ffSHBb

i# :

•"/> '.<S

' ^ ^ * 4- - ' ' ' - . * '■*>’ aN

a\^«3 ' sQ $ ^ V

/ \\w.' „,

* ■ I V v V S • • . ^

V S . v

\<p- w w
/% lyw: ^VvW*V^ =-“

* A o. ''< ,vV£ <* . '° « ^ ?/

... C0

* , A v ^

-I sC> TV O

* %% a\

' ^./v\ ^ •* ^ K

oo

\0 o. , ■ - -r‘ — ^ * A -f „

+ , w>s# oc v*$** A S %

*'"■' v'Vl^V**0

<v

tOV V. 1 " M / "C‘

/V ^ rt '* . * -fy

. - <>

Vf> .>

v,

$%

■* '■+ .;* .

' '+j- <$ :

« o \° °^. )• ■»•'

A r ft 5* %

•>> c'> «* rA o* /k r ^V. A'

*V , c.^ AV *:• © Pr 0\V

« <sx\ ,vfer\y/ 7 >. ^ «\v

6 ft

’ ^ V*

n> « ' 1 » * '<*

0 ,
>O0X .“/V -;j‘ ,> ' ^ V*

A

V^X c 0 N r; *
iN >'v~ -/ *

*

jnfcSSC"^ C 'V. ' " o5 ^ ' ^Lll> \>A s. A «<• V^Iliy x cK, ' «X.

a % '..A-- a a, V'*a^’ / %.‘ata-\/ ,.., V*

•° k 0 ^ * ° a y ks s***''' ^ v<y ^ ^ • ° / c> v . s -^7 * ^

* .V. &. ,4* vV *' 5^S% "» '>

,V '

oN *

•/

\ '^s' ,0* ^ ' /

.‘iLcy,> cP v

<V> 'Cj. o '!■, * jV^ ,J?/».

^ v

A?/' aA' " 0^'^. ' ^Lj^* ^ -V* '%■

s'

' o „ k * .A 'o. ’ '/ ; y \ <.' ,ox

1 « O « S *Ni » |

A

0 + \

A «\

'W

o c

^ "C

\0ox.

* A

a\N c (' j ,:

s Ct', ' i.

v 1 1# „ , f>. * n o

'* V A ° A/

■p s ,*1^. ' . \

<■- -*> * ^

o'" "o '■ vv 'VV'* A0' v

^ /jAW/k*' ^ A- ^ A-

- JT . v ' . vr>t - I <, - >• i~j ? e.

V ^

. '-?/> - •* v> y> =

* 'V % V? ^

• . a >' \ <?• ,'AAa*. ^

s' * ^yy/nam „ ^ ^ , tku\\\xs» S

N - ' *• / I

* ^ A^’ o 0 s 0 i rto.

■i sfj ^ o'

^ ^ ' 'V '\'s .y C

-. : jp®* »sg_v,

,0° ^O, ^ .❖y '' s 0 >

,o> »'*», ';o \> s'*/ '* > .0- »’■•- V,-

.^’ s s & «. ^ * y^«o^ ', •>- ^ rr -' s.v ^

<y •■ \ -- <$> t

f %,

s'

\

'V

A V

•\V <*\

■*,%. "''* v'V'"'

■'

va2_Iy / ' /

c$% *

«? Vf'

,0

b ^ & g . ^ V

V * > v

<^' «\>
«V

fc— . -' r»

, 0 ^ O' /-

■>boN

H -vA

> o xO O

*5 ^

A °Zr-

co N' ^

.0^' «■ ' * " / lc
A r , :■' v V.

^ A ' * jM \ °

v ^ * Ksao^ % '

o v ■‘sf'viy < \^'' ^

S ^ oN1' <r' ',. .•V1%1'* A '-^ %

■* a s /y an , , <y y o • >• •* 4a ?. */

s” «s5^y^ ' ^ ^ * ’<* An s ^ ,<v '

s„r,\ a -t V ® y voox

,GX

> <^,

-0k' x*' ' * * ,'% U" V’' ‘

N £ vf/y*- * <o a) ^

^ox J- >

, V° °X.

s- „r % y>!- • ■ ^ «' n ■ r. %

~ * o s o 5 i®’ *..V*

s’ '*, s' ,<r *. ' o a C*

' ^ ^ % ,a

z .

^ x .

. _ . 0 *kM

vV ^ '

S' * <t n V v .'A ^ ''/,.> s .'I' C. * « * V

* *<>

% ^ v. I .

itaf® "■ ^ V

■tof , ,*■* HB ., V'* .sTT’* A0° %!*■ » ^, . 's\\^' .. ^ ,0 v.®'

81 vy A N .-ioo, s 1 VN ^ > jf %'°n

- - -sta^.r ^ •.

/ «f?\ \fym: **’% '■

+ * y ■ ^L~/^r v ^ - * ^e1, 4 Kj, s ^bi^s x . >

y 0 " \A 0 N c ^ s \p^° *' 1 ^ \ y ° ’ * .c 0 * c ♦ \o^\ I '

y **

•'•'x\* ^ "a ->

X*

<v + ' ft 0 yo

5J

y'£t, %■

A ■%. >*

<S‘ .0

o >> C,'

V

A - * ' o-

<V r $-„ 'V, r. ''ti.

A^V

^ V2

•v * fA* ".p. «> «r-r ^

■4 'p x x, . V

1 8 A ., Xf. 7 " * 1 \^A X " N c * \. ' 6 4 ' ' .0^ x* ' 1

\0^.

V

x»v%.

•Ss '»7^> ', 6V

^Z* y » « s .VI

' '^-" * ''*'*% % /' :M\\\ *'

^ V, 4

z .: z

v'

‘s ^OOvVlV^ y '

«/• ,v v

^ 0X

^ 4 -/^

<Xr rds *
* r *1* ^ Jv '

V^'' oOV

fi >1 r V> 7 * * ■ S l 1 f] r,

' 0 ^ *

: \\ " ^ **v
o.v ' *'■■*■•*

^ cON0 ^ 0^ x

^:' ^•^y< *, ^ y-

; 05 ++

‘.v

'«ysss».' Oo '° 0"' <v '\^s, X

> v0^ ^ ^ VX

* ■>p <1 - <v rA^sf ^ xA *rr>s^y%

- <^s . A\M/A 0 Sp <<v

p v* %, y ■

xA ^ '

^ '• *,

■» A O. -V

V « B

"boX

A^' r . g3 ^

• * * y ' \ o % x-A

v , ^ A'

>l|l*' ^ s , , , 0 K 0

A^’% •

<* ^ 0 v ^ -A 0 ** , . s

y^u / c»n '■ ♦, ^0. ** .0

♦ A ^ iTV^ ✓

r ^ o*

' z

? ^ Cx /»

<y> o

, v 1 . " / \cjy;!* /ybJ '**''.

*. ^ c, * 'P v A0tP>X^. s

' v^V ^

■V «

0 * \ v A 3> » >s a<

A •

i{\r -y

AN v 0 K c ♦ '0_

V* » «» ~S^ x

A> x ' '

^ r •>*-

O 0

\ sC<

"*,,,*\y;l,..

■

rjy /

• ^ %**m*f-,* aV ^ ^ “ ^ Av

;' .•".,V"*' y.''i*.*,V"*V'.*' *,x /’>•«**/'

-'-■^' ' w ' >0 ^ i-mv

n ♦ «o

7 -V > O ^*0/ ^C8

• A ^ .V,

% ;/ .'y c **4$'

: ,$> *+. :UV '

/• *c

> v

.,» *•' if ^ W • «>•%. i i, • .*>4,' ' A-i | ; ■' i*%i = A^V ■V -j

A00^

0 o X

O V / s v

°o * py^'f *+ ,

V °

:* ^ \. V / % , ",w/ #v Sflmj / % „

y to»«, t‘0 •**' X(c«sc» ”*o. /4“' o° - ’ 1 ' * "

> K » «■ V- ,

-, V*

/ .<-:;■ % '%
a\ ^ r >5^ '

*j- v* = . r-> '

, . h -JrA «

° / * y I A * \v , 0 ^ * '» H o J -£/* * « i A * yV

t . o , '/£. * 1 \> ' ' ' * > «. 11 * ° / *c- \>

H:\YW •■

■r — “t '" - /■ •■' TTSX \\ / - „ r*— -v , > 2

^ ^ c> <2k

o « v

s* .** "

’?/. 7 / ♦ * S S V • B k V*

% °j j<'

^ £'* - - + >r* \\ s

' ■n, A . A» ■< ^ ®

* 0O(# * ^ ,

s a" ’ O *,S>\* r%?U - o

S \?*""S'1 -.;% *-'• vn /

■*. \ i ♦♦v% '.j&H: .$\ lK^* **v%. *

oo

* ¥prv * x°

✓ v' > - *

*o.

3> N

^y <p^ - r _; -

i ‘^Ll, , O^' ^ ^ *4 S .V. O ' a

* A '^''**''/,>"/^ ',,lV toM<V/,*''o^ /

0 ' ~ ^ r ™ 1

"oo' .sin^" v

> ,1 ^ /■ ■

' *9K0,O^ \,.o„% V>V ■'•.0^,'-

♦ ^ S

1 « o ^ *(1's \^ ^

O' ^ 1 A C‘ \ x ^ ,

^ -r.r,^A.V ^ .V. - ^SS0Sl'. ^ ^

«-* c, *

^ ;"i;: 4>% V'^v' ; :.W: / %

V ^0, 0 >>' ■ SV° - ‘ >\o - • ^/:'S\^\'.*%

: § j£E i ^ ^ f ,

o' ,0-

A

r v, *

’ ° ^ ;

* 0 . J. 14 A

, * , %

"2-, •?>

;

X0^ .

\5£

9 I

'fO ' Ar _ -■ . 0

V. ^ 0 N 0 v
. > ,0’ ^
^ <\V r

^ V -

x * o

<\v </> ^ ■ v j j * v> ^ -

. .. I .* ^ ^ ^

J*l\\' oN^ /^ ° * S « ' 1 8 * <

^ / --^ - ?- ' '-*■*■
Sia.'. '••'

\ Ss '^>-' »■ % %7ir^y. r*v«. - ’ '-

BE -%•. ^ :imA'‘ \ «?' ‘/{Ip,'-

v\> - V%#/ " ■V’% " .\^' %

lla ^ ,1*- Av' .»N(. '

y*d8&£ * y_^r,k j

,0 o

sy % ' * 0 / c- v ,

x'

»

V%*'To’0^

y J s ,9V ^ ' c »

^ S - I

S % o «

.... c^ ^ "o

<\ y ° » x ^ N c o « s s ^ - t i fi .

^ ^ °Nc^ c» 0^\ '^,

-f* *»^ c^v X. ° 0 ^ ^;r^%2L,% A

p .« ,'.'ys«&.- ...v *-

x° °^ .

5l; >

©
t-

:' .'’* ‘ ^ .^*r*% * 9^' 0^ „.,

/// X-6’ T-^rt ft »e -^J., O,

'f.K : ^ r ^ fxX .; / f. J ^ ^ % * T J

rfo.w-* <A- )M,.i1'''*sSi6 ,i'.4 "vj,'"11 V\\A c°N '• *'^o ' “*S cr° *v"«

.f

X

/

VN

Or

o •-

y %

° \
u» 1

r

% ^ *
<y ^

A^V ^

' / > ' ;; * :%, .#' • " •

0 f -y. a\ ^ ^

>bo'i * A v :-v

V ■ ® x° °X. >• ^

coy \7,.' / „ ^

'*0/ V

r

c« _

^ /,^>r x .v
f ^o. ,0l . *v 1 * ♦

•V

' 0 , X * A

N

^ ♦ -. s 0 ’ a0-

/C& ^

■- %/

: .^% \

-X ’ c^ >

° * ' ■* \N .0N»,\

o

• XA

0> . X *0, ^ tf ‘ v> O • • ^

^ '***».*

! ^ = x° o.

Z
O

o ** j *oT ^ y

V 0 N r* n **b 6 * ' 0> o' 1 B * <p

v* >b0^ :£mu*+

* ,0J °0 *■* , , A»" ^V

<A

c*

•f *^p

^ ^ V^X - - .

® %.^\/^.rS>4 ;r:,^-<»rV

% X A^' A* <C? > ^

E. W “*<w-' %/

' ,|»^ iw#.* V'»^' *»•»

-to ^V.

-* ^ '*is A^” \ . V»

o . X ■* A n S c %■ * ' * . ' ' A6 V
VV t 0 N c * , *Q. .0 x *

f ° . 0 :

; ^ o^

A +*

9v ^ o*

.V ^ ' r/N h A Ar. A^'

>'S y y r .0- Cb <$■ %, '» a s o

N‘ A’ r /, '> ** J> V V\^I'

B^il a a a - ^ c

s '

g> ^
kv vVc. /■

MEMOIRS

OF THE

GEOLOGICAL SURVEY OF INDIA.

ijJiTta'ontaloght Jndtia,

BEING

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

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

S e r , X.

INDIAN TERTIARY & POST-TERTIARY VERTEBRATA.

Vol . IV.

Part I. SIWALIK MAMMALIA - SUPPLEMENT I.

V

vk4

By R. LYDEKKER, B.A., F.G.S., etc.

TH t> PLATES.

i

I

CALCUTTA

SOU) AT THE

GEOLOGICAL SURVEY OFFICE, AND BY ALL BOOKSELLERS.
LONDON : TRUBNER & CO.

MD.CCCLXXXVI.

PRINTED BY GIBBS AND BAAIFORTH, ST. AiBANS, HERTS.

G) B

Palawntotogia Jmlijpt.

series x.

VOL. IV.

//

Fcrtfc*-,

MEMOIRS

OF

THE GEOLOGICAL SURVEY OF INDIA.

JJalinnitoIonia Jiulict,

BEING

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

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

Series X.

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

VOL. IV.

Part 1.— SIWALIK MAMMALIA— SUPPLEMENT 1.

By R. Lydekker, B.A., E.G.S., Etc.

Part 2.— THE EAUNA OF THE KARNUL CAVES
(and addendum to Part 1).

By R. Lydekker, B.A., E.G.S., Etc.

Part 3.— EOCENE CHELONIA FROM THE SALT-RANGE.

By R. Lydekker, B.A., E.G.S., Etc.

CALCUTTA :

bold at the

GEOLOGICAL SURVEY OFFICE.

LONDON : KEGAN PAUL, TRENCH, TRUBNER & CO.

MDCCCCII.

mvm bt ibb auPBRiNiBVDEirt o» aoTHUtUKar pbiktuto, isdja, 8, eastiwgs itbibi, Calcutta.

Calcutta :

GOTBBNMFNT OF INDIA CENTBAX PRINTING OFFICE,
8, BASTINGS STREET.

CONTENTS.

/

PART I.

Siwalik Mammalia— (Supplement I).

1 NTR0 DU CTORY ....

Tlie Parnates . . ,

Family Simiidas

Troglodytes sivalensis
Simia, sp. . .

Family Ccrecpitheeidce

Semnopithecus palaeindicus
Macacus sivalensis
Cynocephalus sub-himalayanus
„ falcon eri

The Antelopes

Oivas (?) latidens
Strepsioeros (?) falconeri
ldcselaphus, sp. .

Palmoryx .

Hippotragos sivalensis
Gazella porrecticornis
„ sp.

Cobus (?) palseidicus
„ patulicornis
Genus non. det.

Alcelaphus palaeindicus
The Genus Merycopotamus
Mesycopotamus dissimilis
„ nanus

„ pusillus

Addendum with part II .

Genus, letraoeros, Leach

„ Daviesi, n. sp

Postscript

nobis

Introductory Observations
Mamma’ ia . .

Semnopithecus entellus
Cynocephalus, sp.

Felis tigi is (.or ? leo)
paidus

Cbaus . .

rubiginosa .
Hyiena ciocula . .

"Viverra kavnuliensis .

Prionodon. sp.

Herpestes griseus .
fusous

Ursus labiatus
Sorex, sp.

Taphozous saccolsemus .

Phyliorhina diadema .

Sciutus macrurus .

Gerbillus indicus

PART II

The Fauna of the Kabnul Cates.

tags.

1

„

••

*•

4

i,

6

,,

6

7

II

8

II

8

10

11

i>

12

>■

11

ii

ii

16

>i

17

18
» i
11
>i

22

23

26

28

I*

21

#•

M

£0

M

31

32

»»

33

31

f*

*

11

CONTENTS.

Nesokia bandicoots
kok . .

Mus mettada

platythrix .
sp.

Golunda ellioti .
Hystrix crassidens .
Atherura karnnliensis
Lepus [cf. nigricollis)
Eqtius asinus
ap. a

Rhinoceros karnnliensis
Bos, or Bubaius.sp.
Boselaphus tragooamelus
Genus, non. det. .
Gazella benneiti .
Antilope cervicapra
Tertraceros quadricornis
Cervus aristotelis
axis

Cervulu8 mnntjac
Tragulus {cf. meminna)
Sus cristatus . .

karnuliensis
Manis gigantea
Aves ....
Neophron percnopterus
? Milvui, or Circus, sp.
Ketupa ceylonensis
Bubo coiomandus .
Francolinus pictus .

pondicerianu
Grus \cf. communis)

Ibis melanocephala
Reptilia . . .

Crocodilus, sp. .
Varanus dracaena .
Python molurus .

Naia tripudians .
Ptyas mucosus
Amphibia .

Bufo melanostictus .
Mollusca . . .

Addendum . .

Herpestes nipalensis
List of the Fauna .

PAGB.

35

it

36

it

it

it

37

38

it

39

it

40
14

it

45

it

46

a

it

47

»

49

50

51
»
»
53

53

n

9t

54

) i

55

66

*

»

57

• »

it

63

■w

TAUT III.

Eocene Chelonia fkom the Salt-Range

Introductoey ......

Pleurodira ......

Family Carettochelydidaa .

Genus Hemichelys, nobit. 6 .

„ warthi, nobis. 1

Family Chelydidae

Genus Podocoemis, Wagler 6 .

„ Indica, n. sp. nobit

59

61

62

63

it

it

MEMOIRS

OF THE

GEOLOGICAL SURVEY OF INDIA.

|M'd'ontologia Judicit,

BEING

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

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

Ser. X.

INDIAN TERTIARY & POST-TERTIARY VERTEBRATA.

Vol. IV.

Part I. SIWALIK MAMMALIA - SUPPLEMENT I.

By R. LYDEKKER, B.A., F.G.S., etc.

WITH 6*’ PLATES.

CALCUTTA :

SOLD AT THE

GEOLOGICAL SURVEY OFFICE, AND BY ALL BOOKSELLERS.
LONDON : TRUBNER & CO.

MDCCCLXXXVT.

PRINTED BY GIBBS AND BAMEOETH, ST. ALBANS, MEETS.

GfE/SZ

CONTENTS.

PAGE

Introductory . . . . . . . i

The Primates

Family Simiid» „

Troglodytes sivalensis ,,

Simia, sp. ........ 4

Family Cercopithecidse ,,

Semnopithecus palaeindicus ..... 5

Macacus sivalensis . . . . . . ,,

Cynocephalus subhimalayanus .... 6

,, falconeri 7

The Antelopes

Oreas (?) latidcns ....... 8

Strepsiceros (?) falconeri ,,

PAGE

Boselaphus, sp. . . .... 9

Pakeoryx . . 10

Hippo tragus sivalensis. . . . . . .11

Gazella porrecticomis . . . . . „

.. sp 12

Cobus (?) palseindicus „

,, patulicomis 14

Genus non. det.

Alcelaphus palajindicus ,,

The Genus Merycopotamus . . . . . .16

Merycopotamus dissimilis . . . . . . ,,

, , nanus . . . . . .17

, , pusillus . . . . . .18

/

INDIAN TERTIARY & POST-TERTIARY VERTEBRATA.

SIWALIK MAMMALIA— SUPPLEMENT I.

By R. LYDEKKER, B.A., F.G.S., etc.
(WITH PLATES I. TO VI.)

INTRODUCTORY.

The present memoir contains descriptions and figures of the remains of three
groups of Siwalik Mammals, which have either not been fully treated of in the
preceding volumes of this work, or in which new species have been founded since
such descriptions were written.

The three groups here treated of comprise the Primates, the Antelopes, and
the genus Merycopotamus. The first two groupsuare of especial interest in regard to
distributional zoology on account of their containing a large number of forms either
definitely or provisionally referred to genera now characteristic of the Ethiopian
region,1 'which have not been hitherto found in the tertiaries of Europe. In the
case of some of the fossil antelopes where the generic reference is provisional it is
quite clear that these - species have no affinity to the numerically small antelopine
fauna of modern India, and if they be not really generically identical with the
Ethiopian forms with which they are grouped, they are evidently very closely
allied, and probably indicate the source whence such genera took their origin. In
other instances, however, there can be no doubt as to the generic identity of the
Siwalik and Ethiopian forms.2 It may be observed that since the Pikermi beds
have yielded certain antelopes ( Palceoreas and Proiragelaphus ) allied to existing
Ethiopian forms, it may be inferred that the ancestors of a considerable part of the
existing Ethiopian fauna ranged over a large area of the Euro-Asiatic continent,
but as none of the Pikermi antelopes (excepting the widely distributed Gazella) have

1 Some of the existing baboons and antelopes of the Ethiopian also enter the adjacent parts of the Palsearetic region.

2 In regard to the migration of Pliocene Siwalik genera into Africa see ‘ Quart. Joum. Geol. Soc.’ vol. XLII. p. 175 (1886).

A

2

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

been generically identified with existing African forms, while Troglodytes , Gyno-
cephalus and Struthio are unknown from those deposits, the affinity of the fauna of
Africa to that of the Pikermi beds does not appear to be so close as it does to that
of the Siwaliks. The genus Merycopotamus needs no introductory comment.

I. THE PRIMATES.

Family I. SIMIIDJD.

Genus I. TROGLODYTES, E. Geoffroy.1

Dentition. — The upper dentition of the existing representative of this genus is
characterized by the small size of m. 3 and the relative shortness of the antero-
posterior diameter of the upper premolars, in both of which respects it differs
markedly from Gorilla and Simla and approximates to man. The Siwalik representa-
tive was formerly referred to a distinct genus under the name of Palceopithecus.2

Species. Troglodytes sivalensis, nobis.3

Syn. Palceopithecus sivalensis, nobis.4

History. — The palate on which this species is founded was collected in 1878 by
Mr. Theobald near Jabi in the Punjab, and has been already described and figured
in the paper cited.

Palate. — The type palate is represented in pi. I. figs. 1, la; on the right side
the canine and the last four cheek-teeth are preserved, while pm. 3 and the outer
incisor are shown by their broken bases : the summit of the canine and the postero-
external angle of m. 2 have been broken off, and the centre of the crown of the latter
has been affected by decay during the life of the animal. On the left side m. 2 is in
a perfect condition, but the crowns of m. 1 and m. 3 are broken off. All the teeth are
partially worn. It was originally considered that the specimen belonged to a
female, but a comparison with skulls of the Chimpanzee (in which the canine is
relatively smaller than in the Orang and Gorilla) indicates that it should rather be
referred to a male.6

That the affinity of the specimen is with the larger Simiidce is self-apparent.
The two series of cheek-teeth have a slight convergence anteriorly, in which respect
they differ from Gorilla, Simla, and Troglodytes niger (in which the inclination is in
the opposite direction), and agree with Hylobates. The outer border of the pre-
molars is placed somewhat internally to that of the true molars, while the corre-
sponding border of m. 2 is placed externally to that of m. 1 and m. 3, in both of
which features the specimen differs from Gorilla 7 and Simia8 and agrees with Troglo-
dytes niger8 although in the excessive development of these features it is still closer

1 ‘ Ann du Museum.’ vol. XIX. p. 87 (1812).

2 ‘ Rec. Geol. Surv. Ind.’ vol. XII. p. 33 (1879). The writer then followed Owen’s arrangement of including Gorilla in

Troglodytes. 3 ‘ Rec. Geol. Surv. Ind.’ vol. XII. p. 33 (1879). — Palceopithecus. < Loc. cit.

