In the third ridge the dentine surfaces still form islands ;
in this ridge the dentine of the inner column does not present a trefoil-shaped
pattern, as in the anterior ridges at a corresponding state of detrition. The enamel
of the fourth ridge is only just perforated by wear. The hind-talon is lower than
the last ridge, and is still untouched by detrition; it presents a large median
tubercle, fianked by a smaller tubercle on either side. The length of this tooth is
5'4 inches, and its greatest width 2‘8 inches. This tooth shows well the great
thickness of the enamel, a very characteristic point in the molars of the species.

Comparison with M. longirostris. — If the figured second upper molar of
M. perimensis be compared with the corresponding iooth of M. longirostris in the
palate specimen figured by Kaup,^ it will be seen that the two teeth are almost exactly
in the same relative condition of wear, and therefore well adapted for comparison.
The Indian tooth differs from the European in possessing a very large talon, which
appendage never seems to attain any size in the second molar of the latter. The

* Loc, cit., PI. XVI, fig 5.

243—62

SIWALIK AND NADBADA PEOBOSCIDIA.

two teeth are further distinguished hy their wear ; in the European tooth the dentine
surfaces of the last ridge are much larger than in the Indian tooth, showing that
the plane of wear is less oblique in the former than in the latter. Again, in the
tooth of M. perimensis, the dentine surfaces of the two columns of each ridge
seem to unite at an early period of wear, and form a straight hand across the tooth ;
and there is consequently no distinct trefoil in the outer column of the second
ridge, which is so conspicuous in the tooth of M. longirostris when well worn, as
in Kaup’s specimen.

Last upper molar. — No specimen of the last upper true molar of Mastodon
perimensis is figured either in the “ Palseontological Memoirs” or in the “ Pauna
Antiqua Sivalensis,” although the cranium represented in Plate XXXVIII of the
latter work shows the base of the last true molar, the crown having been hammered
off. The magnificent molar of a Mastodon represented in Plate XLII of this
memoir was collected by Mr. Theobald in the Punjab, and is the last left upper
true molar, as may be determined from its form, size, and number of transverse
ridges. The tooth is long and narrow, which character alone is sufficient to dis-
tinguish it from the corresponding tooth of M. latidens ; the anterior ridge is the
widest, the transverse diameter of the crown gradually decreasing from this point
to the posterior extremity, where the tooth is rounded regularly off. This tooth carries
five transverse ridges, and a small posterior talon ; there is also a small step or
ledge at the anterior extremity, which doubtless represents an anterior talon. The
ridges are very tall, and are divided by the median cleft into conical inner and
outer columns. The two first ridges have been partly worn, and exhibit well the
uneven rate of wear of the two sides of the crown, which we have seen to be
characteristic of the molars of this species. The general arrangement of the
columns of this tooth is much the same as we have seen to prevail in the earlier
teeth of the species ; the inner column of each ridge wearing into a trefoil, and
the outer into an oblong. A few accessory tubercles are developed at the bases of
the outer extremities of the transverse valleys, which are not found in the anterior
teeth. The fifth ridge is narrower and lower than any of the preceding ridges,
and the hind-talon consists of three small low tubercles, the largest of which is
placed in front and on the inner side of the .others. The tooth shows the thick
enamel, and the small amount of cement so characteristic of the species. In the
following table the dimensions of this tooth are given, together with those of the
corresponding tooth of the opposite side of M. longirostris represented in fig. 9
of Plate XVIII of Kaup’s “ Ossements Eossiles du Museum de Darmstadt.”

M. perimensu. M. longirostris.

Length 9'6 8'9

Greatest width . . . . . . . , . . . 3'7 3'9

Height of third ridge 2'8 2’6

Comparison with M. longirostris. — Prom these measurements it wiU be seen
that the last upper true molar of M. longirostris is shorter and broader and has

SIWALIK AND NAEBADA PEOBOSCIDIA.

63—244

lower ridges than the corresponding tooth of M. perimensis. A comparison of the
two figures will, however, show that in general characters the two teeth are exceed-
ingly alike. The European tooth is distinguished, in addition to the difference in
relative size, by diminishing less rapidly in width posteriorly, and by the consequent
greater size of the fifth ridge and hind-talon : the columns of the ridges are also
somewhat less distinctly separate than in M. perimensis.

Second lower true molar. — Of the lower dentition of Mastodon perimensis the
first tooth with which I am acquainted is the second true molar, of which a speci-
men is represented in fig. 1 of Plate XLI, collected by Mr. Theobald in the Pun-
jab. The tooth has a concave profile, showing that it must belong to the lower
jaw ; the ridges present a concave surface to the broken end, showing that this must
be the posterior extremity, and that the tooth must consequently belong to the left
side of the jaw, since the median cleft is nearest the upper border of the figure.
Prom its size the tooth is inferred to be the second of the true molar series. The
crown is slightly narrower in front than behind and carries four transverse ridges,
(of which the last is broken), and a large posterior talon (also broken) ; a ledge on
the first ridge represents an anterior talon. The ridges are high, and are divided by
a median cleft, which, however, is not so distinct as in the upper molars. The sum-
mit of each of the four ridges (all of which are quite unworn in the specimen) ter-
minates in about six blunt cusps ; the portion of each ridge placed externally to the
median cleft (lower half of figure) would wear into trefoils, owing to the presence of
accessory tubercles in the valleys, along the median line of the tooth, which block the
vaUeys. Cement is present in the valleys. The length of this tooth is 5'1 inches, and
its width 2-8 inches. The size of this tooth agrees, therefore, with that of the
second upper true molar from the Punjab, wliich has been described above ; from
this, and from the presence of a large hind- talon in both, the lower molar has been
considered as the second of that series. The somewhat greater simplicity in struc-
ture of this lower molar, as compared with the corresponding upper tooth, is similar
to that which occurs in M. longirostris. A well-worn specimen of the second left
lower true molar of M. perimensis from Perim Island is represented in fig. 10 of
Plate XXXI of the “ Pauna Antiqua Sivalensis.” That tooth has a length of 4’ 8
inches, and a width of 2 ’6 inches. Allowing for the one tooth being a germ, and the
other much worn, the two teeth in question seem to have the same form : in
Palconer’s specimen the hind- talon is seen to consist of two large cones, only one of
which remains in Mr. Theobald’s specimen.

Last lower true molar. — A specimen of the last lower true molar of the present
species from Perim Island is represented in fig. 11 of the last quoted plate of the
“Pauna Antiqua Sivalensis.” The tooth carries five ridges and a large hind-talon,
and agrees in all essential characters with the last upper molar figured in the plates
accompanying this memoir. The length of the lower molar is 8’4 inches, and its
greatest width 3'5 inches. A very similar tooth, collected by Mr. Theobald, is in
the collection of the Indian Museum, but I have not thought it necessary to give a

245—64

SIWALIK AND NAEBADA PROBOSCIDIA.

figure of that specimen. Another last lower molar in a nearly complete mandible
from tlie Punjab has a length of 10'7 inches, and a width of 4‘4 inches.

Mandible. — In a notice published in the “ Records, I have recorded the
acquisition by the Indian Museum of a complete specimen of the adult mandible of
M. perimensis, obtained by Mr. Theobald in the Punjab. The specimen has not
been figured here, owing to its very large size, and the consequent difficulty of
getting an accurate drawing of it with the means at my disposal. The specimen
shows the last true molar and a portion of the penultimate tooth. The ramus is
sub-circular in cross section with a slight lateral compression ; it bulges out below
the last molar ; the lower border of the ramus is very slightly convex posteriorly,
and as slightly concave below the symphysis. The latter is produced into a short
trough-like symphysis similar to that of the specimen represented in fig. 2 of
Plate XLIII. In the figured specimen there is seen at the anterior fractured extre-
mity, in transverse section, a pair of small incisors, with a pyriform section. In the
complete mandible, of which the symphysis is entire, there are no incisors ; the
figured specimen probably belonged to a male, and the other to a female. The
dimensions of the complete mandible are as follows : —

Length from hinder border of last molar to distal end of symphysis 29‘0

Ditto from anterior border of penultimate ditto to ditto ........ 13'5

Ditto of last molar 10'7

Interval between condyle and coronoid process 6'0

The form of the symphysis of the mandible is of itself quite sufficient to dis-
tinguish M. perimensis from M, longirostris^ in which the symphysis is longer,
wider, and more deflected, with a shallower spout. The lower border of the ramus
of the latter is much bent, whereas that of the former is nearly straight. In
M. longirostris there is a considerable lateral compression of the rami, totally
wanting in M. perimensis.

Undetermined mandible. — In , fig. 1 of Plate XLIII, I have had lithographed
another symphysis of the mandible of a Mastodon^ bearing a considerable general
resemblance to that of M. perimensis, but also presenting some points of difference.
Its main point of difference is that in place of having a deeply hollowed spout-like
symphysis, it has only a very slight hollowing out of the produced symjDhysis, and
swells out anteriorly, while the symphysis of the typical tuskless form of M. peri-
mensis (as shown in the large unflgured mandible) tapers somewhat more to a point.
The trough in the latter mandible is, however, somewhat less deep than in the
figured tusked specimen, and I am inclined to think that the specimen represented
in fig. 1 must also be referred to a variety of M. perimensis. That specimen
certainly cannot belong to either M. pandionis (Plate XXXVI fig. 1) or M. siva-
lensis ; ^ its depth, when broken posteriorly, is too great to have fitted on to the

> Eec. Geol. Surv. India, Yob XI, p. 71.

^ “ Fauna Antiqua Sivalensis,” PI. XLV, fig. 10.

» Ibid., PI. XXXV, fig. 1.

SIWALIK AND NAEBADA PEOBOSCIDIA.

65—246

slender jaw of M. latidens noticed above, and I do not think from its form that it
belonged to Jf. falconeri ; if it does not belong to 31. ^erimensis, it therefore must
belong to a new species.

Cranium. — In noticing the cranium of M. perimensis figured in Plate XXXVIII
of the “Pauna Antiqua Sivalensis,” Dr. Palconer makes the following remarks : —
“ The cranium is in many respects singularly perfect, although it has suffered from
a crushing force, which has forced in the temples, so as to have contracted to a few
inches the inter-temporal portion of the forehead. The ascending ramus of the lower
jaw on either side is in situ, with the coronoid process and condyle, and what is more
remarkable, the greater part of the hyoid bone lies upon the sphenoid. The atlas
also was found attached to the condyles. The teeth are completely hammered down
to the margin of the alveoli. The most remarkable character of all about this head
is the low height of the pterygoid processes of the sphenoid, which are very little
higher than the condyles, and the comparatively little elevation of the condyles
above the palate. The interval between the plane of the lower surface of the
condyles and that of the palate is only 5 inches, the height of the occiput
being 22 inches. This is very much as in the North American Mastodon, and
even more so, so that the plane of the grinder does not differ much from that of
the condyles, thus showing a tendency in the direction of Dinotherium and the
Trilophodon Mastodon ohioticus {maximus). The pterygoids rise with a sharp
posterior border, aad do not spread out into a flap over the posterior border of the
maxillary. They are not rugous as (in) M. ohioticus, nor are they so far (proportion-
ably) extended behind. There are two large palatine foramina near the end of the
molar. The molars (aUov ing perhaps for some distortion from pressure) run parallel
and do not at any rate diverge in the remarkable way exhibited by M. ohioticus ;
perhaps they are less divergent even than in M. sivalensis. The palate looks long.
On either side are two molars, the penultimate and last true. The tusks exhibit an
oval outline in section.” '

The length of the last molar is only 7‘4 inches, so that this cranium must have
belonged to a much smaller individual than that to which the last upper molar figured
in the accompanying plates (Plate XLII), or than that to which the mandible noticed
above, belonged. It is probable that the Perim Island skull, together with the last
lower molar figured on the adjoining plate of the “ Pauna Antiqua Sivalensis,”
belonged to a female. I beheve that no complete specimen of the cranium of Mas-
todon longirostis has hitherto been discovered, so that we are unable to compare the
cranium of the latter and that of M. perimensis. Kaup has, however, figured^ a speci-
men of the palate of the former species with the penultimate and last true molars ;
and if that figure be compared with the palate of M. perimensis represented in the
“ Pauna Antiqua Sivalensis,” it will be seen that the molars in the former are mueh
less closely approximated than in the latter, and appear to be somewhat proportion-
ately wider.

Loc. cit. PI. XVI, fig. 5.

1^47—66

SIWALIK AND NARBADA PEOBOSCIDIA

General cliaraeters and Summarizing what is known regarding this

species, we find that, as far as we are at present acquainted with it, the dentition
follows the normal tetralophodont law in respect to the number of ridges carried on
the crowns of the molars. Premolars were developed in the upper jaw, and, therefore,
probably also in the lower ; the mandible is produced into a trough-like symphysis,
intermediate in length between that of the living Indian elephant and that of M-
pandionis ; it was furnished with incisors in some individuals, which were wanting in
others. The molars carry tall ridges, of which the columns develope accessory tubercles
blocking the transverse valleys ; one of the columns of each of the anterior ridges
when worn down, presents on its summit a trefoil-shaped islet of dentine. Cement is
present in small quantity in the valleys. The characters of the cranium have already
been pointed out.

In the form of its molars this species approaches nearest to the European
M. longirostris, but is at once distinguished by the form of the mandible and by
certain characters of the molars already pointed out.

The molars of the tetralophodont Mastodon dissimilis of Jourdan,^ from the
upper Pliocene of the Ehone valley, are of the type of those of M. perimensis, but.
as is well shown in the last upper molar, the ridges have a greater tendency to an
alternate arrangement of their columns and show less distinct trefoils of dentine, than
in M. perimensis. The last upper molar of the European species is more oblong in
form than that of the Indian, not contracting posteriorly as the latter does, and the
hind-talon consequently being of greater relative width. Other smaller differences
in the teeth of the two species will be seen on comparing the figures. There are no
figures of the cranium or of the complete lower jaw of M. dissimilis given in the plates
quoted 'in the note.

It will probably have struck the reader that in regard to the shape of the
ridges, the molars of M. perimensis present a striking resemblance to those of M.
falconeri, the two being, however,. broadly distinguished by the one having a tetra-
lophodont, and the other a trilophodont arrangement of the ridges in the interme-
diate molars. The two species, as far as I can judge from the broken mandible of
M. falconeri, are further distinguished hy the form of the mandible, which is more
convex interiorly and thicker in M. falconeri ; in the latter species there are no
premolars and no cement, both of which occur in the other species. It is unfortu-
nate that I have not been able to trace M. perimensis through its milk-molar
dentition, and it might be argued from this that both species had trilophodont milk-
molars, and both tri-and tetralophodont true molars. Apart, however, from the
important distinctions already pointed out, the last milk-molar of M. falconeri
(Plate XXXII, fig. 3) (which is exactly like the first true molar in the same young

’ Lortet and Chantre : “Mastodontes du Bassin du Rhone.” Archiv. du Mus. d’Hist. Nat. de Lyons, Vol. II,
Pis. I-VII Lyons, 1878. Only the plates of the molars of this species have yet reached India, the letter-press being
in a succeeding number.

SIWALIK AND NADBADA PEOBOSCIDIA. 67—248

skull) is so utterly unlike tlie first true molar of M. perimensis (Plate XL), that
this alone would he sufB.oient grounds of distinction.

An analogous instance of two species of Mastodon, the one with a trilopho-
dont and the other with a tetralophodont formula, is afi^orded by the European
M. angustidens and M. longirostris, in both of which the true molars are formed
on the same general plan, and present trefoil islets of dentine on their worn
columns.^

Distribution. — Remains of this species have been obtained from the Siwaliks
of the Punjab to the west of the Jhelum and in Perim Island. I am not quite
sure whether this species occurs in Sind, as I have some fragmentary molars from
that district which may belong to M. perimemis. As no specimens of the teeth of
this species were obtained by Ealconer and Cautley in the more easterly Siwaliks
or Burma, it is probable that this species did not extend its range much to the east
of the Jhelum.

Species 5 : Mastodon sivalensis, Ealconer & Cautley. Pis. XLI, fig. 2, XLIV.

Ristory .---ThQ first notice of Mastodon sivalensis seems to have appeared in a
paper read by the late Sir Proby (then Captain) Cautley before the Asiatic Society
of Bengal in June 1836.^ In that paper it appeared that the author considered
certain molars of a Mastodon a dents etroites from the Siwalik hills to belong to a
variety of the Mastodon angustidens ® of Cuvier ; to this variety he gave the name of
M. sivalensis. In a paper published in December of the same year in the “Journal
of the Asiatic Society of Bengal by the same author, the same conclusion was
arrived at as to the specific relations of this Siwalik Mastodon.

It thus appears that the name Sivalensis was given by Sir Proby Cautley alone.
In Dr. Ealconer’s subsequently published table of the species of Mastodon and
Elephant,® Mastodon sivalensis appears as a distinct species, with the names of
both Ealconer and Cautley affixed to it. In that table the species is placed among
the tetralophodont Mastodons, with the description “ Colliculi numero 5, obtusi
alternatim mammillee, valliculse interruptse,” wdth the remark “ the only known
species indicating a Pentalophodon-type.” In the same memoir® (which was
published in April 1857), Dr. Ealconer remarks : — “ Mastodon sivalensis is regarded
as having five ridges to the intermediate molars, instead of four; but this re-
markable character being restricted at present to a single species, it was deemed in-
expedient to form a systematic section for it alone, and it is ranged at the end of the

* Pal. Mem., Vol. II, PI. III.

* Jour. As. Soc., Bengal; Vol. V, p. 294 ; Pal. Mem., Vol. I, p. 126.

* M. angustidens included the tetralophodont M. anernensis, as well as the trilophodont form to which the
me is now restricted.

Pal. Mem., Vol. I, p. 127.— J. A. S. B. Vol. V, p. 768.

* Ihid., Vol. II, pp. 14, 15.

‘ Ibid., p. 18.

249- 68

SIWALIK AND NARBADA PROBOSCIDIA.

TetralopJiodons. — In a subsequent paper communicated (but not then published)
to the Geological Society of London in June 1857/ Dr. Falconer^ says : “ In the

preceding part, when discussing the conditions of the ' ridge-formula ’ in Trilophodon
and Tetralopliodon, it was stated that while the penultimate milk-molar always
presents one ridge less than the intermediate molars, the last true molar presents
one ridge more. Conformably, the last true molar in M. sivalemis presents six
ridges, besides the hind ‘ talon,’ thus maintaining throughout, so far as the denti-
tion is known, the numerical characters to be inferred from the ridge-formula, as
ascertained in Trilophodon and Tetralophodon. I consider it suflB.cient, on the
present occasion, to caU attention to this as a point of some interest and importance
in the systematic and palaeontological relations of the Proboscidian family, in
reference to the indications they present of an order of successive serial develop-
ment, without entering in detail upon the evidence in support of the view here
taken. That the species is a distinct form is abundantly borne out by the marked
characters of the skull, independently of the strong dental characters. The ridge-
formula for the true molars in Mastodon sivalensis is inferred to be —

5+5+6
5+5+(6— 7)

and when the dentition is fuUy made out, it is anticipated that the complete ridge-
formula will be nearly thus —

Milk-molars.

True molars,

5+5+6
5+5+(6— 7).”

2+4+5

It thus appears that in his last notice. Dr. Falconer considered M. sivalensis
to have a complete pentalophodont ridge-formula.

In Plates XXXII and XXXIII of the “ Fauna Antiqua Sivalensis” a figure of
a fairly complete cranium of the species is given, while in Plates XXXIV to XXXIX
of the same work a considerable series of the teeth are figm^ed. From these figures
and from various notes by Dr. Falconer, we gather that the molars of this species
are characterized by being narrow, and with complex crowns, on which the columns
of the ridges are placed alternately so as to completely block the valleys. The
columns are tall, and when worn present irregularly shaped discs of dentine,
and there is no cement. The mandible is represented in fig. 1 of Plate XXXV
of the “Fauna Antiqua Sivalensis” and is characterized by a very short spout-
like symphysis.

First loiver milk-molar. — I cannot find any specimen of the first upper milk-
molar of M. sivalensis, and the first tooth we have, therefore, to consider, is the first
lower milk-molar, a specimen of which is contained in each of two specimens of the

* Pal. Mem., Vol. II, p, 1.

* Ihid, p. 87.

SIWALIK AND NARBADA PROBOSCIDIA.

69—250

mandible of calves represented in figs. 1 and 2 of Plate XXXVII of the “ Pauna
Antiqua Sivalensis.” In both these specimens the crown of the tooth has been
hammered off. The length of the more perfect tooth is 0*6 inch, and its greatest
width 0*4 inch.

Second uppermilJc-molar. — Two specimens of the second upper milk-molar of
M. sivalensis are represented in figs. 1 and 2 of Plate XXXVI of the “ Pauna
Antiqua Sivalensis.” The first of these specimens is detached, the second is in a
fragment of the maxilla, which also contains the third milk-molar. The second
milk-molar seems to be somewhat oval-shaped, and carries three ridges, the hind-
most of which is shorter than the others. The length of this tooth is 2’ 6 inches,
and its greatest width 1‘8 inches.

Second lower milk-molar. — Each of the two fragmentary mandibles referred to
above as containing the first milk-molar, also contains a succeeding tooth, which
must accordingly be the second milk- molar. This tooth, like the corresponding
upper tooth, carries three transverse ridges, and has a length of 1-9 inches and a
width of 1*2 inches. These dimensions are slightly smaller than those of the
corresponding upper milk-molars, both of which, as we shall see below, probably
belonged to male animals.

Third upper milk-molar. — A specimen of the third upper milk-molar is re-
presented in fig. 2 of Plate XXXVI of the “ Pauna Antiqua Sivalensis.” This
tooth carries four ridges with alternately arranged columns, and has a length of
4 inches and a width of 2’3 inches. This tooth, as far as regards form, agrees
exactly with the smaller specimen represented in fig. 2 of Plate XLI of this
memoir. The latter specimen was transferred from the collection of the Asiatic
Society of Bengal to the Indian Museum, and is described on page 38 of the Catalogue
of the Possil Vertebrata in the Society’s collection by Dr. Palconer as follows : —

“ Pine specimen of the upper maxilla, left side, comprising the greater portion
of the palate and two molars embedded in the jaw, with four empty pits, marking the
situation of the fangs of the second milk-molar, which had fallen out. The
third milk-molar (figured tooth) is shown nearly entire with the enamel (of the)
crown broken off at the outside of the first two ridges ; the first three ridges are
seen to be touched with wear, forming depressed cups. The crown of the tooth is
bisected longitudinally into an outer and inner division, and the groups of mammillse
are seen to alternate with accessory mammillse in the valleys. The surface of the
enamel is deeply grooved vertically, so that the ridges when worn present a very
complex pattern. Behind the fourth ridge is a talon consisting of a complicated
group of small mammillse. The four empty fang cavities in front are (arranged so
as to form) nearly a square, showing that the crown of the tooth had a similar
form.”

The length of the third upper milk-molar is 3'2 inches and its greatest width
2 inches. The alveolar space of the second milk-molar occupies a square of about
1-3 inches in diameter.

251-70

SIWALIK AND NAEBADA PROBOSCIDIA.

It is probable tliat the third milk-molar figured here belonged to a female,
while the specimen figured in the “Fauna Antiqua Sivalensis” belonged to a male
animal.

In the jaw in which the figured third upper milk-molar is implanted there are
no traces of any premolar, either above that tooth, or having replaced 'the second
milk-molar, which has fallen out ; from this we may conclude that premolars were
not developed in Mastodon sivalensis.

The serial position of the figured third upper milk-molar is of considerable
importance, as from it we are enabled to determine the position of some other teeth,
and I will therefore recapitulate the grounds upon which its position is fixed. The
figured tooth corresponds exactly in form with the larger tooth represented in fig.

2 of Plate XXXVI of the “Fauna Antiqua Sivalensis,” which again has in front of
it a tooth of such a large size that it must be the second milk-molar ; hence the
larger tooth must be the third of that series.

The third upper milk-molar figured here has behind it three ridges of a larger
tooth, which must consequently be the first true molar, to which we shaU have to
refer again.

Third lower milk-molar. — The tooth represented in fig. 2 of Plate XLIV is im-
planted in a fragment of the right ramus of the mandible, and was obtained by Mr.
Theobald in the Siwaliks of the Kangra district. As it is larger and contains one
ridge more than the above noticed second lower milk-molar figured in the “ Fauna
Antiqua Sivalensis,” it must be the third lower milk-molar, or the corresponding
lower tooth to the third upper milk- molar, represented in fig. 2 of Plate XLI.
The last ridge of this tooth has been somewhat damaged, and the posterior talon com-
pletely broken away. The crown is long and narrow, with the outer border concave
and the inner convex. The inner extremity of each ridge is considerably higher
than the outer. Each of the three ridges which are uninjured are worn, and exhi-
bit irregularly shaped hollows of dentine. The ridges are of considerable relative
height, and each is mesially divided by the cleft traversing the longitudinal axis
of the crown into an inner and an outer column. These columns are arranged
alternately so that each ridge is set obliquely to the long axis of the crown, in such
a manner that its inner column is considerably in advance of its outer column. Bv
this arrangement, aided by the presence of some outlying tubercles, the transverse
valleys are completely blocked. The length of this tooth is 3*4 inches, and its
greatest width 1’8 inches. The height of the somewhat worn penultimate ridge is
1-3 inches. In fig. 3 of Plate XXXVII of the “ Fauna Antiqua Sivalensis,”
there is figured a worn third lower milk-molar of M. sivalensis which seems to have
carried four ridges, and which, as regards form and size, agrees almost exactly with
the specimen figured here.

Ridge -formula of the milk-molars. — From the specimens of milk-molars figured
here and in the “ Fauna Antiqua Sivalensis,” the serial positions of which are fixed
beyond the possibility of doubt, it is quite clear that these teeth always follow the

SIWALIK AND NARBADA PROBOSCIDIA.

71—252

normal tetralophodont law, and their ridge-formula may accordingly l)e expressed
as follows ; —

Milk-molars

Falconer’s conjectural ridge-formula of the milk-molars of the species noticed above
is consequently incorrect. With the evidence of the specimens figured in the “ Fauna
Antiqua Sivalensis,” and described in the “ Catalogue of the Fossil Vertehrata in
the Museum of the Asiatic Society of Bengal” before him, it appears, to say the
least, very curious that Falconer should have put forw^ard the pentalophodont
formula for the milk-molars of M. sivalensis. The only possible exception to this
rule would be the case of a lower jaw in the Museum of the College of Surgeons
referred to by Falconer in a note to the description of Plate XXXVII of the
“ Fauna Antiqua Sivalensis.” It is, however, probable that the tooth there considered
as the last milk-molar is really the first true molar, as we shall see below.

First upper true molar. — At page 47 of the “ Catalogue of the Fossil Vertehrata
in the Asiatic Society of Bengal,” Dr. Falconer shortly notices a palate of M. siva-
lensis containing the last milk-molar and the first true molar (No. 13). Tlie position
of these teeth is determined from their identity with the two corresponding teeth
in the jaw which contains the third upper milk-molar figured in Plate XLI, fig.
2, of this memoir. In the latter jaw, as already said, the first true molar is imperfect,
hut in the former it is complete, and is shown to have four ridges and a hind-
talon. It has a length of 4’ 2 inches and a width of 2*4 inches. I have not figured
this tooth, because it is a good deal damaged and chipped. I cannot find any
specimen of a five-ridged tooth w''hich I can certainly refer to the first true molar.
The characters of the first upper true molar are so exactly similar to those of the
milk-molars that I have not thought it necessary to give any description of that
tooth.

First lower true mslar. — -To the description of Plate XXXVII of the “ Fauna
Antiqua Sivalensis” Dr. Falconer adds a note (1) in which he describes a lower jaw'
of Mastodon sivalensis in the College of Surgeons which contains a complete six ridg-
ed tooth followed by a larger incomplete tooth. The former has a length of 4’ 2 inches
and a width of 2'4 inches. The width of this tooth is considerably greater than that
of the third lower milk-molar figured here, from which fact, coupled with its high
ridge-formula, I think that this tooth is probably the first true molar and not the
last milk-molar, as it is considered by Falconer.

I cannot find any other complete tooth which I can certainly class as the first
lower true molar of this species, though I tlunk it probable that specimens wfill
eventually he found with either four or five ridges.

Second upper true molar. — In fig. 4 of Plate XXXVI of the “ Fauna An-
tiqua Sivalensis ” there is figured a tooth which seems evidently to be the second
upper true molar of M. sivalensis, and w^hich seems to have four ridges and a
large hind-talon. Its length is 5*6 inches, and its width 2*9 inches. In fig. 5

253—72

SIWALIK AND NAEBADA PEOBOSCIDIA.

of the same plate there is another specimen of the same tooth which has five
ridges and a talon. Its length is 6’5 inches and its width 2’9 inches.

In the palate specimen figured in Plate XXXIV, fig. 1, of the “ Pauna An-
tiqua Sivalensis,” containing the secoijd and third true molars, the former tooth
has four ridges and a talon, and has a length of 4' 7 inches and a width of 2' 8
inches. A very similar palate in the Indian Museum also shows the second true
molar with four ridges and a hind-talon.

The number of ridges in this tooth may, therefore, be four or five.

Second lower true molar. — There is no complete specimen of the second
lower true molar of this species either in the collection of the Indian Museum,
or among the series figured in the “ Pauna Antiqua Sivalensis.” The imperfect
specimens, however, seem to indicate that this tooth, in some instances, carried five
ridges.

A specimen of the homologous tooth in the Museum of the College of Sur-
geons, referred to by Palconer in a note (2) to the description of Plate XXXVII
of the “ Pauna Antiqua Sivalensis,” has only four ridges and a talon.

This tooth, therefore, like the corresponding upper molar, may have either four
or five ridges.

Third upper true molar, — In fig. 6 of Plate XXXVI of the “ Pauna An-
tiqua Sivalensis” there is represented a complete last upper true molar of Mastodon
sivalensis. This tooth has six ridges and a hind-talon, and a length of 7'8 inches
and a width of 3 ‘3 inches. It has the alternate arrangement of the columns
very conspicuously displayed.

In the palate specimen represented in fig. 1 of Plate XXXIV of the same
work, the last molar has five ridges and a hind-talon. On this tooth Dr. Palconer
remarks : —

“ The front ridge has two confluent ridges on the outside, to the inner of which
tlie intermediate pillar is attached, joining on with ' the innermost large point of
the third ridge. This third ridge shows but one thick point on the inner divi-
sion, and two on the outer, with the intermediate pillar connecting the outer
division of the third with the inner division of the fourth ridge, and so with the
last ridge. This diagonal connection of the posterior surface of the outer division
with the anterior surface of the inner points of the next following ridge causes
in wear the alternate-like discs of detrition, which characterize the teeth of Masto-
don sivalensis^^

In fig. 1 of Plate XLIV of this memoir, there is drawn a very fine specimen
of an almost unworn last upper molar of a Mastodon, taken from the left side
of a palate specimen collected by Mr. Theobald in the Siwahks of the Punjab ^
It is inferred to belong to M. sivalensis from its narrowness, general complexity
of pattern, and somewhat alternate arrangement of the columns, though this latter
character is not so well marked as in Palconer’s specimens. The tooth is preceded
in the palate by a much worn and smaller tooth, in which the number of ridges

SIWALIK AND NARBADA PROBOSCIDIA.

73-254

originally carried on the crown cannot now be determined. The figured tooth
carries six transverse ridges and small fore-and-aft talons ; its ridges are divided by
a medhm cleft into inner and outer columns. Accessory columns are present in the
valleys. When worn down, the columns would present irregularly shaped discs
of dentine on their summits, and not trefoils (the preceding molar in the palate
shows these irregular discs). As before said, the columns are not arranged in our
specimen in nearly so alternate a manner as in Balconer’s specimens of the last
upper molar of M. sivalensis ; it will, however, be seen from other teeth figured
in the “ Eauna Antiqua Sivalensis” (and also from the specimens in the Indian
Museum), that there is considerable variation in this respect, as also in the number
of accessory tubercles. In consequence of this different arrangement of the columns
of the ridges in our specimen and Ealconer’s specimens, the valleys are much less
blocked in the former than in the latter.

The last upper molar of the species may, therefore, carry either five or six
ridges.

Third lower true molar. — Specimens of the third lower true molar of M. sivalensis
are represented in the “Eauna Antiqua Sivalensis” (Plate XXXV, fig. 1, Plate
XXXVII, figs. 7 and 8), aU of which carry six ridges and a hind-talon. The
ridges are disposed alternately, and a few accessory tubercles, which may vary
in number, are present in the valleys. The specimen represented in fig. 8 of
the latter plate has a length of 8'8 inches and a width of 2’9 inches.

In fig. 3 of Plate XLIV of this memoir, a tooth is figured, which is implant-
ed in a nearly complete, though somewhat crushed mandible, which was collected
by Mr. Theobald in the Punjab. This tooth, though differing in the number of
ridges from Ealconer’s specimens, seems probably to be the last lower molar of
M. sivalensis. The figured tooth is from the right ramus of the mandible, and as
it is well worn and has no tooth succeeding it from behind, must be the last of
the series.

This tooth carries five ridges and a very small hind-talon. The last ridge is the
smallest. This last ridge consists of a very distinct inner and outer column, separ-
ated by a median cleft. The worn summit of each column presents an irregularly shap-
ed and depressed islet of dentine. The columns have only a very slight alternate
arrangement. In each valley there is a cluster of accessory tubercles, which are
in general attached to the preceding ridge. The first valley has four tubercles, the
second three, the third two, and the fourth one.

If this tooth be compared with the corresponding tooth of Mastodon pandionis
represented in fig. 4 of Plate XXXV, it will be seen, as already noticed, that the two
are almost identical in form, if the large talon in the latter be counted as a fifth
ridge ; and indeed if the two teeth had been found detached, they would almost cer-
tainly have been referred to the same species. It will, however, be remembered that
the tooth of M, pandionis was implanted in a fragment of the mandible to which
belonged the enormous symphysis represented in fig. 1 of Plate XXXVI. The figured

‘ \

255—74 SIWALIK AND NAKBADA PEOBOSCIDIA.

tooth of M. sivalensis, on the other hand, is implanted in a complete mandible,
which, though slightly broken at the symphysis, shows that such symphysis was short
like that of the living Indian elephant, and in fact agrees exactly with the man-
dible of M. sivalensis represented in figs. 1 and la of Plate XXXV of the “ Pauna
Antiqua Sivalensis,” which carries a six-ridged tooth with alternately disposed
columns. Prom this identity in form of the two mandibles, and from the similar
general structure of the two teeth, I have referred the specimen figured here to
Jf. sivalensis.

If the latter tooth he compared with the tooth of M. pandionis, it will be found
that there is a little difference in the disposal and wear of the outlying tubercles in
the two, while the hinder ridges are relatively highest in M. pandionis. The most
important mark of distinction, however, is that the wear of the tooth of Jf. pandionis
is almost in one plane, whereas the tooth of Jf. sivalensis is deeply concave in the
middle.

The figured tooth of Jf. sivalensis differs from the specimen figured in Plate
XXXVII, fig. 8, of the Pauna Antiqua Sivalensis,” in having one ridge less, less
alternately disposed columns, and more accessory tubercles.

Relationship to Mastodon arvernensis.—'WQ have already seen that Sir Proby
Cautley thought that the molars of Jf. sivalensis could not be distinguished speci- “
fically from those of Jf. arvernensis of the English Crag, ^ and the subsequent re-
searches of Palconer confirmed this close relationship. Both species agree in the
alternate arrangement of the mammillse of the ridges, but the mammillse of the
European molars are generally less alternate than those of the Indian. The Euro-
pean species does not seem to have any tendency to a pentalophodont type, and is
further distinguished by having premolars, ^ which seem to be absent in the Indian
species. Both species seem to have had a short elephantine mandible, without
tusks. ^

The cranium of Jf. arvernensis is, I believe, unknown, and cannot therefore be
compared with the very peculiar cranium of Jf. sivalensis.

There can be no question but that the two forms are very closely allied. The
differences indicated above afford, however, amply sufficient grounds for the speci-
fic distinctness of the two forms.

Relationship to M. pandionis. — Mastodon sivalensis may be regarded as present-
ing the same kind of relationship to the trilophodont Jf. pandionis, as we have already
shown to exist between the tetralophont Jf. perimensis and the trilophodont M. fal-
coneri. In both the two first-named species the pattern of the crown of the molars
is complex, with alternate columns, accessory outlying tubercles, blocked valleys, and
irregularly shaped discs of dentine. The early molars of the two species are very
different.

' Included at that time under the name of M. angustidens.

2 Pal. Mem., Vol. II, p. 34,

3 Ibid., p. 44.

SIWALIK AND NARBADA PROBOSCIDIA.

75—256

Distribution. — Remains of M. sivalensis have been obtained throughout the
sub-Himalayan Siwaliks, and not, as far as I am aware, elsewhere.

General conclusions regarding Mastodon molars.

The foregoing survey of such a large series of Mastodon molars has led to the con-
clusion that the very regular ridge- formula given by Ralconer will not always hold
good in regard to the true molars, though in the Indian species, at aU events, it
appears to be always constant in the milk-molars. We have seen that there is a ten-
dency in the true molars of some of the Trilophodons {M. falconeri) to develope the
talon into a fourth ridge, and in the Tetralopliodons [M. latidens and M. sivalensis),
a similar talon is developed into a fifth ridge, in the intermediate true molars. We
thus see the passage from a Triloyhodm to a Tetralophodon, and from the latter to
a Stegodon.

We have no instance of a tooth with a certain number of ridges ever being suc-
ceeded by a tooth with a smaller number : neither have we at present any instance of
a trilophodont being mixed with a tetralophodont formula in the same species.^

Genus 2 : ELEPHAS, Linne.

Elephants, whose intermediate molars always have more than five ridges, and
in which cement is present in considerable quantity.

Sub-Genus 1 : Stegodon, Ealconer.

Elephants, whose intermediate molars do not generally have more than eight
low ridges, and in which the cement does not form a continuous surface with the
ridges in the germ molars.

Species 1 : Stegodon CLiPTn, Ealconer and Cautley. Plate XL V, figs. 1 & 2.

Synonyms : 8. sinensis, Owen.

Mastodon elephantoides, Clift.

History. — Remains of this species were first obtained by the late Mr. Crawfurd
from strata near Ava, in 1826, and a specimen of the palate, and of the first upper
true molar, were figured by the late Mr. Clift in the “Transactions of the Geologi-
cal Society of London,”^ under the names of Mastodon latidens and Mastodon
elephantoides. In the only part of the letter press of the “ Eauna Antiqua Siva-
lensis,” which ever appeared, and which was published in 1846, the authors pointed

* l am here referring only to Indian species, and do not therefore mention M. andium.

* Second Ser., Vol. II, Pis. 36 and 39, fig. 6.

257—76

SIWALIK AND NARBADA PROBOSCIDIA.

out that Clift had confused the remains of several species under the name of M,
elephantoides, and they accordingly gave a new name to the two species of “ transi-
tional Mastodons,” which had been partly mixed up with Mastodon latidens. To
the elephant teeth with six ridges they gave the name of Elephas eliftii, and to the
teeth with a greater number of ridges E. insignis} In a subsequent paper, ^ these
and other species appeared under the suh-generic name of Stegodon, proposed by
Dr. Balconer.

Eistribution. — All the specimens of the molars of Stegodon cliftii figured in
the “ Eauna Antiqua Sivalensis ” were obtained from Burma, and it is not certain
whether the authors of that work had seen any authentic specimens from the suh-
Himalayan Siwaliks, though one specimen in the collection of the Asiatic Society of
Bengal was doubtfully said to have come from the Indian Siwahks. Subsequent
researches have, however, shown that the species is of not uncommon occurrence
in the Siwaliks of the Punjab, and a considerable number of specimens of the
teeth and jaws are now exhibited in the Indian Museum, obtained from that district.
One of these specimens is a palate with the two penultimate molars, each of which
has six ridges, and which is almost exactly similar to the specimen figured by Mr.
Clift in Plate XXXVI of his above- quoted memoir. Another specimen, the last
lower true molar, has been previously noticed by myself in the “ Records.”® We
shall subsequently see that the species probably occurs in China.

Characters of molars. — The characters of the molars of Stegodon cliftii, as
given by Dr. Palconer in various notices,^ are as follows : The first and second true
molars have only six ridges, continuous across the crown, and chevron- shaped, with
numerous mammillae or cusps. The cement does not fill up the interval between
these ridges, hut occurs only, in inconsiderable quantities, in the hollows. The outer
side of the upper molars is higher than the inner, in which character the species
differs from all the other Stegodons named by Palconer, in the molars of which the
central part of the ridges is the highest. In all these characters the molars of
S. cliftii are intermediate between those of the Mastodons and those of the other
Stegodons. A complete milk-molar was not known to Palconer, but he inferred
that the last milk-molar would carry five or six ridges.

Second upper milk-molar. — In the “ Quarterly Journal of the Geological Society
of London” for 1870,® Professor Owen described a penultimate upper milk-molar
of a Stegodon obtained from China, under the name of S. sinensis, considering it to
l^e distinct from any of Palconer’s species. Of this tooth the Indian Museum has
lately obtained a cast, through the courtesy of the Trustees of the British Museum,
which is figured in Plate XLV, fig. 2, of tliis memoir. In regard to this tooth
I have come to a conclusion directly opposed to that of Professor Owen, and con-

* “ Paltcontological Memoirs,” Vol. I, p. 82.

2 Ihid., Vol. II, p. 14.

^ Eec. Geol. Surv. of India, Vol. XI, p. 72.

^ Pal. Mem., Vols. I, p. 82, II, p. 84.

‘ Vol, XXVI, p. 417, PI. XXVII.

SIWALIK AND NARBADA PROBOSCIDIA.

77—258

sider that the tooth in question really belongs to Stegodon cliftii. Before I had
obtained the cast of the tooth, I accepted the conclusions at which Professor Owen
had arrived concerning it, and from the evidence of a single common character,
referred the tooth of another Stegodon to the same species, although I now find that
such identification was completely erroneous^

The Chinese tooth is a small oblong molar, narrower in front than behind,
carrying four complete ridges, and an anterior talon, connected with the centre of
tlie first ridge. The grinding surface of the crown is slightly convex antero-
posteriorly, thereby showing that the tooth belongs to the upper jaw, while the
lower border (in the figure) of the ridges being the most worn, show^s that it belongs
to the left side ; its size shows that the tooth must he the second or penultimate
milk-molar. All the ridges have been abraded by wear, and the inner portion of
each ridge is the lowest and the most worn ; the dentine surfaces consequently
present a pear-shaped form. The first ridge is nearly straight, but the succeeding
ridges are somewhat bent, the inner halves being slightly in advance of the outer.
There is a slight constriction at the point where the ridges are bent, probably repre-
senting the mesial cleft which occurs in the molars of the Mastodons. Each ridge
when unworn probably carried on its summit a considerable number of cusps or
mammillge. The length of this tooth is 2-8 inches, and its greatest width 2T inches.
Professor Ov^en in describing the tooth reckons the imperfect anterior ridge as a
true ridge, and not as a large talon, as I have done.

In referring the tooth to its proper species, Professor Owen proceeds to com-
pare it with the teeth of Stegodon insignis, S. ganesa, and S. bombifrons. This com-
parison is, however, unnecessary, because the Chinese tooth differs in to to from the
molars of the three above-named species, in all of which the centre of the tooth, and
not the outer side, is the highest point. The corresponding tooth of the first-named
species differs also in having a greater number of ridges.

The upper molars of Stegodon cliftii, on the other hand, as we have already seen,
are characterized by having the outer sides higher than the inner, and by the den-
tine surfaces of the ridges being wider internally than externally, as is shown in the
figure of the first upper molar given by Clift.^ We thus see that in one very im-
portant character the Chinese milk-molar agrees with the molars of S. cliftii, and
with those of no other species of the genus. The latter molars, moreover, have a
low ridge-formula, in which respect also they agree with the Chinese tooth. It now,
therefore, remains to enquire on what grounds Professor Owen referred the Chinese
tooth to a distinct species. The Professor appears to have overlooked the manner of
wear of the ridges, and observes that the Chinese tootlP cannot belong to S. cliftii,
because in the molars of that species there is “ no mark of a longitudinal line bisecting

* Kec. Geol. Surv. of India, Vol. XI, p. 73. See also below. In a paper published in the IXth Volume of the
Reeords (p. 49), when showing that 8. orientalis was founded on an insufficient character, I joined to it the name of 8.
sinensis, whose specific distinctness does not depend on the same character.

* loc. cit., PI. XXXIX, fig. 6.

^ loc. cit., pp. 419-20.

259—78

SIWALIK AND NARBADA PROBOSCIDIA.

the tooth into an outer and an inner division,” and because the ridges in the latter
are not bent. Now, with regard to the first of these two negative characters, Dr.
iPalconer himself, in describing the upper molar of S. cUftii, represented in fig. 1 of
Plate XXX of the “ Pauna Antiqua Sivalensis,” says, that the “ mesial line of divi-
sion of the ridges is not very distinct,'’ clearly implying that in ttat specimen there is
a median division : other teeth are, however, described as showing no trace of this cleft.
It is quite true that the mesial cleft is very indistinct in the upper true molars
of S. cliftii, but it appears to me from an examination of the cast of Clift’s specimen,
and other molars in the Indian Museum, that there is almost or quite as much trace
of this median cleft in them as in the Chinese tooth j it may also not be improbable
that this cleft is more distinct in the earlier teeth, which not unfrequently show an-
cestral characters more distinctly than the later ones. There now, therefore, only
remains the question of the curvature of the ridges. Professor Owen, in describing
tlie Chinese tooth, himself admits that he has observed a curvature of the ridges in
certain specimens of the molars of S. bombifrons, and this curvatm^e will be subse-
quently shown to occur occasionally in molars both of that species and of S. insignis..
If, therefore, a curvature may occasionally occur in the ridges of the molars of one
species of Stegodon which are normally straight, there appears to be no valid reason
why it should not equally well occur in those of another.

Since, therefore, in my opinion, there seems to be no character by which the
Chinese milk-molar can be distinguished from the molars of S. cliftii, and since the
milk-molar agrees with the true molars of that species in the manner of wear of the
ridges, and in having a low ridge-formula, there appears to me to be every pro-
bability that the Chinese tooth is really the second upper milk-molar of Stegodon
cliftii. If this interpretation be correct, the name of S. sinensis must be expunged,,
and the distribution of S. cliftii be extended to China. As a character of perhaps
minor import the Chinese tooth agrees with the first true molar of S. cliftii in having
a large number of cusps on the ridges, and in the worn enamel being much cre-
nulated.

Third upper milk-molar. — The tooth represented in fig. 1 of Plate XLV, collec-
ted by Mr. Theobald in the Punjab, is a left upper molar of a Stegodon, which I for-
merly' considered as the first true molar of S. bombifrons, but which, as will be seen
below, I now find cannot belong to that species. The tooth is considerably worn
and carries six ridges and a large hind-talon ; from the number of ridges this
tooth cannot belong to S. insignis or S. ganesa, in which the two first “intermediate”
molars never carry less than seven ridges each {vide infra). There now, therefore,
only remains S. cliftii, to which the specimen can belong ; but before going further,
it will be well to describe the specimen. The two first ridges of the specimen are
so worn that their dentine surfaces have become united, and a trace of the enamel
of the first valley remains only on the inner side. The remaining four ridges are
also well worn ; they extend straight across the crown, and their outer is slightlv

‘ Rec. Geol. Surv. of India, Vol. XI, p. 74.

SIWALIK AND NARBADA PROBOSCIDIA.