6 In the original figure (plate facing p. 52 fig. 5) the two fragments of the specimen were not put in apposition.

6 In the small size of its canine the male Chimpanzee resembles the female Orang.

1 Compare Owen ‘ Trans. Zool. Soc.’ vol. IV. pi. XXVI. 8 Ibid. pi. CT .XV.

9 Owen “ Odontography” pi. CXVIII. fig. 1.

SIWALIK MAMMALIA

3

to Ihjlobates} The true molars differ from those of Simla by the simple structure
of the cusps and the absence of the rugosities on the crown surface, as well as
by the small size of m. - and are also widely different from those of Gorilla in which
m- 3 is much larger than either of the other two molars ; the premolars are also of
very different form from the corresponding teeth of those genera. Compared with
Troglodytes the fossil agrees in the form of the cusps of the true molars, in the
relatively small size of m. 3 and the slight development of its posterior cusps and the
oblique connecting ridge,2 as well as in the presence of an indistinct cingulum on the
internal surfaces of these teeth3 ; the relatively small antero-posterior diameter of
the premolars and the larger size of their external as compared with their internal
cusps are also features characteristic of Troglodytes ,4 In some specimens of T. niger
the antero-posterior diameter of the premolars is relatively greater than in the fossil
but in others5 it is very nearly the same. The canine agrees precisely with that of
the male T. niger? The outer (and therefore probably the inner) incisor is relatively
narrower than in the latter, and the diastema is thus somewhat larger, but this need
not be more than a specific difference. In absolute size the fossil is rather larger
than the male T. niger 7 ; its dimensions in inches being as follows :

Interval between outer borders of second molars . . ■ .

y y y y inner 9J fy ...

.

1-5

Antero-posterior diameter of base of outer incisor

0-3

Transverse ,, ,, .

.

0-19

Antero-posterior diameter of canine

0-53

Transverse ,, ,,.....

0-51

Length of series of cheek-teeth .....

Antero-posterior diameter of pm. 3 (broken)

0-3

Transverse • ,, ,,

0-5

Antero-posterior diameter of pm. 4 ....

Transverse ,, ,,

0-35

Antero-posterior diameter of m. 1

Transverse ,, ,, .....

Antero-posterior diameter of m. 2

. . 0*5

Transverse ,, ,,

Antero-posterior diameter of m. 3

0-41

Transverse ,, >>•••••

0-46

Affinities. — The above comparisons shew that the specimen under consideration
indicates an ape generically distinct from both Gorilla and Simla , but so close to
Troglodytes as to leave little doubt of its identity — an identity rendered the more
probable by the occurence of Gynocephalus in the same region. In those respects in
which the Siwalik Troglodytes differs from the existing African species it shows
in a still more marked degree the approach to the human type of dentition
presented by the latter, and serves, in a small degree, to bind still closer the
connection between the Simiulce and the ITominidce. In the inclination of the two

1 The fossil is distinguished from this genus not only by its superior size, but by the relatively narrower premolars.

2 See Owen “ Odontography,” p. 446. 3 Owen “Anatomy of Vertebrates,” vol. II. p. 320.

4 Owen “ Odontography,” p. 446. 5 Compare Blainville “ Osteographie,” Genus Pithecus. pi. V.

6 Owen. op. cit. p. 445. 7 The length of the three molars of the fossil is 131, the corresponding length in a

male T. niger being 1*2.

4

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

series of cheek-teeth and the relative lateral position of the premolars to the true
molars the genus Hylobates makes an approach to the human type which is wanting
in all the larger existing Simiidce , and it is very noteworthy to find a similar relation
obtaining in the Siwalik Troglodytes , accompanied by a more human-like structure of
the upper premolars.1

The occurrence of Troglodytes and Simla in the Siwaliks and the existence of
the latter in Borneo and Sumatra indicates that the later ancestral home of the
larger existing Simiidce was probably in the Oriental region, although in earlier times
the family ranged over Southern Europe ; and it is a very significant fact from an
evolutionary point of view that the Siwalik Troglodytes appears to be more specialized
than any of its existing allies.

The existence of large apes in the Siwaliks of the Punjab indicates that the
physical condition of . that region in the pliocene epoch must have been widely
different from that obtaining at the present day.

Genus II. SIMIA, Linn.2

Distribution. — The genus is now confined to Borneo and Sumatra. Remains of
the existing species are found in a subfossil condition in the caves of the latter
island, but, with the exception of the Siwalik form, no fossil species have been
described.

Simia, sp.

History. — The first notice of this form was published in 1837 by Falconer and
Cautley,3 and later notes by Falconer and Prinsep4 indicated its close affinity to
Simia satyrus. The present writer5 when describing the Siwalik Troglodytes as
Palceopithccus regarded the type specimen of the present form as belonging to a male
of the Punjab ape.

Tipper canine. — The one known specimen of this form is a worn left upper
canine,6 which was obtained from the Siwalik Hills, but is now unfortunately lost.
It is fully equal in size to the corresponding tooth of a large male S. satyrus , the
resemblance being stated to be so close that it was impossible to distinguish between
the two. There is therefore every probability that the specimen indicates the
existence of a Siwalik species of Simia.

Family II. CERCOPITHEOID M
Genus I. SEMNOPITHECUS, F. Cuvier.7

Distribution. — The genus is at the present day characteristic of the forest regions
of the whole Oriental region ; and was represented by S. monspessulanus , Gervais,
in the lower pliocene of Montpellier (H^rault) and Casino (Tuscany.)8

1 The human cranium with the most ape-like palate is the one figured by O. Thomas in the ‘ Journ. Anthrop. Inst.’
vol. XIV. pi. XIV. (1885). 2 “Syst. Nat.” ed. 12. vol. I. p. 34 (1766).

3 1 Journ. As. Soc. Beng.’ vol. VI. p. 359. pi. XVIII. fig. C.

4 “ Palaeontological Memoirs,” vol. I. pp. 304-307. fig. 11 ; vol. II. p. 578.

5 ‘ Bee. Geol. Surv. Ind.’ vol. XII. p. 38 (1879). 6 “ Palaeontological Memoirs,” vol. I. p. 304. fig. 11.

7 “ Hist. Nat. d. Mammiferes ” (1821). Semnopitheque. 8 See Major ‘ Atti. Soc. Tosc. Sci. Nat.’ vol. I. p. 224 (1876).

SIWALIK MAMMALIA.

5

Species. Semnopithecus pal^eindicus, nobis.1

History. — The history of the specimens referred to this species (which were all
obtained from the typical Siwalik Hills) is given in the u Cat. Foss. Mamm. Brit.
Mus.” pt. I. pp. 2, 3.

Mandible. — The type specimen is figured in pi. I. fig. 7, and consists of part of
the hinder portion of the right ramus of the mandible showing the last four cheek-
teeth in a half-worn condition ; the right ramus of the mandible of a large male of
S. schistaceus (Hodgson) being represented in fig. 6. A comparison of these two
figures shows such an extremely close resemblance as to indicate that the two
specimens are generically identical.2 The absolute size of the teeth is the same in
the recent and fossil jaws, and both exhibit a large hind talon to mT3.3 The fossil
jaw is, however, readily distinguished from the mandible of 8. schistaceus and the
rather smaller 8. entellus by the inferior vertical depth of the ramus.4

Other specimens. — Another fragment of a right mandibular ramus and a right
astragalus are provisionally referred to this species in the “ Cat. Foss. Mainm.
Brit. Mus.” pt. I.

Affinities. — The specific distinctness of this form from 8. schistaceus and S.
entellus indicates that it is equally distinct from all the other existing species,
none of which attain the size of the former ; the fossil 8. monspessulanus is also of
considerably smaller dimensions. That S. palceindicus may have been the ancestor
of 8. schistaceus is highly probable.

Genus II. MACACUS, Cuv. and Geoffroy.5

Distribution. — The genus is now spread over the whole of the Oriental region,
and also occurs in north Africa, Gibraltar, Tibet, north China, and Japan. If the
generic determination of M. pliocenus , Owen,6 be correct the genus occurred in the
pleistocene of England, and it was represented in the upper pliocene of the Val
d’Arno by M. florentinus ) Cocchi, and M. ausonius, Major,7 and in the lower pliocene
of Montpellier and Casino8 by M. priscus, Gervais.

Species. Macacus sivalensis, nobis.9

History. — This species is founded on two fragments of maxillae obtained in 1877
by Mr. Theobald, from the Siwaliks of the Punjab, which have been previously
described and figured in the ‘ Records.’10

Maxilla. — The two type specimens are represented in pi. I. figs. 9, 1 0, and the
right half of the palate of a male of M. rhesus in fig. 8. Fig. 9 is a fragment of

1 Supra, vol. III. p. 123 (1884).

2 A comparison of these figures with figs. 4, 5, will show the distinctness of the present fossil from Cynocephalus.

3 Some specimens of S. entellus show a much smaller talon to this tooth. It is not improbable that the jaw with which
Falconer and Cautley compared the fossil may have belonged to S. schistaceus rather than to S. entellus.

4 See “ Palaeontological Memoirs,” vol. I. pi. XXIV. fig. 5.

5 ‘ Magasin Encyclopedique,” 1795. — Macaque 6 See “ Cat. Foss. Mamm. Brit. Mus.” pt. I. p. 4.

7 1 Atti. Soc. Tosc. Sci. Nat.’ vol. I. p. 39 (1876). The writer is unacquainted with a description or figure of the remains
of this species or of M. florentinus. 8 See Major, op. cit. p. 224. 9 ‘ Rec. Geol. Surv. Ind.’ vol. XI. p. 70 (1878).

10 1 Loc. cit. and vol. XII. plate facing p. 52. figs. 2, 3.

B

6

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

the left maxilla showing the unworn m. 3 and the broken bases of the three pre-
ceding teeth, while fig. 10 belongs to the opposite side and exhibits the base of pm. 4,
the almost unworn m. 1 and m. 2, and m. 3 in alveolo. The curved line of the molar
series ( fig . 10) together with the relatively small size of the hinder cusps of m. 3
( fig. 9), and the absolutely small dimensions of the teeth indicate that the fossils
belong to Macacus rather than Semnopithicus, while the small size of m. 1 in
comparison to m. 2 forbids their reference to Cercopithecus }

Affinities. — With regard to the question of species, the fossil jaws are smaller
than those of M. rhesus , M. silenus, M. radiatus, and (apparently) M. sinicus, while
the anterior zygomatic root is placed more anteriorly. The specimens have not been
compared with the non-Indian existing species of the genus, from which their
specific distinction may be assumed: they are decidedly smaller than the fossil
M. pliocenus and M. prisons , but for the reason already stated the writer has been
unable to compare them with the two Italian fossil species.

Genus III. CYNOCEPHALUS, Lac^p^de.2
Distribution. — This genus is now mainly characteristic of the Ethiopian region,
although it also ranges over the whole of Africa, C. hamadryas extending from
Abyssinia to Arabia. A fossil species is recorded by M. P. Thomas3 from the upper
pliocene of Algeria, and has been provisionally named G. atlanticus. In India the
genus occurs in the pliocene of the Siwalik Hills, and persisted into the late
pleistocene of Madras.4 It may be added that for palaeontological purposes the
genus Theropithecus must be included in Cynocephalus.

Species 1. Cynocephalus subhimalayanus (H. von Meyer6).

Syn. Setnnopithecus subhimalayanus , H. von Meyer.6
History. — The history of the only known specimen of this species is given in
the writer’s “ Cat. Foss. Brit. Mamm. Mus.” pt. I. pp. 4, 5.

Cranium. — The type specimen, which was obtained from the Siwalik Hills, is
represented in pi. I. figs. 3, 3a ; and in order to exhibit more clearly its characteristic
features it has been partially restored in outline, and a figure given (fig. 2) of the
right half of the palate of a female of the existing C. babouin in which the cheek-
teeth are unworn. The fossil comprises the greater part of the right half of the
facial part of the cranium, showing the root of the zygomatic arch and the inferior
border of the orbit (or.), and containing the five cheek-teeth in a half-worn condition,
and part of the base of the canine. The small size of the canine shows that the
specimen belongs to a female, while the contour of the maxilla and the characters of
the teeth indicate conclusively its generic position. The general contour is not
unlike that of the cranium of C. babouin , but the lateral surface of the maxilla is

1 See ‘ Rec. Geol. Surv. Ind.’ vol. XI. p. 69.

2 “Nouv. Table Method.” (1799), in ‘Mem. d. 1’ Institut.’ vol. III. p. 490 (1801).

3 ‘Mem. Soc. Geol. France.’ ser. 3. vol. III. art. 2. p. 14. pi. X. fig. 4 (1884).

4 The Madras form 'will he described in part 2 of this volume.

5 In Bronn’s “Index Palseontologicus,” p. 1133 (1848). — Semnopithecus.

6 Loe. eit.

SIWALIK MAMMALIA.

7

convex instead of concave, and the teeth are of very much larger size. In regard
to the contour of the maxilla C. anubis comes nearer to the fossil. The length of
the space occupied by the five cheek-teeth is the same as in a full-sized male of the
large C. porcarius, but the true molars of the fossil are considerably wider. The
writer has been unable to compare the fossil with female skulls of all the existing
species of the genus, but judging from the last comparison it would seem probable
that the Siwalik form attained a larger size than any existing species ; and this,
together with the apparent extinction of all the contemporary mammals, indicates
that there can be little doubt as to its specific distinctness.

Species 2. Cynocephalus falconeri, n. sp. nobis.

History. — The history of the type and only known specimen of this smaller
species is recorded in the u Cat. Foss. Mamm. Brit. Mus.” pt. I. p. 6.

Mandible. — The type mandible was obtained from the Siwalik Hills, and is
represented in pi. I. fig. 4, the right ramus of a mandible of a female of G. babouin
being represented in fig. 5. The fossil comprises the entire symphysis and the
greater portion of the right ramus, and shows the three true molars in a well-worn
condition, together with portions of the two premolars ; the canine is wanting, but
the protuberance of the ramus at the point where this tooth should be, together with
the generally elongated form of the jaw, indicates that the specimen probably
belonged to a male.

That this mandible belongs to Cynocephalus is at once evident from the elon-
gated symphysis, and the flatness of the oral surface of the latter,1 together with
the relatively broad cheek-teeth and the small size of the hind talon of m~3.
The length of the space occupied by the three true molars is very much less than in
the palate represented in fig. 3, which at once indicates that the present mandible
(even if it belong to a female) is specifically distinct from C. subhimalayanus .2 The
length of the space occupied by the three true molars is the same as in the female of
G. babouin, but the length in front of pm. 1 is much greater.

Specific distinctness. — Although it cannot be certainly affirmed that the present
form is specifically distinct from all the existing species of the genus, yet the
apparent extinction of all the true Siwalik mammals renders it probable that this is
the case, and the name of C. falconeri is accordingly proposed for it.

II. THE ANTELOPES.

Genus. OREAS, Desmarest.3

Distribution. — The two existing species inhabit all tropical and southern
Africa. Palceoreas is an allied form occurring in the lower pliocene of Pikermi, in

1 Semnopithecus and its allies (fig. 6) have the symphysis deeply depressed immediately behind the incisors.

2 In the Cercopithecida the length of the lower molars always exceeds that of the upper ; compare figs. 5, 2.

3 “ Mammalogie,” p. 471 (1822).

8

INMAN TERTIARY AND POST-TERTIARY VERTEBRATA.

the upper pliocene of the Yal d’Arno1 and of Algiers,2 and the pleistocene of
France ;8 the species of the latter area being regarded as connecting the older forms
with Tragelaphus.

Species. Oreas (?) latidens, nobis.*

Syn. Cervus latidens, nobis.5

Summary. — This species is known by the dentition, which is described and
figured in vol. III. pp. 111-114. pi. XIII. figs. 12, 13. The upper molars agree very
closely with those of 0. canna, but have a larger internal accessory column ; — a cir-
cumstance of importance in regard to the teeth of the next species, as it may indicate
a character common to the earlier forms. The enamel is smooth, and there is a
distinct median costa on the outer surface of the anterior lobes of the upper true
molars.

Genus. STREPSICEROS, Gray.6

Distribution. — The two existing species range over tropical and southern Africa ;
the smaller S. imherbis being found in Somali-land, which is on the route by which the
Siwalik genera may have passed into Southern Africa.7 The allied Tragelaphus
is a still smaller form, which is closely related to Pro tragelaphus, Dames,8 of the
Pikermi beds.

Species. Strepsiceros (?) falconeri, nobis.9

History. — This species is founded on an immature cranium in the British
Museum,10 which has not hitherto been figured.

Cranium. — The type cranium is represented in plate II. figs. 2, 2a, a cranium of
the existing S. kudu being represented in figures 1, la, of the same plate. The
fossil, which was obtained from Perim Island, wants the nasals, the premaxillae, and
part of the maxillae; the palate is obscured, the teeth partly broken, and only a
fragment of the right horn-core remains : the milk-molars are still retained, m. 1 is
in use, m. 2 in alveolo,11 while m. 3 is not seen at all. It agrees with the existing
species of Strepsiceros and Oreas 12 in the depressed frontals, with large supraorbital
pits, but differs from the latter by the absence of a protuberance on the lachrymal,
in the position of the notch in the rim of the orbit, in the wider interval between
the horn-cores, the closer approximation of the latter to the orbits, and their greater
distance from the supraoccipital ridge. As will be seen by a comparison of the

1 F. Major, ‘Quart. Joum. Geol. Soc.’ vol. XLI. p. 2 (1885). — P. montis-caroli.

2 Thomas, ‘ Mem. Soc. Geol. Soc. France,’ ser. 3. vol. III. art. 2. p. 16 (1884). P. gaudryi.

3 See Deperet, * Bull. Soc. Geol. France.’ 1884. p. 278. — P. tortieomis (Aym).

4 Supra, vol. I. p. 65 (1876). — Cervus. 5 Loc. cit.

6 List of Mammalia in British Museum, p. 155 (1843).

7 Lydekker * Quart. Joum. Geol. Soc.’ vol. XLII. p. 175 (1886). 8 Sitz. Ges. nat. Berlin. 1883. pp. 95-97.

9 ‘ Geol. Mag.’ dec. 3. vol. II. p. 170 (1885).

10 See “ Cat. Foss. Mamm. Brit. Mus.” pt. II. p. 47. No. 37262 (1885).

11 In the “ Cat. Foss. Mamm. Brit. Mus.’ loc. cit. this tooth was mistaken for m. 3.

12 A figure of the cranium is given by Rutimeyer, ‘ Hinder der Tertiar-Epoche.’ pi. VT. (‘ Abh. schw. pal. Ges.’ vol. V.)

SIWALIK MAMMALIA.

9

figures, the fossil agrees with the kudu in all the above-mentioned respects, and
also in the contour of the characteristic facial profile, and the position of the horn-
core relatively to the orbit. The position of the large lachrymal fissure is still
indicated in the fossil. The horn-cores are proportionately somewhat smaller than
in the kudu, but their direction may have been approximately similar. This
smaller size may be partly due to immaturity, but may also indicate a resemblance
to & imberbis. The upper true molars of the existing species of Strepsiceros are
relatively broad, the outer surface of each lobe being wide and without a distinct
median costa ; the enamel is rugose, and there is no accessory internal column in
the median valley. The teeth of the fossil agree very closely with those of the
kudu, with the exception that they possess a distinct internal accessory column,1 2
but it is very doubtful if this difference necessarily implies generic distinction. An
upper molar of a Siwalik ruminant has been described and figured in vol. III. p. 114.
pi. XIII. figs. 9, 10, which it was suggested might possibly belong to the Pikermi
genus Palceoryx ; since, however, this tooth appears to agree precisely with the
molars of the present form it is probable that it should be referred to it : the tooth
presents the characteristic antelopine (as opposed to cervine) feature of the accessory
column being attached entirely to the posterior crescent. When adult it is probable
that the fossil species was nearly as large as the kudu.

Affinities. — The resemblance of the cranium under consideration to that of the
existing Strepsiceros is so close that it may safely be regarded as indicating an allied,
and probably ancestral form, which may be at least provisionally included in the
same genus. This form is certainly specifically distinct from the two existing
species, and is larger and otherwise different from Trayelaphus, and therefore
probably also from Protragelaphus (of which the skull is not figured) ; it is moreover
widely different from Palaeoreas?

Distribution. — If the molar figured in vol. III. pi. XIII. figs. 9, 10, belong to
the present form, its distribution will have extended from Perim Island to the
Punjab.

Genus. BOSELAPHUS, Blainville.3

Syn. P or tax, Hamilton Smith.

The genus is confined to India, where it has existed since the Siwalik epoch.

Boselaphus, sp. nobis.*

Summary. — This form is known by teeth from the Siwaliks of the Punjab
described and figured in vol. III. pp. 114-116. pi. XIII. which with the exception of
their superior size are indistinguishable from those of B. trayocamelus. The upper

1 The first true molar of one side has been cleaned in order to exhibit this structure ; the accessory column was
unfortunately cut through, hut its section is visible.

2 Compare Gaudry, “Les Enchainements, &c. — Mamra. Tert.” p. 82. fig. 91.

3 ‘ Bull. Soc. Philom. Paris.’ 1816. p. 75.

4 Supra vol. III. p. 114 (1884).

C

10

INDIAN TERTIARY AND POST-TERTIARY YERTEBRATA.

molars of this genus ( woodcut fig. 1 ) are characterized
by their broad and tall crowns, the narrow external
surfaces of the lobes carrying distinct median costae,
and the tall, narrow, internal accessory column,
which is attached entirely to the posterior crescent,
and is not expanded antero-posteriorly at its inner
termination.1

Genus. PALiEORYX, Gaudry.

The upper molar described and figured in
vol. III. p. 114. pi. XIII. figs. 9, 10, which it was
suggested might possibly belong to this genus has
been shown on the last page to be indistinguishable
from the molars of Strepsiceros (?) falconeri.

Fig. 1. Boselaphus tragocamelus (Pallas). The
second left upper true molar, in an almost
unworn condition : recent, India.

Genus. HIPPOTRAGUS, Sundevall.2

Distribution. — The existing species of this genus inhabit tropical and south Africa.

Species. Hippotragus sivalensis, nobis.3

Syn. Antilope sivalensis , nobis.4

History. — The so-called Antilope sivalensis was founded on a very imperfect and
crushed cranium from the Siwaliks of the Kangra district, which is preserved in the
Indian Museum, and is figured in vol. I. pi. XXV. figs. 1, 2 of the present work.
A more perfect cranium in the British Museum, was identified by the writer with this
species, and at first provisionally,5 and subsequently more definitely6 referred to the
present genus. The Calcutta specimen has been sent to England and compared
with the one in the British Museum ; the two are very similar, but the former has a
depression on the lachrymal, which is not improbably due to crushing.7

Cranium. — The British Museum cranium is represented in pi. II. figs. 4, 4a ; a
cranium of the African H. niger being represented in figs. 3, 3a of the same plate.
The fossil, which probably belonged to an immature male, shows the bases of the
horn-cores, but has lost the nasals, premaxillae, and a small portion of the occiput ;
the last two milk-molars are still retained, the third true molar not being protruded
from its alveolus. A comparison of the figures will show how extremely close is the
resemblance between the recent and fossil crania. Thus both show the absence of
a distinct lachrymal depression, the existence of a small elongate lachrymal
fissure, and similar supraorbital foramina without pits at their entrance. Both

1 Unworn teeth of Cervus aristotelis present a considerable resemblance to those of Boselaphus, but are readily distinguished
by their broader and shorter crowns, and by the circumstance that the inner accessory column is attached to both crescents.
The distinctive features of the molars of Hippotragus will be noticed under the head of that species.

2 ‘ K. Svenska Vet-Akad, Handl.’ for 1844. p. 196. 3 Supra vol. I. p. 154 (1878) — Antilope. 4 Zoc. cit.