79—260

liiglier than their inner sides ; they are, however, so much worn that the exposed
dentine surfaces have almost the same width throughout. The fifth and sixth
ridges show traces of two constrictions. The hind-talon is just perforated by wear.*
The crenulations of the enamel show that the ridges carried a considerable number
of cusps ; while the extremities of the ridges show that they were low and broad.
The length of the tooth is 4'9 inches and its greatest width 2-9 inches.

In all the above characters, as well as in its great proportionate width, the tooth
agrees with the first true molar of Stegodon cliftii : and I have little doubt that it
belongs to that species. It is true that the first true molar of /S', cliftii has no
hind -talon ; the second true molar of that species, however, has a large hind-talon,^
the absence or presence of which must accordingly be considered as of no import-
ance. It now only remains to consider the serial position of the tooth. If we turn
to the posthumous description of the plates of the “ Banna Antiqua Sivalensis,”^ we
shall find that the first upper true molar of S. cliftii has a length of 6T and a width
of 3*3 inches. The Punjab tooth must consequently be the last milk-molar, as it is
so much smaller than the first true molar : all the previously known teeth of the
species are of unusually large size, in which respect the second and third milk-molars
agree with them. The third milk-molar of this species, in having six ridges, accords
with the prediction of Dr. Falconer that such would probably be the case when
the tooth was discovered. This tooth agrees with the Chinese second milk-molar in
the crenulations of the exposed ridges of the enamel.

A specimen of the palate of Stegodon cliftii figured in Plate XXX, fig. 1, of
the “ Fauna Antiqua Sivalensis” shows a very much worn third milk-molar, the
ridges of which have been completely ground down. The length of this fragmentary
tooth is 3 ’3 inches and its width 2 inches, but when complete, it must have been
much larger, and would probably then have agreed in size with the tooth which I
have described above.

First upper true molar. — As we have already seen, a complete specimen of the
first upper true molar of Stegodon cliftii is represented in fig. 2 of Plate XXX of the
Fauna Antiqua Sivalensis,” and is described in the following words by Dr. Falconer :

“ This is the first or antepenultimate true molar, upper jaw, left side, entire, detach-
ed, and beautifully preserved. It has six ridges and a small hind-talon. The tooth
is nearly oblong. The enamel is somewhat fluted, and there is very little cement.
There is very little convexity of the tooth across, and no distinct indication of the
mesial dividing line. There are as many as eleven to twelve denticles or points
(cusps). The tooth is compressed and angular in front, and the three front ridges are
worn. This tooth is also represented in fig. 6 (Plate XXXIX) of Mr. Clift’s memoir.

It is there described as Mastodon elephant oides. Its elephantine afiinities are
indicated by the absence of a longitudinal line of division along the crown, and by
the greater number of points (cusps) that enter into the composition of the ridges.
Length 6‘1, width in front 3", width behind 3‘3 inches.”

» F. A. S., PI. XXX, fig. 3.- * Hid., p. 42 (Index).

261—80

SIWALIK AND NARBADA PROBOSCIDIA.

Second upper true molar. — Two specimens of the second upper true molar of
this species are represented in fig. 3 of the above quoted plate of the “ Fauna Antiqua
Sivalensis” and are described as follows : “ Superb palate specimen containing

the penultimate true molar in situ on both sides. The tooth is proved to be the
penultimate by its large dimensions, and by the circumstance that two or three
ridges of another tooth (third true molar) are seen behind it. The crown ridges are
all more or less worn and partly damaged by fracture, but enough remains to show
that the tooth was composed of six ridges and a hind-talon. The palate is very flat,
and the teeth on either side (in the erect position of the skull) slope upwards from
without inwards. The teeth have very little cement. The slope of the grinding
surface from without inwards is a mastodontoid character. In JEleplias {Stegodon) in-
signia the tendency of the grinding surface is to shelve upwards from the inside
outwards. Length of molar 8'2 inches, width 4’ inches.”

A very similar specimen to the above is exhibited in the Indian Museum,
collected by Mr. Theobald in the Hushiarpfir district. The penultimate molar
in that specimen is well worn, and carries six straight and low ridges, and a hind-
talon. The tooth, except in the matter of size, agrees in every detail with Falconer’s
specimen. Its length is 7’3 inches, and its greatest width 3‘8 inches.

Third upper true molar. — Several specimens of the last upper molar of S. cliftii
are contained in the collection of the Indian Museum, all of which have been
obtained from Burma. They carry either seven or eight ridges and a hind-talon,
and agree in general characters with the teeth noticed above. One of these specimens
is described by Dr. Falconer on page 113 of the first volume of the “ Palseontological
Memoirs,” from which the following extracts are taken ; “ The plane of wear slopes
obliquely from the outside inwards, so that the interior side of the first three ridges
IS ground down much lower than the outer side. There is no mark of a longitudinal
line bisecting the tooth, as in the true Mastodons, into an outer and inner division.
The ridges are a little convex in front and concave behind, determining a similar
form to the valleys between them, which run across without any interruption.
The mainmillge are obtuse and closely packed, and nine or ten of them may he
counted on each ridge. Hardly any crusta petrosa (cement) is visible, except
between the fifth, sixth, and seventh ridges. Length of molar 9'3 inches ; greatest
width 4*3 inches.”

Last lower tome molar. — I cannot find any record of any of the lower molars
of Stegodooi cliftii, with the exception of the last true molar, of which a specimen
from Burma is represented in fig. 5 of Plate XXX of the “Fauna Antiqua
Sivalensis,” described by Dr. Falconer as follows : “ Last true molar, lower jaw,

left side, consisting of eight ridges and a talon. Five ridges are touched by wear,
and the two anterior ridges are nearly worn out. The crown is very flat ; the
points (cusps) are large and few in number, and there is no very distinct mesial
dividing line (and) hut little cement. Length 12-7 inches, width 4'5 inches.”

Another specimen of the same tooth, collected by Mr. Theobald in the Siwahks

SIWALIK AND. NARBADA PROBOSCIDIA.

81—262

of the Punjab, lias been already referred to by me in the “ Records,” as noticed
above. That tooth carries seven ridges, each of which has only a small number of
cusps, and exhibits a slight trace of a mesial cleft, with the plane of wear sloping
from the inner to the outer side. It seems from these two specimens that the
lower molars of S. cliftii carry a smaller number of cusps tlian the upper.

Conclusion. — The skull and complete mandible of S. cliftii are unfortunately
still unknown, and \re must, therefore, depend entirely on the teeth for its specific
characters. The whole of the upper molar series, with the exception of the first
milk-molar, are now known, while of the lower molars only the last is known. The
ridge-formula of the species may probably be represented as follows -

True molars.

6 -h 6 +(7—8)

+(7— 8)

Milk-molars.

? + 4 + 6
? + 4 F + 6?

Species 2 : Stegodon bombifuons, Palconer & Cautley. Plates XLV, fig. 3,

XLVI, figs. 1 & 3.

History.— The first mention that I can find -of Stegodon hombifrons occurs in
the Introduction to the “ Pauna Antiqua Sivalensis,”^ where the name only is
mentioned. Purther on in the same w'ork^ mention is made of the cranium of the
species, and some of the characters of the molars are pointed out. The molars are
said to differ from those, of S. insignis [vide infra) in having broader and less
elevated ridges, and more open valleys, and by the number of ridges being fewer,
though greater than in S, cliftii. In Plates XXIXA & B of the same work
various specimens of the molars are figured, showing that the last upper milk-
molar had six ridges, and the corresponding lower tooth either five or six ridges,
Avhile other specimens of the “ intermediate” molars had seven ridges, and the
last lower molar never carried more than nine ridges. In the “ Records”^ I alluded to
a six-ridged tooth, which I then thought to be the first upper true molar of S. bombi-
frons, but which, as noticed above, I now find probably belongs to S. cliftii; the cor-
responding tooth referred to on the same page of the “ Records” as belonging to the
so-called S. sinensis of Professor Owen I now find belongs to the present species.

Molar series. — With these preliminary remarks, I wiU now consider in serial
order such of the molar teeth of this species as are at present known to me.

First milh-^molars. — Of the first milk-molar, either of the upper or lower jaw,
I can find no specimen.

Second upper milk-molar. — Of the second upper milk-molar I likewise cannot
find any specimen.

‘ “ Palseoiitological Memoirs,” Vol. I, p. 20.

Ibid., p. 81.

Rec. Geol. Surv, of India, Vol. XI, p. 73,

263—82

SIWALIK AND NAEBADA PEOBOSCIDIA.

Second lower milk-molar. — Of the second lower milk-molar of S. bombifrons
there is no specimen figured in the “ Bauna Antiqua Sivalensis,” neither is any
mention of this tooth made in any of Balconer's subsequent memoirs. The speci-
men represented in Plate XLVI, fig. 3, of this memoir, is undoubtedly the second
lower milk-molar of a Stegodon, and in all probability of S. bombifrons. It was
obtained by Mr. Theobald in the Siwaliks of the Punjab. The tooth is a germ
specimen of the crown, entirely untouched by wear : it is much narrower in front
than behind, and very concave on one side, from which characters I infer that the
tooth belonged to the lower jaw. The concave side will he the outer, and the tooth
will consequently belong to the left ramus of the mandible. Its size proves it to
he a second milk-molar, while the unbroken transverse ridges, with very little
cement, shows that it must belong to a Stegodon.

This tooth carries four complete transverse ridges, increasing in width regularly
from the first to the last. There is a very small anterior talon carrying ‘two cusps,
and a larger posterior talon, the number of cusps on which is concealed by matrix.
The first ridge carries six cusps, and the second seven : the number of cusps on the
two succeeding ridges cannot he clearly ascertained. The two first ridges are not
quite straight ; each ridge is low and blunt, and very broad at the base : cement
is present in the valleys in small quantities. The central point of each ridge is
the highest. In the following table the dimensions of this tooth are compared
with those of the second lower milk- molar of Stegodon insignis, which will he
described in the sequel.

' S', bombifrons, S. insignis.

Length of tooth ............ 2'2 2'7

Width of first ridge I'O I'Oo

Ditto of last ditto 1‘4 1‘55

Height of last ditto 07 1‘05

Antero-posterior diameter of base of ridge 0'5 O'o

Interval between summits of penultimate and last ridges 0'46 0'35

Prom this table it will he seen that the tooth under consideration differs from the
corresponding milk-molar of S. insignis (including S. ganesa) by the ridges being
lower and wider, and placed farther apart from one another. The two teeth further
differ in that the second lower milk-molar of S. insignis has a larger quantity of
cement, and five in place of four ridges (see below). In its smaller number of
ridges and in their form the Punjab tooth agrees with the characters of the molars
of S. bombifrons as given by Palconer, and I have, therefore, no doubt in referring
it to that species.

The second upper (and probably therefore the corresponding lower) milk-molar
of Stegodon cliftii (see above, and Plate XLV, fig. 2) corresponds, as far as regards
the number of ridges carried by the crown, with the corresponding tooth of S. boni-
bifrons, but is distinguished, among other characters, by one side of each ridge being
higher than the other, by the still greater thickness of the ridges, as well as by the
form of the anterior talon, and the greater size of the whole tooth.

SIWALIK AND NAEBADA PEOBOSCIDIA.

83-264

Third upper milk-molar. — I have no complete specimen of the third upper
milk-molar, though an incomplete specimen of this tooth will be noticed when I
come to the first upper true molar.

Third loioer milk-molar. — Two much- worn and somewhat mutilated specimens
of the third lower milk-molar of Stegodon bomhifrons are figured in the “ Eauna
Antiqua Sivalensis” (Plate XXIX A, fig. 1, Plate XXIXB, fig. 5). The first of
these teeth carries six ridges and a talon, and has a length of 4 inches, and a width
of 2 inches. The second carries five ridges, and fore-and-aft talons, and has a
length of 3'6 inches and a width at the widest point of 1’8 inches. No description
of these teeth has ever appeared.

In fig. 3 of Plate XLV of this memoir I have had drawn a perfect specimen
of a tooth which certainly appears to he the third lower milk-molar of the right side
of Stegodon bomhifrons. This tooth is now in the Indian Museum, and was obtained
by exchange with the Dahore Museum. It was found in the Siwahks of the
Punjab. The crown carries five transverse ridges, and two small talons, the most
anterior one being somewhat broken. The first three ridges are worn, but the two
last are still untouched by attrition.

The penultimate ridge carries about nine cusps. The ridges are very low and
blunt, and there is but a small quantity of cement in the valleys. The dimensions
of the specimen are as follows : —

Length 4i‘l

Width of first ridge 17

Ditto of penultimate ridge 2‘2

Height of ditto ditto I'l

Antero-posterior diameter of base of penultimate ridge .075

Interval between sumnodts of penultimate and ante-penultimate ridges 075

The unworn ridges are lowest at their centres ; the two first ridges have a
slight obliquity in the plane of wear, which slopes from the inner to the outer side :
in the third ridge, which is less worn, and in the two unworn ridges, the centre
of each ridge is the lowest point, showing that the tooth cannot belong to
S. cliftii. In its blunt ridges, low ridge-formula, and small quantity of cement,
the tooth presents all the characters of the molars of Stegodon bomhifrons. It will
be seen below that the corresponding tooth of Stegodon insignis carries a larger
number of ridges, which are also thinner and taller. The corresponding tooth of
S. cliftii is unknown, but its ridges would probably be considerably higher on the
inner than on the outer side, and the whole tooth would also be wider than the
specimen under consideration. The figured tooth seems to agree very closely with
the five-ridged tooth figured by Palconer, but is of slightly larger size, and the
hind-talon relatively rather smaller.

It seems that the third lower milk-molar of Stegodon bomhifrons may have
either five or six ridges, and it is therefore inferred that the corresponding upper
tooth, when discovered, will present either one or both of these numbers.

265—84

SIWALIK AND NAEBADA PBOBOSCIDIA.

First upper true molar. — There is no specimen of the first upper true molar of
Stegodon bombifrons figured in the “ Fauna Anti qua Sivalensis,” neither do any of
the crania of that species in the Indian Museum contain this tooth.

The six-ridged tooth represented in fig. 1 of Plate XLVI of this memoir, is I
think, in all probability, from its low ridge-formula, the first upper true molar of
Stegodon bombifrons ; the tooth certainly does not belong to S. cliftii, and I cannot
find any instance of the corresponding tooth of S. insignis carrying less than seven
ridges ; moreover, the ridges are rather lower in the figm’ed specimen than in the
corresponding tooth of S. insignis.

The figured tooth is the specimen referred by me to Stegodon sinensis
in the “ K-ecords.”^ This determination was made from the curvature of the
ridges, before I found that this was a valueless character, and before I had ob-
tained the cast of the tooth on which the Chinese species was founded by Professor
Owen.

The figured tooth was obtained by Mr. Theobald in the Kangra district, and is
implanted in a fragment of the left maxilla, which also contains the much-worn last
milk-molar. The crown of the tooth carries six complete ridges, and a large anterior
talon. The first four ridges are worn, while the two last are still intact. The
penultimate ridge carries twelve or thirteen tubercles, and, like the other ridges, has
its highest point in the centre. The first three ridges are nearly straight ; but the
three last are much curved. One of the tubercles of the last ridge is placed con-
siderably in advance of the others. There is a slight trace of a median division in
the first two ridges. The length of this tooth is 5 ’9 inches, its greatest width 3
inches, and the height of the penultimate ridge 1‘6 inches.

The height of the ridges in this tooth is slightly less than is generally the case
with those of the corresponding tooth of Stegodon insignis, though the ridges in
some specimens of the latter are hot higher than in the present specimen. It is
therefore only from the low ridge-formula that I consider it probable that this tooth
belongs to S. bombifrons, in which we know that the last molar has a lower formula
than in S. insignis. Another very similar tooth in the Indian Museum, which I
also refer to S. bombifrons, has six ridges, but no trace of a hind-talon. Prom
the convexity of the ridges the figured tooth cannot belong to S. cliftii, of which
the first true molar has been already described.

The last milk-molar in the same jaw is worn away to such an extent that only
the two last ridges are differentiated from the common surface of dentine to which
the others have been reduced. Still, however, from the shape of the crown of this
tooth, I am nearly certain that it could not have carried more than either five or six
ridges, in which case the jaw may pretty safely be referred to the present species,
as the last milk-molar of S. insignis never has less than seven ridges.

I have figured this tooth chiefly to show that a great degree of curvature may
occur in the ridges of species of Stegodons, which are usually quite straight. In the

Kec. Geol. Sui v. India, Vol. XI, p. 73.

SIWALIK AND NARBADxi PEOBOSCIDIA.

85—206

}'>eiiultimate upper true molar of a cranium of S. insignis^ in the Indian Museum,
the ridges are still more curved than in the present specimen.

First loiver true molar. — The two lower molars of a Stegodon represented in
Plates XXIXA, fig. 2, and XXIXB, fig. 6, of the “ Pauna Antiqua Sivalensis,”
are classed by Dr. Palconer as the first true molars of Stegodon homhifrons. Each
of these teeth carries seven low broad ridges and a talon. These teeth carry one
ridge more than the corresponding upper molar, and the same number as the cor-
responding tooth of S. insignis.

Second upper true molar. — In the Indian Museum there are two palate speci-
mens of a Stegodon, with the nearly complete penultimate molars, each of which
carries seven ridges and a talon. These ridges are thick and low, and the valleys
have but little cement. In the smaller specimen the ridges are quite straight, but
in the larger they are bent and irregular, like those of the first upper molar des-
cribed above. Erom the breadth and lowness of the ridges there is no doubt
but that these teeth belong to S. hombifrons. The corresponding tooth of S. insignis
has nearly always eight ridges, which are much higher, and which contain more
cement in the valleys. The length of the tooth in the smaller specimen is 7'4<
inches, and its greatest width 3'4 inches ; in the larger specimen the length of the
corresponding tooth is 8-3 inches, and its greatest width 3*9 inches. The teeth are
much worn and somewhat broken, and are not, therefore, calculated to afford a
good figure.

Second loioer true molar. — A second lower true molar of this species is figured
by Palconer in Plate XXIXA, fig. 3, of the “ Eauna Antiqua Sivalensis.” That
tooth carries seven ridges and a hind-talon, the front one having probably been
worn away. The ridges are low and blunt, and seem to carry but few cusps. A
corresponding tooth in the Indian Museum shows the same number of ridges, and
small fore-and-aft talons. The ridges are much lower and blunter than those of
the corresponding tooth of S. insignis, in the same collection, and the valleys
contain much less cement. The tubercles on the two last ridges are only five or six
in number, while those on the corresponding ridges of the tooth of S. insignis carry
eight or nine. The length of the second lower molar of S. hombifrons is 7' 7 inches,
and its greatest width 3’3 inches. The tooth being somewhat battered would not
afford a good figure ; Ealconer’s figure, though on a small scale, is, however, amply
sufficient to distinguish this tooth.

Third tpper true molar. — The cranium figured in Plates XXVII and XXVIII
of the “ Eauna Antiqua Sivalensis” exhibits the last true molar of the present
species. The tooth of either side carries nine ridges and a talon ; its length is 10' 2
inches and its width 3 A inches. The antero -internal extremities of the molars in
this cranium are only separated by an interval of 1 inch; another skull in the

' This cranium (No. ) is described by Dr. Falconer at page 41 of the “ Catalogue of the Fossil Vertebrata in
the Museum of the Asiatic Society of Bengal.” No notice is, however, taken of the curvature of the ridges, probably
.showing that the author thought it a point of no importance.

267—86

SIWALIK AND NAEBADA PEOBOSCIDIA.

Indian Museum, in which the last molars carry eight ridges and a hind-talon, shows
that this extreme apjDroximation of the molars is due to lateral crushing, since in
the latter skull the antero-internal angles of the molars are separated by more
than 2 inches.

Third lower true molar. — Specimens of the last lower molar of this species are
represented in Plate XXIX A (fig. 8) and XXIX B (fig. 7) of the “ Pauna Antiqua
Sivalensis,” showing that this tooth carried nine ridges and a hind-talon.

Ididge-formula. — Prom the foregoing details of the teeth of Stegodon bombifrons,
theii ridge-formula may he approximately represented as follows : —

Milk-molars. True molars.

.P+4P-K5— 6)? 6+7+(8— 9)

P+4 +(5-6) 7+7+9

This formula wiU he seen to he intermediate between the formulae of S. cliftii
(see above) and S. insignis below). The last milk-molar in some cases presents
a lower number of ridges than in S. cliftii, while the true molars have generally a
liigher number than those of the latter. The last true molar of S. bombifrons has
always a smaller number of ridges than the corresponding tooth of S. insignis, while
the two preceding teeth present, on the whole, a smaller number of ridges, though the
highest number of ridges (seven) which occurs in the teeth of S. bombifrons equals
the lowest which occurs in those of S. insignis.

Cranium.- — A very perfect cranium (now in the British Museum) is figured in
Plates XXVII and XXVIII of the “Pauna Antiqua Sivalensis,” and another less
perfect specimen (also in the British Museum) in the preceding plate of the same
work. A third cranium, lacking a portion of the left frontal region, is in the col-
lection of the Indian Museum, obtained by Mr. Theobald in the Siwaliks of the
Kangra district. This specimen, though of rather smaller size, agrees precisely in
form with the British Museum specimens, so that this form of cranium may be
considered as characteristic of the species. The characters of the skull, as given by
Palconer, are as follows : The skull " is convex from occiput to front, and also
across, and is very narrow at the temporal contraction. The bounding ridges sweep
round by a bold curve into the post-orbitary processes, as in B. meridionalis . There
is a deep furrow between the tusks (which are of small size). The nasal opening
for the trunk is above the line (or nearly so) of the post-orbitary process of the
frontal bone. Above the infra-orbitary foramen on the right side there is another
small opening.”

distribution. — Bemains of this species have not, as far as I am aware, been
found elsewhere than in the sub-Himalayan Siwaliks, and I doubt if they occur to
the west of the Jhelam river.

SIWALIK AND NARBADA PROBOSCIDIA. 87—268

Species 3 : Stegodon insignis, Palconer & Cautley. Plates XLV, fig. 4, XLVI,

figs. 2 & 4.

Synonym ; (?) S. orientalis, Owen.

B-istory. — The first figured tooth of this species seems to he the specimen ob-
tained from Ava, and figured by Mr. Clift in the “ Transactions of the Geological
Society of London,”^ under the name of Mastodon elephantoides. Dr. Palconer
subsequently showed ^ that the latter name really comprehended more than one
species, and gave to the teeth with a high ridge-formula the name of Elephas {Ste-
godon) insignis. Several skulls and a large series of teeth appear under that name
in the “ Pauna Antiqua Sivalensis.” In the same work a considerable number of
teeth also appear under the name of Elephas (Stegodon) ganesa : these latter teeth,
however, cannot he distinguished from those named Stegodon insignis, and as we
shall see subsequently,* it is only the adult skulls of these two very closely allied
species that can he distinguished. The molars of Stegodon insignis also hear a very
close resemblance to those of S. bomhifrons, hut are distinguished by a generally
higher ridge-formula, and by the ridges being narrower and taller, as well as by the
greater amount of cement which is usually present on them. Dr. Palconer^ consi-
ders that the following formula represents the average number of the ridges of the
molars of this species, though he remarks that there is a greater tendency to varia-
tion in this respect than there is among the true Mastodons. Dr. Palconer’s formula
is as follows: —

Milk-molars, True molars.

2-t-5-l-7 7-1-8 -h (10-11)

2 + 54-7 7 + (8—9) + (11—13)

W e shall subsequently see that this formula only requires very slight alteration
as far as our present knowledge goes. As a very large series of the teeth of this
species are figured in the “Pauna Antiqua Sivalensis,” I have only thought it
necessary in this memoir to figure a few specimens of the milk-molar dentition,
which exhibit well the general dental characters of the species, as distinguished
from S. bomhifrons.

First milk-molars. — Of the first milk-molars of Stegodon insignis I know of no
perfect specimens, though a fragment of the upper tooth is attached to the young
maxilla represented in Plate XIX, fig. 1, of the “ Pauna Antiqua Sivalensis.”

Second upper milk-molar. — Of the second upper milk-molar of this species two
specimens are represented in the “Pauna Antiqua Sivalensis;” one in fig. 1 of
Plate XIX and the other in fig. 3 of Plate XXIX (misnamed in plate S. bombi-
frons). The first of these teeth is complete and well worn ; it exhibits five complete
ridges, in front a large talon or incomplete ridge, and behind a small talon. The

* Ser. 2, Vol. II, PI. XXXVIII, fig. 2.
2 Pal. Mem., Vol. II, p. 85.

^ Ibid., p. 86.

269—88

SIWALIK AND NAEBADA PROBOSCIDIA.

ridges are straight, fairly broad, and appear to have carried a large number of cusps
on their unworn summits ; the number of cusps on the yet unworn last ridge seems
to be twelve. The tooth is considerably narrower in front than behind ; its length
is 2’ 7 inches and its greatest width 1’6 inches : the width of the base of the central
ridge being -6 inch.

On page 43 of the “ Catalogue of the Possil Vertebrata in the Museum of the
Asiatic Society of Bengal,” Dr. Palconer describes a second upper milk-molar of an
elephant (No. S. 7) as belonging to S. insignis. This identification appears to me
to be a mistake ; the tooth really belonging to Loxodoii planifrons. The tooth
differs from both the second upper molars of Stegodon insignis figured in the
Pauna Antiqua Sivalensis,” by being less narrowed in front, by the ridges being
six in number, by their being narrower and higher, by the number of cusjds on the
last ridge being only five or six, by there being no hind-talon, and by the much
greater quantity of cement present in the valleys, which completely overlaps the
last unworn ridge. In all the above-mentioned points this tooth agrees with the
teeth of Loxodon planifrons, and in the Indian Museum, where it is now deposited,
it has accordingly been classed as belonging to that species.

Stegodon orientalis, Owen.^ — The fragment comprising the two last ridges and
the hind- talon of a (probably) last upper milk-molar of a Stegodon from China,
figured and described by Professor Owen in the “ Quarterly Journal of the Geologi-
cal Society of London’”^ as the second upper milk-molar of a new species of the
genus, under the name of S. orientalis, does not appear to me to be sufficiently
distinguished from the corresponding tooth of the present species.

Professor Owen, in describing the Chinese milk-molar, and a fragmentary true
molar, which he also refers to the same species, lays great stress on the number of
cusps carried on the ridges of these teeth, which in the larger tooth he estimates at
about a dozen. N ow, I have already shown elsewhere^ that there is a considerable
tendency to variation in the number of cusps carried by the ridges of the molars of
Stegodon ganesa and S. insignis, and in the last lower milk-molar of the latter
species represented in fig. 4 of Plate XL VI of this memoir, some of the ridges
carry at least seventeen cusps, a greater number than occurs in the larger Chinese
tooth. Again, the figured third lower milk-molar of S. insignis in two of its ridges
exhibits a most distinct median cleft, which character Professor Owen considers
distinctive of his Chinese specimen. If the figure of the latter be compared with the
figure of the second upper milk-molar of S. insignis, given in fig. I of Plate XIX
of the “Pauna Antiqua Sivalensis,”^ it will be seen, as far as can be gathered from
the small size of the latter figure, that the two teeth have very much the same
general form and size. The Chinese tooth cannot belong to S. cliftii, as its ridges
are too tall, are highest in the middle, and carry too much cement. The ridges of

' Vol. XXVI, p. 421, PI. XXVIII, figs. 3 and 4.

Rec. Geol. Surv. India, Vol. IX, p. 49.

* The corresponding tooth of the same species represented in fig. 3 of Plate XXIX of the Fauna Antiqua
Sivalensis (wrongly referred in plate to S, bomhifrons) is somewhat broader posteriorly.

SIWALIK AND NARBADA PROBOSCIDIA.

89—270

the Chinese tooth are likewise too high for Stegoclon homhifrons. Although it may
not be a matter of complete certainty until we obtain the entire Chinese milk-molar,
it appears to me that there is every probability that that tooth is really the second
upper milk-molar of Stegodon insignis, and that the imperfect true Chinese molar
referred by Professor Owen to his new species, S. orientalis, also belongs to
S. insignis.

Second lower milk-molar. — Of the second lower milk-molar of Stegodon in-
signis, two imperfect specimens are represented in figs. 1 and 3 of Plate XX of the
“ Pauna Antiqua Sivalensis their distinctive characters are not, however, very
well displayed, neither is there any description of these teeth in the explanation of
the plates. To give a good idea of this tooth, I have had drawn in fig. 2 of
Plate XLVI, a very nearly perfect specimen from the collection of the Asiatic
Society of Bengal, now in the Indian Museum. The tooth was determined by Dr.
Palconer, and a short notice of it will he found on page 44 of his Catalogue of
the Possil Vertehrata in the Museum of the Society. The specimen is implanted
in a fragment of the left ramus of the mandible, and has been somewhat damaged
anteriorly ; most of the ridges have been well worn, hut of the last only the summits
of the cusps are touched by wear. The crown of the tooth is considerably narrower
in front than behind, and carries five ridges (the anterior broken) and a small hind-
talon. Relatively to the corresponding tooth of Stegodon homhifrons,, which has
been already described, the ridges are high and slender ; there is a consideral)le
quantity of cement in the intervening valleys, and the enamel is thrown into deep
vertical groovings and puckerings. The last unworn ridge carries ten cusps. The
length of this tooth is 2’7 inches, its greatest width 1’6 inches, and the height of
its ridges 1 inch. The important points in which this tooth dilfers from the corres-
ponding tooth of Stegodon homhifrons have been already sufficiently pointed out
when describing the dentition of that species.

Third upper milk-molar. — Of the third upper milk-molar of Stegodon insignis,
specimens are figured in the “ Pauna Antiqua Sivalensis,” Plates XIX, figs. 2 and 3,
and XIX A, fig. 1, which show that this tooth carried seven ridges and small fore-
and-aft talons. The very perfect tooth of a Stegodon represented in Plate XLV,
fig. 4, of the present memoir agrees with the last-mentioned specimens in tlie
characters referred to, and as its other characters are also those of the molars of
S. insignis, it has been classed as the third upper milk-molar of that species. The
tooth was collected by Mr. Theobald in the Siwaliks of the Punjab : only its two
first ridges have been affected by attrition, but portions of the summits of most of
the others have unfortunately been broken away. The tooth is narrower in front
than behind, and, as before indicated, carries seven ridges,^ and very small fore-and-

' The last ridge is smaller than the others, and might he reckoned as a talon, hut that it carries a still smaller ridge
behind. This last ridge is evidently a talon in process of passing into a true ridge. The hind-talon in the tooth drawn
in fig. 1 of the same plate will he seen to be relatively shorter in proportion to the true ridges, and is evidently only
a large talon. It will, of course, be seen that the distinction is really an arbitrary one.

'271—90

SIWALIK AND NAEBADA PEOBOSCIDIA.

aft talons. The anterior talon is connected by a longitudinal ridge with the inner
extremity of the first true ridge. Each transverse ridge is tall and slender, and is
highest along the median line ; there is an indistinct trace of a longitudinal median
cleft across the anterior ridges. Owing to the broken condition of the summits of
the ridges, it is not easy to estimate the precise number of cusps borne by each,
hut they appear to he from thirteen to fifteen. The last ridge is more closely
approximated to the penultimate, than is the latter to the antepenultimate. The
hind-talon merely forms a kind of ledge on the last ridge. Both these characters
are shown in Ealconer’s specimen. The length of the specimen is 4’6 inches, its
greatest width 2'4 inches, and the height of the central ridge 1'3 inches. The
number and height of the ridges, together with the great number of cusps borne on
them, are distinctive characters of the dentition of Stegodon insignis as opposed to
8. bomhifrons.

Thwd lower milh-molar. — -Of the third lower milk-molar of the present
species, two imperfect specimens are represented in figs, 1 and 2 of Plate XX of
the “ Eauna Antiqua Sivalensis” ; one of these shows that the tooth has seven
ridges. In Plate XLVI, fig. 4, of the present memoir, I have had figured a quite
complete specimen of this tooth, obtained by Mr. Theobald in the Shvaliks of the
Punjab. This very beautiful specimen is implanted in a fragment of the right
ramus of tfie mandible, which also contains the two last ridges of the penultimate
milk-molar. The figured tooth is oblong in shape, and nearly as wide in front as
behind : its outer border is concave and its inner convex. The crown carries seven
ridges and fore-and-aft talons. Only the first and second ridges have been affected
by wear ; the ridges are tall and slender, and hear a large number of cusps on their
summits ; thus the fourth ridge carries no less than seventeen cusps, a considerably
larger number than occurs in the so-called Stegodon orientalis of Professor Owen,
referred to above. The anterior talon is closely applied to the first ridge ; the pos-
terior talon forms a small distinct ridge, carrying five large cusps. The length of
the tooth is 4‘9 inches, and its greatest width 2’1 inches. .Another similar specimen
of this tooth, also from the Punjab, is contained in the collection of the Indian
Museum : from this we may fairly infer that the number of ridges in this tooth is
generally constant. The tooth is widely different from the corresponding tooth of
S. bomhifrons described and figured above (Plate XLV, fig. 3) ; the distinctive points
are the greater number of the ridges, their thinner and more elongated form, and
the greater number of cusps which they carry on their summits.

Mrst upper true molar. — Of the true molar series of Stegodon insignis, a very
large number of teeth are figured in the “ Eauna Antiqua Sivalensis,” and I have,
therefore, thought it unnecessary to figure any in this memoir, and have contented
myself with merely noticing the salient points of each tooth. A specimen of the
first true molar is shown in the skull depicted in fig. 4 of Plate XIX of the
above-mentioned work. It carries seven ridges and a talon ; the ridges are high,
and the valleys well filled with cement ; each ridge carries a large number of

SIWALIK NAEBADA PEOBOSCTBIA.

91-272

cuspsl Of two young skulls in the Indian Museum containing the third milk- and
the first true molars, the first true molar in the one carries seven ridges and tAvo
talons, and has a length of 5‘7, and a width of 2’6 inches, while in the other the
corresponding tooth has eight ridges with a length of 6*5, and a width of 2'9 inches,
A specimen of a skull from the Asiatic Society of Bengal, now in the Indian
Museum, has seven ridges in the first true molar. It would appear, therefore,
that while seven is the normal number of ridges in this tooth, this number may
occasionally he increased to .eight.

First lower true molm\ — Of the first lower true molar a worn specimen in the
Indian Museum seems to have seven ridges, and has a length of 5‘2 inches, with
a breadth of 2'8 inches ; a very similar specimen also carries the same number of
ridges. A specimen of the first lower true molar of this species represented in
fig. 4 of Plate XXV of the “ Pauna Antiqua Sivalensis” also has seven ridges
and a hind-talon. As far, therefore, as is at present known, seven seems to he tlie
constant number of ridges in this tooth.

Second upper true molar. — A specimen of the second upper true molar repre-
sented in fig. 5 of Plate XIX of the “ Eauna Antiqua Sivalensis” exhibits eight
ridges and a front talon : another specimen in fig. 4 of the succeeding plate of
the same work has also eight ridges, with fore-and-aft talons. A cranium described
on page 41 (Xo. S. 4) of Dr. Ealconer’s “ Catalogue of the Eossil Vertehrata in the
Museum of the Asiatic Society of Bengal,” and now in the collection of the Indian
Museum, exhibits the first and second true molars, the latter of which has only
seven ridges ; the penultimate tooth in the same skull has only six ridges, hut an
anterior ridge may have disappeared in wearing. In the skull figured in fig, 2 of
Plate XIX A of the “ Eauna Antiqua Sivalensis,” there are only seven ridges in the
penultimate true molars. Other specimens of the second upper molar of S. msignis
in the Indian Museum exhibit eight ridges. Eight may, therefore, he taken as
the normal number of ridges in this tooth, with seven as an occasional variety.

Second lower true molar. — Three perfect specimens of the second lower true
molar of Stegodon msignis in the collection of the Indian Museum exliihit respect-
ively, eight ridges and two talons. Two specimens of this tooth represented in figs. 4
and 9 of Plate XX of the “ Eauna Antiqua Sivalensis’’. exhibit, respectively, nine
and ten ridges. Another specimen figured in the same work has seven ridges
(Plate XXA, fig. 2, S. ganesa)\ while yet another (Plate XXIV A, fig. 3) has up-
wards of twelve ridges. The specimen drawn in Plate XXXVIII, fig. 2, of Clift’s
memoir, has nine ridges and a large hind talon. Eight or nine may therefore he
taken as the normal number of ridges in this tooth, with occasional variations from
seven to twelve.

Last true molars. — The number of ridges in the last upper true molar seems to
vary from ten to eleven, and in the lower from eleven to thirteen.

1 The tooth figured in Plate XXIV, fig. 1, of the Fauna Antiqua Sivalensis as the first true molar of S. ganesu.
has only six very broad ridges, with few cusps, and probably belongs to S, bomhifrons.

273—92

SIWALIK AND NARBADA PROBOSCIDIA.

Conclusion. — Prom the above descriptions the ridge-formula of this species
may be tabulated as follows : —

2 + 5 + 7 (7—8) + (7 — 8) + (10 — 11)

2 + 5 + 7 7 + (7 — 12) + (11 — la)

The ridge-formula of this species is constants as far as we at present know, in
the milk-molar series, but is liable to very considerable variations in the true molar
series. In the succeeding sub-genera {Loxodon, Eueleplias) , it will be found that
this variability extends back into the milk-molar series, which seems to be the most
constant in its ridge-formula. In Stegodon insignis the number of cusps on the
ridges seems to be greatest in the milk-molar series, sometimes attaining as many
as seventeen, while in the second true molar, according to Dr. Palconer, ten seems
to be the average number.

Cranium. — The adult cranium of this species is remarkable for the peculiar
flattened form of the vertex, from which character it derives its specific name ; a
further discussion on the skull will be found under the head of the next species.
Dr. Palconer mentions as a character of this species the great height of the ptery-
goids, which form a ridge running up to the base of the orbits. The lower jaw is
much like that of the Indian elephant in general shape ; the rami, however, diverg-
ing to a greater extent posteriorly.

Distribution. — Remains of this species has been obtained from the Siwaliks,
and not improbably from the Narbada gravels.

Species d : Stegodon ganesa, Palconer & Cautley.

Mistory.' — 'This fourth species of Siwalik Stegodon appears to have been mainly
founded on an imperfect cranium with long tusks, presented by the late Colonel
Baker to the British Museum. This cranium is represented on Plates XXI and
XXII of the “ Pauna Antiqua Sivalensis,” and in a restored state on Plate XXIII
of the same work.

Distinctness of the species. — Subsequently to the publication of the Pauna
Antiqua Sivalensis,” Dr. Palconer^ seems to have had considerable doubts as to the
specific distinctness of this cranium, and to have thought that it might be merely
a variety of Stegodon insignis. This conclusion appears to have been arrived at
because no distinction could be found between the molars of Stegodon insignis and
S. ganesa, and also because there appears to be somewhat of a transition from the
typical form of the cranium of S. insignis (P. A, S., Plate XV) towards that of
.S', ganesa. The typical cranium of the former species is characterized by the extra-
ordinary depression of its fronto-parietal region, and by the very small width of the
temporal fossae. S. ganesa, on the other hand, is characterized by a development of

Pal. Mem., Vol. II, p. 84.

SIWALIK AND NAEBADA PEOBOSCIDIA. 93—274

the parieto-frontal region in much the same manner as in the living elephants, and
by the consequently great width of the temporal fossse.

Less typical skulls of S. insignis, as the specimen represented in fig. 1 of
Plate XVII of the “Pauna Antiqua Sivalensis,” and another in the Indian
Museum, have a somewhat greater vertical development of the parieto-frontals.
The temporal fossa in these skulls is, however, still very narrow. Young skulls of
the species, however, have the fronto-parietal region well developed, as in S. ganesa
(P. A. S., Plate XVIII, figs. 1, 2, 3), and it appears, therefore, that the aborted
frontal region of N. insignis is only a characteristic of the adult.

That the one large cranium ascribed to Stegodon ganesa is not a single abnor-
mality is proved by the existence of another very similar cranium, with the tusks
broken off, obtained by Mr. Theobald in the Punjab, and now in the Indian Museum.
This cranium was described by myself in the “ Eecords,”^ under the name of
ganesa. The resemblance of this cranium to that named S. ganesa in the ‘‘ Pauna
Antiqua Sivalensis” is so close that I have thought it unnecessary to give a figure
of it, almost the only difference between the two being that the former has the
frontal region somewhat hollowed mesially in place of being quite flat as in the
latter.

It thus appears that there are two very distinct forms of adult crania of 8te-
godons, which have the same dentition, and which in the young state also appear to
be indistinguishable. On the whole, I now think it very probable that S. ganesa
may be merely a very well marked race of 8. insignis,, or, in other words, may be
a species in process of evolution. The adult race is, however, so well marked
and so distinct that I prefer to continue to apply to it a separate specific name, with
the above proviso.

On page 31 of the tenth volume of the “ Eecords,” I have noticed a speci-
men of a huge tusk of an elephant from the Narbada valley, which from its size
and form seems to belong to 8, ganesa. Prom the same deposits teeth which agree
with those of 8. insignis have been obtained, but no specimen of the cranium, and
it is therefore not certain whether they might not also belong to 8. ganesa. The
discovery of the cranium of either of the 8tegodons in the Narbada valley will be a
matter of great interest, as we shall then see whether they had undergone any
further differentiation in the period immediately succeeding that of the Siwaliks. I
have already commented upon the similarity of certain of the molars of some of
the Indian Mastodons as bearing upon the case of 8. insignis and 8. ganesa^ in
which no distinction can be found in the molars.^

1 Eec. Geol. Surv. of India, Vol. IX, p. 42.

2 It may be mentioned that the molars of two well defined species of Himalayan marmots {Arctomys caudatus
and A. himalayanus) are absolutely indistinguishable, and if found in the fossil state would be referred to one species.

275—94

SIWALIK AND NARBADA PROBOSCIDIA.

Sub- Genus 2 : Loxodon, Dalconer.

Elephants whose molars in most species have a higher ridge-formula than in the
Stegodons, and in which the ridges are thinner and higher, and the valleys more
completely filled with cement.

Species I : Loxodon planifrons, Ealconer & Cautley.

General characters. — This species, together with the African elephant, in the
characters of its molars, forms a link between the higher-ridged Stegodons and the
extinct European Loxodons, the two species having a lower ridge-formula than any
other species of the suh-genus Loxodon. We shall find that the number of ridges
in the molars of this species are liable to a much greater degree of variation than
is the case in any of the preceding species, — a variability correlated with the greater
complexity of the molars of this group. Dr. EalconeE makes the ridge-formula of
this species as follows : —

3 + 6 + 7 7 + 8 + 10

3+6 + 7 7+(8-9)+(10— ]1)

This formula does not give a sufficiently wide range of variation for the number
of ridges, and another will he proposed after a short survey of the teeth.

The enamel ridges on the molars of this species are intermediate in height be-
tween those of Stegodon insignis and those of Euelephas hijsudricus ; when worn the
crowns present lozenge-shaped cross-sections of the ridges, often with detached cylin-
ders of enamel near the median line of the tooth ; the enamel is of great relative
thickness, and much crenulated or crimped in the higher portions of the ridges,
hut interiorly this crimping is absent ; this causes a great difference in the appear-
ance of the crown-surface of a little- worn and a much-worn tooth. The molars of
the species are readily distinguished from those of the Stegodons by the eement
completely filling up the intervals between the enamel ridges. The present species
is further distinguished from all other species of elephants, both recent and fossil,
as far as is at present known, by having been furnished with two pairs of premolars
in both upper and lower jaws,^ in which respect it agrees with many species of
Mastodon.

Crankm. — Eigures of the cranium of L. planifrons will be found on Plates IX
and X of the “Eauna Antiqua Sivalensis.” The cranium is readily distinguished
from that of any other species of elephant by the perfectly flat and expanded
frontal region (whence the specific name), and by the small degree to which the
temporal fossse extend on to this surface. The nasal aperture is of relatively small

^ Pal. Mem., Vol. II, p. 9l.

Fauna Antiqua Sivalensis, PI. VI, figs. 4, 5, 6 ; PI. XII, fig. 8,

SIWALIK AND NARBADA PROBOSCIDIA.

95—276

extent, and the nasals themselves form a short protuberance. The pterygoids are
small, and the incisive sheaths diverge to a considerable extent anteriorly.

Distrihiition. — Dossil remains of this species have been obtained only from the
sub- Himalayan Siwaliks, and it appears that the geographical range of both this
and the succeeding species did not extend into Sind, or into Perim Island, or Burma.
A very large series of the molars of this species are contained in the collection of
the Indian Museum, but as they do not differ from the beautiful series figured in
the “ Pauna Antiqua Sivalensis,” none have been figured in this memoir.

First upper milk-molar. — A specimen of the first upper milk-molar is repre-
sented of the natural size in fig. 1 of Plate XII of the “ Pauna Antiqua Sivalensis
this tooth has three ridges and a talon.

First lower milk-molar. — Of the first lower milk-molar I cannot discover any
specimen : Dr. Palconer gives the number of ridges in this tooth as three, probably
from the number in the corresponding upper tooth.

Second upper milk-molar. — A detached specimen of a second upper milk-molar
of this species in the Indian Museum (formerly in the collection of the Asiatic
Society of Bengal, where it was catalogued by Dr. Palconer as belonging to Stegodon
insignis), has six ridges and a talon: its length is 2‘8 inches, and its greatest width
1-6 inches. Another specimen of this tooth, in a young cranium in the Indian
Museum, seems to have only five ridges.

Second lower milk-molar. — There is no specimen of this tooth in the Indian
Museum ; the specimen represented in fig. 7 of Plate XII of the “ Pauna Antiqua
Sivalensis” has six ridges and a talon: its length is 2-4 inches, and its greatest
width 1’4 inches.

Third upper milk-molar. — The number of ridges in the third upper milk-molar
represented in fig. 4 of Plate VI of the “Pauna Antiqua Sivalensis” is six, with
talons : the length of this tooth is 4‘0 inches, and its greatest width 2’4 inches. A
specimen in the Indian Museum has seven ridges.

Third lower milk-molar. — The specimen of the third lower milk-molar repre-
sented in fig. 8 of Plate XII of the “ Pauna Antiqua Sivalensis,” shows seven
ridges, with talons : its length is 4-4 inches, and its greatest width 2-4 inches. Of
two specimens in the Indian Museum, one shows eight ridges, and the other nine.
The latter specimen was presented by the Boorkee Museum, and is but little worn ;
it exhibits well the characteristically thick enamel plates, widely separated, and
enclosed in a large mass of cement.

First upper true molar. — The germ specimen of the first true molar in the
young cranium represented in fig. 4 of Plate VI of the “ Pauna Antiqua Siva-
lensis” exhibits seven ridges, and has a length of 5‘5 inches with a width of 2-7
inches. I have not seen any other specimens of this tooth ; it is, however, possible
from the number of ridges in the last milk-molars, that the number of ridges in
both upper and lower first true molars may sometimes be more than seven.

277—96

SIWALIK AND NAEBADA PEOBOSCIDIA.

First lower true molar. — Of the first lower true molar, two specimens in the
Indian Museum have seven ridges each : the length of one of these teeth is 6 inches,
and its greatest width 2'8 inches. The specimen of the same tooth represented in •
fig. 10 of Plate XII of the “Pauna Antiqua Sivalensis” also shows the same
number of ridges.

Second upper true molar. — In a very perfect palate specimen in the Indian
Museum, showing the penultimate and last true molars, the penultimate tooth has
eight ridges, with a length of 5-8 inches and a width of 4' 2 inches : a smaller
specimen in the same collection has nine ridges. The three specimens drawn in
figs. 5, 5«, and 6 of Plate XII of the “Pauna Antiqua Sivalensis” have each
eight ridges.