5 ‘ Geol. Mag.’ dec. 3. vol. II. p. 170 (1885).

6 “ Cat. Foss. Mamm. Brit. Mus,” ,pt. II. p. 49. No. 39558 (1885).

7 If this depression should be normal it would indicate that the Calcutta is distinct from the British Museum specimen ;
in view of which contingency comparisons will be confined to the latter.

SIWALIK MAMMALIA

11

show the same narrow muzzle, with a sudden expansion of the maxilla at the
infraorbital foramen ; in both the orbit is large with a notch below the level
of the supraorbital foramen ; the horn-cores are similarly situated in respect to
the orbits, with a considerable interval between them, and directed at their bases
upwards and backwards ; there is also the same concave facial profile in the orbital
region, and the same strongly-marked ridge descending from the orbit to the first
true molar. The palate is generally similar in the two, but the anterior pair of
basioccipital tubercles are more approximated in the fossil ; both present the same
deep concavity in the median line in advance of the premolars. The upper molars
of H. niger have very broad and nearly square crowns, presenting a remarkably
bovine structure ; the outer surface of each lobe is relatively wide, with a strongly-
marked median costa, and there is a large internal accessory column, of which the
inner termination is expanded antero-posteriorly.1 The imperfect protrusion of the
molars of the fossil specimen renders their distinctive features obscure, but mm 4
agrees in general characters with those of the existing species. The united length
of m. 1 and m. 2 is nearly the same as in II. niger, and thus indicates that adult male
specimens of the fossil were about equal in size to that species. On the assumption
that the Calcutta specimen is specifically the same as the British Museum cranium,
it will probably indicate a female individual.

Affinities. — The resemblance between the British Museum Siwalik cranium and
that of H. niger is so close as to leave little doubt that the two species are generically
the same, and in any case that they are very closely allied. No other fossil species
of the genus appear to have been recorded ; and of the existing species II. niger and
E. leucophceus are confined to south Africa, while H. equinus ranges into Nubia and
the adjacent districts.

Genus. Gazella, Blainville.2

Distribution. — This genus has a wide distribution at the present day, being most
abundant in the deserts on the borders of the Palaearctic, Oriental, and Ethiopian
regions; its range in Europe extends from the lower pliocene of Pikermi to the
Norwich Crag.3

Species. Gazella porrecticornis, nobis!

Syn. Antilope porrecticornis , nobis.5

History. — This species was founded upon the left half of a frontlet and horn-
core figured in vol. I. pi. XXV. fig. 4, which was there stated to be allied to the
gazelles, and was subsequently referred to the present genus.6

Frontlet. — The type specimen is represented in the accompanying woodcut
(fig. 2), the right half having been restored. The frontal is complete with the

1 The inner expansion of this column at once distinguishes the molars of Hippotragm niger from those of Boselaphus, in
which the crowns and the base of the outer surfaces of the lobes are relatively narrower.

2 ‘ Bull. Soc. Philom. Paris.’ 1816. p. 75. 3 Q. anglica, Newton.

4 Supra vol. I. p. 158 (1878) — Antilope.

6 See “Cat. Foss. Mainm. Brit. Mus.” pt. II. p. 50.

5 Loc. cit.

12

INDIAN TERTIARY AND POST-TERTIARY YERTEBRATA.

exception of the process descending on the outer side of the nasal, while the horn-

core has lost about one-quarter of its
length. In the laterally compressed
horn-core, the form and position of
the supraorbital foramen and sulcus,
and the relative shortness of the
frontal,1 the specimen agrees so
exactly with the gazelles, that its
reference to the present genus may
be considered certain. It indicates
a considerably larger species than the
Indian G. bennetti, and in this respect
agrees with the east African G. granti ,
although differing slightly in the form
of the supraorbital sulcus and the
curvature of the horn-core. It agrees

Pig. 2. Gazella pwrecticomis. The frontlet ; from the Siwaliks of with that Species in the extreme
the Punjab. $. Indian Museum (No. B. 229). shortness of the frontal ; but Unfortu

nately it cannot be determined whether it also agreed in the absence of the
lachrymal depression, which is present in G. bennetti.

Gazella, sp.

Frontlet. — The frontlet represented in pi. IV. fig. 6 was obtained from Niki in
the Punjab,2 and has hitherto been included in the preceding species ; it agrees
exactly in size with the frontlet of a male of G. bennetti , and if adult, as appears to
be the case, is too small to be the female of G. porrecticornis. The specimen is too
imperfect to afford a specific diagnosis, but it not improbably indicates the existence
in the Siwaliks of the Punjab of a gazelle closely allied to the living Indian species.

Genus. COBUS, A. Smith.3

Including Adenota and Onotragus, Gray.

Distribution. — The genus includes a considerable number of species inhabiting
tropical Africa. An antelope has been described by M. P. Thomas4 under the name
of Antilope tournoueri from the upper pliocene of Algeria, which is regarded as allied
to the present genus.

Species 1. Cobus (?) pal^indicus, nobis: 5

Bistory. — This species was founded on an imperfect cranium in the British
Museum (No. M. 2402) from the Siwalik Hills, which was provisionally referred to the
present genus.

1 These, among other characters, distinguish Gazella from Cobus.

2 “ Cat. Siwalik Vert. Ind. Mus.” pt. I. p. 20. No. B. 228. 3 “ Illustra. Zool. of S. Africa.” No. 12 (1840), Kobus.

4 ‘Mem. Soc. Geol. France.’ ser. 3. vol. III. art. 2. p. 15. pi. VII. fig. 1 (1884).

6 “Cat. Foss. Mamm. Brit. Mus.” pt. II. p. 53 (1885).

SIWALIK MAMMALIA.

13

Cranium. — The type cranium is represented in pi. III. figs. I, la; a cranium
of the existing Cobus sing-sing being represented in figs. 4, 4a. The fossil, which
wants the horn-cores, and is otherwise damaged, indicates a species about one-third
smaller than C. ellipsiprgmnus and C. sing-sing. It agrees with the figured recent
cranium of the latter in the large frontal depression and foramina, the presence of
an infraorbital fissure and absence of a lachrymal depression, as well as in general
contour, position of the orbits, setting on of the horn-cores, and the marked ridges
bordering the temporal fossae; it also agrees in the form of the basioccipital
tubercles, of the tympanic bullae, and of the hinder part of the palate. The molars
of the existing species are narrow, and have narrow external lobes with a strongly-
marked median costa, and no inner accessory column ; in all of which respects those
of the fossil cranium agree. The latter is, however, narrower than the cranium of
C. sing-sing , its post-cornual portion relatively longer, and the angle formed by the
meeting of the nasal and parietal planes considerably less open. In the two latter
respects the species generically separated by Gray under the name of Adenota 1 make
a nearer approach to the fossil, while Cobus leucotis comes perhaps still nearer in these
points, and also agrees in size.

A second Siwalik cranium in the British Museum (No. M. 487)2 is in very nearly the
same condition as the type, with which it agrees almost exactly, although rather wider.

The hinder portion of a cranium represented in figs. 2, 2a of plate III. is also
from the Siwalik Hills, and is preserved in the British Museum (No. 39559) ; it
agrees with the type cranium in essential characters, and is therefore provisionally
referred to the same species. The horn-cores are widely separated, rounded, and
directed upwards, backwards, and somewhat outwards ; the horns, as in C. sing-sing ,
may have been lyrate. The relatively large size of the post-cornual portion of the
cranium and the comparatively sharp angle at the junction of the nasal and parietal
planes are well exhibited in this specimen.

The females of the existing species of Cobus are hornless, and as there is in
the British Museum the hornless cranium of an antelope (No. 39569s) from the
Siwalik Hills, which agrees with the present form in size, general contour, and
structure of the molar teeth, it is highly probable that it belongs either to this or the
next species.

Affinities. — The close resemblance existing between the specimens described
above and the crania of the genus Cobus , together with the strong evidence in favour
of the female of the Siwalik form having been hornless, indicates that the recent and
fossil forms are closely allied, and justifies at least the provisional reference of the
latter to the existing genus.

Distribution. — All the specimens referred to this species were obtained from the
typical Siwalik Hills.

1 Compare “ Cat. Mamm. Brit. Mua. — pt. III. Ungulata Furcipeda,” pi. XI. fig. 1 (1852).

2 This specimen was overlooked when part II. of the Brit. Mus. Catalogue was written.

3 “ Cat. Foss. Mamm. Brit. Mus.” pt. II. p. 58.

D

14

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

Species 2. Cobus (?) patulicornis, nobis.1

Syn. Antilope patulicornis , nobis}

History. — This species was originally described on the evidence of an imperfect
frontlet figured in vol. I. pi. XXY. fig. 3, which was provisionally referred to the
present genus in the u Cat. Foss. Mamm. Brit. Mus.” pt. II. p. 54, where several
imperfect specimens of the cranium are recorded.

Cranium,. — The best preserved of the British Museum specimens (No. 39559, a)
is represented in plate III. figs. 3, 3a, and consists of the hinder part of the cranium
bearing the basal halves of the horn-cores.3 In general characters this specimen
agrees with the cranium of the preceding species, but is readily distinguished by the
more widely divergent horn-cores ; in which respect it also differs from existing
species of Cobus.

Affinities. — The resemblance of the cranium of the present form to that of the
preceding species renders it probable that it should be referred to the same genus,
but more perfect specimens are required before its affinities can be further elucidated.

Distribution. — The type specimen was obtained from the Hushi&rpur district,
while the British Museum examples were derived from the typical Siwalik Hills.

Genus, non. det.

Frontal. — The frontal and horn-core represented in pi. III. fig. 5 was obtained
from Perim Island, and is described in the “ Cat. Foss. Mamm. Brit. Mus.” pt. II. p. 54
as of the left side, but it is very difficult to say to which side it really belongs, and
it is here figured as belonging to the right.4 It has the depressed frontal region and
the widely separated horn-cores of Cobus , and if it belong to the left side the
direction of the horn-cores is nearly the same as in existing species of that genus ;
if, however, it is here figured in the right position their curvature is reversed. More
perfect specimens are required to elucidate the full affinities of this large antelope.
Genus. ALCELAPHUS, Blainville.6

Including Damalis , Gray.

Distribution. — There are some nine or ten existing species, ranging over the
whole of Africa, and extending north-eastward into Syria. A. bubalis occurs fossil
in the upper pleistocene of Algiers.6

Species. Alcelaphus pal^eindicus (Falconer7).

Syn. Antilope palceindica , Falconer.8

Alcelaphus balceri, Lydekker.9

History. — The history of this species will be found in vol. III. p. 117, and in
the “ Cat. Foss. Mamm. Brit. Mus.” pt. II. p. 55.

i Supra, vol. I. p. 157 (1878) — Antilope. 2 Zoo. eit.

3 The artist has unfortunately not drawn fig. 3 in the same position as the other specimens, the occipital portion having
been too much elevated.

4 If the specimen belong to the left side the right side of the figure is the inferior border.

5 ‘Bull. Soc. Philom. Paris.’ 1816. p. 75. 6 Thomas. ‘ Mem. Soc. Geol. France.’ ser. 3. vol. III. art. 2. p. 38 (1884).

7 “Cat. Foss. Vert. As. Soc. Beng.” p. 154 (1859). 8 Loe. cit. 9 ‘ Geol. Mag.’ dec. 3. vol. II. p. 170 (1885).

SIWALIK MAMMALIA.

15

Cranium. — The type cranium is represented in pi. IV. figs. 4, 4a, and a second
specimen in fig. 5 ; the former shows the greater part of the face, but lacks the
hinder part of the occiput, while the latter shows the whole of the occiput, but has
lost the anterior part of the face ; there are some slight differences in these two
specimens, which cannot, however, be regarded as of more than individual value.

The general contour of the skull indicates without doubt that the species belongs
to Alcelaphus, and in order to exhibit this affinity the crania of the existing A. tora
(fig. 1) and A. pygargus (fig. 2) have been figured for comparison.1 The fossil agrees
with A. torn (with which may be grouped A. caama and A. bubalis) in its extremely
elongated face, characterized by the straight infraorbital profile, and the long
terminally-expanded nasals, which form a bold median ridge bounded by deep
hollows in the lachrymals and maxillae. The horn-cores are, however, placed
directly over the orbit as in A. pygargus (which agrees in this respect with A. albifrons
and A. senegalensis ), instead of being pushed back to the vertex of the cranium as in
the A. tora group,2 and the occipital and parietal planes are consequently distinct, as
in the former group.3 The direction of the horn-cores is nearly the same as in the
A. pygargus group, which also agrees with the fossil in the concavity of the facial
profile at the orbit, although differing in the perfect straightness of the infraorbital
portion.

Young cranium. — The young cranium represented in figs. 3, 3a, is the type of
A. ba/ceri, which was regarded as allied to the A. pygargus group. The opportunity
of comparing this specimen with the one represented in fig. 5 has, however, con-
vinced the writer that the two cannot apparently be specifically separated ; the
shorter face of the young specimen being probably due to its immaturity.

Molars. — The second and third left upper true molars of the Calcutta specimen
are represented in fig. 5a. These teeth agree with those of existing species of the
genus in their extremely narrow form, and the absence of any trace of the internal
accessory column.

Affinities. — The present species may apparently be regarded as one intermediate
between the A. tora and the A. pygargus group, being allied to the former in the
characters of the face, and to the latter in those of the postorbital portion of the
cranium. Judging from the analogy of the genus Bos , the A. tora group is the more
specialized of the two, and as it is probable that the facial would be more readily
modified than the occipital region of the cranium, it is interesting to find that in the
Siwalik form it is this part which has assumed the specialized character. The
intermediate nature of A. palceindicus fully confirms the conclusion now generally
entertained as to the generic unity of the A. tora and A. pygargus groups.

1 The absence of a lachrymal fissure and of supraorbital pits, the presence of a well-marked lachrymal depression, and
the narrow crowns of the cheek-teeth are characteristic features of the genus.

2 Compare Gray “ Cat. Mamm. Brit. Mus. — pt. III. Ungulata Furcipeda.” pi. xvi. fig. 1 (1852).

3 When describing the type specimen in the “ Cat. Foss. Mamm. Brit. Mus.” pt. II. the writer thought that the post-
cornual portion of the cranium would be much shorter than it really turns out to be, and he therefore considered the species
nearer to the A. tora group than now seems to be the case.

16

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

Distribution. — All the known remains1 of this species have been obtained from
the typical Siwalik Hills.

HI. THE GENUS MERYCOPOTAMUS.

Present notice— The chief characters of this genus have been already described
in the present work2 and in the “ Cat. Foss. Mamm. Brit. Mus.,”3 and the object of
the present notice is to give figures on a scale sufficiently large to illustrate some of
the characteristic features of the two larger typical species. The structure of the
upper molars of the small M. pusillus renders it probable that Hemimeryx 4 must be
included in the present genus.

Species 1. Merycopotamus dissimilis, Falc. and Caut.6

Characters. — This species is of comparatively large size, and characterized by
the wide mandibular symphysis, the long jaws, the comparatively straight profile
.and slight vertical depth of the cranium, the presence of a deep fossa on the outer
side of the mandible of the male behind the canine, and the nearly flat inner surface
of the third lobe of mTh in the same sex. The canines are proportionately smaller
than in the next species.

Cranium. — The type cranium is represented in the “ Fauna Antiqua Sivalensis,”
pi. LXVII. figs. 1, la, on a reduced scale; a full-sized view of the last four cheek-
teeth being given in pi. V. fig. 1 of this memoir. As the anterior part of this
specimen is unfortunately wanting the size of the canine cannot be determined, but
the large dimensions of the cranium indicates that it belonged to a male. The
straightness of the cranial profile is well shown in the original figures ; the length of
the space occupied by the three true molars (of which the last is almost unworn) is
3’2 inches.

In pi. V. fig. 2 there is given the palatal view of the anterior part of a cranium
collected by Mr. Theobald near Bhimber, which may provisionally be regarded as
belonging to a female6 of the present species. The specimen is considerably damaged,
but exhibits nearly all the cheek-teeth, portions of the canines, and the broken bases
or alveoli of the incisors. In its general contour and low elevation the cranium
agrees with the type specimen, but is of considerably smaller dimensions, the length
of the space occupied by the three true molars being 2-8, and the interval between
the canine and the hinder border of m. 3 6-2 inches. The structure of the cheek-
teeth is similar to that of the type specimen7 ; the canines are comparatively small,
and have an antero-posterior diameter of 0-7 inch. The structure of the palate
agrees generally with that obtaining in Sus, and exhibits similar large foramina ; it
is, however, relatively wider, and the incisors are of subequal size. In the two latter
respects the genus agrees more nearly with Hippopotamus ; and similar features are

1 See “ Cat. Foss. Mamm. Brit. Mus.” pt. II. pp. 65-6, and “ Cat. Siwalik Vert. Ind. Mus.” pt. I. p. 20.

2 Vol. II. p. 165. The feet were tetradactylate. 3 Part II. p. 209.

4 Supra, vol. II. p. 167 pi. XXIII. fig. 5. 5 < Asiatic Researches.’ vol. XIX. p. 51 (1836) — Hippopotamus.

6 It is entered in the “ Cat. Siwalik Vert. Ind. Mus.” pt. I. p. 38 (No. B. 110) as helonging to a small male.

7 Those of the tyye spcimen are viewed more from the external surface.

SIWALIK MAMMALIA

17

exhibited by the reentering angle formed by the anterior termination of the nasals,
the backward position of the fronto-nasal suture, and the general contour of the
facial aspect of the cranium.

Mandible. — The right ramus of the mandible of a male of this species figured
by Falconer and Cautley in the “Fauna Antiqua Sivalensis,” pi. LXVII. figs. 4, 4a,1
is refigured on a larger scale in pi. V. fig. 3. The length of the space occupied by
the three true molars is 3-45, the interval between the canine and the hinder border
of mT3 7*8, and the antero-posterior diameter of the canine 1T5 inches; these
dimensions agreeing very closely with those of the male cranium. The hippopotamine
features presented by this specimen have been described in previous notices.

There are no complete specimens known of the mandibular ramus of the female
of this species ; and it is very difficult to make specific determinations in the case of
imperfect specimens. The left half of a symphysis has been referred to a female in
the “ Cat. Foss. Mamm. Brit. Mus.” pt. II. p. 211. No. 18444, and it is not impro-
bable that the immature specimen No. 17002 noticed on the same page may also
belong to this sex. The hinder part of a right mandibular ramus from the Punjab
figured in pi. VI. fig, 3. agrees with the latter specimen, and also accords in size with
the female cranium figured in pi. V. fig. 2. The third true molar is narrower than
the corresponding tooth in the jaw of the male of the present species, and in the
two figured jaws of M. nanus (pi. YI. figs. 4, 5).

Distribution. — A young cranium from Burma is referred to this species in the
“Cat. Siwalik Vert. Ind. Mus.” pt. I. p. 38. No. B. 112, and the range of the
species seems therefore to have extended from the Punjab to Burma.

Species 2. Merycopotamus nanus, nobis.2

Characters. — This species is smaller than the last, and is characterized by the
narrower mandibular symphysis, the relatively shorter jaws, the convex profile
and great vertical depth of the cranium, the absence of a deep fossa behind the
lower canine of the male, and the deep concavity of the outer surface of the third
lobe of mTB. The canines are of great relative size.

Cranium.— The cranium of a male of this species figured in the “Fauna
Antiqua Sivalensis,” pi. LXVII. figs. 3, 3a,3 is refigured on a larger scale from the
palatal aspect in pi. VI. fig. 1 of this memoir. The length of the space occupied by
the three true molars is 2-6, the interval between the canine and the hinder border of
m. 3 5-55, and the antero-posterior diameter of the canine 086 inches. Contrasting
this figure with that of the female cranium of M. dissimilis (pi. V. fig. 2), the
characteristic shortness of the jaw is well exhibited, and this would be still more
marked if the male cranium of which the cheek-dentition is represented in pi. V.
fig. 1. were sufficiently complete to exhibit the whole of the dental series.

A female cranium is figured on a scale of one- third in the 1 £ Fauna Antiqua

1 Also in pi. LXII. fig. 16, on a scale of

2 < Geol. Mag.’ dec. 3. vol. I. p. 545 (1884). The specimens of this species figured in the “ .fauna Antiqua oivaiensis,”

were all referred to M. dissimilis. 3 See “ Cat. Foss. Mamm. Brit. Mus.” pt. II. p. 211. No. 16551.

18

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

Sivalensis,” pi. LX VII. figs. 5, 5a,1 the left half of the palate being also figured on
a scale of one-half in pi. LXII. fig. 15.2 The length of the space occupied by the
three true molars is 2-6,3 the interval between the canine and the hinder border of
m. 3 5-3, and the antero-posterior diameter of the canine 0*7 inches. The charac-
teristic convex profile and great vertical height of the cranium are well shown in this
specimen.4 The middle part of a very old cranium from the Punjab in the Indian
Museum, of which the left dentition is represented in pi. VI. fig. 2, agrees very
closely with this specimen. The teeth are very much worn, and the premolars
are broken ; the three true molars occupy a space of 2-46 inches. The dimenions
of these two specimens differ very greatly from those of the female cranium of
M. dissimilis represented in pi. V. fig. 2.

Mandible. — The left ramus of the mandible of a male figured in the “ Fauna
Antiqua Sivalensis,” pi. LXVII. figs. 8, 8a5 is refigured on a larger scale in pi. VI.
fig. 4. The inner side of the specimen is not cleared from matrix, and the necessary
foreshortening of the canine (c.) does not give a good idea of the large size of this
tooth. In relative dimensions this specimen agrees precisely with the male cranium
represented in fig. 1, the length of the space occupied by the three true molars being
2*7, the interval between the canine and the hinder border of m. 3 6*4, and the
antero-posterior diameter of the canine 1T8 inches. The remarkable difference
between this specimen and the male mandible of M. dissimilis represented in pi. V.
fig. 3 will be apparent from a comparison of the figures and dimensions.

In plate VI. fig. 5 a larger view is given of a female mandibular ramus figured
in the “Fauna Antiqua Sivalensis,” pi. LXVII. figs. 7, 7a.6 The specimen belongs
to the right side, and shows the last true molar in a half- worn condition, the bases of
the preceding cheek-teeth, and a section of the canine (c.). The interval between
the latter and the hinder border of m. 3 seems to have been nearly the same as in
the male specimen, but the antero-posterior diameter of the base of the canine is only
0-78 inch.

Distribution. — Assuming that the specimen represented in pi. VI. fig. 2 is rightly
referred to the present species, the range of the latter will have extended from the
Siwalik Hills to the Punjab.

Species 3. Merycopotamus pusillus, nobis.6

The characteristic features of the third upper true molar on which this small
species is founded have been already described in the preceding volume of this work
(pp. XII.-XIII).

i “ Cat. Foss. Mam. Brit.’ pt. II. p. 212. No. 16552. 2 Wrongly entered as the lower jaw in the description of the plate.

3 Owing to the worn condition of m. 1 in the male cranium the length of the molar series is abnormally short.

4 See the figures in the “ F. A. S.” 6 See “ Cat. Foss. Mamm. Brit. Mus.” pt. II. p. 212. No. 18407.

6 “ Cat. Foss. Mamm. Brit. Mus.” pt. II. p. 212. No. 15349.

POSTSCRIPT.

It should have been observed that the reference of Antilope torticornis, Aymard, to Paloeoreas is on
the authority of M. P. Thomas (‘ M6m. Soc. G6ol. France.’ ser. 3. vol. III. art. 2. p. 16), and that it is
referred by M. Depdret in the passage quoted in the text to Tragelaphus. According to Prof. Riitimeyer’s
description (‘Die. Rinder der Tertiar-Epoche, etc.’ p. 85) the molars of this form are very similar to
those of Strepsiceros (?) falconeri. Since the text was in type the writer has noticed that a tooth from the
‘ Bohnerzgruben ’ of Wiirtemberg has been described by Prof. Riitimeyer (see p. 88 of the work cited)
as Hippotragus fraasi. If this determination and the reference of A. torticornis to Tragelaphus rather
than to Paloeoreas be correct, there is evidence of the occurrence of two existing African genera of
antelopes in the tertiaries of Europe.