Second lower true molar. — Of two specimens of this tooth in the Indian
Museum, one has eight, and the other nine ridges : the length of the latter tooth is
9 inches, with a width of 3' 5 inches. The corresponding tooth represented in fig. 6
of Plate XI of the “ Pauna Antiqua Sivalensis” carries nine ridges ; the specimen
represented in fig. 8 of the same plate has eleven ridges.

Third upper true molar. — The last molar in the large palate specimen referred
to above has ten ridges, with a length of 11'5 inches and a width of 4'6 inches.
The last true molar in the cranium figured in Plate X of the “Pauna Antiqua
Sivalensis” has eleven ridges.

Third lower true molar. — A perfect specimen of this tooth in the Indian
Museum has twelve ridges, with a length of 10' 5 inches and a width of 4 inches.
The same tooth in the lower jaw represented in fig. 2 of Plate XI of the “ Pauna
Antiqua Sivalensis” has thhteen ridges, while the specimen represented in fig, 12
of Plate XII of the same w^ork exhibits only ten ridges.

Ridge-formula. — The foregoing survey of the molar series of Loxodon planifrons
affords the following as the nearest approximation I can at present give to the
ridge-formula of the species, riz. : —

H/r n/r ^ P P i\/r ^ P P
m. m. 3 + 6 + (7-9) 7 p (8-11) 4- (10-13)

Possibly some alteration will have to be made in the number of ridges in the
first true molar, hut otherwise the number of ridges are probably nearly correct.
It will be observed that the “ intermediate” molars are hepta-and octolophodont,
and that the whole ridge-formula is very close to, though generally slightly higher
than that of Loxodon africanus, as exhibited in the table given in the sequel. Both
these species have a tendency to the prevalence of seven and eight ridges in the
intermediate molars, and are doubtless links between the Stegodons and the higher ■;
ridged Loxodons.

97—278

SIWALIK AND NARBADA PROBOSCIDIA.

Stjb-Genus 3 ; Euelephas, Falconer.

Elephants in whieh the ridges of the molars are developed into tall and nearly
parallel plates, the intervals between which are completely filled with cement.
The number of plates in the last lower molar may reach as many as twenty-four.

Species 1 : Euelephas hysedhicus, Falconer & Cautley.

General characters. — This species is the only representative of the thin-plated-
toothed, or elasmodont, elephants found in the Siwaliks. The cranium and the
dentition are well illustrated in the “ Fauna Antiqua Sivalensis,” and, therefore, I
have not thought it necessary to figure the dentition in this memoir.

The cranium ^ is nearest in form to that of Etielephas indicus, hut is readily
distinguished by the alveoli of the tusks being straighter, as well as by the greater
lateral development of the parieto-frontal protuberances, and by the greater incision
of the temporal fossae.

In describing briefly some of the specimens figured in the ‘‘ Fauna Antiqua
Sivalensis,” Dr. Falconer remarks of the cranium — that the young tusks are oval
in cross-section, that they diverge slightly, and are very near in size to those of the
young Indian elephant, but are narrower in front and more convex. The palatal
bones diverge in front and the infra-orbital foramen is of unusually large size. The
ridge-formula of the molars is much lower than that of the Indian elephant, indi-
cating that the fossil species forms a link connecting the latter with the loxodont
elephants. The plates of enamel and ivory in the molars are lower than in other
species of elasmodont elephants, but are narrow and vertical ; the interspaces occu-
pied by the cement are generally wider than the enamel and ivory plates themselves.
The enamel is usually thicker than in the African elephant and much crenulated.
The molars are at once distinguished from those of Loxodon planifrons by the plates,
being more numerous, narrower, and extending further down towards the root of
the tooth, not being in fact mounted upon a common base of dentine as in that
species. The enamel is also thinner, and the included ellipses on the worn crown
surface much narrower. Specimens of the molars are not unfrequently found,
from which the investing cement has decomposed and fallen off, leaving the enamel
plates standing out separately in a manner which might lead the inexperienced
observer to think that they belonged to a species distinct from that to which the
complete teeth belonged. Great variability exists in the number of plates of the
molars, there having not improbably existed two races, one with considerably more
plates in the molars than the other.

A nearly complete specimen of the mandible is represented in fig, 7 of Plate
XIII A of the “Fauna Antiqua Sivalensis;” it is characterized by a short
symphysis, with a very small and narrow spout : a similar specimen is in the Indian
Museum.

' P. A. S., PI. XII B, fig. 4; PI. XLIII, figs. 20 A and B.

A 1

279—98

SIWALIK AND NARBADA PROBOSCIDIA.

Remains of this species have been found only in the suh-Himalayan Siwaliks
to the eastward of the Jhelum river.

First milk-molars. — Of the first milk-molars, both upper and lower, I can find
no specimens.

Second upper milk-molar. — In one young skull in the Indian Museum the
second milk-molar has five ridges, and in another seven. The same tooth in the
two skulls drawn in Plate VI, figs. 1 and 3, of the “ Pauna Antiqua Sivalensis,” has
five ridges.

Second lower milk-molar. — The second lower milk-molars drawn in Plate VII,
figs. 5 and 7 of the last-named work, have either seven or eight ridges, according
as one is reckoned as a true ridge or a talon : a specimen doubtfully classed as
the same tooth, drawn in fig. 6 of that plate, has nine ridges.

Third upper milk-molar. — In three young crania in the Indian Museum, the third
milk-molar exhibits, respectively, eight, ten, and eleven ridges. The same tooth in
the skull figured in plate VI, fig. I, of the “ Pauna Antiqua Sivalensis ” has
seven ridges, with fore-and-aft talons. Other specimens figured or referred to have
the number of ridges between seven and nine.

Third lower milk-molar. — Two specimens of the third lower milk-molar drawn
in figs. 8 and 9 of plate VII of the “ Pauna Antiqua Sivalensis ” have each nine
ridges : a specimen of the same tooth in the Indian Museum (presented by the
Roorkee Museum) has eleven ridges and two talons.

First upper true molar. — In a specimen of the cranium of F. hysudricus
referred to in a note (No. 3) to the description of Plate VI of the “Pauna Antiqua
Sivalensis ” the number of ridges in the first true molar is given as eight ; Dr. Pal-
coner remarks that this tooth is remarkable for its unusual width, and the small
number of ridges which it contains ; a detached specimen of the same tooth figured
in Plate VII, fig. 2, of the same work, has twelve ridges.

First loioer true molar. — A specimen of the first lower molar in a jaw figured
in Plate VIII, fig. 4, of the “ Pauna Antiqua Sivalensis ” has ten ridges, with a
small hind- talon; another specimen drawn in fig. 10 of the preceding plate of
the same work has twelve ridges.

Second upper true molar. — The second upper molar in the cranium drawn in
Plate VIII, fig. 1, of the “Pauna Antiqua Sivalensis,” has ten ridges and a large
heel ; the specimen of the same tooth drawn in fig. 3 of the preceding plate of
the same work has twelve ridges.

Second lower true molar. — Two lower jaws in the Indian Museum, in which
the second true molar is in wear, exhibit twelve and thirteen ridges, respectively, in
that tooth.

Third upper true molar. — In two crania in the Indian Museum, in which the
last true molar is in use, the tooth in one instance has thirteen ridges and talons,
and in the other seventeen ridges.

SIWALIK AND NAEBADA PEOBOSCIDIA.

99-280

Third lower true molar. — Three specimens of the last lower molar in the Indian
Museum exhibit, respectively, fourteen, fifteen, and sixteen ridges. The specimen
drawn in fig. 12 of Plate VII of the “Pauna Antiqua Sivalensis ” shows seventeen
or eighteen ridges.

Ridge-formula. — Prom the above series of teeth, the ridge-formula of the
molars of E. hysudricus may approximately be represented as follows : —

ivr AI ? + (5-7) + (7-11) ^ (8-12) + (10-12) + (13—17)
ivx. lU. p (9—11) (10-12) + (12—13) + (14—18)

Prom the general symmetry of this formula, it is probable that it presents a very
fair approximation to the truth. It will he found from the table of formula given
below, that the position assigned to the species from this formula is between
E. mnaidriensis and E. antiquus, or the second in the list of true elasmodont ele-
phants : as being a pliocene species this ridge-formula, as being lower than that of
the pleistocene species (with one exception, which will be referred to again below)
is in accordance with the theory of evolution, and the less specialization of the older
forms of a genus. The height of the ridges of the molars of this species is much less
than in the Indian elephant : the height of the eighth ridge of the last upper molar
of the former species averaging about 5-| inches, and in the latter upwards of 8
inches ; the plates of the one are also much thicker than those of the other.

Species 2 : Eublephas namadicus, Palconer & Cautley.

The remains of this species of elephant have been obtained only from the Pleis-
tocene deposits of the Narbada valley, there being no traces of its existence in the
earlier Siwalik rocks. A considerable number of the true molars are figured in the
“Eauna Antiqua Sivalensis,” but, with the exception of the last lower milk-molar,
no complete specimens of the milk-molars are there figured. The collection of the
Indian Museum unfortunately contains no specimens of the milk-dentition, with the
exception of the last lower milk-molar ; so that the milk-molar dentition is vir-
tually still unknown. A second upper true molar in the Indian Museum has twelve
complete plates. , The specimens in the laiter collection, and those figured in the
“ Eauna Antiqua Sivalensis,” enable me to make the following approximation to the
ridge-formula of this species, viz. : —

Milk-molars. True molars.

? + ? + ? .P + 12 + 18

? + ? + (9—10) (i2— 13) + (12—15) + (19—20)

This ridge-formula will be found to be extremely close to that of the dentition
of the Em’opean Elephas antiquus, as given by Professor Leith Adams and from
this resemblance of the ridge-formula, and from the great similarity of the teeth of
the two species, that writer has come to the conclusion that the Narbada elephant
is probably a variety of E. amtiquus. With regard to this conclusion, in the first

'“Dentition and Osteology of Elephas antiquus,” p. 47. Palaeontograph. Soc., 1877.

281—100

SIWALIK AND NAEBADA PROBOSCIDIA.

place, as we have seen, the milk-molars of the Narbada species are, with one excep-
tion, unknown, and it is possible that their ridge-formula may differ somewhat
from that of the milk-molars of E. antiquus. Secondly, there is the very peculiarly
shaped cranium of E. namadicus, characterized by the great supraorbital ridge, well
exhibited in the large cranium figured in plate XIIB of the “ Bauna Antiqua Siva-
lensis,” in another craniuni in the British Museum, and in two crania in the
Indian Museum. All the other crania of E. namadicus in the latter collection, and
I believe also in the former, are imperfect superiorly, so that in all the complete
crania known to me this very characteristic frontal ridge occurs, and it may, there-
fore, he fairly considered as characteristic of the Indian species. In noticing this
peculiarity of the cranium of the Narbada elephant. Professor Leith Adams suggests^
that it may he due to distortion, a view which, I think, is disproved by the facts
given above.

Of Elephas antiquus Professor Leith Adams, in the above quoted passage, says,
he is unacquainted with any English cranium, but refers to one in the museum at
Borne, in which I infer, though it is not clearly expressed, that there is no ridge,
like that of the Narbada elephant. If the cranium of the European elephant have
no such ridge, while it is constant in the Indian form, I cannot think it by any
means proved that E. namadicus and E, antiquus are varieties of one species, though
there can be no doubt but that, as was admitted by Ealconer, they are extremely
closely allied, and possibly that the molars of the two are indistinguishable in a
large series.

With regard to the great similarity of the teeth in the two species, we have
already seen in the case of Stegodon ganesa and S. insignis that the teeth of two
forms of elephant may be indistinguishable, while the crania are very widely
different. In the lower true molars of E. namadicus there seems to be a tendency
to a somewhat higher ridge-formula than in E. antiquus?

As the crania and molars of Eleplias namadicus in the collection of the Indian
Museum do not differ from those figured in the “ Eauna Antiqua Sivalensis,” I have
not considered it necessary to give figures of any of them on the present occasion.

The molars of this species are stated by Ealconer to differ^ from those of E.
hysudricus, in the greater height of the ridges or plates, in the slight amount of their
thinning superiorly, and in their nearly vertical direction. There is also no loop near
the middle of the plates on the grinding surface in the enamel, and the cement is
thinner. The molars differ from those of E. indicus (apart from the difference in
the number of ridges), by the worn dentine surfaces being thicker and presenting no
curve towards the apex ; the enamel is also thicker. In the crimping of the enamel
plates the two species are very much alike.

' loc. cit., p. 52.

^ At page 68 of his memoir on E. antiquus, Prof. Adams seems to be under the impression that E. namadic.us
is found in the Siwaliks of Northern India and extends back into the Miocene. Eemains of the species have' only been
found in the presumably Pleistocene deposits of the Narbada valley ; E. namadicus is probably later in time than E.
antiquus, and not earlier, as is stated by Prof. Adams.

SIWALIK AND NARBADA PROBOSCIDIA. 101—282

TABLE SHOWING DIMENSIONS OF MOLARS OP INDIAN MASTODONS.

Mastodon

palconeei.

Mastodon

PANDIONIB.

Mastodon

LATIDBNS.

Mastodon

PEEIMBNSIS,

Mastodon

SlYALENSIS.

Upper.

Lower.

Upper.

Lower.

Upper.

Lower.

Upper.

Lower.

Upper.

Lower.

Length of first milk-molar

1-2

1-3

0-6

Width of ditto ditto ....

• ?

1-0

1-1

. 0-4

Length of second ditto ....

1-8

2-12

2-2

1-6

2-05

2-3

1-9

Width of ditto ditto ....

1-6

1-45

1-8

1-5

1-7

1-8

1-2

Length of third ditto ....

3-04

3-4

3-7

3-2

3-4

Width of ditto ditto ....

2-3

2-2

2-0

1-8

Length of first true molar

3-5

4'1

4-0

4-2

4-0

4-2

3-8

4-2

Width of ditto ditto . ‘ .

2-2

2-3

2-5

2-3

2-4

2’2

2-4

Length of second ditto

5-2

6-4

4-2

6-3

5-4

6-5

Width of ditto ditto ....

3-5

3-7

2-6

3-6

2-8

.2-9

2-5

Length of third ditto . . . . i

7-7

7-4

8-6

9-6

10-7

7-8

8-8

Width of ditto ditto . . . .

3-3

3-0

4-2

3-7

4-4

3-3

2-9

Length of penultimate premolar

1-6

Width of ditto ditto

1-4

Length of last ditto

2-05

2-2

2-4

Width of ditto ditto

1-5

1-8

1'7

TABLE SHOWING DIMENSIONS OP MOLARS OF STEGODONS.

SlEGODOl

ff CLIETII.

StEGODON BOMB!-
FEONS.

SlEGODON INSIGKIS
and GANESA.

Upper.

Lower.

Upper.

Lower.

Upper.

* Lower.

Length of first

milk-molar

Width of ditto

ditto

Length of second

ditto

2-8

2-2

2-7

Width of ditto

ditto

2-1

1-4

1-6

1-55

Length of third

ditto

4-9

• P

4-1

4-6

4-9

Width of ditto

ditto ......

2-9

2-2

2-4

2-1

Length of first

true molar.-

6-1

5-9

6-4

5-7

5-2

Width of ditto

ditto

3-3

3-0

2-9

2-6

2-8

Length of second

ditto

8-2

7-4

8-5

9-7

9-3

Width of ditto

ditto

4-0

3-4

3-2

3-5

3-2

Length of third

ditto

9-3

12-7

10-2

13'4

11-0

12-2

Width of ditto

ditto

» . ..

4-3

4-5

3-7

4-2

3-8

3-5

283—102

SIWALIK AND NARBADA PROBOSCIDIA.

TABLE OF RECENT AND FOSSIL PROBOSCIDIAN
Family I. — FlNOTSEBIDuF.

Genus: DINOTHERIUM (Kaup).

1. D. giganteum (Kaup.)

2. D. indicum (Falconer)

3. D. honigii (Kaup.)

? D. bavaricum (Meyer)
? D. cuvieri (Kaup.)

4. D. fentafotamicB (Falc.)

5. D. sindiense (Lyd.)

. Europe. Miocene.

. India. Mio-Pliocene.
. Europe. Miocene.

. India. Mio-Pliocene.
. India. Miocene.

Family II.—FLEFEANTID^.
Genus I : MASTODON, (Cuvier).

Teilophodon

Tktralophodon

3. M. andium (Cuvier)

4. M. angustidens (Cuvier)

M. simorrense (Lartet.)

Jll. cuvieri (Pomel.)

5. M. horsoni (Hays)

M. huffonis (Pomel.)

6. M.falconeri (Lyd.)

7. M. humboltii (Cuvier)

8. M. maximus (Cuvier)

M. oliioticus (Blum.)

M. giganteus (Auct.)

9. M. obscurus (Leidy.)

? M. chapmani (Hays)

10. M. 'pandionis (Pale.)

11. M. pentelici (Lart. & Gaud.)

12. M. productus (Cope)

13. M. goyrenaicus (Lart.)

114. M.'tapiroides (Cuvier)

M. turicensis (Schinz.)

15. M. virgatidens (Meyer)
f 16. , M. arvernensis (Croiz & Jobert)

IM. brevirostre (Gervais).

17. M. dissimilis (Jourdan)^

18. M. latidens (Clift)

19. M. longirostris (Kaup.)

20. M. mirificiis (Leidy)

21. M. sivalensis (Falc. & Caut.)
22. M. perimensis (Falc. & Caut.)

. S. America. Pliocene.
. Europe. Miocene.

. Europe. Pliocene.

. India. Mio-Pliocene.

. S. America. Pleistocene.
. N. America. Pleistocene.

. N. America. Miocene(?).

. India. Mio-Pliocene. Pleistocene(?)
. Europe. Miocene (?)

. N. America. Pliocene.

. Europe. Miocene.

Europe. Miocene.

. Europe. Upper Pliocene.

. India. Mio-Pliocene.

. Europe. Miocene.

. N. America. Pliocene.

. India. Mio-Pliocene.

* I believe this table to include all the species of which descriptions have reached India ; there may be other
species described lately which I have not seen. Only a few of the most important synonyms are given. — E. L.

* M. dissimilis and M. longirostris seem to have a tendency to the development of an extra ridge in the third

and fourth molars. ’ •

SIWALIK AND NAEBADA PEOBOSCIDIA.

103—284

Genus II: ELEPHAS (Linn^).
Sub-Genus 1 : Stegodon (Ealconer).

23. S. homhifrons (Falc. & Caut.)

24. 8. cliftii (Falc. & Caut.)

? S. sinensis (Owen).

25. 8. ganesa (Falc. & Caut.)

26. 8. insignis (Falc. & Caut.) .
? 8. orientalis (Owen).

India. Mio-Pliocene.

India and China. Mio-Pliocene.

India. Mio-Pliocene and Pleisto-
cene.

India and China. Mio-Pliocene
and (?) Pleistocene.

Sub-Genus 2 : Loxodon (Ealconer).

27. L. africanus (Blum.)

28. L. melitensis (Falc.)

L.falconeri (Busk.)

29. L. meridionalis (Nesti.)

30. L. planifrons (Falc. & Caut.)

. Africa. Eecent.

. Europe. Pleistocene.

. Europe. Pliocene.

. India. Mio-Pliocene.

Sub-Genus 3 : . Euelephas

31. E. americanus (DKay)

E. columbi (Falc.)

32. E. antiquus (Falc.)

? E. intermedius (Jourdan, MSS.)

E. prisons (Gold.)

33. E. armenaicus (Falc.)

34. E. hysudrieus (Falc. & Caut.)

35. E, indicus (Linne )

E. sumatranus (Temminck.)

36. E. mnaidriensis (Adams.)

37. E. namadicus (Falc. & Caut.)

38. E. primigenius (Blum.)

(Ealconer).

. . N. America. Pleistocene.

. . Europe. Pliocene.

. Europe. — ?

. . India. Mio-Pliocene.

. . India. Kecent.

. . Europe. Pleistocene.

. . India. Pleistocene.

■. Europe, Asia, and N. America.
Pleistocene.

Table showing the number of ridges in the molars of the Pboboscidia.

Genus : DINOTHERIUM.

Milk-molars.

True molars.

All species

... H-2-f3

3+2 + 2

1 + 2 + 3

3 + 2 + 2

Genus: MASTODON.

Section A. — Trilophodon.

Normal species

... l + 2-f3

3 + 3 + 4

1 + 2 + 3

3 + 3 + (4 — 1

286—104

SIWALIK AND NARBADA PEOBOSCIDIA.

Section B. — Tetralophodon.

Milk-molars.

True molars.

Normal species

... 2+3 + 4

4 + 44- 5

2 + 3 + 4

4 + 4 + 5

Af. latidens ...

... 2 + 3 + 4

(4-5) + (4—5) + (5—6)

2 + 3 + 4

4 + (4—5) + (5— .P)

M. sivalensis

... 2 + 3 + 4

4 + (4—5) + (5-6)

2 + 3 + 4

(? 6) + (4-5) + (5-6)

Genus: ELEPHAS.

Sub-Genus 1: Stegodon.

S. cUftii

? + 4 +6

6 + 6 + (7-8)

? + 4(?) + 6(?)

6? + 6? + (7—8)

S. bombifrons

... ? + 4 (?) + (5-6 (?) )

6 + 7+ (8-9)

? + 4 +(5-6)

7 + 7+ 9

S. insignis ^ )

8. ganesa j

^2 + 5+ 7

(7-8) + (7- 8) + (10-11)

12 + 5+7

7 + (7-12) +(11-13

Sub-Genus 2 : Loxodon.

L. africanus

... (2-3) + (5-6) + (4-7)

7+8 +10

(2-3) + (5-6) + (4-7)

7 + (8—9)+ (10—12)

L.planifrons

... 3 + (5-6) + (6—7)

7 + (8— 9) + (10—11)

3 (?) + 6 + (7-9)

7 + (8-11) +,(10—13)

JL. meridionalis

... 3 + (5-6) + 8

(8—9) + (8—10) + 13

3+ 6 +8

8^ + (8- 9) + (11-15)

L. melitensis

...

... 3 + 5+8

8 + 9 + 12

3 + 5+ 8

8+9 + 12

Sub-Genus 3 : Euelephas.

JE. mnaidriensis

... (2-3) + (4-6) + (8—9)

(8— 9) + 10 + (12—13)

(2-3) + v5-6) + (8-9)

(8—10) + 10 + (12—13)

hgsudricus

... ? + (5-7)+ (7-11)

( 8—12) + (10—12) + (13—17)

? + (7—9) + (9-11)

(10—12) + (12—13) + (14^18)

E. antiquus ...

... (2-3) + (5-7) + (8-10)

( 9—12) + (12-13) + (15—20)

3 + (6-8) + (9—11

(11—12) + (12—13) + (16-19)

E. namadicus

...? + ?+ ?

? + 12 + 18

? + ? + (9—10)

(12—13) + (12—15) + (19—20)

E. americanus

4 + 8+12

12 + 16 + 20

4 + 8 -k 12

12 + 16 + 20

E. indicus ...

... 4 + (7-9)+ (11-13)

(12—14) + (16—18) + 24

4 + (7—8) + 12

(12—14) + (16—18) + (24—27

E. armeniacus

? + ? + ?

? + ? + 24

? + ? + ?

E- primigenius

...

... ? + (3—4) + (6—9) + (9—12)

(9—15) + (14-16) + (18—27)

2 + (3—4) + (6—9) + (9—12)

(9-15) + (14—16) + (18-27)

SIWALIK AND NAEBADA PEOBOSCIDIA.

105—286

Note. — In the foregoing table the ridge -formulae of the molars of the Indian fossil Proboscidia
have been chiefly compiled from specimens in the Indian Museum, from those figured in the
Fauna Antiqua Sivalensis/^ and from Dr. Falconer’s notes. In the other species the formulae have
been taken from the' valuable memoirs of Professors Busk^ and Adams ^ in the publications of the
Zoological and Palaeontographical Societies, and from the notes of Dr. Falconer. When, among the
elephants with a high ridge-formula, a species appears (like E. americanus) with only a single cipher in
each term, the formula must be regarded as only an approximation to the true one.