« PLATE I.

Primates — Simiidce and Gercopithecidce.

. 1, la, Troglodytes sivalensis, Lyd. The imperfect palate of a male from Jabi, Punjab.
Indian Museum (No. D! 1). Page 2.

2. Cynocephalus babouin, Desm. The right half of the palate of a female; Africa.
British Museum (No. 36. C).

3. 3a. Cynocephalus subhimalayanus (Meyer). The imperfect right half of the facial portion

of the cranium of a female; from the Siwalik Hills. British Museum (No. 31157).
Page 6.

4. Cynocephalus falconeri, Lyd. The imperfect mandible of a male ; from the Siwalik
Hills. British Museum (No. 15709). Page 7.

5. Cynocephalus babouin, Desm. The right ramus of the mandible Belonging to the same

individual as the palate represented in fig. 2.

6. Semnopithecus schistaceus (Hodgson). The right ramus of the mandible of a male ;
Kashmir.

7. Semnopithecus pal^eindicus, Lyd. Part of the right ramus of the mandible of a
male (?) ; from the Siwalik Hills. British Museum (No. 1 o 7 10). Page 5.

8. Macacus rhesus (Audebert). The right half of the palate of a male ; India.

9.

Macacus sivalensis, Lyd. Part of the left maxilla; from the Siwaliks of the Punjab.
Indian Museum (No. D. 2). Page 5.

10.

Macacus sivalensis, Lyd. Part of the right maxilla of an immature individual ; from
the Siwaliks of the Punjab. Indian Museum (No. D. 2, a). Page 5.

%

All the figures natural size.

Vol.IV. Pi. I.

Geol. Surv. of India.

TERTIARY VERTEBRATA.

C B erj e a.u del et litk .

West, Newman &C° rmp

*

*

' *

i

PLATE II.

Artiodactyla — Bovidce.

1, la. Strepsiceros kudu, Gray. Cranium; S. Africa. J. British Museum (No. 646, n).

2, 2a. Strepsiceros (?) falconeri, Lyd. Immature cranium ; from the Siwalik Hills. British

Museum (No. 37262). Page 8.

3, 3a. Hippotragus Niger (Harris). Cranium; S. Africa. British Museum (No. 1038,a).

4, 4a. Hippotragus sivalensis, Lyd. Immature cranium ; from the Siwalik Hills. £. British

Museum (No. 39558). Page 10.

Geol. Svirv. of India. .

T E RT I ARY VE RT EBRATA

Vol.IV. PI. II.

4? CL

2 CL

YTect.Newnisiu&Co imp.

■

V.

PLATE III.

Artiodactyla — Bovidce.

1, la. Cobus (?) pal^indicus, Lyd. Cranium; from the Siwalik Hills. British Museum (No.

M.2402). Page 13.

2, 2a. Cobus (?) palajindicus, T.yd. Hinder half of the cranium ; from the Siwalik Hills.

British Museum (No. 39559). Page 13.

3, 3a. Cobus (?) patulicornis, Lyd. Hinder half of the cranium ; from the Siwalik Hills.

British Museum (No. 39559, a). Page 14.

4, 4a. Cobus sing-sing (Bennett). Cranium ; W. Africa. British Museum (No. 744, e).

5, Genus ? ion. del. A frontal ; from Perim Island. British Museum (No. M.2402, a).

Page 14.

Figs. 4, 4 a I, the rest $ nat. size. In fig. 3 the occiput has been more elevated than in figs. 1, 2, 4.

Geol. Surv. of India..

TERTIARY VERTEBRATA

Vol.IV. PI. HI

C.Berjeau olel ©t litH.

We»t NewmanXCo unp

-

.

■

-

■

PLATE IV.

ARTIO DACTYL A — Boi'idcV.

Fig. 1. Alcelaphus tora, Gray. Cranium ; Abyssinia. British Museum (No. 1647, d).

,, 2. Alcelaphus pygargus (Pallas). Cranium ; South Africa. British Museum (No. 644, a).

,, 3, 3a. Alcelaphus pal^indicus (Falconer). Immature cranium; from the Siwalik Hills.

British Museum (No. 39598). Page 15.

,, 4, 4a. Alcelaphus pal^indicus (Falconer). Cranium ; from the Siwalik Hills. British Museum

(No. 39594). Page 15.

,, 5, 5a. Alcelaphus pal.eindicus (Falconer). Cranium and rn. 2 and m. 3 ; from the Siwalik

Hills. Indian Museum (No. B. 331). Page 15.

„ H. Gazella, sp. Frontlet ; from the Siwaliks of the Punjab. Indian Museum (No. B.228,a).
Page 12.

* ■> 2, t ; fig*- 3, 3a, 4, 4a, 5, £ ; fig, 6, £ ; fig. 5a, *

Geol. Sunv. of India.

TERTIARY

VERTEBRATA.

Vol.IV.Pl.IV.

C.Beirjeau del et lifch .

We st "Newman Sc. Co imp

*

' .

4

PLATE V.

Artiodactyla — Merycopotamidce.

Merycopotamus dissimilis, F. and C. The last four left upper cheek-teeth of a male ;
from the Siwalik Hills. British Museum (No. 18411). Page 16.

Merycopotamus dissimilis, F. and C. The anterior portion of the cranium of a female ;
from the Siwaliks of Bhimber, Jamu. Indian Museum (No. B. 110). Page 16.

Merycopotamus dissimilis, F. and C. The right ramus of the mandible of a male ;
from the Siwalik Hills. British Museum (No. 18142). Page 17.

; Jigs. 2, 3, l. In jig. 2 the teeth are viewed directly from the grinding surface, but in fig. 1 the
viav is taken obliquely in order to show the external surface.

Geol. Surv. of India.

TERTIARY

VERTEBRATA

Vol.IV.Pl.V

Berjeaia. Sc HigTaley del etlith

West , Kev.raa.n Sc. Co. imp .

.

*

-

4

'

PLATE VI.

Artiodactyla — Mery co p o tamidce .

Fig. 1. Merycopotamus nanus, Lyd. Cranium of a male; from the Siwalik Hills. British
Museum (No. 16551). Page 17.

,, 2. Merycopotamus nanus, Lyd. The worn left upper cheek-dentition of a female ; from the

Siwaliks of the Punjab. Indian Museum (No. B. 111). Page 18.

,, 3. Merycopotamus sp. Part of the right ramus of the mandible ; from the Siwaliks of the

Punjab. Indian Museum (No. B. 118). Page 17.

,, 4. Merycopotamus nanus, Lyd. Greater part of the left ramus of the mandible of a male ;

from the Siwalik Hills. British Museum (No. 18407). Page 18.

,, 5. Merycopotamus nanus, Lyd. Greater part of the right ramus of the mandible of a

female: from the Siwalik Hills. British Museum (No. 15319). Page 18.

*

All the figures l nat. size.

imp .

Berj ea.u &. Highley delet lrtli.

West, Newman &, Co

.

MEMOIRS

OF THE

r “'LOGICAL SURVEY OF INDIA.

BEING

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

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

S e r . X.

INDIAN TERTIARY & POST-TERTIARY VERTEBRATA.

Vol. IV.

Part II. THE FAUNA OF THE KARNUL CAVES.

C And Addendum to part l.J.

By R. LYDEKKER, B.A., F.G.S., etc.

SOLD AT THE

GEOLOGICAL SURVEY OFFICE, AND BY ALL BOOKSELLERS.
LONDON : TRUBNER & CO.

MDCCCLXXXVI.

PRINTED BY GIBBS AND 1&AJIEORTH, ST. ALBANS, HERTS.

MEMOIRS

OF THE

GEOLOGICAL SURVEY OF INDIA.

IMa'oiMlojw Jtulicit,

BEING

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

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

Ser. X.

INDIAN TERTIARY & POST-TERTIARY VERTEBRATA.

Vol. IV.

Part II. THE FAUNA OF THE KARNUL CAVES.

By R. LYDEKKER, B.A., F.G.S., etc.

WITH 5 PLATES.

CALCUTTA :

SOLD AT THE

GEOLOGICAL SURVEY OFFICE, AND BY ALL BOOKSELLERS.
LONDON : TRUBNER & CO.

MDCCCLXXXVI.

FEINTED BY GIBBS AND BAMFOETH, ST. ALBANS, HEETS.

( Substitute for slip previously issued ).

CORRIGENDA.

Vol II. p. ix. note 1, after' Amer. Nat,’ add 1880. The formmla of the cutting-teeth of the American
rhinoceroses has been amended ; see “ Cat. Foss. Mamm. Brit. Mus.” pt. III. p. 141.
,, 144, note \,for Nat. read G£ol. (The volume is dated 1882, although the article quoted,

appeared in 1881).

,, 145, ,, 2, for Nat. read Geol., for VII. read VIII., and for 1878 read 1877. (In the reprint

quoted the articles from the two vols. are paged consecutively).

,, 148. The age of Anthracolherium dalmatinum is upper eocene.

,, 154, note 'l, for 143 read 103.

,, 157. If the Hempstead group be reckoned as equivalent to the Ronzon beds the age of both

Hyopotamus bovinus and II. velaunus will be lower miocene (mid. oligocene) only.
For synonomy see “ Cat. Foss. Mamm. Brit. Mus.” pt. II.

,, 164, line 11 from bottom, for 1839 read 1836.

,, 169. Sivameryx is the same as Chceromeryx ; vide ' Geol. Mag.,’ vol. II., p. 73 (1885).

,, 175, line 13 from top, for 'Kill. pi. V. read XII., art. 3.

,, 176, ,, 15 ,, ,, ,, XIV. read XIX.

,, 185, after Mellivorodon palceindicus add n. sp.

„ 208. The relative lengths of the molars of Ursus arctos are not constant.

,, 236, note 2, and 354, line 19 from top, add ser. 4 (1855).

,, 314. The premolar dentition of Felis is — — .

,, 348, line 1 1 from bottom, for Laizier and Pariere, read Laizer and Parieu.

After Hyopotamus palceindicus (p. 158), Hycenarclos paler indie us (232), and /Elurogale sivalensis (317), dele
n. sp., and after Hemimeryx (167) and Sivameryx (169) dele n. gen.

For amended nomenclature of non-Indian Mammalia mentioned in this volume see the writer’s “ Cata-
logue of Fossil Mammalia in the British Museum.”

Vol. III. p. 237. The vertebra of Python described on this page and figured in pi. XXXV., figs. 7, 7 a,
as a caudal is a dorsal.

SIWALIK MAMMALIA.

19

ADDENDUM.

Genus. TETRACEROS, Leach.1

Distribution. — The genus is represented at the present day only by T. quadri-
cornis ,* which occurs over the whole of India.

Tetraceros daviesi, n. sp. nobis.

Definition. — This species is slightly larger than Cephalophus madoqua, and
distinguished from T. quadricornis by its inferior size, and the narrower anterior
premolars.

Cranium. — In the course of a recent examination of the duplicate Siwalik
collection in the British Museum, Mr. W. Davies brought to light the middle portion
of the cranium of a small ruminant which he at once recognized as new, and kindly
handed over to the writer for description. The specimen (woodcut fig. 3) comprises

Fig. 3. Tetraceros daviesi, Lyd. The right half of the middle portion of the cranium and of the palate ; from
the Siwalik Hills. British Museum. (No. M. 3492).

the portion of the cranium between the middle of the orbit and the posterior ex-
tremity of the premaxilla, and indicates an animal slightly larger than the existing
African Cephalopus madoqua , Gray ; the whole of the palate and cheek-dentition is
preserved, but the crowns of some of the teeth (which are in an early stage of wear)
are either broken off, or more or less damaged. All the teeth have high crowns, and
the true molars are relatively wide, and have no internal accessory column. The
fourth premolar has a very strongly-marked anterior costa, and the superior border
of its inner crescent is incomplete anteriorly ; pm. 3 is much elongated, narrow
transversely, with a very incomplete inner crescent, and only a small inner root ;
the crown of pm. 2 is broken off, but the section of its base shows that this tooth
was very long and narrow, and had no inner root. The palate is narrow and

l ‘Trans. Linn. Soc.’ vol. xiv. p. 524 (1823). 2 The so-called T. quadricornutus is only a variety.

F

20 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

concave, and the median notch on the posterior border extends as far forward as the
anterior border of m.3. The fragment of the fronto-nasal region still remaining
shows that in the neighbourhood of the orbit this portion was flattened and triangular,
with a slight sulcus in advance of the small frontal foramen ; there is a shallow
lachrymal depression, of which the whole is visible from the frontal aspect of the
cranium, but no lachrymal vacuity ; there was no horn immediately above the orbit.
The length of the series of cheek-teeth is 2T, that of the true molars 2T, and that
of the premolars D04 inches.

Mandible. — The fragment of the right ramus of the mandible of a small ruminant
from the Siwaliks represented in woodcut fig. 4 agrees in relative size with the
cranium, and not improbably belongs to the same species. It contains the last two
lobes of mm. 4, and the complete m. 1 ; on the outer aspect of
the latter there is seen the broken base of a slender accessory
column, like that occurring in some examples of Tetraceros
quadricornis.

Affinities. — The structure of the true molars shows that
the only groups to which the cranium under consideration
could belong are the Tragulidce, the Cervidce, or the antelopes.
From the existing genera of the former1 it is distinguished
by the taller teeth, the presence of an inner crescent to pm- 3, the form of the palate,
and the presence of a lachrymal depression. That it does not belong to the smaller
members of the second family, like Cervulus and Moschus, is evident from the hypso-
dont dentition, the elongation of pm. 3, and the incompleteness of its inner crescent,
together with the small size of the inner root of the latter and its total absence in
pm. 2 5 while the absence of a lachrymal vacuity, the concavity and narrowness of
the palate, and the form of the fronto-nasal region are equally distinctive characters.

Among the smaller antelopes the specimen is at once distinguished from Gazella,
Nemorheedus, Neotragus , Nanotragus ,2 and their allies by the squareness of the crowns
of the true molars ; but when compared with the group comprising the African genus
Gephalopus and the Indian Tetraceros a close agreement is found in this respect. The
smaller species of the former, like C. madoqua , agree nearly in size with the fossil, but
are distinguished by the much deeper lachrymal depression, of which only a portion
is visible from the frontal aspect, by the wider frontal region, and the greater anterior
extent of the median palatine notch. The premolars are moreover less elongated,
and have less incomplete inner crescents, while the outer surfaces of the true molars
are much narrower. In the upper molars of the larger species of that genus, like
C. sylvicultriz , there is a well-developed inner accessory column, while some of the
smaller species show a projection from the hinder border of the anterior crescent in
little-worn true molars, which is wanting in the fossil.

1 In the genera Prodremotherium. and Baehitheriim, ■which have been provisionally referred to this family (“ Cat. Foss.
Mamm. Brit. Mus.” pt. II. pp. 150, 155), there is an imperfect inner crescent to pm. 3. The molars are of a very brachydont
structure.

2 Including Calotragus, Scopophorus, Nesotragus, and Oreotragus ; see Brooke ‘Proc. Zool. Soe.’ 1872. p. 642.

Fig. 4. (?) Tetraceros daviesi, Lyd.
Fragment of the right ramus
of the mandible ; from the
Siwalik Hills. British Museum
(No. 16535).

SIWALIK MAMMALIA.

21

Turning now to Tetraceros and taking for comparison a female skull1 of which
the first four upper cheek-teeth are figured in the accompanying woodcut, it will be

found that there is a very close resemblance in
almost every particular. The general contour of
the orbital and fronto-nasal region, of the palate,
and of the lachrymal depression, is almost identical
in the recent and fossil crania. The true molars of
the recent form agree with those of the fossil in the
breadth of their external lobes, and the absence of
an inner accessory column ;2 while the premolar
series °* the former bears near,y the same relative

of a female, from the grinding and proportion to the true molar series as obtains in

outer aspects. Recent, Madras. British

Museum (No. 884a). the latter, and also shows the same incompleteness

of the inner crescents of pm. 2 and pm. 3 ; the length and narrowness of the two latter
teeth is, however, rather less marked in the recent skull, and the anterior costa in
pm. 3 and pm. 4 is somewhat less prominent, while there is a minute inner root to
pm. 2, and the lateral palatal notches are deeper than in the fossil. The length of
the whole series of cheek-teeth in the recent cranium is 2*35 inches. The fragment
of the mandible mentioned above agrees very closely with the lower jaw of the
recent species, and, as already mentioned, exhibits the slender accessory external
column characteristic of certain specimens of the latter.

As the result of the foregoing comparisons there appears no doubt that the
cranium under consideration indicates an antelope so closely allied to the existing
T etraceros , that it may be pretty safely referred to the same genus ; but whether the
specimen indicates a female individual, or that the anterior horns (which are incon-
stant in the existing species) were not developed in the fossil form, cannot be
determined. The inferior size of the latter and the characters of the premolars
indicate its specific distinctness, and it may be appropriately named T. daviesi.

The present form is of considerable interest as exemplifying more fully than
hitherto the remarkable mingling in the Siwaliks of genera now respectively confined
to India and Africa. A fragment of the maxilla with the three true molars of a small
Siwalik antelope in the British Museum (No. M. 3493) differs from the corresponding
part of the cranium of Tetraceros daviesi by the narrower outer surfaces of the teeth,
and may perhaps belong to a small species of Gephalopus.

1 This specimen belongs to the variety specifically separated by Gray under the name of T. subquadricornutus ; see
Hand-list of Edentate, Thick-skinned, and Ruminant Mammals in British Museum,” p. 89. No. 884a (1873).

2 This character is not a constant one, since the male crania of the type form in the British Museum exhibit a slender
accessory column in the true molars ; similar variations occur in the corresponding lower teeth.

POSTSCRIPT.

It should have been observed that the reference of Antilope torticornis, Aymard,
to Palceoreas is on the authority of M. P. Thomas (‘ M&n. Soc. G£ol. France.’ ser 3.
vol. III. art. 2. p. 16), and that it is referred by M. Dep^ret in the passage quoted in
the text to Tragelaphus. According to Prof. Riitimeyer’s description (‘ Die Rinder
der Tertiar-Epoche, etc.’ p. 85) the molars of this form are very similar to those of
Strepsiceros (?) falconeri. Since the text was in type the writer has noticed that a
tooth from the ‘ Bohnerzgruben ’ of Wiirtemberg has been described by Prof.
Rutimeyer (see p. 88 of the work cited) as Hippotragus fraasi. If this determination
and the reference of A. torticornis to Tragelaphus rather than to Palceoreas be correct,
there is evidence of the occurrence of two existing African genera of antelopes in
the tertiaries of Europe.

INDIAN TERTIARY & POST-TERTIARY VERTEBRATA.

THE EAUUA OF THE KARNUL CAVES.

BY R. LYDEKKER, B.A., F.G.S., etc.
(WITH PLATES VII. TO XI.)

I. INTRODUCTORY OBSERVATIONS.

Locality. — The history of the exploration of the caves in the Karnul district
of Madras is given in three papers by Mr. R. B. Foote,1 and does not need
recapitulation on this occasion The most important of these caves are those of
Billa Surgam, a spot lying on the south of a valley opening on the east of the Yerra
Konda, a range of hills forming the western side of the Karnul basin, and situated
three miles to the east-south-east of Betumcherru, in the south-eastern corner of the
Nandial taluk ; the nearest place of any importance being Banaganpilli.

The caves — According to Mr. Foote’s description, Billa Surgam consists of three
deep and short ‘canons,’ joined by natural arches; the various caves opening into
the canons at different levels, and the canons themselves having once been caves also ;
in wet weather a stream flows through the canons. The accompanying plan2 will
obviate the need of further description ; it being only necessary to mention that the
four main caves are those named the ‘ Charnel-House,’ ‘ Purgatory,’ the ‘ Cathedral,’
and the ‘ Chapter-House.’

Sections of cave-deposits. — Two sections of the cave-deposits may be quoted
from Mr. Foote’s papers. The first is from the Charnel-House, and is as follows.

Al. Surface bed.

A. Rubble bed.

B )

Stiff red clay, with sandy partings.

1 ‘ Rec. Geol. Surv. Ind.’ vol. XVII. pp. 27-34, 200-208 (1884), and vol. XVIII. pp. 227-234 (1885).

2 The oblique -shading indicates the area of the existing caves.

G

24 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

BILLA SURGAM CAVES

FAUNA OF THE KARNUL CAVES.

25

E j Rubble bed.

H. Red cave-earth, stony above.

I. Red and mottled cave-earth.

J. Red-brown cave-earth with patches of calcareous sand.

K )

k' | Red sandy cave-earth with blocks of limestone.

M.\

q' > Stiff marly clay.

P. J

In this section bed A1 is of extremely recent origin, and may be neglected ;
human teeth were found in bed A, pottery occurred in B. C., and Mr. Foote1 records
implements from K and L.

The second section is in the Cathedral, and is as follows.

C. Surface bed.

C. Grey sandy bed.

Stalagmite in irregular masses.

Ca. Red sandy cave-earth.

Cb. j

Cc. J Stiff red clay.

Cd. )

| Stiff dark marl.

Ch. Dark loamy marl.

£j- | Grey marl.

qJ1' | Grey marl.

In this section the beds Cc. and Cd. yielded the most important specimens.

Condition of the bones. — The majority of the bones and many of the teeth are
stained of a full brown colour, and strongly impregnated with mineral matter.
Other specimens, however, which were obtained from beds containing extinct species,
are scarcely altered, and are almost indistinguishable from the bones of recent
animals ; and since these specimens belong to fossorial rodents and carnivores it is
pretty evident that they are of later age than the highly mineralized specimens with
which they are associated.

The teeth of Rhinoceros from bed Cc. in the Cathedral are but little altered ; and
many of the incisors of Ilydrix and other rodents from bed Cd. still retain their
original orange colour. Of the larger mammals no complete skulls were found, the
majority of the remains consisting either of detached teeth, fragments of the jaws,
or more or less imperfect limb-bones. Of the smaller mammals skulls were found in
some instances ; but in many cases the only determinable remains are fragments of
the jaws and limb-bones. Vast quantities of bones of Chiroptera and small Rodentia —
probably introduced, as Mr. Foote suggests, by owls — were obtained in many of the
beds, but these are totally valueless as not being even generically determinable.
A considerable number of the larger bones have been gnawed by porcupines.

l ‘ Rec. Geol. Surv. Ind.’ vol. XVII. p. 206.

26

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

Age of the fauna— The comparatively large number of species either totally
extinct, or which are not now found living in India, renders it probable that the age
of a considerable part of the Karnul cave deposits is not newer than the pleistocene ,
and the fauna, as being almost certainly more recent than that of the Naibada beds,
may be provisionally assigned to the later part of that period.

II. MAMMALIA.

List of species. — Three preliminary and provisional lists of the Karnul mammals
have been already published — two by Mr. Foote,1 * and one by the present writer.
The following list gives the result of the final determinations, but there is no evidence
that Mus platythrix, and Golunda elhoti are contemporaries of the extinct forms.

Primates. — Semnopithecus entellus ( Dufresne ). Rodentia. Golunda ellioti, Gray.

Hystrix crassidens, nobis.

Atherura karnuliensis, nobis.

Lepus (cf. nigricollis, F. Cuv .)

Ungulata. — Equus asinus, Linn.

sp. a.

Rhinoceros karnuliensis, nobis.

Bos or Bubalus, sp.

Boselaphus tragocamelus {Pall.)
Genus non. del.

Gazella bennetti (.Syto).

Antilope cervicapra {Linn).
Tetraceros quadricornis {Blainv.)
Cervus aristotelis, Cuv.
axis, Erxl.

(?) Cervulus muntjac (Zimm.)

Tragulus (cf. meminna \_Erxl.f)

Sus cristatus, Wagner.
karnuliensis, nobis.

Edentata. — Manis gigantea, Llliger.