Conclusion.

At the close of this survey of the fossil Indian Froboscidia, a few general
thoughts present themselves, which we will now consider.

We find that in the Siwalik period there lived in India three species of Bino-
therium, five of Mastodon, four of Stegodon, one of Loxodon, and one of EuelepJias.
In the succeeding Narbada period, this proboscidian fauna, which is the richest in
the world, was greatly reduced in number, and was only represented by possibly two
species of Stegodon and one of Euelephas, the two former being Siwalik species and
the latter new. It is, however, possible that Mastodon pandionis may have lived in
the same period, as its teeth are said to have been found in the Deccan. In modern
India this fauna has dwindled down to one species of Euelephas, — a species which is
not known before the period of the recent alluvium.

We find that the whole of these species of Frohoscidia (unless we accept Pro-
fessor Leith Adams’ identification of JE. namadiciis with E. antiquus) are peculiar to
India, but that many of them are represented by allied species in the Tertiaries
and Post-Tertiaries of Europe.

The following table exhibits this relationship : —

Indian.

Binothenum, 3 sp.

Mio-Pliocene.

Mastodon pandionis .

Mio-Pliocene, Pleistocene (?) .
M. falconeri
Mio-Pliocene.

M. latidens
Mio-Pliocene.

M. perimensis .

Mio-Pliocene.

Eueopean.

. No very closely allied species.

Genus of upper Miocene age.

. M. angustidens.

Upper Miocene.

. No closely allied form, though approaching
last species in form of molars.

. No aUied species.

f M. dissimilis^.

Upper Pliocene.

j M. longirostris.
h Upper Miocene.

' Trans. Zool. Soc., London, Vol. VI.

2 “ British Fossil Elephants” Pal. Soc,, 1877-79. “ Maltese Fossil Elephants” : Trans. Zool. Soc,, London,

Vol. IX.

^ Of M. dissimilis I have only seen the figures of the molars given in Vol, II of the “ Archives du Museum
D’Histoire NatureUe de Lyon the letter-press having not yet reached India, The molars seem to he of the type of
M. longirostris, hut have less distinct trefoils.

C 1

287—106

SIWALIK AND NAEBADA PEOBOSCIDIA.

Indian.

M. sivalensis
Mio-Pliocene.

Stegodon ....

Mio-Pliocene and Pleistocene.
Loxodon jplanifrons
Mio-Pliocene.

Hueleplias hysudrious
Mio-Pliocene.

Enelephas namadims .

Pleistocene.

Enelephas indicus
Recent.

Eueopean.

. M. arvernensis.

Upper Pliocene.

. No European representative of the sub-
genus.

. No closely allied European species.

Sub-genus Pliocene to Recent.

. No closely allied European species.

Sub-genus Pliocene to Reeent.

. Enelephas antiquus.

Pliocene.

. Enelephas primigenins.

Pleistocene.

With regard to the age of the Siwaliks as indicated hy the Prohoscidia, we
find that these deposits contain the genns Dinotlierium, which is characteristic of
the upper Miocene of Europe. Of the Mastodons we find a mingling of Miocene
and Pliocene forms in the Siwaliks. Stegodon is unknown in Europe, while Loxo-
don and Enelephas do not occur before the Pliocene in Europe.

Another fact, however, presents itself. It will he found from the distribution
given in the text, that the whole of the species and genera of Indian Prohoscidia
which are represented hy allied species, in the Miocene of Europe (with the possible
exception of M. pandionis) are found only in the Siwaliks of Sind, the Western
Punjab, and Perim Island. On the other hand, the species which are represented hy
allied species in the post-Miocene of Europe are mainly found only to the east of
the Punjab ; the only exception being M. sivalensis, which is also found somewhat
to the west of the Jhelum, but not in Sind.

These facts therefore, as far as the evidence of one group of fossils goes, would
show that the whole mammaliferous beds of Northern India range from the upper
Miocene well into the Pliocene period, and that the strata on the east of the Jhelum
are mainly of Pliocene age, while those to the west of that river are partly, at all
events, of Miocene age.

These conclusions precisely agree with those which I have arrived at elsewhere,
from other evidence, as to the age of the Siwaliks, and there is, therefore, a strong
presumption of then? correctness.^

* The view that the Siwaliks are in great part of Pliocene age seems to be gradually gaining ground among
palaeontologists. In a lecture on the “ Extinct Mammals of North America,” delivered by Professor Flower, before
the Eoyal Institution, on March 10th, 1876, the author speaks of the Siwaliks as belonging to a transitional period
between the Miocene and Pliocene (p. 12). Professor Riitimeyer in his memoir on “ Die Hinder der Tertiar-Epoche”
(Abhand. der Schweiz Pal. Gesell., Vol. V, 1878, pp. 184 — 89,) adopts a similar view. Professor 0. C. Marsh in his
paper on the “ History and Methods of Palaeontological Discovery” (Amer. Journ. of Sci. and Arts, p. 348, Novem_
her 1879), refers to the Siwaliks as being of Pliocene age.

Professor Le Conte, however, in his recently published “Elements of Geology,” p. 498 (New York, 1879), totally
ignores the partially Pliocene age of the Siwaliks, and classes them altogether as Miocene. He also ignores the nu-
merous additions to the mammalian fauna of those rocks made hy myself, and noticed in the previous parts of this
volume, and in the “ Records” of the last few years. In the last edition of Professor Nicholson’s “ Palaeontology”

SIWALIK AND NARBADA PROBOSCIDIA.

107—288

In the list of Prohoscidia given aliove, out of a total of thirty-eight species,
no less than sixteen, or nearly one-half, belong to India, and fourteen of these to
the Siwalik period. There is no better instance than the Prohoscidia of the enor-
mous richness of the Siwalik fauna in large forms, and of their almost total
extinction at the present day. The cause of this extinction has been attributed to
the glacial period by Professor Huxley,^ who was followed by Mr. Wallace.^

In the table of ridge-formulge of the sub-genus Bueleplias given above, one
very noticeable point presents itself, which I will shortly refer to. It will be ob-
served that, with the remarkable exception of E. mnaidriensis, the arrangement of
the species from their ridge-formulse, corresponds very closely with the order of
their appearance in time, the two Pliocene species coming first, and then the
Pleistocene and Recent species. Euelephas mnaidriensis occrn^s somewhat out of its
position in time, in the ridge-arrangement : this, however, must probably be re-
garded as a special instance of an earlier development of the higher type.

Again, E. primigenius attained a higher development of its ridge-formula than
its successor in time, the Indian elephant.

With regard to the last-named species, it may be well to notice that Professor
Boyd Dawkins® has recently expressed his opinion that E. primigenius is the parent
of E. indicus, and that the latter is specifically identical with E. americanus {colum-
hi) and E. armeniacus. I have not myself had an opportunity of examining a large
series of the molars of these species, and the following remarks are, therefore, submit-
ted with deference. In the first place, it seems to me somewhat improbable that the
whole of the fossil Indian elephants were entirely swept away and replaced by a
western form, which E. indicus must be according to Professor Dawkins, this replace-
ment having probably taken place within the human period, as we have no evidence of
the existence of the last-named species previous to that period in India. Purther,
according to Dr. Palconer, the molars of E. primigenius are broader, and have thinner
and more numerous plates than those of E. indicus, showing that the supposed descent
would be retrogressive instead of progressive. Dr. Palconer ^ even goes so far as to
place the two forms in distinct groups, E. primigenius belonging to his eurycoronine
and E. indicus to his stenocoronine group. Again, the highly curved tusks of the
mammoth are strangely unlike those of the Indian elephant.® I cannot at present
indicate the line of descent of the latter species from any of the fossil Indian forms,
but it may be observed that both the Narbada E. namadicus and the Siwalik

(1879), the Siwaliks are generally classed as upper Miocene, and none of the mammals discovered since Falconer’s time
are noticed.

' Presidential Address to Geological Society of London, 1870, P. G. S. L., p. Ivii, 1870.

^ “ Geographical Distribution of Animals,” Vol. I, p. 150. In the “ Manual of the Geology of India” (p. 587),
Mr. W. T. Blanford attributes the origin of the theory of the extinction of the great mammals by the glacial period to
Mr. Wallace, whereas it seems to have been first put forth by Professor Huxley in the passage quoted above.

^ Quar. Jour. Geol. Soc., London, 1879. p. 145.

Pal. Mem., Vol. II, p. 13.

* I am informed by Mr. A. 0. Hume that a pair of tusks of the Indian elephant in the possession of one of the
native princes are curved like those of the mammoth.

289—108

SIWALIK AND NARBADA PROBOSOIDIA.

E. hysudricus have simpler molars than those of the Indian elephant, and that any
descent from these forms would he in the order of regular progressive development .
With regard to the first named species, I do not think that the Indian elephan^
could he descended from it, because the Narbada elephant lived to a very recent
date in India, and because of its very peculiarly shaped skull. E. hysudricus^ on the
other hand, lived at an earlier period, and has a skull which only requires a little
modification to bring it to that of the Indian elephant. It must also he remem-
bered that in Northern India, where all the Siwalik elephants lived, we have (with
one or two possible exceptions) no record of the animals which lived subsequently
to the Siwahk (Pliocene) period, and we have therefore no means of saying whether
E. hysudricus then became extinct, or whether it lived on and became developed
into a more specialized form. I do not put forward these suggestions as having
any certainty, hut merely as points for consideration before the two species can he
classed together. I may add that in the discussion which followed the reading of
Professor Dawkins’ paper. Professor L. Adams expressed his opinion that E. primi-
genius and E. indicus were decidedly distinct, hut that E. americanus {columbi),
E. armeniacus, and E. indicus might possibly he the same species.

The growing tendency which there appears to be among mammahan palae-
ontologists to take Httle heed of more or less minute points of difference, and to
unite distinguishable forms under one species, appears to me to he a retrograde
movement, which is especially confusing to the geologist, as it deprives him of all
assistance in identifying strata by their included remains. The identity of the Indian
elephant and the mammoth is no new idea, hut was originally adopted by De Blain-
ville,^ who said that he could not distinguish between the two. The elaborate
labours of Palconer subsequently showed the points which distinguish the mammoth
from other allied elephants, labours which seem in a fair way of being disregarded
by modern writers.

References to non-Indian species oe fossil Proboscidia.^

Dinotherium giganteum.

Kaup : “ Ossements Fossiles des Mammiferes qui se trouvent au Museum grand-ducal de Darmstadt,
Darmstadt, 1832.

Dinotherium honigii.

Kaup ; Art. 15, pp. 5 — 14.

Meyer : Jahrbucli, 1831, p. 296, 1836, p. 59.

{D. havaricum).

Mastodon andium.

“ Hugh Falconer, Palaeontological Memoirs of the late, ” by C. Murchison, London, 1868, Vol. I
p. 99, II, p. 14.

1 “ Osteographie” des Elephants, p. 222.

* This list of references is not meant to make the least approach to a complete bibliography of the fossil Proboscidia, but
is merely intended to enable the reader to make comparisons between the Indian species figured in this volume with the non-
Indian species. .

SIWALIK AND N AHEAD A PEOBOSCIDIA.

109—290

M. angustidens.

M. Vacek : “Abhand. der k. k. Raichs.,” Wien, Vol. VII, pt. 4, 1877. Biedennau : “ Abliand.

der Schweiz pal. Gesell.,” Vol. Ill, 1876.

Falconer: “ Pal. Mem.,” Vol. I, p. 89, II, pp. 14, 21.

E. Sismonda : “ Osteographia di un Mastodonte angustidente.” Mem. Acc. real, di Torino, Ser. II, Vol.
XII, 1852.

Meyer; “ Palseontograpbica,” Vol. XVII, Cassel. Lortet and Chantre, Archiv. du Mus. d’Hist. Nat. de
Lyon, Vol. II. Lyons, 1878.

M. horsoni.

M. Vacek -. loc. cit.

Lortet and Chantre : loc. cit.

M. JiumhoUii,

Meyer : loc. cit.

Lortet and Chantre : loc. cit.

M. maximus.

Warren: “Descrip, of skeleton of M. giganteus (maximus).” Boston, 1852.

Lortet and Chantre : loc. cit. {M. ohioticus.) Leidy : “ Contributions to Extinct Vertebrate Fauna of the
Western Territories.” Washington, 1873, Vol. I, p. 237. {M. americanus).

M, ohsctirus.

Leidy : “ Contributions to Extinct Vertebrate Fauna of the Western Territories.” Washington, 1873,
Vol. I, PI. XXI, p. 231.

M. pentelici.

Gaudry: “ Aniraaux Fossiles et Geologic de I’Attique.” Paris, 1862.

M. productus.

Cope: “U.S.Geograph. Survey, W. of 100th Meridian,” Vol. IV, Pt. II, p. 306, Pis. LXX— LXXI.
Philadelphia, 1877.

M, pgrenaicus, Lartet, MSS.

M. tapiroides.

Gaudry : “ Animaux Fossiles et Geologie de I’Attique.” Paris, 1862. {M. turicensis).

M. Vacek : loc. cit.

Meyer : loc. cit. {M, hiricensis).

Lortet and Chantre : loc. cit.

M. virgatidens.

Meyer : loc. cit.

M. arvernensis.

Croizet et Jobert, “ Eecherches sur less Ossemens Fossiles du departement du Puy-du-dome,” Paris, 1828.
M. Vacek : loc. cit.

Falconer : Pal. Mem., Vol. II, pp. 14, 26, 39.

M. dissimilis.

Lortet and Chantre : loc. cit.

M. longirostris.

Kaup ; loc. cit.

M. Vacek : loc. cit.

Lortet and Chantre ; loc. cit.

Falconer : Pal. Mem., Vol. I, pp. 159, 107, 468, 472, II, pp. 14, 23.

M. mirificus.

Leidy : loc. cit., p. 237.

“ Extinct Mammalian Fauna of Dakota and Nebraska.” Jour. Acad. Nat. Sci. Philadelphia,

Ser. II, Vol. VIII, p. 249, PI. XXV, figs. 1, 2. Philadelphia, 1869.

JLoxodon melitensis.

L. Adams : “ Maltese Fossil Elephants,” Trans, Zool. Soc., London, Vol. IX.

Busk : Trans. Zool. Soc., London, Vol. VI.

D 1

291—110

SIWALIK AND NARBADA PROBOSCIDIA.

L. meridionalis.

Ducrot and Lortet : Archiv. du Mus. d’Hist. Nat. de Lyon, A^ol. 1, 1876.

Falconer: loc. cit., Vol. II, p. 118.

Euelephas americanus .

Leidy : “Extinct Mammalian Fauna of Dakota and Nebraska,” p. 251.

Falconer : Pal. Mem., Vol. II, pp. 14, 211, 212.

E. antiquus.

Ducrot and Lortet : loc. cit.

L. Adams : “ Dentition and 0.steology of E. antiquus," Palseontographical Society, 1877.
Boyd-Dawkins : Quar. Jour. Geol. Soc., London, Vol. XXVIII, p. 413.

Falconer: Pal. Mem., Vol. II, pp. 14, 147.

E. mnaidriensis .

L. Adams.

“Maltese Fossil Elephants,” Trans. Zool. Soc., London, Vol. IX.

Busk: Trans. Zool. Soc., London, Vol. VI, p. 251 {E.falconeri).

E. primigenius.

Ducrot and Lortet : loc. cit.

Falconer : Pal. Mem., Vol. I, pp. 79, 421, II, pp. 14, 158.

Boyd-Dawkins: Quarterly Journal of the Geological Society of London, Vol. XXXV, p. 138.

L. Adams: “Dentition and Osteology of Elephas primigenius" Palseontographical Society, 1879.*

1 Only jnst received in India : the presumed occurrence of the pre-antipenultimate milk-molar in the mammoth, would have been noticed in
the note on the dentition of the Elephants, if Prof. Adams’ memoir had .arrived in time. B.L. March 1880.

SIWALIK AND NARBADA PROBOSCIDIA.

111—292

APPENDIX.

Since the above was in type, several specimens of Siwalik proboscidian remains have been received
in the Indian Museum, some of which require a short notice, and have engendered the necessity
of issuing an additional plate (XXXV A) .

Dinotherium sindibnse, nobis.

Of this species, a single ramus of the mandible, collected by Mr. Wynne in the Siwaliks of
the Kohat District, has been received by the Indian Museum. The crowns of the molars in this
jaw have, unfortunately, all been broken off, but the size and form of the jaw agree precisely with
those of the Sind specimen (Plate XXXI, fig. 4), and confirm the conclusions deduced therefrom.

Mastodon pandionis. Falconer.

Mr. Theobald has remitted to the Indian Museum a magnificent specimen of the palate of this
species, from Nila in the Punjab, showing the penultimate and last true molars of either side. In
Plate XXXY A, the two teeth of the left side have been figured. These two teeth are very
important, as completing our knowledge of the true molar dentition of this species ; no specimen
of the penultimate upper molar was known to me when describing the dentition in detail, and only
one bad specimen of the last molar, alluded to on page 218. I shall not describe the new teeth in
detail, but merely mention their leading features.

The penultimate tooth is considerably worn, carries three transverse ridges, and may have had
a small hind-talon. The last molar is only slightly worn, and carries four ridges and an anterior
talon : the ridges of this tooth dimmish in size gradually from the first to the last. A large
quantity of cement is present in the valleys, which are completely blocked by accessory columns or
tubercles. The ridges when worn present irregularly shaped, and depressed surfaces of dentine.

In all the above characters, these two teeth agree with the molars of M. pandionis described
above, and unquestionably belong to that species ; they show that the species was a true Trilophodon.

In the palate, the two teeth of either side converge to a very small extent anteriorly : their
plane of wear slopes very considerably from the outer to the inner side. The dimensions of the

specimen are as follows : —

In.

Length of two molar teeth ......... 11'3

Interval between outer sides of last molar at second ridge ..... 10'5

Interval between inner sides of ditto at same point . . . . . .3-0

Interval between antero-internal angles of penultimate molars .... 2’8

Length of penultimate molar . . . . . . . . . 5'1

Width of ditto ditto .......... 3’6

Length of last ditto . . . . . . . . . . 6'2

Width of ditto ditto . . ........ 3‘9

293—112

SIWALIK AND NARBADA PROBOSCIDIA. .

This specimen shows that in its last upper molar, M. pandionis makes no approach to M. siva-
lensis (Plate XLIV, fig. 1),. and in this respect differs from the lower molars, as I have noticed
above (page 226),

The penultimate molar of the new specimen is very important, as showing the difference of
this tooth from the corresponding molar of M. falconeri (Plate XXXII, fig, I). The two teeth
are in different conditions of wear, that of M. falconeri being the least worn, but they still permit of
comparison. It will be observed that the molar of M. pandionis presents irregular, and not trefoil-
shaped surfaces of dentine on its inner columns, and that its valleys are completely blocked. If the
molar of M. falconeri were worn as much as that of M. pandionis, the dentine surfaces of the two
first ridges, as well as those of the two columns in each ridge, would have been united, instead of
remaining quite distinct : this is shown in the first upper true molar of M. falconeri, drawn in fig. 4
of Plate XXXII. Again, the molar of M. pandionis must have had only a very small hind-talon,
as the width of the united hind-ridge and talon is very little in excess of the second ridge : in
M. falconeri, on the other hand, the width of the united hind-ridge and talon is equal to the width
of the second ridge, plus half that of the first. Finally, the two teeth are broadly distinguished by
the presence of cement in the one and its complete absence in the other.

Mr. Theobald has also sent there detached specimens of the last upper true molar of M. pan-
dionis, two of which agree with the figured specimen, while the third has a hind-talon, like the
specimen referred to on page 218.

Since the whole of the plates illustrating this memoir have been lithographed, a comparison of
the figures of the • last lower molars of M. pandionis and M. sivalensis has shown me that I had
overlooked certain differences between them. The right last lower molar of the former species drawn
in fig. 4 of Plate XXXV will be seen to be convex externally and straight internally, the reverse
being the ease with the corresponding tooth of M. sivalensis drawn in fig. 3 of Plate XLIV ;
the latter tooth is also narrower than the former. The corresponding tooth of M. pandionis drawn
in fig. 2 of Plate XXXIV, which succeeds an undoubted trilophodont tooth, appears in the engraving
to be slightly concave on the outer side ; this is, however, an error on the part of the native artist;
the outer border of the base of the crown is in reality perfectly straight, a rod laid along the base
of the outer columns touching them all, whereas in M. sivalensis a similarly placed rod would only
touch the two end columns. The last lower molar of M. pandionis has, therefore, its external
surface either convex or straight, and in the former case, its hinder extremity inclines towards the
inner side. In M. sivalensis, on the other hand, the last lower molar (as is shown in my figure and
in the specimen with more ridges drawn in fig. 8 of Plate XXXVII of the “ Fauna Antiqua
Sivalensis^'’) has its outer surface always concave, and its hinder extremity inclining towards the
outer side. The specimen represented in fig. 2 of Plate XXXIV approaches in form to the tooth
of M. sivalensis, but is widely distinguished by the presence of a large quantity of cement. Another
specimen of the last lower molar of M. pandionis in the Indian Museum, is intermediate in character
between the two figured specimens. We thus see that in this species there is a form with narrow,
and another with broad molar teeth.

Mastodon pykenaicus, Lartet.

On page 212 1 stated that I could not find any description of the molars of this species. I have
subsequently come across a figure of the last upper molar given by M. Gaudry.i It seems that that
writer is inclined to consider M. pyrenaicus merely as a variety of M. angustidens.

* “ Lea Enchainements du Monde Animal,” page 174.

SIWALIK AND NARBADA PROBOSCIDIA.

113—294

Mastodon dissimilis, Jourdan.

The text of MM. Lortet and Chantre's monograph on the Mastodons of the Rhine basin,
alluded to on pages 2.86 and 290, has just been received in India : from this memoir I find
that M. dissimilis is merely a synonym of M. arvernensis, and not a distinct species. The name
dissimilis is a manuscript one proposed by the late M. Jourdan in 1840 for specimens in the
Lyons museum. The name of M. arvernensis was applied by Croizet and Jobert in 1826, and has
always been recognized ^by subsequent writers. I cannot but think it extremely ill-advised to rake
up an utterly forgotten name for a well known species, and still more so to publish the plates under
that name without the explanatory text.

March 1880.

INDEX

N. B. — The numbering of the pages in this Index refers to the volume paging, and not to the separate
paging of each part. Synonyms are in italics.

Acerotherium

periinense .

Acotherulum

Alcelaphus

Amphibos

acuticornis

antilopinus

elatus

Amphicyon

palseindicus .

dominans

helveticus

gracilis

minor

vetus

major

Anoa depressicornis
Anthracotheridse .
Anthracotherium

silistrense

Antilope cervicapra
gyi-icornis
palseindica
patulicornis
picta

porrecticornis .
sivalensis
Antoletherium
Aphelops
Atelodus

Bibos banting ,

frontalis
gaurus
gavseus

Bison

grunniens

sivalensis

prisons

Bos

Page.

xiii, xiv, 20, 51
77
158

150, 172, 174
)2, 174, 176, 180
88, 178, 180
88, 178, 180
. 19, 20

. 21, 84
84
84

84
84
84
89, 176
78

79, 81
XV, 78

92, 154, 180
92, 157, 180
89

92, 158, 180
92, 154, 180
185

. xiii

viii

. xvii, 89
xvii,89, 105, 120
xvii, 89, 105, 120

xvii
122
. xvi

88, 90, 92, 122, 180
xvi, 90, 126
. 95, 173
vi, 90, 92, 112, 173, 180
. 89, 106

acntifrons

chinensis

293

INDEX.

Bos — continued.

falconerianus .....
indicus .....

namadicus .....

occipitalis .....

planifroTis . . ...

platyrhinus .....

primigenius . . . • .

Bovidffi, table of frontal measurements of—

Bramatherium .....

perimense .....

Bubalus . . . . .

arni .....

huffelus .....

cafFer .....

pallasii .....

palreindicus . . . . .

platyceros . ...

sivalensis . ... .