Relations. — The most remarkable feature in this list is the occurrence among a
number of existing Indian species of a Cynocephalus which may be identical with
a living African species, of Jlycena crocuta , of a small Equus indistinguishable from
E. asinus , and of a Manis apparently identical with the existing west African M.
gigantea ;3 while scarcely less noteworthy is the occurrence of a peculiar species of
Rhinoceros , and of a Hystrix and a Viverra specifically distinct from the species now
living in India, as well as of the non-Indian genus Atherura. The occurrence
of the genus Cynocephalus and of forms identical with African species of Jlycena

1 ‘Rec. Geo. Surv. Ind.’ vol. XVII. p. 202. XVIII. p. 231. 2 Ibid. vol. XVIII. pp. 120-121.

3 The importance of the occurrence of these forms would not he diminished even if it should ever he discovered that

some of them present slight differences from their existing African representatives which might entitle them to specific
distinction.

c^ynucepiiaius, sy.

Carnivora. — Felis tigris (or ? leo) Linn.

(?) pardus, Linn.
chaus, Guldens/.
rubiginosa, Geoffr.
Hyaena crocuta (Erxl.)

Viverra karnuliensis, nobis.
Prionodon (?), sp.

Herpestes griseus, Desm.

fuscus, Waterh.
Ursus labiatus, Blainv.
Insectivora. — Sorex, sp.

Chiroptera. — Taphozous saccolsemus, Temm.

Phyllorhina diadema (Geoffr.)
Rodentia. — Sciurus macrurus. Hardw.

Gerbillus indicus (Hardw.)
Nesokia bandicoota (Bech.)

kok, Gray.

Mus mettada (Gray).
platythrix, Sykes,
sp. var.

FAUNA OF THE KARNUL CAVES.

27

Equus and Manis is extremely important in supplementing the evidence afforded by
the Siwalik fauna as to the probable derivation of many of the existing Ethiopian
mammals from those of the later tertiaries of India ; and it is interesting to trace
the gradual dying out in the latter country of genera and species which are now
dominant forms in Africa. In many instances such forms have totally disappeared,
while in others the modern Indian representatives are either few in respect of
individuals or inferior in size to their tertiary congeners. Thus Cynocephalus was
represented by two species in the Siwaliks,1 persisted to the pleistocene, and then
finally disappeared. Similarly the crocutine group of Hycena is represented in the
Siwaliks by II. felina and II. colvini, and in the pleistocene by II. crocuta (which was
probably the descendant of H. colvini), after which period the group entirely died out
in India. Again in the case of the edentates we find the gigantic Siwalik Macro-
therium sindiense ,2 which presents features connecting it with Manis, succeeded by the
smaller Manis gigantea of Karnul, which has now migrated to Africa, while all the
species of the latter genus now inhabiting peninsular India are of greatly inferior size.
Atlierura offers an instance of a genus now totally unrepresented in India proper but
occurring in Africa, and also in the east of the Oriental region. The lion, which is
known to occur in the pleistocene of Europe and probably existed at the same date
in India, may be cited as an analogous instance, since it is a comparatively rare
animal in the latter country, although very abundant in Africa. Equus again, though
still poorly represented in the north-west of India, has entirely died out as a wild
genus in the greater part of the country, but has attained a great development in
Africa. Still more striking examples are to be found in the total disappearance from
India of Hippopotamus , Giraffa , and Alcelaplius and other antelopes of modern African
genera.3 There is at present no satisfactory explanation of this total extinction in
India of genera and species which appear equally as well suited to exist there at the
present day as those which remain.4 On the other hand Felis chaus may be cited as
an example of a species which probably originated in India, and is still common
there, although having extended its range to northern Africa.

The Viverra and Hystrix of Karnul are probably the progenitors of the existing
Indian species and the descendants of the Siwalik representatives of those genera ;
while it is also probable that the existing Sits cristatus was derived from the
Siwalik S. falconeri , from which may also have originated the extinct S. Icarnuliensis.
The extension of this group of pigs into the pliocene of Africa is indicated by S.
phacochceroides. The Karnul Bhinnceros and the other pleistocene Madras species
(R. deccnnensis) belong to the less specialized section of the atelodine group,5 of
which there is no representative either in pliocene6 or recent India. The former
range of the existing B. unicornis into Madras is noticed in the sequel.

l Vide supra, pp. 6-7. 2 Syn. Manis sindiensis, vide infra. 3 Vide supra , pp. 7-16.

4 The glasial epoch can scarcely be brought forward in the case of southern India..

5 Comprising It. bicornis and all the European pleistocene species except R antiquitatis. See “ Cat. Foss. Mamm. Brit.
Mus.” pt. III. pp. 101-123. (1885).

6 R. platyrhinus belongs to the more specialized section.

G

28 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

Specimens figured. — "With the exception of some of the smaller rodents, of which
the pleistocene age is doubtful, remains of nearly all the Karnul mammals are figured
in the accompanying plates and woodcuts; the figured specimens generally com-
prising more or less incomplete crania, mandibles, or teeth. Of limb-bones, specimens
of the humerus, and more rarely the femur, of a considerable number of genera
have been figured, as it has been thought such figures may be useful to future
explorers of Indian caverns and aid them to identify some of the commoner forms
with which they may expect to meet.

Semnopithecus entellus (Dufresne) : var.

Maxilla and mandible. — In the writer’s preliminary notice the remains of this
form were referred, after Mr. Foote’s provisional determination, to S. priamus,
but subsequent examination has shown that they are too large for that species.
The fragments of an associated right maxilla and mandibular ramus represented
in plate VII., figs. 2, 3 were obtained from bed M. of the Charnel-House,1 and
exhibit the teeth in an early condition of wear ; the inferior portion of the mandible
has been crushed and bent on one side. The teeth of these specimens are rather
larger than those of any specimen of S. entellus in the British Museum, and agree
in this respect with the Himalayan S. schistaceus f it is, however, practically certain
that they do not belong to the latter, and they may accordingly be provisionally
regarded as belonging to a large variety of the former. In the upper jaw the length
of the space occupied by the three true molars is IT, and in the lower P22 inches.
The specimen represented in fig. 1 of the same plate comprises the greater part of
the right half of the palate, and shows all the teeth except iT and m.3 ; it was
obtained from Red A. in the Charnel-House,3 and from the small size of the canine
evidently belonged to a female ; the cheek-teeth agree in size with those of the
■preceding specimens. Two canines (No. F. 200) belonging to male individuals have
been obtained from beds Cc. and Ce. in the Cathedral, and numerous detached cheek-
teeth from other beds.

Calcaneum. — The right calcaneum represented in plate VII. fig. 4 was obtained
from bed C. in the Cathedral, and agrees in relative size with the teeth. An
associated right calcaneum and astragalus (No. F. 201, c) from the Purgatory cave
may belong to an immature individual of the present species, or may indicate the
occurrence of a second form.

Horizon and range. — The present form appears to have existed during the whole
period of the Karnul cave deposits. There appears some uncertainty as to the
southern limit of the range of the existing race, but it is probable that it embraces
at least a portion of the Deccan.4

Cynocephalus, sp.

Loiver molar. — The unworn second left lower true molar of a large monkey

1 See Eoote 1 Rec. Geol. Surv. Ind.’ vol. XVII. p. 207.

2 Compare the mandible represented in pi. I. fig. 6. of tlie preceding part of this volume.

3 See Foote, op. cit. p. 206. 4 See Jerdon, ‘ Mammals of India,” pp. 5, 6.

FAUNA OF THE KARNUL CAVES.

29

represented in pi. VII. figs. 5, 5a was obtained from bed M. in the Charnel-House.1
In its very large size, tall crown, and form of the cusps, this specimen differs widely
from the teeth of Semnopithecus and Mcccacus , and agrees so exactly with those of
Ci/nocephalus that there can be no doubt as to its belonging to that genus.

Affinities. — The specimen accords very closely with the corresponding tooth of
the existing C. annbis, but as it might apparently have equally well belonged to the
Siwalik C. falconeri ,2 it appears impossible to make any specific determination. The
intimate relationship existing between some of the Karnul mammals and those of
Africa suggests, however, that the present form may be more nearly allied to the
existing species of that country than to those of the pliocene of India.

Felis Tigris (or ? leo), Linn.

Limb-hones. — Since it appears impossible to distinguish the limb-bones of the
tiger from those of the lion it is by no means certain, in view of the occurrence of
Ili/cena crocuta in the Karnul caves, that at least some of the under-mentioned
specimens may not belong to the latter rather than to the former species.3 The first
phalangeal of the third digit of the right pes represented in pi. VII. fig. 20 was
found in association with a first and second lateral phalangeal, the fragment of a
metatarsal, and the greater portion of the tibia of the same side, in the Purgatory
cave.4 These specimens indicate an individual equal in size to a large tiger. Another
first phalangeal (No. F. 224, b) was obtained from bed Cf. of the Cathedral.

Carnassial.5 — The anterior portion of the outer half of a left upper carnassial
(No. F. 224) apparently agreeing with that of the tiger was obtained from bed
Ce. in the Cathedral.

(?) Felis paedus, Linn.

Limb-bones. — The only evidence for the existence of this species in the Karnul
cave-deposits is afforded by the first phalangeal represented in pi. VII. fig. 19, which
was obtained from bed Cc. in the Cathedral, and by a metacarpal (No. F. 225, a)
from bed 0. in the same cave. These specimens are, however, of comparatively
small size and can hardly be regarded as conclusive evidence.

Felis chaus, Giildenstadt.

Mandible. — This cat is represented by the imperfect left mandibular ramus
figured in pi. VII. fig. 17, which was obtained from bed Ca. in the Cathedral. The
carnassial has been broken, but the two premolars are perfect, and the posterior half
of the alveolus of the canine still remains.

1 This is apparently the specimen noticed by Foote in the ‘ Rec. Geol. Snrv. Ind.’ vol. XVII. p. 207, and regarded as
belonging to the same species as the Semnopithecm.

2 Supra, pi. I. fig. 4. 3 Jerdon, “ Mammals of India,” p. 92. mentions the occurrence of the lion south of the Narbada.

4 See Foote, ‘ Eec. Geol. Surv. Ind.’ vol. XVII. p. 207.

5 Mr. Foote op. cit. mentions an upper carnassial rather smaller than that of a tiger, but the writer was unable.to identify

the specimen in the collection.

30

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

Canine. — The left upper canine1 represented in fig. 18 of the same plate exhibits
on its outer surface2 the grooves characteristic of the genus, and from its size may
be referred to the present species.

Distribution. — This species occurs at the present day throughout India, and to
the eastward ranges into Assam and Burma, and to the westward into Persia, the
borders of the Caspian, and north Africa. Its occurrence in the pleistocene of India
points to the probability of that country having been its original home.

Felis rubiginosa, Geoffr.

Maxilla and mandible. — This very small cat, which is peculiar to Madras and
Ceylon, is represented by four specimens in the Karnul collection. The imperfect
left maxilla figured in pi. VII. figs. 14, 14a shows pm. 3 and pm. 4, and the alveoli of
the canine and m4, and also exhibits the characteristic absence of pm. 2. This
specimen was obtained from bed Ca. of the Cathedral. The slightly imperfect right
mandibular ramus represented in fig. 15 of the same plate was obtained from bed
Cb. of the latter cave ; and the collection also contains a less perfect specimen of
the same side from bed Ca.

Humerus. — The right humerus represented in fig. 16 of the above-mentioned
plate was found in bed Cc. of the Cathedral. The specimen has lost the head and
part of the lateral expansion of the entepicondyle (en. c ) ; it exhibits the well-marked
entepicondylar foramen (cf) and the absence of a supracondylar perforation, which
are characteristic features of the humerus of Felis as distinguished from that of
Viverra i3 and Herpestes, and since it agrees in relative size with the jaws it may be
pretty safely referred to the present species.

Hyaena crocuta (Erxleben).

Carnassial. — The partially- worn left lower carnassial of a hyaena represented in
pi. VII. figs. 13, 13a was obtained from bed Ce. in the Cathedral, and agrees precisely
with the corresponding tooth of the present species. It is somewhat smaller than is
usually the case with the pleistocene European race of the species4 (although some
examples of the latter are scarcely larger5), and agrees more nearly in this respect
with the existing African race. The characteristic features of the lower carnassial
of this species are the total absence of an inner cusp, the very small size of the
hind talon, and the frequent presence of a small cingulum on the external aspect of
the anterior half of the blade ; this cingulum being well-developed in the present
specimen. The lower camassials of the Siwalik H. felina 6 and II. colvini 7 are mainly
distinguished from the corresponding tooth of II. crocuta by the decidedly larger

1 The -writer could not determine from which cave this specimen was obtained ; it is marked Pop.

2 The inner surface is shown in the figure.

3 The perforation is absent in Oenetta and Paradoxurus, but the humerus of the latter is distinguished by the great width
of the distal expansion.

4 Compare Gaudry, “Histoire des Temps Quatemaires,” pi. TV. fig. 9. (1876).

5 Ibid. pi. I. fig. 1. 6 Supra, vol. II. pi. XXXVIII. fig. 1.

7 Ibid. fig. 3.

FAUNA OF THE KARNUL OAVES.

31

hind talon ; the cingulum is, however, absent in the first, although well-developed in
the second Siwalik species, being in some instances1 considerably larger than in the
existing one.

Canine. — An imperfect upper canine (No. F. 223) from bed Ce in the Cathedral
may not improbably be referred to the present species.

Origin of the species. — The occurrence of II. crocuta in the Karnul caves is of
extreme importance in regard to the history of the species, since it connects its
distributional area with that of the primitive crocutine Siwalik hyaenas, and leaves
but little doubt that the species originated in India from H. colvini and thence spread
westward into Europe at the close of the pliocene or commencement of the pleis-
tocene epoch2 (when it attained its largest dimensions), and subsequently reached
Africa. The strongly-marked cingulum in the Karnul carnassial is especially note-
worthy as indicating the close connection of this race with H. colvini rather than
with H. felina.

VlVERRA KARNULIENSIS, nobis.3

Definition. — The species may be defined as equal in size to V. zibetha, but
distinguished by the much longer space occupied by the premolar series, in which
respect it appears allied to the Siwalik V. baJceri.

Mandible. — The type specimen of this species was obtained from bed L
in tlie Charnel-House,4 and is represented in pi. VII. figs. 6, 6a. The specimen

consists of a fragment of the alveolar portion of the left ramus of the mandible

showing the alveoli of all the premolars except pm. 1, and the complete mTI in a

partially-worn condition. The latter tooth agrees very closely, both in size and

structure, with the carnassial of Viverra zibetha ; almost the only observable difference
being that the crown is relatively rather wider, more especially in its talon half.
The premolar alveoli, as will be seen by the measurements, occupy, however, a
much longer space, and the ramus when complete was evidently much deeper. In
the following table the dimensions of the specimen are compared with those of the

mandible of V. zibetha.

V. karnuliensis. V. sibetha.

Length of mTl 0'59 0-6

Wid*h of do. at mandible 0'3 0*3

Length of space occupied by alveoli of last three premolars . . . . 1’28 1-08

Affinities. — Viverra megaspila agrees so closely with V. zibetha that what applies
in one case applies in the other, and the lower carnassial of F. civetta is quite
different from that of the fossil. None of the larger existing species of Herpestes
agree with the latter, so that its distinctness from all living species may be taken for
granted. Of the Siwalik species V. balceri 5 (of which the mandible is unfortunately
unknown) agrees with the present form in relative size, as well as in the important
character of pm. 3 being longer than in the existing species.6 This probably indicates

l Supra, vol. II. p. 296. fig. 14. 2 It occurs in the Norfolk Forest bed. 3 ‘Rec. Geol. Surv. Ind.’ vol. XIX. p. 120 (1886).

4 See Foote * Rec. Geol. Surv. Ind.’ vol. XVII. p. 206. 6 Supra, vol. II. pi. XXXIII. fig. 1. 6 Ibid. p. 270.

H

32

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

that the Karnul civet is allied to this species, although the distinctness of the other
Karnul mammals from those of the Siwaliks renders it unlikely that it is specifically
the same. Compared with Iditherium the fossil agrees nearly in size with I. rob us turn,1
and approaches it in the depth of jaw and elongation of the premolars, but differs
by the blade of the carnassial being longer in proportion to the talon. Since the
two Siwalik species decidedly belong to Viverra 3 and not to Iditherium, it appears
most probable that the Karnul form should likewise be referred to the existing genus.
The Karnul mandible indicates, however, that the three extinct Indian species (like
V. pepraxti, Dep^ret,3 of the lower pliocene of France) were probably intermediate
between V. zibetlia and I. robustum ; and it is also probable that the origin of the
existing specialized Indian species may be traced directly through V. Jcarnuliensis to
the Siwalik V. bakeril The apparent occurrence of Viverra 8 in the upper eocene
(oligocene) of England indicates that the genus is an old one, and Iditherium must
probably, therefore, be regarded as an offshoot from one of the later species which
has assumed characters almost precisely intermediate between the less specialized
species of Viverra and Ilgcena.

Prionodon (?), sp.

Humerus. — The imperfect left humerus of a small carnivore represented in pi.
VII. fig. 12 was obtained from bed Cb in the Cathedral, and may probably be
referred to the Viverridce. It indicates a species agreeing approximately in size with
Herpestes griseus (fig. 9), but differs from the humerus of that genus by the absence
of the supracondylar perforation and the smaller lateral expansion of the entepi-
condyle ( en.c .).6 It agrees in these points with Prionodon, and its resemblance to the
corresponding bone of the Nipalese P. pardicolor is so close as to indicate the pro-
bability of its belonging to the same genus, although it is of considerably larger
size than the corresponding bone of that species, and would probably, therefore,
agree more nearly with P. maculosus of Darjiling, Tenasserim, etc. The African
genus Poiana is, however, very closely allied to Prionodon ,7 and since the writer is
not aware that the limb-bones of the two can be distinguished, the reference of the
present specimen to Prionodon must be regarded as purely provisional ; to whatever
genus it really belongs the specimen is important as indicating a form apparently
different from any now found in Madras.

Herpestes griseus (Desmarest).

Skull. — The slightly imperfect cranium and left mandibular ramus represented
in pi. VII. figs. 7, 8 were obtained, in association with a considerable portion of the

1 See Gaudry, “ Animaux Fossiles et Geologie de l’Attique,” pi. VII.

2 This is shown by the shorter pm. 9 and larger m_l and m.‘2.

3 Theses. Faoult. Sci. Paris, ser. A. No. 67. (Bassin Tertiaire du Rousillon) p. 137. pi. IV. figs. 1-6 (1885).

4 Supra, vol. II. p. 271. 6 Y. hasting site, Davies. See Lyd. “Cat. Foss. Mamm. Brit. Mus.” pt. I. p. 101. figs. 11-12.

6 The humerus of Paradoxurus is distinguished by its wide distal expansion. In Mustela the humerus has no supracon-
dylar perforation, but in M . Jlavigula that bone is much larger and its distal extremity relatively wider than the present
specimen, while in the smaller species of that genus the whole bone is considerably smaller.

7 See Mivart, ‘Proc. Zool. Soc.’ 1882. p. 159.

FAUNA OF THE KARNUL CAVES.

33

appendicular skeleton, from bed Ch in the Cathedral; they are remarkable for their
extremely fresh and unaltered appearance, from which circumstance, together with
the fossorial habits of the species, it is probablo that they are of later age than the
bed in which they occur. A left mandibular ramus without teeth (No. F. 230, b) in
a more mineralized condition from bed Ca in the same cave, and another belonging
to the right side from the Purgatory cave (No. F. 230, c) may be apparently referred
to the same species, and indicate its existence among the proper cave fauna.

Humerus. — A left humerus from bed Ca in the Cathedral is represented in fig.
9 of the same plate. This specimen is slightly smaller than the corresponding bone
associated with the skull, but the difference is probably merely an individual one.
Other specimens exhibit slight variations in both directions from the figured speci-
mens, and it is not improbable that some of the larger ones may belong to U. smithi.

Hespestes fuscus, Waterhouse.

Mandible. — Of this small species, which occurs in Madras and Ceylon,1 two man-
dibular rami have been obtained from bed Cd in the Cathedral, of which the most
perfect is represented in pi. VII. figs. 10, 10a. The figured specimen, which has lost
all the teeth with the exception of pm. 4 and mT, agrees precisely with the mandible
of the existing form.

Humerus. — The left humerus represented in fig. 11 of the same plate is one of
several specimens agreeing in relative size with the mandible, which are accordingly
referred to the same species. These specimens were obtained from beds Ca and Cb
of the Cathedral cave.

Uksus labiatus,2 Blainville.

Syn. Melursus ursinus (Shaw).

Humerus. — The imperfect distal extremity of a right humerus of this species
represented from the palmar aspect in pi. VII. fig. 21 was obtained from the Chapter-
House cave in a bed numbered Ab.3 This specimen, which has lost the greater
portion of the entepicondyle (m.c.), agrees so exactly with the humerus of the
existing form that there can be no reasonable doubt as to its specific identity.4

Distribution. — The species is spread over all southern and central India, and is
probably descended from the Siwalik U. theobaldi .5

Sorex, sp.

Crania. — The superficial layers of the caves have yielded several crania of shrews,
measuring IT inches in length, which probably belong6 to S. serpentarius, Geoffroy,
and it was at first thought that three imperfect crania from bed Ca in the Cathedral

1 The Ceylon form has been separated by some writers as H. niacearthice.

2 The reasons for adopting this name will be found in vol. II. p. 207.

3 Mr. Foote has not published a section of the beds in this cave.

4 It was suggested in the preliminary list that the specimen might perhaps belong to TJ. malayanus or U. namadicus.

5 Supra, vol. II. p. 211. 6 In the preliminary list these specimens were provisionally referred to S. caruletcens.

34 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

and bed X in the Charnel-House were specifically identical. A subsequent examina-
tion of these specimens has, however, shown that they are of considerably smaller
size ; but the extreme difficulty of distinguishing the crania of the members of this
genus renders it unsafe to attempt their specific determination. The well-marked
hook to the first incisor indicates, however, that the specimens belong to Sorez
proper.

Taphozous saccol^emus, Temminck.

Skulls. — Three imperfect crania and several fragments of the mandible of this
species were obtained from beds C and Ci in the Cathedral, and also from the
Charnel-House. The most perfect example of the cranium, which was obtained
from bed Ci in the former cave, is represented in pi. VIII. figs. 10, 10a; the
characteristic elongated postorbital process of the frontal is well shown in fig. 10,
while the enlarged view of the cheek-teeth in fig. 10a exhibits the three true molars,
the large pm. 4, and the minute preceding premolar.

Distribution. — The species is found over a considerable part of India, and also
in Ceylon, Burma, the Malay Peninsula, Sumatra, and Java.

Phyllorhina diadema (GeofEroy).

Skull. — Of this widely distributed species imperfect crania and mandibular
rami are comparatively common in both the Cathedral and the Charnel-House caves.
The middle portion of a cranium from the latter cave is represented in pi. VIII. fig.
1 1 , and the cheek-dentition of another specimen from bed C/i in the Cathedral in
fig. 11a. The upper cheek-dentition of this species is numerically the same as in
Tajpliozous, but the penultimate premolar is placed more externally, and the inner
borders of the true molars are more rounded ; there is no supraorbital process to the
frontal. A left mandibular ramus from bed Ci in the Cathedral, in which the
anterior teeth are wanting, is represented in fig. 12 of the same plate ; and there
are numerous similar specimens from the Charnel-House.

Humerus. — The right humerus from bed Ci in the Cathedral represented in pi.
VIII. figs. 9, 9a may probably be referred to the present species. There are other
similar specimens in the collection.

Sciurus macrurus, Hardwicke.

Mandible. — The fragment of the left mandibular ramus of a squirrel represented
in pi. VIII. fig. 5, which was obtained from bed Ga in the Cathedral and contains
the partially worn mTT, agrees so closely with the mandible of S. macrurus, now
inhabiting Madras, that it may be pretty safely referred to that species.

Gerbillus indicus (Hardwicke).

Mandible. — This species is represented by several mandibular rami from the
Cathedral and Charnel-House caves. The mineral condition of these specimens
indicates that they are of the same age as the other fossils with which they are

FAUNA' OF THE KARNUL CAVES.

35

associated. The specimen represented in woodcut fig. 1 A is from bed Ca in the
Cathedral, and belongs to a fully adult male, the molars being well worn. The

A. B.

Fig. 1. Gerbillus indicus. The left ramus of the mandible (\) and grinding surface of the lower molars (f).

A. belongs to an adult male and is from bed Ca in the Cathedral ; while B. belongs to an immature female,
and is from bed X in the Charnel-House. (Indian Museum, No. F. 338).

smaller specimen represented in fig. 1 B has the molars much less worn, and may
probably be referred to an immature female ; it was obtained from bed X in the
Charnel-House. The length of the space occupied by the three molars in this
specimen is 0*24 inch, against 0*28 in the larger example j1 the size of the teeth in
the former being not greater than in G. hurriance of north India and Persia. Since,
however, the structure of the molars and incisors is similar to that obtaining in G.
indicus and quite distinct from that of G. hurriance, 2 there seems no doubt that the
specimen must be regarded as belonging to a small example of the former species.

Nesokia bandicoota (Bechstein).

Syn. Nesokia, gigantea , Auct.

Maxilla and mandible. — Three fragments of maxillae and three left mandibular
rami, the majority3 of which were obtained from bed H in the Charnel-House,
belong to this species. Although the burrowing habits of these animals might
account for the occurrence of their bones among the pleistocene fauna, yet the con-
dition of the specimens of this and the next species indicates that they were probably
contemporaries. One of the mandibular rami, in which the incisor has been pushed
back in its socket, is represented in pi. VIII. figs. 15, 15a. The great width of the
incisors characteristic of the genus4 is well shown in this specimen.

Limb-bones. — To this species5 may probably be referred the imperfect humerus
and femur represented in pi. VIII. figs. 1 and 2 ; while the larger femur represented
in fig. 3 of the same plate may perhaps belong to an unusually large individual of
the same species. These bones are relatively stouter than the corresponding elements
of the skeleton of Mus decumanus.

Nesokia kok, Gray.

Crania. — The imperfect cranium of this species represented in pi. VIII. figs. 13,

1 In the enlarged figures the artist has made the teeth of the two specimens too nearly equal in size.

2 See Blanford, “ Zoology and Geology of Eastern Persia,” p. 69. (1876).

3 Owing to the breaking of these specimens in transit their numbers were partially destroyed.

4 The genus occurs in the Siwaliks ; see ‘ Cat. Foss. Mamm. Brit. Mus.’ pt. I. p. 226. (1886).

5 There is no example of the skeleton of a Nesokia in the British Museum.

I

36

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

13a was obtained from bed Cb in tbe Cathedral ; and there is another specimen from
bed B in the Charnel-House.

Mandible. — Mandibular rami are comparatively common, and have been obtained
from beds well down in the series in both the Cathedral and Charnel-House caves ;
the specimen1 figured in pi. VIII. figs. 14, 14a is from bed Ch in the former.

Limb-bones. — vThe imperfect femur represented in fig. 4 of the above plate, as
being smaller than the corresponding bone provisionally referred to N. bandicoota,
probably belongs to the present species. A humerus (No. F. 347) from bed Ca in
the Cathedral agrees in relative size with the figured femur.

Mus mettada2 (Gray).

Syn. Golunda mettada , Gray.

Skull. — An imperfect skull and numerous mandibular rami (No. F. 209) of this
species have been obtained from the Charnel-House, but the extremely fresh ap-
pearance of these specimens, coupled with the burrowing habits of the species,
renders it doubtful whether they are really contemporaneous with the pleistocene
fauna. The mandible of this species is readily distinguished from that of Golunda
ellioti by its inferior vertical depth, and a specimen of a left mandibular ramus (No.
F. 209a) from bed Cd in the Cathedral, which is well fossilized, probably indicates
the occurrence of the species among the pleistocene fauna.

Mus platythrix, Sykes.

Syn. Leggada platythrix , Gray.

Skull. — Of this small burrowing mouse an imperfect cranium and several man-
dibular rami (No. F. 210) have been obtained from the Charnel-House, but their
fresh condition points to their recent introduction.

Mus (?) sp. var.

Mandible.— Three species of Muridce apparently distinct from any of those
specifically named in this memoir are indicated by mandibular rami, which from
their well-mineralized condition are apparently contemporaneous with the pleistocene
fauna. The first specimen (No. F. 339) belongs to the right side, and the length of
the three molars is 029 inch ; it was obtained from bed Cd in the Cathedral. The
second (No. F. 340) is from bed X in the Charnel-House, and is rather larger than
Mus mettada; a fragment of the right maxilla probably belongs to the same in-
dividual. The third form is represented by a left mandibular ramus (No. F. 341)
from the same bed, which is slightly larger than Mus platythrix.

Golunda ellioti, Gray.

Skull. — Several specimens (No. F. 211) of the cranium and mandible of this
species have been obtained from the Charnel-House, but their unaltered condition
indicates their recent origin.

1 This specimen belongs to the right side, but has been reversed in order to facilitate comparison with fig. 15.

2 Amended from meltada ; see W. T. Blanford, ‘ Joum. As. Soc. Beng.’ vol. XLV. pt. 2. p. 170. note (1876), where it
is shown that the species is not a Golunda.

FAUNA OF THE KARNUL CAVES.

37

Hystrix crassidens, n. sp., nobis.

Definition. — This form1 is considerably larger than II. hirsutirostris , from which
it is also distinguished by the circumstance that the upper incisors are much wider
than the lower.

Tipper incisors. — The extremities of a pair of upper incisors are figured in pi.
VIII. fig. 19 ; another left tooth being represented from the lateral aspect in fig. 20.
The measurements of the latter and of an upper incisor of H. hirsutirostris are as '
follows.

H. crassidens. E. hirsutirostris.

Antero-posterior diameter 0-44 0-30

Transverse do 0-32 0-22

Mandible. — The left ramus of a complete immature mandible is figured in pi.
VIII. figs. 17, 17 a ; mm. 4 is still in situ, m.3 has not come into wear, and the incisor
is pushed partly out of its alveolus. The contour of this specimen agrees very closely
with the mandible of H. hirsutirostris, but the space occupied by the molar series is
longer ; the latter feature being still better exemplified in a rather older example
represented in fig. 16, where m.3 is more protruded, and pm. 4 is seen in alveolo. In
the latter specimen the length of the space occupied by the three true molars is 1 *22
inches, the corresponding length in a somewhat older example of H. hirsutirostris
being 1*05 inches. The incisor of the figured mandible agrees in width with that of
the living species, and is much narrower than the upper incisor of the fossil, and
since the same feature is exhibited by all the lower fossil incisors (the extremities of
a pair being represented in fig. 21), it may be taken as a character of the species.
The concave surface of the lower incisors of the fossil is less narrowed than in
the living species. The dimensions of the recent and fossil lower incisors are as.
hollows.

H. crassidens. E. hirsutirostris.

Antero-posterior diameter 0-32 0-28

Transverse do. ............ 0-25 0'25

Cheek-teeth. — Beyond their superior size it does not appear that the cheek-teeth
of the fossil can be distinguished from those of the recent species.2 A detached
right upper premolar is represented in fig. 18 to show the large dimensions which
these teeth attain, the antero-posterior diameter of the base of the crown being 05
inch.

Humerus. — A specimen of the right humerus, wanting the proximal epiphysis, is
represented in pi. VIII. fig. 6 ; it is of considerably larger size than the corresponding
bone of H. hirsutirostris, from which it is also distinguished by the shorter deltoid
ridge {dr).

Distinctness imd affinities. — The difference in the size of the upper and lower
incisors appears tq be a character by which the Karnul porcupine is distinguished

l Referred in ‘ Rec. A01. Surv. Ind.’ vol. XIX. p. 120. to H. hirsutirostris. Tlie writer follows Prof. Flower in
adopting the latter name in piq.ce of II. leucura.

e molars in the specimen represented in fig. 17, 17a is due to their very early stage of wear.

38 INDIAN TERTIARY AND POST-TERTIARY YERTEBRATA.

not only from H. hirsutirostris , but from all other existing species. In respect of
size the Siwalik H. sioalensis1 agrees with the fossil, but is distinguished by its low-
crowned and rooted molars ; and since the writer has been unable to identify the
Karnul form with any other fossil porcupine it is regarded as a new species, which
may be known as H. crassidens. It is probable that this species is the connecting
link between the Siwalik and the existing Indian porcupines.

Horizon. — The majority of the remains of the present species were obtained
from the Cathedral cave, and mainly from the bed Cd, although some were met
with in beds C, Gc, Ce, and Cf.

Atherura karnuliensis, n. sp., nobis.

Distribution of the genus. — The genus Atherura is now confined to west Africa
and the regions on the eastern side of the bay of Bengal. The Oriental A . fasciculata
is rather larger than the Ethiopian A. africana , but is closely allied.

Definition. — The present species is rather larger than A. fasciculata, and distin-
guished by the greater bevelling of the lateral borders of the anterior surface of the
incisors.

Incisors. — The species is represented only by a few upper and lower incisors, two
of which, obtained from bed Cd in the Cathedral cave, are figured in pi. VIII. figs. 22,
23. Compared with the incisors of A. fasciculata these teeth indicate a slightly larger
form, and are distinguished by the more-marked bevelling of their outer edges, and
a difference in their curvature. They are still larger than the incisors of A. africana ,
but much smaller than those of any existing species of Hystrix. Their resemblance
to the incisors of Atherura is indeed so close as to leave little, if any, doubt as to
their generic identity, and since they appear to indicate a form decidedly distinft
from either of the existing species it has been thought well to give to this form tne
name of A. karnuliensis. The rarity of the remains of Atherura as compared v/ith
those of Hystrix seems to indicate that the former genus was dying out in Madras in
the pleistocene, but its occurrence there is of considerable importance as tending
to bridge over the enormous interval separating the areas respectively inhabited by
the two existing species.

Lepus (cf. nigricollis, F. Cuvier).

Mandible. — In the absence of any evidence to the contrary the remains of
the Karnul hare are provisionally referred to the existing south Indian species,
although they are really insufficient for specific determination. There are three
specimens of imperfect right mandibular rami (No. F. 218), of which the least im-
perfect was obtained from bed Ca in the Cathedral, while the other two were found
in bed A a in the Chaper-House.

Limb-bones and vertebrae. —In the Cathedral, Charnel-House, and Purgatory caves
limb-bones and vertebrae of the hare are very common, and exteud to beds low down

i Supra, vol. III. p. 109.

FAUNA OF THE KARNUL CAVES. 39

in the series. A humerus from the Purgatory cave is figured in pi. VIII. fig. 7.

Equus asinus, Linn.

Definition of the term. — Under this specific name may be included both the
domestic races, and the wild asses of Nubia and Somali-land, which are referred to
two races by Mr. Sclater.1 2

Molars. — The much-worn third right upper true molar, of which the crown
surface is represented in pi. IX. fig. 15, was obtained from bed Cf in the Cathedral,
and agrees so exactly with the corresponding tooth of the domestic E. asinus that
there is every probability of its specific identity.* This specimen, of which the
antero-posterior diameter is only 089 inch, is in a thoroughly fossilized condition. A
little-worn left upper tooth belonging to the middle of the cheek-series (No. F. 253)
from bed Cb in the same cave agrees in relative size with the preceding specimen ;
its antero-posterior diameter being 0*85, and the transverse 0*79 inch. The third
right lower true molar from bed Cf in the Cathedral represented in fig. 1 1 of the
same plate agrees so exactly with the corresponding upper tooth that it might have
been thought to have belonged to the same individual, were it not that it is in a
less worn condition. The antero-posterior diameter of this specimen is T08 inches.
The much-worn left lower tooth represented in fig. 12 of the same plate was
obtained from bed Ca in the same cave, and from its very small size is evidently
m.l ; its length is 0*8 inch.

Metatarsal. — A right third metatarsal3 (No. F. 258) from bed Ca in the Cathedral
agrees in relative size with the teeth. The length of this specimen is 8-2, and the
transverse diameter of the distal extremity T54 inches.

Distribution. — The existing wild races of E. asinus being confined to north
Africa it is extremely interesting to find evidence in the pleistocene of southern
India of a form which there is every reason to regard as specifically the same, and
which not improbably indicates that the African races originally migrated from
India. A fragment of the right ramus of the mandible of a small species of Equus
from the pleistocene of the Narbada (British Museum No. M. 26904) may not im-
probably indicate the existence of the present or an allied species in that area.

Equus, sp. a.

Upper molars. — Several molars of an Equus of superior size to E. asinus , and
agreeing approximately in this respect with E. onager of north-western India and
Persia have been obtained from the Cathedral cave. Three of these specimens
belonging to the upper jaw are represented in pi. IX. figs. 14, 16, 17 ; those in figs.

1 ‘ Proc. Zool. Soc.’ 1884. p. 542. Equus taniopus, Heuglin, is a synonym of one of these forms, and the inclusion of
both in E. asinus renders unnecessary the note in reference to E. tamiopus given in vol. II. p. xi. of this work.

2 There are apparently no means of distinguishing the teeth of E. asinus from those of a very small pony, but the
absence of any wild diminutive species of true horses in India and Africa renders it highly improbable that the Kamul Equus
belonged to such a form.

3 Noticed by Mr. Foote in the ‘ Rec. Geol. Surv. Ind.’ vol. XVII. p. 204 (No. 2).

* “ Cat. Foss. Mamm. Brit. Mus.” pt. III. p. 73 (1885).

J

40 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

14 and 17 being from the middle of the cheek-series of the left side, and the one in
fig. 1 6 the last upper true molar of the same side ; two are from bed Gc, while the
third is from Gd. These teeth agree with those of existing species in the excessive
length of the antero-internal pillar (e), and are inferior in size to E. namadicus of
the Narbada pleistocene.1 The specimen represented in fig. 16 appears scarcely
larger than the corresponding opposite tooth of E. asinus represented in fig. 15 ;
but this is accounted for by its much earlier stage of wear, its antero-posterior
diameter at the horizontal plane corresponding to that of the latter specimen being
0-98 inch.

Lower molar. — The third right lower true molar from bed Cb in the Cathedral
represented in fig. 13 exhibits the difference in size between the corresponding tooth
of E. asinus (fig. 11).

Metatarsal. — A right third metatarsal (No. F. 259) from bed Gf in the Cathedral,
which has lost a portion of its distal extremity, agrees in relative size with the fore-
going teeth ; its extreme length being 9*3 inches. This specimen is much smaller
than the average of the metatarsals of the fossil races of E. caballus (and therefore
than those of E. namadicus), and apparently indicates a species of the size of E.
onager.

Affinities. — The foregoing specimens are insufficient for specific distinction, and
all that can be said about them is that they indicate a species superior in size to E.
asinus , which is certainly distinct both from the larger E. namadicus of the earlier
pleistocene, and E. sivalensis2 of the pliocene of India. This species was about equal
in dimensions to the existing Indian E. onager and some of the south African
species ;3 and, judging from the marked Ethiopian facies of a considerable portion
of the Karnul fauna, and the absence at the present day of any existing wild Equus
in southern India, it is not improbable that its affinities may be with the latter.

Rhinoceros karnuliensis, nobis.4

History. — The remains on which this species is founded were provisionally
identified by Mr. Foote5 with B. sondaicus, but their distinctness was shown by the
present writer in the paper cited above. The remains comprise numerous detached
upper and lower cheek-teeth, many of which are imperfect, the greater portion of
the left ramus of the mandible, a fragment of a right ramus with one true molar,
the greater part of a humerus, three imperfect cervical vertebrae, and the distal half
of a metapodial. The mandible and the more perfect teeth, to which comparisons
will be mainly confined, are figured in pi. X.

Mandible. — It will be convenient to commence the description with the left
mandibular ramus, of which two views are given on a scale of one half in pi. X.

1 Vide supra, vol. II. pi. XTV. fig. 3 — on the assumption that some of the specimens are premolars.

2 Distinguished hy the antero-posterior shortness of the antero-internal pillar of the upper cheek-teeth.

3 In the upper molars of E. zebra figured by Riitimeyer in the “Pferde der Quaternar-Epoche” (‘ Abh. shweiz. pal. Ges.’
vol. II.) pis. I. and II. fig. 7 (1877), the antero-internal pillar is more elongated antero-posteriorly.

4 ‘ Rec. Geol. Surv. Ind.’ vol. XIX. p. 120 (1886). 5 Ibid. vol. XVIII. p: 232 (R. javanicuej .

FAUNA OF THE KARNUL OAVES.

41

figs. 4, 4a. This specimen comprises the greater portion of the horizontal ramus
and the hinder part of the symphysis ; the last five cheek-teeth are in situ , and from
their worn condition indicate that their owner was an adult individual ; the alveolus
of pm. 2 still remains, but as there is no trace of that of pm. I the latter tooth must
have totally disappeared. In size the specimen corresponds with the mandible of
R. sondaicus. In the broken extremity of the symphysis there is no trace of alveoli
for canines, and this circumstance, together with the backward extension of the
symphysis to the anterior border of pmV3 (fig. 4), the convexity of the inferior
border of the ramus, the sudden inward curvature of the external border of the
ramus in advance of the same tooth, the backward position of the mentary foramen
(for.), and the narrow, deep, symphysial channel, indicate that the specimen belongs
to that group1 of rhinoceroses in which the canines are usually absent,2 and all the
known forms are bicorn. The length of the space occupied by the five cheek-teeth
is 7*6 inches, and the length of m. 3 1*9 inches. The cheek-teeth are remarkable for
the extremely thick coat of cement which invests the bases of their crowns ; they
have no trace of any external cingulum, but do not present any other well-marked
specific characters. The fragment of a right mandibular ramus (No. F. 238) con-
taining the slightly worn ul3 is of rather larger size, the length of the tooth being
2*0 inches.

Upper true molars. — Of the upper true molars the crowns3 of the associated left
m.2 and m.3 are represented in pi. X. figs. 1, la, ib ; the collection also contains the
crown of the right m.3 of the same individual, and a less perfect specimen of the
right m. 1 or m.2 (No. F. 234). The figured specimens are in a middle condition of
wear, and belong to the more common type of structure, as exemplified in R.
sondaicus. The first and second costae (c, d) are prominent and form a well-marked
buttress,4 and the external surface is deeply curved ; there is a distinct cingulum on
the anterior and inner surfaces of the anterior collis (a), which is totally absent on
the posterior collis ( b ) ; the two colies are separated by a considerable interval : the
crochet ( e ) is well developed and has a separate accessory tubercle in the median
valley, which occurs in all the specimens in the collection ; there is no combing-plate,
no antecrochet, nor any trace of a tubercle at the entrance of the median valley ( g ),
and when more worn the crowns would present only two fossettes. Compared with
the molars of B. sondaicus the crowns appear to have been relatively rather shorter,
and may be described as being probably of a sub-brachydont type.5 The length of
the outer surface of m-2 is 1*8, and that of its anterior surface 2*1 inches.

l The Atelodine group (to which R. deecanensis belongs) ; see “ Cat. Foss. Mamm. Brit. Mus.’ pt. III. p. 92.

a In R. persia, Pohlig (‘ Quart. Joum. Geol. Soc.’ vol. XLII. p. 178), of Maragha lower canines are present.

3 The roots of all these teeth have been gnawed off by porcupines.

4 The absence of a buttress is seen in the third right upper true molar of R. unicornis represented in figs. 3, 3a of the
same plate.

5 In the preliminary notice these teeth were described as decidedly brachydont. Subsequent examination has, however,
shown that owing to their partially worn condition and the fact of the base of the crowns having been gnawed away, it is
very difficult to come to a certain conclusion on this point.

42

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

Upper premolars. — There are unfortunately no perfect upper premolars in the
collection ; hut the inner half of a well-worn right pm. -3 is represented in pi. X.
fig. 2. This specimen1 shows that there is no trace of a cingulum on the anterior
collis (a), and merely an oblique ridge running downwards and backwards on the
anterior aspect of the posterior collis (b) ; the two colles unite at their junction for a
considerable part of their height. The corresponding portion of the left pm. 4 of
the same individual (No. F. 135a) presents precisely similar features.

Affinities. — That the present form is specifically distinct from all the existing
Indian species of Rhinoceros is self-apparent ; and it will not be difficult to show its
apparent distinctness from all the fossil species of the same country. In Madras
two other species occur in a fossil condition ; the first of which appears identical
with the existing R. unicornis , and is known by the slightly-worn third right upper
true molar represented in pi. X. figs. 3, 3a, which was obtained several years ago by
Mr. Foote from a turbary, and is interesting as showing the extensive range of this
species in past times.2 The second species, R. deccanensis, Foote,3 is of pleistocene
age, and, although of somewhat superior size, agrees with the present form in the
absence of lower canines, and in the general plan of structure of the upper true
molars. The teeth are, however, described as being markedly hypsodont, and
without any appreciable quantity of cement, while in the upper true molars the
external surface is nearly flat, and the colles are approximated and show no trace of
any internal cingulum. In the premolars, however, there is a very strongly-marked
cingulum completely surrounding the inner half of the crown,4 and the inner half
of the anterior colles appears less flattened. The premolars are moreover larger in
proportion to the true molars, the antero-posterior diameter of pm. 4 being 1*55 and
that of m.2 1*9, while in the present form the corresponding dimensions are 1T5
and 1'7. In the mandible the arcuation of the inferior border, and especially the
upward inclination of its anterior moiety, is very much more strongly marked,5 and
the bases of the crowns of pm. 2 and pm. 3 are placed on a considerably higher level
than that of pm. 4, instead of in the same horizontal line. The symphysial channel
appears wider, much less distinctly defined, and more open ; pm. 2 is apparently
larger and more widely separated from its fellow of the opposite side, while there is
a distinct cingulum at the two extremities of the outer surfaces of the premolars,
but no trace of a mentary foramen below pm. 3. From the structure of that portion
of the symphysis still remaining it appears probable that this part was rather shorter
in the fossil. With the fossil rhinoceroses of the Siwaliks the present form has no
affinity ; the only bicorn species ( R . platyrhinus ) having upper cheek teeth of a
totally different type of structure. The Maragha R. persice 8 is also an entirely
different form.

1 It is highly probable that this and the next specimen belong to the same individual as the true molars.

2 The artist has foreshortened the inner surfaces of the colles in fig. 3, which makes the crown look lower than it really

is, but its true height is shown in fig. 3a. 3 Supra, vol. I. pp. 1-17. pis. I. -III.

4 Compare Foote, op. cit. pi. I. 5 Compare Foote, pi. II. fig. 3. 6 Vide supra, p 41, note 2.

FAUNA OF THE KARNUL CAVES.

43

Of the fossil European members of the Atelodine group occurring above the
Pikermi horizon the one which apparently comes nearest to the Karnul rhinoceros is
R. etruscus. The upper cheek-dentition of that species is, however, apparently
somewhat more brachydont, while the upper premolars usually have a very distinct
horizontal cingulum on their inner aspect,1 and are larger in porportion to the true
molars. The latter are, however, very like those of the fossil, and show a cingulum
on the inner aspect of the anterior collis in the two last of the series,2 although the
second costa does not extend to the base of the crown. The mandible also approaches
the Karnul jaw in general contour, but the symphysial channel is shallower and less
defined, while the outer surface of the horizontal ramus inclines less inwardly in
front of pm. 3, and the mentary foramen is usually double and has its hinder aperture
placed below pm. 2 instead of pm. 3, and nearer to the inferior border of the ramus ;3
there is also a distinct cingulum at the two extremities of the outer surface of the
lower cheek-teeth.4

Of the two existing African rhinoceroses R. simus has no affinity with the
present form, but R. bicornis appears very closely allied. The upper cheek-teeth of
the latter have, however, considerably higher crowns, the second costa in the true
molars does not extend to the base of the crown, and the buttress in the same teeth
appears less strongly marked, while the premolars have a slight cingulum, which
does not present an oblique ridge on the posterior collis. The contour of the in-
ferior border of the mandible is also more curved,5 but the hinder part of the
symphysis is extremely like that of the fossil, many specimens showing the same
well-defined channel. In most examples of the. existing species the mentary foramen
occupies the same position as in the fossil,6 but the symphysis of the latter was
almost certainly somewhat longer anteriorly, and was perhaps intermediate in this
respect between R. bicornis and R. etruscus. The bicorn R. pacliygnathus of Pikermi
appears closely allied to R. bicornis , and differs from the fossil in much the same
respects as the latter.