Bucapra .....

„ daviesii . . . . .

Buffelus .......

Camelopardalis . . . . .

and Sivatheridm, homologj- of horns of —
girafifa .....
sivalensis ....

Camelus sivalensis .....

Capra . ... ,

aegagrus .....

falconeri .....

perimensis .....

sibirica .....

sivalensis .....

SP-

Cervus . . . . .

duvaucellii .....

latidens .....

simplicidens .....

sivalensis ......

triplidens .....

Cbceropotamus .....

Choerotberium .....

Dentition of Elepbantidse ....

and affinities of Dinotberidae .

Dinotheridae .....

Dinotherium .....

cuvieri .....
giganteum .

indicum .....
konigii .....
pentapotamiae ....
sindiense . . ' .

Dorcatberium .....

majus .....

Page.

XV, 90,

. xvii, 89
92, 95, 180
. 88,174
2, 100, 173, 180
92, 119, 180
XV, 90, 104
153

59, 166, 174
2, 160, 178, 180
. xvi, 127
0, 120, 130, 136
xvii
. xvi, 130

90, 92, 132, 180
92, 127, 173, 180
. 173, 180
180

. 180, 181
xvi

xvii, 57, 93
166
. 57, 59

58, 92, 181
61, 92, 139, 181

169

170

171

92, 170, 181
170

92, 169, 181
92, 171, 181
. xvii, 64
xvii

, 21, 65, 92, 180
, 21, 69, 92, 181
xvii

xvii, 21, 67, 92, 181
. 76, 77

77
198
182
182

2, 183, 283, 284
73

.73, 189, 195. 283
72, 189, 192, 204, 283
283

21, 72, 183, 288
196, 283, 292
. 62, 94

31, 63, 92

INDEX.

297

Page.

l^oi'catlierium — continued.

minus

21, 64, 92

moschinum .

Elephantidse

197

dentition of —

198

Elephas

. 256, 283

Equus primigenius

sivalensis

Euelephas

. 278, 285

americanus

284

antiquus

. 280, 284

armenaicus

. 284, 289

columbi

284

li3’'sudi’icus

. 278, 284

indicus

. 284, 288

mnaidriensis .

. , 284

namadicus

. 278, 284, 288

piimigenius

. 237, 284

Eossil and liecent Proboscidia — Table of—

. 283, 288

Helladotherium

xviii, 164

Heinibos

. 145, 172, 174

acuticornis

. 176, 180

antilopinus

. 178, 180

occipitalis

. 174, 180

ti'iquetriceros .

. 88, 92, 174, 180

Hemitragus hylocrinus .

Hexaprotodon iravadicum

namadicum

sivalense .

Hippobyus sivalensis

Hippopotamidse . . .

Hippopotamus

. 79, 81

Hippotherium antilopinum

Homology of horns of Sivatheridse and Camelopardalis .

166

Hydaspitberium . .

92, 159, 179

grande

179

leptognathus

179

megacephalnm

xviii, 92, 159

Hyopotamus

. XV, 77

Hjmtherium

. XV, 76

Hyracotherium .

Indian Mastodons, measurements of molars of —

282

Leptobos acuticornis

. 176, 180

antilopinus

. 178, 180

elatus ,

. 174, 178, 180

falconeri

. 178, 180

frazeri .

. 178, 180

triquetricornis .

. 174, 180

List of recent and fossil Proboscidia

283

fossil Indian Suina

• 78

Listriodon

. 20, 70

pentapcftamise .

. 21, 70

Lophiodon . ;

. 71, 72

298

IKDEX.

Loxodon

Mams

africanus

melitensis

meridiouaiis

planifrons

sindiensis

Mastodon

americanus
andium
angustidens
arvernensis
borsoni
dissimilis
elejpJiantoides
falconer!
giganteus
humboltii
latideus
longivostris
maximus
mirificus
obscurus
ohioticus
pandionis
penteleci
perimensis
productus
pyrenaicus
sivalensis
tapiroides
turicensix
virgatideiis

Mastodons, measurements of molars of Indian
list of—

Measurements of molars of Indian Mastodons
Stegodons

Merycopotamus • • • ■ •

dissimilis . . • •

Non-Indian species of fossil Proboscidla, references to works on
Ovis, sp. . •

Palmomeryx ...•••

Palaeoryx

Peribos

Portax namadicus . . . ■ •

sp. . . • • •

Proboscidia, list of recent and fossil-

references to works on non-Indian species of fossil-
ridge-formulae of molars of —

Probubalus antelopinus . . ■ ■ ■

celebensis . . • ■ •

triquetricornis . . • ■

200

201, 200, 210

. 275, 285
201
284
284

201,237, 238, 275
. 19,82

. 21,82
. 202, 283, 284

283

209, 210, 283
13, 223, 225, 238, 248,255, 283
283

209, 210, 237, 28S
283-
268

. 202, 282,, 283

. 201, 283
209, 210, 238, 283
201, 227, 238, 282
201, 238, 21.1, 243, 244, 283
209, 210, 237, 238, 246, 283
283

. 209, 212, 283

. 246, 283
212, 213, 238, 255, 282, 283, 292
206, 209, 212, 283
. 201, 238, 239, 282, 283
199, 201, 209, 212, 226, 283
209, 212, 283, 293
201, 226, 238, 248, 282, 283
206, 209, 212, 283
206, 209, 212, 283
209, 212, 283
282
283
282
282
79
78
289
. 92,181

. 141, 172, 174

• xvi, 180

• xvi, ISO'

283
289

284
• 178, 180

176

IH 175, 180
79

Fteropus

INDEX.

299

Rhagatherium .....

Rhinoceros Ihjthii ....

crassus ....
crossii ....

deccanensis ....
dental formula of —
mandible of — .

etruscus ....

from Pikermi ....

hemitsechus ....

indicas ....

inermis ....

iravadicus ....

javanicus ....

list of Indian species of —
leptorhinus ....

megarJiinus ....

merkii ....

namadicus ....

palaeindicus ....
perimensis ....
planidens ....
platyrhinus ....
remarks on —

simus ....

sinensis ....
sivalensis ....
sondaicus ....
stenocephalus ....
snmatranus ....
sumatrensis . . . ' .

tichorhinus ....
various species

Ridge-formulm of molars in Proboscidia, table of—

Ruminants, list of Indian Tertiary —
remarks on fossil — .

Sanitherium

scblagintweitii

Sivatheridae .....

Sivatheridae and Camelopardalis, homology of horns of —

Sivatherium ....

giganteom ....

Stegodou .....

bombifrons ....
cliftii ....

ganesa ....

insignis ....

orientalis ....
sinensis ....

godons, measurements of molars of — .
ridge-formulae of molars of — .

Suina, list of fossil, Indian ....

Page.

. . xiv

viii, xiv, 1, 21, 38, 52

28, 31, 36
13

. 12, 25, 28, 31

ii, xiv, 13, 25, 28, 33, 52
. . xiv

xiii, xiv, 20, 21, 36, 25

xiv, 25, 31, 35, 38, 52

25, 28, 31
12
50

viii, 32, 52

ix, xiv, 15, 20, 22, 52

xiv, 15, 52
xiii, xiv, 20, 21, 41, 52
xi, xiv, 15, 20, 29, 35, 52
21
50

xiv, 15, 21, 38, 52
x, xiv, 15, 20, 26, 38, 52
xiv, 14
xiv, 52
. . xiv, 14

xiv, 25, 31
3, 22, 28, 31, 35, 38, 52
43

284
. 92, 180

166

. 160, 166
92

237, 256, 285
238, 262, 282
237, 238, 256, 282
. 273, 282
237, 238, 26^, 282
269
257
282

285
78

300

INDEX.

Sus giganteus

hysudricus . .....

punjabiensis ......

pusillus ......

Table of recent and fossil Proboscidia ....

of measurements of skulls of Bovidse

of references to works on non-Indian species of fossil Proboscidia
of ridge-formulce of molars of Proboscidia
Tapirus .......

Tetraceros .......

TetraConodon .......

magnum . . ...

Tetraconodontidae . . ...

Tetralophodon ......

Tetraprotodon palmindicum .....

Trilophodon ......

Vishnutherium iravadicum .....

Zalahis ........

Page.

283

153

289

71
166
. 71, 78

xviii, 78 79
82

. 227,285
78

. 209,284
xviii, 21, 55, 92
. xii

INDEX

TO THE

PLATES OF MOLAR TEETH AND OTHER REMAINS OF MAMMALIA

EEOM THE

TERTIARY BEDS

INDIA AND BURMA.

EXPLANATION OP PLATES

IIHINOCEROS DECCANENSIS.

PLATE 1.

1. — View of masticatory surface of the upper molar series, left side.

3. — Left upper molars from the inner side.

PLATE II.

1. — Third and fourth upper premolars, left side, view of outer wall.*

2. — Fragment of first true molar, right side, showing the inner wall of anterior valley

(a), to illustrate its great depth.

3. — Outer side of left ramus of mandible. Half natural size.f

4. — Symphysis of mandible seen from above.

PLATE III.

1. — Squamosal bone with meatus auditorius («), post-glenoid process (d), and post-tympanic
process (c) ; — {d) parts of supra-occipital and ex-occipital (?) bones.

3. — Outer sides of left lower molars.

3. — Crown view of ditto ditto.

All the letters not given here will be found given at page 8.

* This view represents the teeth in their normal position, and does not therefore show their full height,
t The hindermost molar should be marked m3 instead of mi.

u ■

'(■ V'

(-•i- ^ : m [

- :--: - ■. ^ ■ ’■ ■

.! •^: ; t' c-.- .^ :; v.^;"’."' . T

.! [ ;l.:a,: [•

; ,/ .' \.yi o'-.‘!-^ :i'\l Q> ^(-0 , . . - ,

j;; :.:r- /•; ' 1 Ij O- '•

(

1 1

X- i

PLATE I.

Rhinoceros deccanensis, Foote.

Fig-. 1 . View of masticatory surface of the upper molar series^ left side.
Fig. 2. Left upper molars, from the iuuer side.

[ Reprint. ]

Qeol: Surv: of iTidia.

FLUVIATILE DEPOSITS. KRISHNA VALLEY.

Pl:l.

Lith-. and prit: a,t Ole GboV. Survey's Office

Calcut.t.a,

PLATE II.

[ Reprint. ]

Rhinoceros deccanensis^ Foote.

1. Third and fourth upper premolars, left side, view of outer wall.^

2. Fragment of first true molar, right side, showing the inner wall of anterior valley (a),

to illustrate its great depth.

3. Outer side of left ramus of mandible. Half natural size.^

4. Symphysis of mandible seen from above.

This view represents the teeth in their normal position, and does not thei'efa>‘e show their full heiglit.

The hindermost molar should be marked m^ instead of m'.

Oeol: Surv of India

FLUVIATILE DEPOSITS. KRISHNA VALLEY

FI IT

Lith. and print: at tlie Greol. Surveys Office

PLATE III.

Rhinoceros deccanensis, Foote.

[ Reprint. ]

Kg'. 1. Squamosal bone with meatus auditovius {a), post-glenoid process [h], and post-tympanic
process (e) ; [d) parts of supra-occipital and ex-occipital (?) bones.

Fig. 2. Outer sides of left lower molars.

Fig. 3. Crown view of ditto.

All the letters not given here and in preceding plates will be found given on page 8.

[ Reprint. ]

PLATE IV.

Rhinoceros & Acerothbrium.

Fig. 1. Rhinoceros ral.®indicuSj Falc. &CaHt. ; fragment of upper molar of the left side :
Siwalik.

Pig. 2. Rhinoceros sivalbnsis, Falc. & Caut. : 3rd upper true molar of the right side : Siwalik.

Fig. 3. Rhinoceros palhsindicus, Falc. & Caut. : upper premolar of the right side : Siwalik
(copied from “ F. A. S.,” Plate LXXIV, fig. 4),

Fig. 4. Rhinoceros platyrhinus^ Falc. & Caut. : 2nd upper true molar of the right side
(copied from « F. A. S.,” Plate LXXV, fig. 11a).

Fig. 5. Rhinoceros indices, Cuv. ; 3rd upper true molar of the left side : Narbada valley ; des-
cribed in second fasciculus as R. namadicus.

Fig. 6. Rhinoceros indices, Cuv. : 2nd upper true molar of the right side : Narbada valley ;
described in second fasciculus as R. namadices.

Fig. 7. Acerothbriem pbrimense, Falc. & Caut. ; upper true molar of the right side : Siwalik ;
described in text as Rhinoceros planidens, 7iobis.

Fig. 8. Rhinoceros sivaeensis, Falc. & Caut. ; 2nd upper true molar of the right side: Siwalik
(copied from Messrs. Baker and Durand’s memoir) .

Fig. 9. Acbrotheriem pbrimense, Falc. & Caut. : upper molar of the left side : Sinalik ; described
in text as Rhinoceros planidens, tiodis.

Figs. 3 and 4 one-half natural.size ; fig. 8 one-quarter natural size ; the rest natural size.

PLATE IV.~RHmOCEROS.

Fiff.

Fig.

Fi-.

1. Rhinoceros pamindicus^ Falc. and, Caid.; upper molar ; left side. Lehri —
Siwaliks. No. 57.

3. Rhinoceros sivalensis, Falc. and Caut. : 3rd upper molar ; right side. Katana —
Siwaliks. No. xl’j"

3. Rhinoceros palaiindicus, Falc. and Caut. : upper premolar ; right. Siwaliks

(copied from “ F. A. S.” plate 74, fig. 4).

4. Rhinoceros platyrhinus, Falc. and Caut. : 2nd upper molar ; right side (copied

from F. A. S.” plate 75, fig. lla).

5. Rhinoceros namadichs, Falc. et nobis : 3rd upper molar ; left side. Narhudda

Valley. No. 471.

6. Rhinoceros namadichs, Falc. et nobis : 2nd upper molar ; right side. Narhudda

Valley. No. 470.

7. Rhinoceros planidens, n. sp. nobis : upper molar ; right side. Gadari — Siwa-

liks. No. 56.

8. Rhinoceros sivalensis, Falc. and Caut.: 2nd upper molar; right side. Siwa-

liks (copied from Messrs. Baker and Durand).

9. Rhinoceros planidens, sjt?. ; upper molar; left side. Gadari — Siwaliks.

Figures 3 and I one-half natural size ; figure 8 one-quarter natural size; the rest natural size.

mamma L I a .

/V jy.

G-eol: Surv-, of Ijiclia.

TERTIARY

PLATE V- RHINOCEROS.

Fig.

1. Rhinoceros iravadicus^ n. sp. nobis : 2nd upper molar; left side : specimen found

in a Pagoda at Prome and presented to the Indian Museum by Col. Phayre.
No. 219.

2. Rhinoceros iravadicus, n. sp. nobis ; 2nd upper molar ; right side. Irawadi

Valley. No. 216.

3. Rhinoceros iravadicus, n. sp. nobis; portion of occiput. Pegu. No. 251.

4. Rhinoceros sivalensis, Falc. and Caut.: 2nd upper molar ; leftside. Asnot— *

Siwaliks. No. 780.

5. Rhinoceros, n. sp. ? upper premolars ; right side. Burma. No. 220.

Figure 3 one-third natural size ; the rest natural size.

bf) 6c

L Reprint- ]

PLATE V.

Fig. 1.
Fig. 2.

Fig. 5.

Rhikocekos iRAVADicus, n. sp. nobis : 2nd upper true molar of the left side : specimen
found in a pagoda at Prome, and presented to the Indian Museum by Sir A. Phayre.
Rhinoceros iravadicus, n. sp. nobis : 2nd upper true molar of the right side : Irawadi
valley.

Rhinoceros iravadicus^ m. sp. nobis: portion of occiput: Pegu.

Rhinoceros iravadicus (?) n. sp. nobis: two upper milk-molars of the right side:

Burma : described in text as being the premolars of a possibly new species.
Rhinoceros sivalensis, Falc. & Cant. : 2nd upper true molar of the left side : Siwalik.
Fig. 3 one-third natural size ; the rest natural size.

G-eol: Surv; of India

T E R T

ARY

MAMMALIA-

PI P

Pi’mted at S-eol: Survey Office.

~!ay TLasick L all Bose,

PLATE YL-RHINOCEROS AND ACEROTHERIUM.

Fig-.

Fig.

1. Rhinoceros sivalensis^ Falc. and Cant, : 3rd lower molar ; right side. Pinjor —

Siwaliks. No.

2. Acerotherium perimense^ Falc. and Cant. : 1st upper molar ; left side. Gadari —

Siwaliks. No. 55.

3. Rhinoceros, w. lower molar ; right side. Attock — Nahans (?).

4. Ehinoceros platyrhinus, Falc. arid Cant. ; lower premolar; right side. Siwaliks.

5. Acerotherium perimense, Falc. and Cant.: upper molars; right side. Perim

Island. A. S. B. No. I'J-.

6. Rhinocebos {spl) : upper premolar ; left side. Manehhar beds — Sind. No. G.

7. Rhinoceros ??. 1st lower molar ; right side. Siwaliks. A. S. B. No. -3!^.

8. Rhinoceros PAUiEiNDicus : lower molar ; left side. Manehhar beds. Sind.

No. 494.

9. Rhinoceros : upper premolar ; leftside. Siwaliks. No. 174.

10. Rhinoceros {sp) : 4th upper premolar ; left side. Siwaliks. A. S. B. No.

All the figures natural size.

iMi\

[ Reprint. ]

Fig. 1.
Fig. 2.

Fig. 3.
Fig. 4.
Fig. 5.

Fig. 6.

Fig. 7.
Fig. 8.
Fig. 9.

Fig. 10.

PLATE VI.

Rhinoceros sivalensis (?) Falc. & Caul. : 3rd lower true molar of the right side : Siwalik.
Acerotherium pekimense, Falc. & Caut.. : Srd upper premolar of the left side : Siwalik :

described in text as first true molar
Rhinoceros, sp. : lower molar of the right side : near Attock.

Rhinoceros platyrhinus (?) Falc. & Caut. : lower premolar of the right side ; Siwalik.
Acerotherium perimense, Falc. & Caut. : 2nd and Srd upper premolars of the left side •
Perim Island : described in text as iast premolar and first true molar.

Acerotherium perimense, Falc. & Caut. : the 2nd (?) upper premolar of the righ
side : Sind : described in text as left premolar of a Rhinoceros.

Rhinoceros, sp., 1st lower molar of the right side : Siwalik.

Rhinoceros PAUiEiNDicus (?) Lower molar of the left side : Sind.

Rhinoceros sp. : upper premolar, or milk-molar of the left side, probably abnormal
Siwalik.

Rhinoceros platyrhinus (?) An anterior upper milk-molar of the right side ; Siwalik.
All the figures natural size.

Geol-. Suvv. of Itidia

T E R r

ARY

MAMMA L I A

PI vr

Prmtad at aeol; Survey Offtce

by Rasiclc L all Bose.

v'tl

'i'

PLATE VIL-CAMIVORA AND UNGDLATA.

Fig. ..

. 1.

ViSHNUTHERiUM lEAVADicuMj %. gen. iiobis : left ramus of mandible. Burma.

No. 224. Inner view.

Fig.

.. 2.

Outer view o£ fig. 1.

Fig. .

.. 3.

Dorcatherium minus, n. sp. nohis : 2nd and 3rd upper molars ; right side.
Asnot — Siwaliks. No. 1301. Inner view.

Fig.

.. 4.

Dorcatherium majus, var. b., n. sp. nohis : upper molars ; left side. Asnot —

Siwaliks. No. 1307. Inner view.

Fig.

.. 5.

Amphicyon PALiEiNDicus, n. sp. iioUs : 2nd upper] molar ; right side. Kushal-
ghar, near Attock. No. 497.

Fig.

.. 6.

Outer view of fig. 4.

Fig.

.. 7.

Outer view of fig. 3.

Fig.

.. 8.

Amphicyon PALaiiNDicus, n. sp. nohis : lower carnassial, and last milk molar .
Nurpur— Siwaliks. No. 656. Outer view.

Fig. ,

.. 9.

Dorcatherium majus, n. sp. nohis: 3rd upper molar; right side. Kushalghar,

near Attock. No. 502. Inner view.

Fig.

.. 10.

Outer view of fig. 9.

Fig.

.. 11.

Dorcatherium, sp. : right upper molar. Kusbalghar, near Attock. Nos. 502-6.

Fig. .

.. 12.

Inner view of fig. 8.

Fig. .

.. 13.

Bramatherium perimense, Falc. and Cant. : 1st and 2nd lower molars, Perim

Island. A. S. B. No. Outer view.

Fig.

... 14.

Camelopardalis sivalensis, Falc. et nohis : last lower premolar ; right side.

Lehri — Siwaliks. No. 39.

Fig.

,.. 15.

Camelopardalis sivalensis, Falc. partim : 1st and 2nd lower molars ; right side.

Lehri — Siwaliks. No. 39, Outer side.

All the figures natural size.

G-eol: Sui-v: of Ividia. T E R T 1 A R MAMMALIA. Th. VIl.

Pi-infcftd a.t OeoL, Siirvejr Office. libLif! by Ra.sio]i L all Bose.

M r^ >/ !'

« ,

S'.

PLATE VIII.-UNGULATA AND EDENTATA.

Fig.

... 1.

Cervus triplidens, n. sp. nobis: 2nd and 3rd upper molars; right side. Siwaliks.

No. Inner view.

Fig.

... 2.

Outer view of fig. 1.

Fig.

... 3.

Cervus simplicidens, n. sp. nobis : 2nd and 3rd upper molars ; left side. Siwaliks.

No. 48. Inner view.

Fig.

... 4.

Cervus latidenSj n. sp. nobis : lower molar ; right side. Padri — Siwaliks. No. 23.

Inner view.

Fig.

Cervus, sp, : 2nd and 3rd lower molars ; right side. Padri — Siwaliks. No. -j-f-.
Outer view.

Fig.

... 6.

Outer view of fig. 4.

Fig.

... 7.

Cervus latidens, n. sp. nobis: last upper molar ; right side. Asnot — Siwaliks.

No. 1420. Outer view.

Fig.

... 8.

Listriodon pentapotami^, Male, sp, : upper molar ; right side. Kushalghar,

near Attock. No. 499.

Fig.

... 9.

Listriodon pentaPotamiad, Falc. sp. : upper molar ; right side. Asnot — Potwar.

No. 813.

Fig.

... 10.

Inner view of fig. 7.

Fig.

... 11.

Manis sindiensis, n. sp. nobis : second phalange third digit of manus. Lateral

view.

Fig.

... 12.

View of distal surface of fig. 11.

Fig.

... 13.

View of anterior surface of fig. 11.

Fig.

... 14.

View of proximal surface of fig. 11.

GrRol: S-ar^r-. of India

TERTIARY

MAMMAL

Fh-.VUl.

I A

Sfe

PLATE IX.-DINOTHERIUM AND SANITHERIUM.

Fig.

,. 1.

Dinotherium pentapotami^, Falc. : first upper premolar ; right side ; crown-
surface, from the inner side. Kushalghar, near Attock. No. 505.

Fig.

.. 2.

Dinotherium pentapotamia:, Falc. : first upper molar ; right side ; crown-surface,
from the inner side. Kushalghar, near Attock. No. 506.

Fig. ..

,. 3.

Dinotherium pentapotamij), Falc. : second upper molar ; right side ; crown-
surface, from the inner side. Kach. No. G062.

Fig.

,, 4.

Dinotherium pbntapotamijE, Falc. : first upper molar j right side Kach.

No. G 062.

Fig.

.. 5.

Dinotherium pentapotamij, Falc. : first lower molar; left side. Sind,

Nos.

Fig.

.. 6.

Sanatherium schlagintweitii, Fon Meyer : second and part of third lower
molars ; left side. Kushalghar. (Palseontographica, vol. 15, pi. II, fig. 9.)

Fig.

.. 7.

Outer view of fig. 6.

Fig.

.. 8.

Sanitherium schlagintweitii. Von Meyer {Sus pusillus, Falc.) : part of third
lower molar; right side. Kushalghar. No. 500.

Fig. .

.. 9.

Outer view of fig. 8.

Print-ed ah CtroI-. Sm'vey- Office

Lithft iiy ,aa.sick L s2\B o sr

..T'.V ■(

PLATE X.-TETRACONODON.

1. Teteaconodon magnum, 'Bale.: rigLt ramus of mandible ; upper view. Asnot —

Siwaliks. No. 828,

2. Inner view of penultimate premolar of same specimen.

3. Outer view of same specimen.

Geol. Siii'v. of India

Printed at &eol-. S-ur^ey Office LilK?. by Rasic-kXallBoae

[Eeissue. ]

PLATE X.

Tethaconodon marnum, Falconer.

Molar and part o£ premolar dentition of rig-ht ramus of mandible : from the Punjab.
a. Upper view.

Inner view of penultimate premolar.
c. Outer view.

^•VYor>;o

■aoijjO »M.'^ Tpn^HJ.

^T.7! }»'pa -ISUiXYir -^Y-T.

plaii

•VIGNI 'S SMDOa Sn0 30 Vl3y3

"K-ipxii je '.AJ-ng ;j0P-g

PLATE XL

Fig-. 1. Bos NAMADicus, F. & C. : front view of eraninm, ;i- nat. size. Nerbndda Valley.
No. 266, Ind. Mus.

Fig-. 2. Lateral view of same cranium, ^ nat. size.

Figs. 3 & 4. Sections of liorn-core, 3 at tip, 4 at base.

Gcdl'. Surv-. of India. TE.RT1ARY MAMMALIA Pl:XT

Printed at Geol- Survey Office "by H asicic LallBo se.

PLATE XIL

Fig. 1. Bos ACUTIFEONS, n. sp. nobis: front view of cranium, ^ nat. size. Padri, Siwalik.
No. yf 7, Ind. Mus. : 1 a, section of horn-core at tip ■, 1 at base.

Fig. 2. Bos PLANiFEONS, n. sp. nobis : front view of cranium, j nat. size. Kaiigra, Siwalik.
No. ,-f^, Ind. Mus.

-Fi-.xn

rr.

lEl

PLATE XIIL

Bos ACUTiFRONS, n. sp. nobis : larger view o£ specimen drawn in Plate XII, L size.

G-eol: S-ar-v: of India

P«nted at Geol-, Sur xre^ Office ^ Lith^VR^sickLadl B.

• s

.Y‘', '

.4'

'^0

)

PLATE XIV.

Bos PLATYRHiNUSj n. sp. nobis ; front view of lower half of cranium, ^ nat. size. Unib, Siwalik.
No. Ind. Mus.

T E R T I A R Y

MAM M ALIA

Fl-.KJV

G-eol Surv-. of India

!EEMi.teP- at Cl-ouL. Offiee .

-i iV.S ^Jy B-as iclc I all B o s e

PLATE XV.

Bison sia^alensis, P. & C., sp. : front Anew of cranium and section of horn-core, | uat. size. SiiA^alik,
No. YT9i AIus,

of Iridia. TERTIARY MAMMALIA, Ph.XV.

f. and print: at die G-eol. Sas-rveys Offlce.

PLATE XVL

Fig-. 1. Bos NAMADicuSj F. fe C. : occipital view of cranium figured in Plate XI, ^ nat. size.

Fig. 2. Bos ACUTiFROJTS, 11. sp. iiobis : occipital view of cranium figured in Plate XII, fig. 1,
^ nat. size.

Fig. 3. Bos NAMADicus, F. & C. : occipital view of sjiecimen figured in F. A. S., Plate G,
fig. 1. a. Nerbudda.

Pig. 4. Bos PLANiFRONs, n. sp. nobis : occipital view of cranium figured in Plate XII, fig. 2,
^ nat. size.

IjILK-. and Print-: GboL Surveys Ofiice

I •'

PLATE XVII.

Fig. 1. Bison SivalensiSj F. & C., sp. ; occipital view of specimen figured in Plate XV,
nat. size.

Fig. 2. Bubalus pala)indicus, F. & C. : occipital view of specimen figured in Plate XIX,
^ nat. size.

Printed -aX. Geol-, S-arvey

Litlfiiliy aasickL, allBosc .

P o xvir.

G-eol; Snrv-. of' India

PLATE XVIII.

Bubaltjs platyceuos, n. sp. nobis : Fig. 1, front view of cranium, I nat, size ; Fig. ’■Z, occipital view,
I nat. siz. Siwalik, No. ;j2o.. Mus.

Fig. 3. Section of horn -core at base.

Geo] : Surv: of India. ’ TERTIARYMAMMALIA. PI; XVlll.

ntli: tr Lrd), Raef.. .Printed 5.t Geol: .Sur-.^ey OEc.i

J-

1 •

V <

f ; '«. • «’• V ti

‘ ‘

{(

. ■/

■it

-...■a- ■

^ V'.'W I- '■/ .^{q/jC' :

■P'

iV.

•vv

■

■ 1 '

PLATE XIX.

Bubalits PAL.SINDICIJS, F. & C. : front view of cranium, ^ nat. size. Nerbudda Valley.

Geol; SxiTv: of India,

Printed at &eolT Survey Office.

PLATE XX.

Peribos occipitalis, Falconer, sp. (?) : front view of cranium, ^ hat. size. No. - (A S B )
Siwalik. ^

PLATE XX.

Hemibos occipitalis, Falc., sp. ; front view of cranium of male trochoceros form ;
size ; Siwalik, Described in text as Peribos occipitalis .

Synonyms : Probubalus triquetricornis, Riit.

Hemibos triquetriceros, Falc.

1 a. Section of base of horn-core.

[ Reissue- ]

one-half natural

FI: XX.

S-urv; of India,

Russick L all Bose.

Printed at Geol; Survey Office.

PLATE XXL

Fig-. J. AmphibOs A.CUTICORNIS, F. & C. : lateral view of cranium, | nat. size. Siwalik.

Fig. 2. Peribos occipitalis, Falconer, sp. (?) : lateral view of cranium figured in Plate XX,
J nat. size.

&eol; Si

I '

PrititecL at Greol; Siirvejy OffloR

Xtth.“rl3y Ksls'i’kXi allB o s e .

PLATE XXI A.

Hemibos occipitalis, Falc. sp. ; front view of nearly iierfeet cranium of male trochoceros form
one-half natural size ; Siwalik.

T B B. T I A 'R Y

MA MM ALIA .

PI •• XXL A

■ G-eol: Surv. of India.

•To tlulTadiP a s .

.frinled at Geol- Sur vej Oftice

PLATE XXI B.

Hemibos acuticorniSj Talc, sp. ; front view of specimen figured in Plate XXI, fig, 1,

&eol. Surv: of Itidia,

TERTIAE. Y MAMMALIA.

PLXXI.B

JoduHatliDas^LitliA ■

PrinLcd al Gcol'. Survey Orflce,

PLATE XXL

[ Reissue- ]

Fig, 1. Hemibos acijticoenis, Falc. sp. ; prorile view of cranium of a young female; one-half
natural size ; Siwalik. Described in text as Ampliihos acuticornis.

Fig. 2. Hemibos occipitalis^ Falc. sp. ; profile view of specimen figured in preceding plate.

Geol; Sur-v; of Tndia.

tertiary YAMYALIA

PI-.' XXL

(Reas sTie^

J. ScTaaumlour^.Iitlii

Pri-ntei at Seol S-arvej Office

■ i'

Si,-.

[ Reissue- ]

PLATE XXII.

Hbmibos acuticorniSj Pale. sp. ; front view of cranium of an old male, of the triangular horned
variety ; one-half natural size ; Siwalik. Described in text as Hemibos triquetriceros.

Synonym : Amphiios actiUcornis, Falc. and Riit.

Geol'. Surv: of India,.

TERTIARY MAMMALIA

P.t-inted at Gaol: Su>-t/: Office

LitL by BasicTc ball Bose

PI X/Ji

"■1 '

[ Reissue. ]

PLATE XXIIL

PIemibos acuticobniSj Falc. sp. ; profile view of specimea figured in last plate.

Geol: Siirv: of India, TERTIARY MAMMALIA

Rasicic Tall Bose. 5^

PLATE XXIII A.

Hemibos acuticorniSj Falc. sp. ; front view of a male skull somewhat
figured in Plate XXIIj and with less markedly triangular
size; Siwalik.

younger than specimen
horns; one-half natural

a. Section of base of horn-core,

Geol: S-arv. of India

TERTIARY

MAMMA L I A

El: XXIII. A.

Jo d.uUath.Das.LithS-

PrintBiat OeoL Survey Office-

PLATE XXIV.

Amphibos acuticornis, F. & C. : front view of cranium, i nat. size. Padri, Siwalik No.
Mus.

PI: XXIV.

Geol; Surv: of India.

TERTIARY MAMMALIA,

LilK: iy ■Rasick LaQl .Bos<

Printed at Geol: Snrv: Office,

[ Eeissue- ]

PLATE XXIV

Hemibo? occipitalis, Falc. sp. ; front view of cranium of normal (?) male form ; one-half natural
size j Siwalik. Described in text as Awjokibos (Remibos) acidicornis.

G-eol' Surv. of India

TER T I A R Y

M AMMALIA

P1-. XXIV

Reissue.

PLATE XXV.

Fig. 1. Antilopb Sivalensis, n. sp. nobis: lateral view of cranium, | nat. size. Siwalik, No. 47,
Ind. Mus.

Fig. 2. Antilope Sivalensis, n. sp. nobis : front view of preceding specimen.

Fig. 3. Antilope patulicounis, n. sp. nobis : front view of part of cranium, | nat. size. Siwalik,
No. 46.

Fig. 4. Antilope pouuecticornis, n. sp. nobis : front view of left frontal and horn-core (right
side restored in outline), J nat. size. Potwar District, Siwalik, No. 1653,

Geol: Surv : of India .

Printed at Geol: Survey Ofiice,

y-.

■' .0

s ik

■■i'

ts'i.

l :■

■jii

r . '

, ,= :,

\ :n -

•:v ';;

'i;

]

'•:V

PLATE XXVI.

Hyuaspitiierium megacephalum, n. gen. nobis: lateral and front views of cranium. Potwar
District, Siwalik, nat. size, Ind. Mus.

Geol; Surv; of India

itVidbya-dLsic-k Lia.n Bose Printed at Geol: Surv: Office .

■".iir/iy : u

^ :.'.y/J .rj;> ./:- ,fTKT;,— ’ ,. ' ;C; ,.^ f

r 'vV-i;,' , • . ,’■"•■/

fj- .

PLATE XXVir.

Figs. 1 & 2. Hydaspitherium megacephaltjm, n. gen. nobis : palatal and occipital views of
cranium figured in Plate XXVI, J nat. size.

Fig. 3. Last left upper molar of Sivatherium giganteum, nat. size.

Fig. 4. Corresponding tootb of llydaspitliemm megacephahm, nat. size.

Geol- Surv, of India . ' *' TERTIARY MAMMALIA. n, XXTVn.

Printed at Geol'. Survey Office LiIh^Ly Basiclt LallBose .

PLATE XXVIII.

Pig. 1. Capra Sivalensis^ u. sp. nobis : front view of portion of cranium. ^ nat. size : original
specimen in the British Museum. Siwalik.

Pig. 2. Lateral view of last specimen.

Pig. 3. Capra, sp. : fragment of left horn-core, | nat. size. Potwar District, Siwalik, No. 1698.

Pig. 4. Capra perimensis, n. sp. nobis : lateral view of portion of cranium and horn-cores, | nat.
size. Perim Island, No. (A. S. B.)

Pig. 5. Pront view of last specimen, | nat. size.

Geoi: S-urv : of India,.

TERTIARY MAMMALIA.

?L XXVJIl

LitT : iy E,a,sicl^ L all Bose,

Printed, at Geol: Survey Office .

PLATE XXIX.

Dinotherium pentapotAmi^, Falconer et nobis.

... 1. The second and third left lower true molars : Siwalik.

2 & 3. The last premolar, the first true molar, and half of the second true molar of the
left side : from Kushalghar on the Indus.

Both specimens are represented of the natural size, and are viewed from the outer
side.

<39ol; Sin-v. of Inclia -

PJXXIX

TERTIARY MAMMALIA.

LitM by Rasicl^ L^UBose

Pi'mtad at Qeol. Srarv. Office

-v-

vii

• ''A/!

4 '

• i

PLATE XXX.

Dinotherium pentapotami^, Falconer et nobis.

A portion of the left ramris of the mandible containin]!^ the last premolar and the three true
molars. The specimen is from the Laki hills of Sind ; and is represented from the outer
side^ of the natural size.

Geol- Siirv: of Iixdia. .

Ill ifVi ]B.^i,siclcXjal]. Bose P'V'i'vi'ted eit Geol: Sixrv: Oi’fxee

i'l

■4

" T ' , §■

■fc-

' ^5

• I'

♦«

- c

:^i

PLATE XXXI.

Tig.

... 1.

Fig.

... 2.

Fig.

... 3.

Fig.

... 4.

Dinotherium tndicum, Falconer. Hinder ridge of second right upper true molar :
from (?) Dehra Ghazi Khan.

Dinotheriitm indicum, Falconer. First left lower true molar : from Sind.
Dinotherium pentapotami^, Falconer et nobis. Last left upper premolar : from
Sind.

Dinotherium sindiense, nobis. Part of right ramus of the mandible : from Sind.
All the specimens are represented of the natural size. The upper molars are viewed
from the inner, and the lower from the outer side.

LitM'jqy Jo d.iii\arrL Las.

Prmlfid ai G-eol; Sur Yev Office.

&eol-. Sixrv. of India.

T E RT I A RY MAMMAL I A

PL. XXXI.

, v

"'l--

f.,^p

#■

i ■'

'm,

.€r

■■ i .,

.. ^

PLATE XXXII.

Fig.

... 1.

Fig.

... 3.

Fig.

... 4.

Mastodon (Teilophodon) falconeri, nobis.

Second left upper true molar ; from the Punjab.

Second left upper milk-molar : from a young- cranium found in the Punjab.

Third left upper milk-molar : from the same cranium.

First left upper true molar : from the Punjab.

All the specimens are represented of the natural size^ and are viewed from the
inner side.

PLATE XXXIII.

Mastodon (Trilophodon) palconeri, nobis.

Fig. ... 1. Part o£ right ramus of mandible, with second true molar: from the Punjab.

Fig. ... 2. Second right lower milk-molar : from the Punjab.

Fig. ... 3. First right lower true molar : from Sind.

Fig. ... 4. Second right lower true molar : from the jaw represented in fig. 1.

All the specimens are viewed from the outer side : fig. 1 is J the natural size : the
rest are of the natural size.

Geol; Siirv ,of Jiadiu .

MAM M A [, T A

I'i . XXXJli .

Lxtli^'by JoduNafliBas

PrinleA atOaol-. Survey Offic

PLATE XXXIV.

Mastodon (Teilophodon) pandionis, Falconer.

Fig. 1. The second right lower true molar : from the Punjab.

Fig. 2. The third right lower true molar, from the same jaw as the last.

Both molai’s are represented of the natural size, and are viewed from the outer side.

&eol Surv. of India. TERTIARY RTAMMALTA . PR. XKXIY.

JoauNathDas.Lithii Pmnted at &eol- .Surrey Office

PLATE XXXV.

Mastodon (Trilophodon) pandionis, Falconer.

Fig. 1.
Pig. 2.
Fig. 3.
Fig. 4.

The first right lower true molar : from the Punjab.

The last left lower premolar ? ; from Sind.

The second left upper milk-molar : from the Punjab.

The third right lower true molar : from the Punjab.

All the specimens are represented of the natural size. The upper molar is viewed
from the inner side, and the lower molars are viewed from the outer side.

mamma

J SAauTnWi.I.a,^.

'i: » ■

i '■

V I

PLATE XXXVA.

Mastodon (Tuilophodon) pandionis. Falconer.

The penultimate and last left upper true molars : from the Punjab : taken from
a palate specimen.

The teeth are drawn of the natural size, and from the inner side.

Geol; Surv: of India .

PLATE XXXYI.

AIastobon (Trilophopon) pandionis. Falconer.

Fig^. 1. The symphysis of the mandible, from the specimen containing- the tooth repre-

sented in Plate XXXV, fig. 4.

Fig. 2. The symphysis of a mandible with tusks : from the Punjab’

Both figures are half the natural size.

Geol; Surv: of India.

TERTIARY MAK¥ALIA,

PI XXXVI.

J . S cTiauinlurg^ir L itb ft

Pcinted at &eoT. S-urvey Office

PLATE XXXVII.

Mastodon (Tetralophodon) latidens, Clift,
and

Mastodon (Trilophodon) pandionis, Falconer.

Fig.

The specimen drawn in fig*. 3 is provisionally referred to M. pandionis : all the other
specimens belong to M. latidens.

The third right lower milk-molar : from Burma ?

The second right lower milk-mol% : from the Punjab.

Tlie second left lower milk-molar : from Sind.

The second right upper milk-molar : from the Punjab.

The first left upper milk- molar : fi’om the Punjab.

The last left upper premolar : from the same jaw as No. 8.

The penultimate right upper premolar : from the Punjab.

The third left upper milk-molar : from the Punjab.

All the specimens are represented of the natural size. The upper molars are viewed
from the inner, and the lower from the outer side.

G-eoV. Surv; of ludia

TEPvTIARY MAMMALIA,

Pl.XXXVH.

Lith.* iy JodulTathDas.

Printed at G-eop. Survey OfAce,

PLATE XXXVIII.

Mastodon (Tetralophodon) latidbns, Clift.

Fig. 1. The second right upper true molar : from the Punjab.

Fig. 3. The second right upper true molar : from a palate specimen from Burma.

Both molars are represented of the natural size : and are viewed from the inner
side.

G-eol-. Surv: of India

MAMMALIA

PI- XXXVIU.

Prmtei at Survey Office.

PLATE XXXIX.

Mastodon (Tetralophodon) latidens, Clift.

The third right upper true molar : fromi Lehri^ in the Punjab. The specimen is
drawn of the natural size, and is viewed from the inner side.

i/>

PLATE XL.

Mastodon (Tetralophodon) peeimensis, Falc. & Caut.

TragTneiit of left maxilla containing the last premolar, and the first true molar:
from the Punjab. The specimen is drawn of the natural size.

Geol.Suw. of India, T E H T I AB.Y IVTAMlVlALlA . PI.XL

Lith.i]Dy J 0 duTSra,^! D as . Frintei at G-eoL Survey Office.

PLATE XLl.

Fig. 1. Mastodon (Tetralophodon) pkrimensis, Falc. & Caut. Second left lower true

molar : from the Punjab.

Fig. %. Mastodon (Tetralophodon) sjvalensis, Pale. & Caut. Third left ujjper milk-

molar : from the Siwaliks.

Fig. 3. Mastodon (Tetralophodon) perimensis ? Falc. & Caut. First right upper milk

molar : from the Punjab.

Fig. 4. Mastodon (Tetralophodon) perimensis^ Falc. & Caut. Second right upper true

molar : from the Punjab.

All the specimens are drawn of the natural size. The upper molars are viewed from
the inner side. The lower molar is viewed from the outer side.

TERTIARY MAMMALIA,

J. Scliaumlui-^Iiith'^

Printel at G-eol-. Survey Office.

G-eol. S-urv: of India

R. XLi

PLATE XLII.

Mastodon (Tetualophodon) perimensis, Falc. & Caut.

Third left upper true molar : from the Punjab : drawn of the natural size, and
viewed from the inner side.

Geol. Surv: of India, TERTIARY MAMMALIA. FLXLII.

J.Sdianimbur^.LiQi^ Printei at GeoL Survey Office.

■A

‘V

'/r. ■

, • *•

• '4-' . «
.> >

.» ■'f

4VJ

«M • '■

.-?i.v.-. ' ?4

■e v *

r'.

■• ■ "V -:

: '.' A n».

'i- . ., ■>

f.v *

f ■ :v4r'^?%

. t*; ♦ ^■*-

. ♦•

••W v ■'

*4.'

4: ^ >% ;•■«:

' ■■• % ■tv.?’' !* -i-

^'A3 •

#?.■

^ ^ 4>.:

.•%

,.. f. "T-'J ■*.''fs’';rf if-*,-* Vi®' *‘? '*^' « '■’■' ••■’** -V'

■e<.4A

Ji- '.' A , f^--- ■
# •#-

%»‘it

■t .

'■'i

J* .• .i;? ^ -V ■

■' "'■ i *’'■'«“*?■

l,i^ ■• fj/f v.

!" V.'-V

.»■ J»^

iir v^

f ^

.t w->

. 4.

'■

(■•^/> 'V*

k' ■\^V'‘’d£
^3:4

-..S' >• •-”

>,; ' ■ ' ‘'i

•1C4

■=■ :. . ■■ ■s* % ■

.50 Vvk^r ••

ir:

-v;

wV'

■ ’%^

IlCii.

jv •» (> ^4

■h .

^--p- ■

-kt'^ '

* .:
'■ >i

»v

« ►

'/ ; ^ • (

’* # " .■* <* . . * .•

)mV''

\ . f , ^ •

■ ‘\

PLATE XLIII.

1. Mastodon (Tetealophodon) perimensis (?) Falc. and Cant. Symphysis of man-

dible.

2. Mastodon (Tetralophodon) perimensis, Falc. and Caut. Symphysis of man-

dible broken anteriorly, and showing in cross section two small tusks.

Both specimens were obtained from the Punjab, and are represented of half the
natural size. ' <

n- XLin.

G-eol'. Surr. of India.

TERTIARY MAMMALIA,

J. Schaumburg .

Prmt&dL at Geol-. Survej^ Ofiice.

Fig.

PLATE XLIV.

Mastodon (Tetralophodon sivalensis), Falc. & Caut.

1. Third left upper true molar : from a palate specimen collected in the Punjab.

2. Third right lower milk-molar : from a fragment of a mandible collected in Kangra.

3. Third right lower true molar : from a nearly complete palate collected in,, the

Punjab.

The specimens are figured of the natural size, the upper molar being viewed from
the inner and the lower molars from the outer side.

Geol'. Sui-v- of iujia

T P. R T I ^ R Y MA MM A.L I A .

PI. XI, IV,

Pi'iDitorl, at Geol fi-.ii'vej/" Otlico

V S rjiaum k uv^ , Li Ua -

PLATE XL¥-

rig. 1. Stegodon cliftii, Falconer and Cautley. Third left upper milk-molar : from the

Punjab.

Fig. 2. (?) Stegodon cliftii, Falconer and Cautley. Second right upper milk-molar : from

China. Stegodon sinensis, Owen,

Fig. 3. Stegodon bombifrons, Falconer and Cautley. Third right lower milk-molar :

from the Punjab.

Fig. 4. Stegodon insignis, Falconer and Cautley. Third upper milk-molar; from the

Punjab.

All the figures are drawn of the natural sbe, the upper molars being viewed from
the inner and the lower from the outer side.

n-.xLV.

G-eol-. Surv; of India.

TERTIARY MAMMALIA.

Printed at Geol. Survey OfRce .

J. ScauiTil)ur|.Li)li.^

'TS • *

Fig.

PLATE XLYI.

1. Stegodon bombipkons. Falconer and Cautley. First left upper true molar.

2. Stegodon insignis, Falconer and Cautley. Second left lower milk-molar.

3. Stegodon bombipkons^ Falconer and Cautley. Second left lower milk-molar.

4. Stegodon insignis, Falconer and Cautley. Third right lower true molar.

All the figures are drawn of the natural size, the upper molar being viewed from
the inner, and the lower molarsfrom the outer side. All the specimens are from
the Punjab.

?■.

TERTIABY ¥AM¥ALrA

PI. XL VI

G-eol; Surv of India

J- ScliauinliuT|',Lithi

Printed at Geol- Survey Offace.

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Indian Tertiary and post Tertiary Vertebrata"

See other formats