Summary — The usual ill fortune of the palaeontologist obtains in the present
instance, for had but the mandibular symphysis of the Karnul rhinoceros been complete
there would have been no question whether its affinities were nearest to U. etruscus and
R. deccanen&is, or to R. bicornis. There is, however, apparently but little doubt as
to its specific distinctness from the first of these three species ; and if it be assumed
that the presence of the cingulum in the upper premolars, the absence of a large
amount of cement in the cheek-teeth, the contour of the mandible, and the position
of the mentary foramen, are constant characters in the second species, it will bo
evident that the present form cannot be identified with the Deccan rhinoceros. The
apparently more hypsodont dentition of R. bicornis and the difference in the contour

1 Compare Boyd -Dawkins “ Quart. Joum. Geol. Soc.” vol. XXIV. pi. VII. fig. 1 (1868).

2 Ibid. pi. VII. fig. 1. and VIII. fig. 4. 3 Compare “Falconer’s Palaeontological Memoirs,” vol. II. pi. XXVII.

4 Compare Dawkins, op. cit. pi. VII. fig. 3. 5 Compare Blainville’s “ Osteographie ” — genus Hhinocei-os, pi. III.

6 It is occasionally situated below pm. 2-

K

44

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

of the inferior border of the mandible, together with the apparently longer sym-
physis, seems to indicate specific distinction in this instance also.

These conclusions entail the necessity of at least provisionally regarding the
Karnul rhinoceros as specifically distinct from all other described forms. Additional
specimens are, however, essential to a fuller comprehension of its affinities, and all
that can be said at present is that the species appears to show characters connecting
it on the one hand with R. etruscus and R. deccanensis, and on the other with It.
bicornis.

Horizon. — The majority of the specimens were obtained from the Cathedral
cave, the more perfect ones being found in the beds C, Ob, Cc, and Gd, and broken
fragments in Ce and Cf.1 Specimens were also found in the Charnel-House, the
right upper true molar (No. F. 234) noticed on page 41 being apparently the one
alluded to by Mr. Foote2 from that cave.

Bos, or Bubalus, sp.

Limb-bones, teeth , and mandible. — The remains of ruminants belonging either to
one or both of the above-mentioned genera are abundant in the Cathedral, especially
in the beds Ob, Cc, and Cd, and comprise limb-bones, detached teeth, and several
imperfect mandibular rami. In the absence, however, of any of the characteristic
portions of the cranium it seems impossible to make a generic determination of these
specimens.

Boselaphus tragocamelus (Pallas).

Syn. . Portax picta, H. Smith.

Upper molars. — Several upper molars of this species were obtained from the

Cathedral in beds Ga, Cb, Cc, and Gd, of which
three are represented in pi. XI. figs. 7, 8, 9.
The slightly worn specimens represented from
the inner aspects in figs. 7, 9, judging from
their comparatively short crowns, are probably
examples of md (of the right side), while the
unworn tooth of which the outer aspect is
shown in fig. 8 is m,2 of the same side. These
teeth agree precisely with the molars of the
existing race, of which the left mdJ of a female
is figured in the accompanying woodcut, and
exhibit the characteristic tall crown, with the
long and slender accessory inner column, which
is attached entirely to the hinder crescent. It is very difficult to point out any
characters by which these teeth can be distinguished from those of the Siwalik
Boselaphus figured in plate XIII. of the preceding volume of this work.

Lower molar and mandible. — An unworn left (second ?) lower true molar, which

Pig. 2. Boselaphus tragocamelus. The second
, left upper true molar, in an almost unworn
condition : recent, India.

l Vide Foote ‘ Rec. Geol. Surv. Ind.’ vol. XVIII. p. 230.

2 Ibid. vol. XVII. p. 204.

FAUNA OF THE KARNUL CAVES.

45

has lost part of the inner column of the anterior lobe, from bed Aa in the Chapter-
House, is represented from the external aspect in pi XI. fig. 10. Mr. Foote1 refers
to this species a fragment of the left ramus of a mandible (No. F. 285) from bed L in
the Charnel-House, containing the much- worn mTT and n TV ; while the right ramus of
a very young animal with the three milk-molars represented in pi. XI. figs. 14, 14a
may also be referred to it. The latter agrees very closely with a slightly older
specimen of the opposite ramus of the Siwalik form represented in vol. III. pi.
XIII. fig. 4.

Genus, non. det.

Upper molar. — The partially-worn and slightly imperfect left upper true molar
from bed Cb of the Cathedral represented in pi. XI. figs. 2, 2a apparently indicates
the occurrence of a large antelope belonging to the group which comprises the
genera Oryx, Palceoryz, Addax, and Hippotragus. The tooth is characterized by the
square form of the base of the crown, the large size of the inner accessory column,
and the well-developed costas on the outer surface. Among living antelopes the
specimen comes nearest to the teeth of Oryx and Hippotragus, but apparently in-
dicates a more brachydont form ; it differs from the former by being narrower, and
by the absence of the antero-posterior expansion of the accessory column. The
molars of the Pikermi species of Palceoryx have a flatter outer surface, and a very
minute accessory column ; but the tooth of the so-called Antilope boodon, from the
pliocene of France, which is referred by M. Dep4ret2 to Palceoryx is exceedingly like
the Indian fossil. The latter differs from the molars of Orecis, Strepsiceros, and
Palceoreas , by the more marked external costae, and from the existing species of the
two former genera by the presence of the large accessory column, although resem-
bling the fossil forms in this respect.3

Gazella benetti (Sykes).

Mandible. — The imperfect left mandibular ramus from bed Ca in the Cathedral
represented in pi. XI. figs. 15, 15a agrees precisely with the mandible of a female of
this species in the British Museum (No. 617d)4; the relative shortness of the pre-
molars, which is a characteristic feature of the genus, being well shown in the fossil.
A fragment of the left ramus of the mandible of a male (No. F. 279) was obtained
from bed / in the Charnel-House ; and it is not improbable that an atlas vertebra
(No. F. 279a) from the Purgatory cave, an axis (No. F. 279 b) from the Cathedral,
and some metapodials and phalangeals (No. 279 c) may also belong to the present
species.

Antilope cervicapra (Linn).

Syn. A. bezoartica, Auct.

Upper molar. — The partially-worn second left upper true molar from bed Gb in

1 ‘ Rec. Geol. Surv. Ind.’ vol. XVII. p. 206 (No. 1).

2 ‘ Theses. Facult. Sci. Paris.’ ser. A. No. 67 — ‘ Bassin Tertiaire du Rousillon.’ p. 247. pi. III. figs. 9, 10 (1885).

3 Vide supra, pp. 8, 9. 4 Gray “ Hand-list of Edentate, Ruminant, and Thick-Skinned Animals,” p. 109 (1873).

46

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

the Cathedral represented in pi. XI. figs. 13, 13a agrees precisely with the corres-
sponding tooth of this species, and may accordingly be taken as indicating its
occurrence among the Karnul fauna.

Tetraceros quadricornis (Blainville).

Mandible. — The imperfect left mandibular ramus containing all the cheek-teeth
except pm. 2 represented in pi. XI. figs. 12, 12a was obtained from bed Cd in the
Cathedral. This specimen, which exhibits the slender external accessory column in
the true molars characteristic of some individuals of this species,1 cannot be distin-
guished from the corresponding portion of the mandible of existing examples of
the four-horned antelope,' and exhibits the relatively large size of the premolars
characteristic of the genus. The relatively large size of these teeth may be seen
by comparing the figures of the specimen with those of the mandible of Gazella
bennetti represented in figs. 15, 15a of the same plate, when it will be observed that
while the length of m73 is nearly the same in the two specimens, the length of the
space occupied by the three preraolars in the latter is rather less than the united
length of the two hinder teeth of that series in the present specimen. Tetraceros
quadricornis is one of the commonest of the larger Madras mammals, and may
probably be regarded as the descendant of the smaller T. daviesi2 of the Siwaliks.

Cervus aristotelis, Cuvier.

Antler. — An imperfect antler from bed Ca in the Cathedral agrees very closely
with that of the existing sambar.

Molars. — Numerous upper molars indistinguishable from those of the living
race have been obtained from beds Gb, Cc, Cd, and Cf, of the Cathedral, of which
two very perfect examples are figured in pi. XI. figs. 5, 5a, and 6, 6a. The
specimen represented in figs. 5, 5a is a first left upper true molar, while that in
figs. 6, 6a is the second molar of the opposite side. Both teeth are partially worn,
and exhibit very clearly the short, squared crowns, with the large accessory column
attached at its base to both the adjacent crescents.

Vertebra and limb-bones. — It is not improbable that an atlas vertebra (No. F
305) from bed Ca of the Cathedral belongs to this species, and Mr. Foote3 refers to
it an imperfect tibia (No. F. 305, a) and an astragalus (No. F. 305, b) from .bed L in
the Charnel-House.

Cange. — The species is found throughout India, and thence through Assam and
Burma to the Malay peninsula, and also occurs in Ceylon ; it has been provisionally
recorded from the pleistocene of the Narbada valley.4

Cervus axis, Erxleben.

Molars. — To this species may be referred the two specimens of third upper true
molars from beds Cb and Cc in the Cathedral represented in pi. XI. figs. 1, la, and 3.

l Vide supra, p. 20. 2 Ibid, p. 19. 3 ‘Rec. Geol. Surv. Ind.’ vol. XVII. p. 207 (Nos. 15, 16).

i “ Cat. Foss. Mamin. Brit. Hus.” pt. H. p. 103 (1885).

FAUNA OF THE KARNUL CAVES.

47

The specimen in figs. 1,1a belongs to the tight side, and is but slightly worn ; it
exhibits the comparative hypsodontism characteristic of this species, and the very
small size of the inner accessory column. The specimen represented in fig. 3 is
considerably more worn, and belongs to the left side. The right first or second
lower true molar represented in fig. 4 of the sairle plate was obtained from bed Cc
of the Cathedral.

(?) Cervulus muntjac (Zimmermann).

Syn. Cervulus aureus , Auct.

Upper molar. — The imperfect tooth from the Purgatory cave represented in pi.
XI. figs. 11, 11a agrees exactly with the corresponding portion of the second right
upper true molar of the muntjac, and may probably be referred to that species,
although it is very difficult to distinguish such an imperfect specimen from the
corresponding tooth of Tetraceros quadricornis.

Tragulus (cf. meminna [Erxleben]).

Metapodials. — Two specimens of the distal half of metapodials wanting the
epiphyses (No. F. 307), from bed Cc in the Cathedral, indicate the presence of a
chevrotain, which is probably identical with the existing Indian species, among the
Karnul fauna.

Sus cristatus, Wagner.

Occurrence. — Detached teeth of Sus occur very commonly in nearly all the beds of
the Cathedral, and in many of those of the other caves. The difference in the size
of the cheek-teeth indicates the occurrence of two forms, one of which may be
identified with the existing Indian species, while the other apparently indicates a
new species. Since the third true molars afford the best distinctive characters, com-
parisons will be in the main confined to these teeth.

Lower molars. — Three specimens of the third left lower true molar from beds
Cb and Cc in the Cathedral are represented in pi. IX. figs. 1, 2, 4, of which the one
in fig. 1 is the most, and that in fig. 4 the least worn.1 The fragment of the right
ramus of a mandible with the three true molars represented in fig. 5 of the same
plate was obtained by Mr. Foote from a turbary in Madras in company with the
molar of Rhinoceros unicornis figured in pi. X. fig. 3. The teeth of the latter speci-
men, which from the complexity of the talon (b, c, d) may
be inferred to have belonged to a male, agree exactly in
size with those of average individuals of the existing
race, of which a third lower molar is figured in the
accompanying woodcut ; but the Karnul specimens are
lover true moiar'of a maieT^Recwit” slightly larger. In regard to the complexity of the
{. British Museum. No. 7i6o. talon of the third lower molar there is considerable

l The lettering designating the different elements of these teeth is the same as that employed in pi. VII. of vol. III. of
the present work.

L

48

INDIAN TERTIARY AND POST-TERTIARY YERTEBRATA.

variation ; the complexity being greatest in male animals. The figured specimen is
an average example, but in some individuals the portion d consists of only a single
column1 ; while in others the complexity is as great as in the specimen represented in
fig. 5, where there are four distinct columns in this portion. With the exception of
one specimen (No. F. 2 66&), which agrees with the one figured in the woodcut, all
the Karnul teeth exhibit the extreme complexity of talon which is only sometimes
met with in the existing race. In the (probably male) specimen represented in fig.
4 the columns are rather taller, and their inner surface more flattened than in any
recent examples which have come under the writer’s notice, in both of which respects
the specimen approximates to the corresponding tooth of the Siwalik 8. falconeri, of
which specimens are represented in vol. III. pi. VII. figs. 1 and 2 of this work. In
male individuals of that species (vol. III. pi. VII. fig. 2) the third molar is con-
siderably larger than the Karnul teeth ; but in the female (vol. III. pi. VII. fig. 1)
this tooth is very nearly of the same length as the latter. The Karnul and other
teeth of 8. cristatus may, however, be easily distinguished from those of females of
S. falconeri , by the smaller interval between the column marked a and the anterior
extremity, which is indicative of the less degree of lateral compression and antero-
posterior extension of the main columns of the former teeth. This feature is,
however, rather less marked in the specimen represented in pi. IX. fig. 4 than in the
other teeth.

JJpper molar. — The slightly-worn third right upper true molar from bed Cd in
the Cathedral represented in pi. IX. fig. 7 agrees in relative size and the great com-
plexity of the talon (a, b, c ) with the lower teeth. There are two equal-sized
specimens (No. F. 260£, 2676) from beds Cb and Gd in the Cathedral, in the latter
of which the talon is still more complex.

Lower canine. — The anterior portion of the left lower canine of a male from
bed Ca in the Cathedral represented in pi. IX. fig. 10 can scarcely be distinguished
from the corresponding tooth of large individuals of the existing race, and indicates
that some of the molars described above belong to the same sex.

Range and affinity. — The Karnul specimens carry back Sus cristatus to the later
pleistocene, and the impossibility of distinguishing the third lower Karnul molars
from the corresponding tooth of the mandible from the Narbada described on page
85 of the preceding volume of this work2 almost certainly indicates the existence of
the species in the earlier part of the same period. The tendency to a greater com-
plexity of structure in the last molar of the fossil race appears to show decided
evidence of affinity with S. falconeri, and to indicate the probability of the living
species being a descendant from the Siwalik form which has lost the elongated facial
portion of the cranium characteristic of the latter ; and in the sequel it is suggested
that the next species may be the survivor of the intermediate form. The existence

1 In vol. III. p. 75 it is stated (from the comparison of an insufficient number of specimens) that the portion d always
consists of only a single column.

2 See “ Cat. Foss. Mamm. Brit. Mus.’ pt. II. p. 266. Nos. 36843, 36725 (1885).

FAUNA OF THE KARNUL OAVES.

49

of an allied form in tlie reputed pliocene of Algiers is apparently afforded by S.
phacochoeroides , Thomas,1 in which the third lower true molar agrees very closely in
size and structure with the specimen represented in pi. IX. fig. 4, although the
development of lateral accessory columns in the talon is somewhat greater.

Sus karnuliensis, n. sp. nobis.

Definition . — This provisional species agrees in size with S. falconeri, but in the
general structure of the molars with S. cristatus, although some specimens of these
teeth approach those of certain examples of the former.

Loiver molar. — The much-worn third left lower true molar from bed Cd in the
Cathedral represented in pi. IX. fig. 3, and the somewhat less-worn tooth of the
opposite side from bed Cb represented in fig. 8 indicate a species allied in the
structure of these teeth to N. cristatus-, but of considerably larger size. It appears,
indeed, very difficult to detect any structural difference between these teeth and those
specimens of the third lower molars of that species in which the talon ( b , c, d ) is of
the most complex type. Compared with the corresponding tooth of the male of the
Siwalik S. falconeri (supra, vol. III. pi. VII. fig. 2) these teeth agree very closely in
length, but differ by their inferior width and less elongation of the main columns,
so that the interval between the anterior extremity and the point a is very much
smaller. This difference is still more marked if the Karnul teeth be compared with
m.3 of the female of the Siwalik species (vol. III. pi. VII. fig. 1), when it will be
seen that in the absolutely smaller Siwalik tooth the above-mentioned interval is
greater than in the Karnul teeth. In one male mandible of S. falconeri (B.M. No.
M. 201 22) the third lower true molar is very like the Karnul specimen represented in
fig. 3 ; the interval between the anterior border and a being but very slightly smaller.

Upper molars. — The well-worn third right upper true molar from bed Cd in the
Cathedral represented in pi. IX. fig. 6, and the almost unworn corresponding tooth of
the opposite side from the same bed represented in fig. 9, agree in relative size with
the lower molars. Both specimens exhibit great complexity in the development of
the talon (a, b, c ), and also of the lateral accessory columns ; in fig. 9 there is a supple-
mental line of small tubercles behind c which is wanting in fig. 6. Beyond the
circumstance that the accessory columns appear more numerous than in m.3 of S.
cristatus (fig. 7) these teeth do not appear to differ structurally from examples of that
species in which the talon of that tooth is unusually largely developed. Compared
with the two specimens of m.3 of S. falconeri represented in vol. III. pi. VII. figs,
o, 7, the Karnul teeth differ by their greater relative width, the shorter interval
between the anterior extremity and the point a, and the lesser antero-posterior
elongation of the discs of dentine formed by the abrasion of the main columns. In
one male cranium of S. falconeri (B.M. No. 153163) m.3 is, however, wider than
usual, and thereby approaches the Karnul teeth, although the interval between a and

1 ‘ Mem. Soc. Geol. France,’ ser. 3. vol. III. pt. 2, p. 10. pi. X, fig. 1 (1885).

2 See “ Cat. Foss. Mamm. Brit. Mus.” pt. II. p. 264. 3 Ibid. p. 263.

50

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

the anterior border is much superior.1 A slightly-worn second upper true molar
from bed Gc in the Cathedral agrees with m.2 of S. cristatus in the squareness of the
crown, and has not the extremely elongated form usually characteristic of slightly
worn specimens of this tooth in S. falconeri (compare vol. III. pi. VII. fig. 5).

Distinctness and affinities. — The structure of the cheek-teeth of the present form
indicates that their affinity is nearer to S. cristatus and S. falconeri than to any other
species. Their superior size apparently, however, forbids their reference to the
former species, while the differences indicated above appear equally conclusive as to
their distinction from the latter ; and it is accordingly proposed to refer them to a
new species, with the name of S. karnuliensis. The resemblance between the teeth
of this species and those of certain examples of S. falconeri may suggest that the
former is the survivor of the form connecting the latter with the existing S. cristatus.

Manis gigantea, Illiger.

Phalangeal. — The terminal phalangeal of the third digit of the right manus of
a large Manis represented in pi. VIII. figs. 8, 8a was obtained from bed Gc in the
Cathedral cave. This specimen indicates a much larger animal than the existing
Indian M. pentadactgla ,2 and agrees so closely with the corresponding bone of M.
gigantea of western Africa that it certainly indicates a very closely allied form, which,
in the absence of any evidence to the contrary, may be regarded as specifically
identical. The fossil specimen is slightly larger than the corresponding bone of a
skeleton in the British Museum,3 of which the total length is 4 feet 6 inches.

The Siwalik edentate. — In the first volume4 of the present work the second
plalangeal of the third digit of the manus of an edentate from the lower Siwaliks of
Sind was described and figured under the name of Manis sindiensis. The opportunity
of comparing a cast of this specimen with the corresponding bones of Manis gigantea
and Macrotherium giganteum has, however, convinced the writer that this determination

is erroneous, and that the Siwalik edentate
should rather be referred to Macrotherium. Com-
pared with Manis the specimen (woodcut fig.
4) differs by the supratrochlear portion being
shorter, and by the lesser projection of the free
borders of the trochlese, as well as by the
superior termination of the posterior surface
being higher than that of the anterior, and the
narrowness of the ridge dividing the two articu-
lar facettes on the proximal surface. In these
respects the specimen agrees very closely with the corresponding bone of Macro-

1 In the Siwalik specimen the talon is less developed than usual, and thereby shows on a superficial examination a closer
resemblance to the Kamul teeth than is really the case.

2 The Kamul form was referred to this species by Mr. Foote in the ‘ Rec. Geol. Surv. Ind.’ vol. XVIII. p. 232.

3 No. 1458*. See Gray “ Hand-List of Edentate, etc., Mammals,” p. 10 (1873).

4 Page 82 (64), pi. VIII. figs. 11-14.

Fig. 4. Macrotherium sindicnse. The second
phalangeal of the third digit of the manus,
viewed from the anterior and distal aspects ;
from the lower Siwaliks of Sind. Indian
Museum (No. D. 99).

FAUNA OF THE KARNUL CAVES.

51

therium giganteum , but in the deep concavity of the proximal articular facettes and
the more perfect rounding of the distal trochlese it resembles Mam's ; in size it is
intermediate between Macrotherium giganteum and Mauds gigantea. On the whole the
characters of this specimen appear -nearer to Macrotherium than to Manis, and it
therefore appears advisable to refer it to the former genus, but the above-mentioned
resemblance to Manis , coupled with the size of the specimen, and the occurrence of
Manis gigantea in the pleistocene of India, renders it highly probable that the species
to which it belonged was a form connecting the generalized Macrotherium giganteum
of the middle miocene of Europe with the more specialized genus Manis.

III. AVES.

General. — The remains of birds are of far less common occurrence in the Karnul
caves than those of mammals, and this paucity of specimens, together with the
intrinsic difficulty of the determination of imperfect bones of birds,1 renders the
number of determined forms but small. In addition, however, to the forms men-
tioned below an imperfect humerus (No. F. 319) from bed Ca in the Cathedral,
agreeing in general characters with the corresponding bone of Corvus cor one, probably
indicates the occurrence of a member of the Corvidae ; while three smaller specimens
of the homologous bone may perhaps belong to other members of the Passeres. Other
humeri (No. F. 316) from the same cave not improbably indicate the occurrence
of Columbidce ; while a right humerus (No. F. 317) from bed Ca indicates a genus of
Anatidce in which this bone is relatively longer and more slender than in Anas boscas.
All the determined fofms belong to existing Indian species, but there is apparently
evidence of a variety of Francolinus pondicerianus. The following list indicates the
systematic position of the forms described.

Accipitres. Neophron percnopterus {Linn).
? Milvus or Circus, sp.

Ketupa ceylonensis {Gmelin).
Bubo coromandus {Lath.)

Galling. Francolinus pictus {Jard. and Selby).

pondicerianus2 {Gmelin).
Alectorides. Grus {cf. communis, Bechst.)
Herodiones. Ibis melanocephala2 {Lath.)

Neophron percnopterus3 (Linn.)

Humerus. — The distal extremity of a left humerus (No. F. 309) from bed Ca in
the Cathedral agrees exactly in contour with the corresponding bone of the skeleton
of this species in the museum of the Royal College of Surgeons, although belonging
to a somewhat larger individual. The characteristic deep fossa on the dorsal aspect
is well exhibited.

(?) Milvus, or Circus, sp.

Tarso-metatarsus. — The distal half of the left tarso-metatarsus of a small raptorial

1 This difficulty is intensified by the want of a really good collection of bird-skeletons in England.

2 The writer follows Sclater (“ List of Animals in Gardens of Zool. Soc.” 8th ed. pp. 411, 468 [1883]) in referring these
two species respectively to Francolinus and Ibis, rather than making them the types of the genera Ortygornis and Threskiomis.

3 The so-called N. ginginianus is included under this name as being at least paleontologically inseparable.

M

52 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

bird from a bed low down in the Cathedral cave, which is figured in the accompanying

woodcut, may be referred either to Milous ,
Circus , or one of the allied genera. The
fossil specimen agrees very closely in size
with the corresponding bone of a skeleton
of the widely distributed Circus cyaneus

Fig. 5. (?) Milvus or Circus, sp. The distal half of the left tarso- (Lillll.) preserved in the M/USeum of the
metatarsus, from the anterior (A) and posterior (B) aspects ; Hoval College of SurtTCOllS.
from tho Cathedral cave. Ind. Mus. (No. F. 310). J ° 6

Ketupa ceylonensis (Gmelin).

Limb-hones. — This species is represented by a right femur1 (No. F. 313a) and

the distal half of the right tibio-tarsus
from bed Ca in the Cathedral, as well as
by the distal portion of the right tarso-
metatarsus from bed Cb , and the third
phalangeal of the third digit of the pes
from bed Ca in the same cave ; the three
last-named bones being figured from the
anterior aspect in the accompanying wood-
cut (fig. 6 a, b, d). These specimens agree
exactly with the corresponding bones of
the existing bird. As characteristic features
in the tibio-tarsus of Ketupa and allied
genera may be mentioned the absence at
the distal extremity of the anterior surface
of a bridge of bone over the extensor tendons, and the presence of a tubercle on
the inner border for muscular attachment.

Fig. 6. A.B.D. Ketupa ceylonensis. The distal half of
the right tibio-tarsus (A), of the right tarso -metatarsus
(B), and the 3rd phalangeal of the 3rd digit of the pes
(D). C. Bubo coromandus. The 2nd phalangeal of the 3rd
digit of the pes. All the specimens are from the
Cathedral cave.

Bubo coromandus (Latham).

Syn. Urrua coromanda , Auct.

Phalangeal. — The second phalangeal of the third digit of the pes of a very
large owl from bed Cc in the Cathedral represented in woodcut fig. 6c, agrees, as
far as can be determined from comparison with a dried specimen, with the corres-
ponding bone of Bubo coromandus , which now inhabits the Carnatic, lower Bengal,
and the outer Himalaya.

Francolinus pictus (Jardine and Selby).

Tarso-metatarsus. — Evidence of the existence of the painted francolin, which
replaces the black francolin [F. vulgaris) in central and parts of southern India, is

l The collection also contains a perfect right tibia (No. F. 311) marked as from bed Ca in the Cathedral, which, however,
appears to be recent.

FAUNA OF THE KARNUL CAVES.

53

afforded by the apparently tarso-metatarsus represented in woodcut fig. 7. This

Fig. 7. Francolinus pictus. The left tarso- Fg. 8. Francolinus pondicerianus. The left

metatarsus ; from the Cathedral cave (bed tarso-metatarsus ; from the Cathedral cave

Ca). ■}. Indian Museum (No. F. 318). (bed Ca). Indian Museum (No. F. 342).

specimen was obtained from bed Ca in the Cathedral, and cannot be distinguished
from the corresponding bone of the existing bird ; it is, however, also difficult to
distinguish it from the tarso-metatarsus of the female of Francolinus pondicerianus.

Other hones. — A left femur and tibio-tarsus (No. F. 342a) from beds Ca and Cb in
the same cave may probably be likewise referred to the present species.

Francolinus pondicerianus (Gmelin).

Syn. Ortygornis pondiceriana , Auct.

Tarso-metatarsus. — The left tarso-metatarsus from bed Ca in the Cathedral repre-
sented in woodcut fig. 8 is furnished with two spurs, of which the greater portion
of the uppermost one is broken away. The specimen is much smaller than the corres-
ponding bone of the Indian Galloperdix , or the west African Francolinus bicalcaratus,
and agrees in point of size with the metatarsus of the grey francolin. In existing
males of the latter species there is, however, normally but one spur, although Jerdon
states that a secondary spur is occasionally found 1 at the base of the normal one.’1
In view of this variation, and it being certain that the fossil bone cannot belong to
any other existing Indian bird, it is provisionally regarded as indicating the occur-
rence of a variety of the grey francolin among the Karnul fauna.

Grus (c/. communis, Bechstein).

Metacarpus. — Two specimens of the left metacarpus (No. F. 314) of a large
wader agree so closely with the corresponding bone of Grus communis that they may
be regarded as indicating the occurrence among the Karnul fauna of a member of
that genus, which is very probably identical with the above-named species.

Ibis melanocepiiala (Latham).

Syn. Threskiornis melanocephala, Auct.

Limb-hones. — Of the black-headed ibis, which is widely distributed in south-

1 Marshall and Hume (“ Game Birds of India ”) state that they have only observed this secondary spur in domesticated
birds.

54

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

eastern Asia, there are several limb-bones. A slightly imperfect right tibio-tarsus

(No. F. 315) from bed Cd in the
Cathedral agrees very closely with
the corresponding bone of a skeleton
in the British Museum, and measures
6*5 inches in length. The distal
portion of a slightly larger specimen
of the homologous bone from bed
Ca in the same cave is represented
in the accompanying woodcut (fig.
9), and shows the well-marked bridge
over the channel of the extensor
tendons. The distal half of the left
tarso-metatarsus represented in the
same woodcut was obtained from
the Purgatory cave, and exhibits
very clearly the trace of the divisions
between the three component bones,
which is very characteristic of the
species.

IV. REPTILIA.

General. — Reptilian remains from the Karnul caves are, with the exception of
those of Varanus , comparatively scarce, and in most cases even generically undeter-
minable. In addition to the genera noticed in the sequel, there is evidence of the
existence of one or more small species of emydine, and perhaps also of other tortoises.
Several fragments of jaws indicate small lizards probably belonging to the Geclconidce
and Agamidce, but many of the specimens appear to be of very recent origin. There
are also some small snakes, which have not been generically determined.

Crocodilus, sp.

Tooth. — This genus is represented by a small anterior tooth, probably belonging
to the lower jaw, from bed Y in the Cathedral cave.1

Varanus dracaena (Shaw).

Abundance of remains. — Remains of Varanus are of exceedingly common occur-
rence in the Karnul caves, and consist mainly of upper and lower jaws and vertebrae.
In the absence of any evidence to the contrary all these remains are referred to the
common species now inhabiting southern India, which at the present day attains a
length of four feet.

Maxilla. — A right maxilla from bed K in the Charnel-House is represented in

Pig 9. Ibis melanocephala. The distal portion of the right tibio-
tarsus (A) and left tarso-metatarsus (B). The former is from
bed Ca in the Cathedral, and the latter from the Purgatory cave.
Indian Museum (Nos. F. 315a, 315b).

l See Foote “Rec. Geol. Surv. Ind.” vol. XVII. p. 207.

FAUNA OF THE KARNUL CAVES.

oo

woodcut fig. 10, and presents no characters by which it can be distinguished from
the corresponding bone of full-sized individuals of the existing form.

Eig. 10. Varanus dracccna. The right maxilla (A), and a dorsal vertebra from the anterior
(B) and posterior (C) aspects ; from the Cathedral cave. Indian Museum (No. F. 334).

Vertebra. — In the same woodcut is represented a dorsal vertebra from bed Ca
in the Cathedral cave agreeing in relative size with the maxilla ; other specimens
indicate somewhat larger individuals, which may have been as much as five feet in
length.

Shvalik species. — In order to exhibit the enormous size attained by the Siwalik
representative of the genus ( V. sivalensis ) an imperfect dorsal vertebra1 from the
Siwalik Hills is represented in woodcut fig. 11. The length of the centrum is 0-8

Fig. 11. Yaranu s sivalensis, Falconer. A dorsal vertebra ; from the posterior and ventral aspects ; from
the pliocene of the Siwalik Hills. J. British Museum (No. R. 739).

inch. An imperfect cervical vertebra in the same collection (No. R. 740) indicates
a still larger individual.

Python molurus (Linn.)

Vertebrce. — Several dorsal vertebrae

Fig. 12. Python molurus. A dorsal vertebra viewed
from the posterior (A) and ventral (B) aspects ; from
the Cathedral cave (bed Cb). {. Indian Museum
(No. F. 336).

indistinguishable from those of the existing
form have been obtained from beds Cb, Cc,
and Cd of the Cathedral. All these speci-
mens (of which one is figured in the
accompanying woodcut) belong to small
individuals ranging from about seven to
ten feet in length. The species has been
provisionally recorded from the Siwaliks

This and the next specimen were brought to light after the description of the humerus was ’published in vol. III.

p. 236 of the present work.

N

56 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

of the Punjab.1 The characteristic shortness of the centrum and the breadth and
bluntness of the haemal ridge, are well shown in the figure.

Naia tripudians (Merr.)

Vertebra. — A dorsal vertebra of a snake (No. F. 336£) from bed Cb in the
Cathedral indicates the occurrence of Naia tripudians among the Karnul cave-fauna.
This specimen agrees precisely with the dorsal vertebra of this species described and
figured in Owen’s “ Monograph of the Reptiles of the London Clay,” pt. II. p. 55,
pi. XIII. figs. 13-16 (1850), where the distinctive features are clearly identicated.

Ptyas mucosus (Linn.)

Vertebra. — The dorsal vertebra of a large snake from bed Cb in the Cathedral

represented in woodcut fig. 13 agrees so
closely in structure with a vertebrae of a
small example of the widely-distributed Ptyas
mucosus in the British Museum,2 that it may
be safely referred to the same genus, and
probably to the same species. The recent
vertebra with which it was compared belonged
to an individual between two and three feet
in length, but living examples are found of seven-and-a-half feet in length, and the
fossil indicates an individual of perhaps somewhat larger dimensions. The characters
distinguishing the vertebrae of the present genus from those of Python (fig. 12) are
the longer centrum, the more strongly developed haemal ridge, and the greater
prominence of the articular facettes for the ribs.

Pig. 13. Ptyas mucosus. A dorsal vertebra, viewed
from the posterior (A) and ventral (B) aspects ; from
the Cathedral cave (bed Cb). Indian Museum
(No. F. 336a).

V. AMPHIBIA.

Bofo {cf. melanostictus, Schneider).

Humerus. — The only amphibian remains that have come under the writer’s

notice3 are several imperfect specimens of the
'Paw humerus of a Bufo from beds Ob and Cc in

the Cathedral. These specimens, of which
U| one is represented in woodcut fig. 14, may

probably be referred to the common Bufo
llH melanostictus , which ranges over the whole

*,vJ of India, and many of the adjacent coun-

Fig. 14. Bufo {cf. melanostictus). The imperfeot left j. •
humerus; from the Cathedral cave (bed Cc). £. Indian tries.

Museum (No. F. 337).

1 i Supra, vol. III. p. 237. pi. XXXV. fig. 4. The specimen represented in fig. 7 is a dorsal and not a caudal vertebra.

2 The writer is indebted to Mr. G. A. Boulenger of the British Museum for having caused this specimen to be prepared
for the purpose of comparison.

3 Mr. Foote ‘ Rec. Geol. Surv. Ind.’ vol. XVIII. p. 232 mentions the occurrence of Rana.

FAUNA OF THE KARNUL CAVES. 67

VI. MOLLUSCA.

General. — For the determination of the very few shells obtained from the Karnul
caves the writer is indebted to Mr. W. Theobald. The five species which have been
determined are all found at the present day in the same region. Several of the
shells still retain distinct traces of their colouration.

Helicidce. — Of Helix (including subgenera) there are three species, namely II.
( Ariophanta ) cysis , Benson, II. ( Hemiplecta ) vitellina, Pfeiffer, and II. ( I o tula) indica,
Pfeiffer; while Bulimus is represented by B. ( Cylindrus ) insularis,1 Ehr. Specimens
of the first three species have been obtained from beds C and Ca in the Cathedral,
while the one specimen of the last is from the Charnel-House.

Cyclostomidce. — This family is represented by Cydopliorus involutus (Mull.), which
appears commoner than any of the other species. Mr. Theobald remarks that the
Karnul shells are not of the Ceylon type, but agree with the variety now found
living at Midnapur, in Bengal.2 Some of the fossil specimens were obtained from
the Charnel-house and others from bed Ca in the Cathedral cave.

ADDENDUM.

The following specimen was overlooked until the text was in type.

Herpestes nipalensis, Gray.

Mandible. — The imperfect right mandibular ramus of a small species of Herpestes

from bed Cb in the Cathedral represented in
the accompanying woodcut may be referred
to the present species. The specimen shows
the broken canine, the complete mTI, and the
alveoli of the other cheek-teeth. The speci-
men agrees precisely with the jaw of the
existing race ; the large size of the alveolus
of pm. 1 distinguishing it from the allied
African H. gracilis , in which this tooth is either wanting, or of much smaller size
than in H. nipalensis.

Distribution. — At the present day the species occurs throughout the outer Hima-
laya, and also in the plains near the hills from the Punjab to Bengal, as well as
in Assam, Burma, and the Malay Peninsula ; the extension of its range during the
pleistocene into southern India is paralleled by that of Rhinoceros unicornis.

Fig. 15. Herpestes nipalensis. The imperfect right
ramus of the mandible ; from the Cathedral cave
(bed Cb). Indian Museum (No. F. 324).

1 Syn. Bulimus pullus, Gray.

2 Mr. Theobald says that the shell figured in Hanley and Theobald’s “ Conchologia Indica,” pi. II. fig. 8 as from
Midnapur does not represent the form obtained there by himself, which is the same as the fossil one.

58 INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

LIST OF THE FAUNA.

The following list comprises all the forms described in the preceding pages.

MAMMALIA.

Primates. — Semnopithecus entellus ( Dufresne ).
Cynocephalus, sp.

Carnivora. — Felis tigris ( or ? leo) Linn.

? pardus, Linn.
chaus, Giildenst.
rubiginosa, Geoffr.

Hyagna crocuta (. Erxl .).

Viverra karnuliensis, nobis.
Prionodon (?), sp.

Herpestes griseus, Desm.

fuscus, Waterh.
nipalensis, Gray.

Ursus labiatus, Blainv.
Insectivora. — Sorex, sp.

Chiroptera. — Taphozous saccolagmus, Temm.

Phyllorhina diadema {Geoffr.).
Rodentia. — Sciurus macrurus, Hardw.

Gerbillus indicus {Hardw.).
Nesokia bandicoota {Bech.).

kok, Gray.

Mus mettada (Gray).
platythrix, Sykes,
sp. var.

Rodentia. — Golunda ellioti, Gray.

Hystrix crassidens, nobis.

Atherura karnuliensis, neb is.

Lepus {cf. nigricollis, F. Cuv.).
Ungulata. — Equus asinus, Linn,
sp. a.

Rhinoceros karnuliensis, nobis.
Bos or Bubalus, sp.

Boselaphus tragocamelus {Pall.).
Genus non. det.

Gazella bennetti {Sykes).

Antilope cervicapra {Linn.).
Tetraceros quadricornis (Blaitivi).
Cervus aristotelis, Cuv.
axis, Erxl.

? Cervulus muntjac ( Zimm .).
Tragulus {cf. meminna \_Erxli]).
Sus cristatus, Wagner.
karnuliensis, nobis.

Edentata. — Manis gigantea, Llliger.

AYES.

Accipitres. — Neophron perenopterus {Linn.).
? Milvus or Circus, sp.

Ketupa ceylonensis {Gmelin).
Bubo coromandus {Lathi).

Galling. — Francolinus pictus (fard.and Selby).

pondicerianus (Gmelin).
Alectorides. — Grus (cf. communis, Becks/.).
Herodiones. — Ibis melanocephala ( Lath. ).

REPTILIA.

Crocodilia. — Crocodilus, sp. Ophidia. — Naia tripudians (Merri).

Lacertilia. — Varanus dracsena (Shaw.). Ptyas mucosus (Linn.).

Ophidia. — Python molurus (Linn.).

AMPHIBIA.

Batrachia. — Bufo (cf. melanostictus, Shneid.).

MOLLUSCA.

Gastropoda. — Helix cysis, Rens. Gastropoda. — Bulimus insularis, Ehr.

vitellina, Pff. Cyclophorus involutus (Mull.).

indica, Pff.

PLATE YII.

Primates and Carnivora.

Fig. 1.

„ 2,3.

„ 4.

,, 5, 5 a.

„ 6.

„ 7.

„ 8.

„ 9.

„ 10, 10a.
„ 11.

Semnopithecus entellus (Dufresne). Part of the right half of the palate of a female ;
from the Charnel-House cave (bed A). No. F. 201a. — Page 28.

Semnopithecus entellus (Dufresne). Part of the right maxilla and mandibular ramus ;
from the Charnel-House cave (bed M). No. F. 201. — Page 28.

Semnopithecus entellus (Dufresne). The right calcaneum ; from the Cathedral cave
(bed C ). No. F. 2015. Page 28.

Cynocephalus, sp. The second left lower true molar ; from the Charnel-House cave (bed
M). No. F. 202. Fig. 5a exhibits the outer aspect. — Page 28.

Viverra karnuliensis, Lyd. Part of the left ramus of the mandible ; from the Charnel-
House cave (bed L ). No. F. 227. — Page 31.

Herpestes griseus, Desm. The cranium ; from the Cathedral cave (bed Ch, but probably
of recent origin). No. F. 230. — Page 32.

Herpestes griseus, Desm. The left ramus of the mandible associated with the preceding
specimen.

Herpestes griseus, Desm. The left humerus ; from the Cathedral cave (bed Ca ). No
F. 230 a.— Page 33.

Herpestes fuscus, Waterh. The left ramus of the mandible ; from the Cathedral cave
(bed Cd). No. F. 231. — Page 33.

Herpestes fuscus, Waterh. The left humerus ; from the Cathedral cave (bed Ca). No.
F. 231a. — Page 33.

„ 12. (?) Prionodon, sp. The imperfect left humerus; from the Cathedral cave (bed Cb). No. F.
2295.— Page 32.

,, 13, 13a. Hyjena crocuta (Erxl.). The left lower carnassial, from the inner (13) and outer (13a)
aspects ; from the Cathedral cave (bed Cb). No. F. 222. — Page 30.

,, 11, 14a. Felis rubiginosa, Geoffr. The imperfect left maxilla; from the Cathedral cave (bed Ca).
No. F. 229. — Page 30.

,, 15. Felis rubiginosa, Geoffr. The right ramus of the mandible; from the Cathedral cave

(bed Cb). No. F. 229. — Page 30.

„ 16. Felis rubiginosa, Geoffr. The right humerus ; from the Cathedral cave (bed Cc). No.

F. 229 a.— Page 30.

„ 17. Felis chaus. Guldens. The left ramus of the mandible; from the Cathedral cave (bed

Ca). No. F. 228. — Page 29.

,, 18. Felis chaus, Guldens. The left upper canine, from the inner aspect; from one of the

Karnul caves. No. F. 228a. — Page 30.

,, 19. (?) Felis pardus, Linn. A first phalangeal ; from the Cathedral cave (bed Cc). No. F. 225.
Page 29.

,, 20. Felis Tigris (or ? leo), Linn. The first phalangeal of the third digit of the pes ; from the
Purgatory cave. No. F. 224a. — Page 29.

,, 21. Ursus labiatus, Blainv. The distal extremity of the right humerus; from the Chapter-
House cave (bed Ab). No. F. 226. — Page 33.

* All the specimens are in the Indian Museum, and are represented of the natural size. The
humeri are all viewed from the palmar aspect, and the lateral views of the mandible are from the outer
aspect, c.f. entepicondylar foramen ; ec. c. ectepicondyle ; en. c. entepicondyle,

Geol. Surv. of India.

TERTIARY

VERTEBRATA.

Vol.IV. PI. VII

&.HigViley del et litVi .

West, Newman & C? imp.

’ '

»

PLATE VIII.

Chiropteka, Rodentia, and Edentata.

Fig. 1. ? Nesokia bandicoota, Bech. The right femur; from the Charnel-House. No. F. 344. —

Page 35.

,, 2. Ibid. The right humerus ; from the Cathedral cave (bed Cc). No. F. 345. — Page 35.

,, 3. Ibid. The imperfect right femur ; from the Cathedral (bed Cb). No. F. 343. — Page Zb.

,, 4. ? Nesokia kok, Gray. The imperfect right femur ; from the Charnel-House. No. F. 346.

— Page 36.

,, 5. Sciurus macrurus, Hardw. Part of left ramus of the mandible ; from the Cathedral (bed
Ca ). No. F. 217. — Page 34.

,, 6. Hystrix crassidens, Lyd. The right humerus ; from the Cathedral (bed Cd). No. F. 221.
— Page 37.

,, 7. Lepus {cf. nigricollis, F. Cuv.) The left humerus (reversed) ; from Purgatory cave. No. F.
218a.— Page 39.

„ 8, 8a. Manis gigantea, Illig. The terminal phalangeal of the middle digit of the right manus ;
from the Cathedral cave (bed Cc). No. F. 232. — Page 50.

„ 9, 9a. ? Phyllorhina diadema (Geoffr.). The right humerus ; from the Cathedral (bed Cc). No.
F. 206a. — Page 34.

,, 10, 10a. Taphozous saccoljemus, Temm. The imperfect cranium and right upper cheek-dentition;
from the Cathedral cave (bed Ci.) No. F. 209. — Page 34.

,, 11, 11a. Phyllorhina diadema (Geoffr.). The imperfect cranium and right upper cheek-dentition ;
from the Cathedral cave (bed CA.) No. F. 206. — Page 34.

„ 12. Ibid. The left ramus of the mandible ; from the Cathedral (bed Cf). No. F. 207. — Page 34.

„ 13, 13a. Nesokia kok, Gray. The imperfect cranium; from the Cathedral (bed Cb). No. F. 213.
— Page 35.

„ 14, 1 la. Ibid. The right ramus of the mandible and teeth (reversed) ; from the Cathedral (bed CA).
No. F. 214. — Page 36.

,, 15, 15a. Nesokia bandicoota, Bech. The imperfect left ramus of the mandible and lower cheek-
teeth ; from the Charnel-House cave (bed H). No. F. 216. — Page Zb.

,, 16. Hystrix crassidens, Lyd. Part of the left ramus of the mandible; from the Cathedral
(bed C). No. F. 219. — Page 37.

,, 17, 17a. Ibid. The left ramus of the mandible and cheek-teeth ; from the Cathedral (bed Cd). No.
F. 219 a— Page 37.

,, 18, 18a. Ibid. The right upper premolar ; from the Cathedral (bed Cd). No. F. 220. — Page 37.

,, 19. Ibid. Anterior aspect of the upper incisors ; from the Cathedral (bed Cd). No. F. 220b. —
Page 37.

,, 20. Ibid. Outer aspect of left upper incisor ; from the Cathedral (bed Cd). No. F. 220a. — Page 37.

,, 21. Ibid. Part of the lower incisors, from the anterior aspect ; from the Cathedral. No. F.
220c. — Page 37.

„ 22, 22a. Atherura karnuliensis, Lyd. Right upper incisor ; from the Cathedral (bed Cd). No.
F. 221 a.— Page 38.

„ 23, 233. Ibid. Right lower incisor; from the Cathedral (bed Cd). — Page 38.

* All the specimens are in the Indian Museum, and, with the exception of figs. 5a, 10a, 11a, 14a, 15a
(which are f) are represented of the natural size. The humeri, except fig. 9a, are viewed from the
palmar aspect, and the lateral views of the mandibles are from the outer aspect, dr. deltoid ridge ; ec. c.
ectepicondyle ; en. c. entepicondyle.

ec.C'

Geol. Surv o£ India..

TERTIARY VERTEBRATA.

Vol. IV. PI VIII.

C.Berjeau del. et lith

We st, Newman & Co. imp

I

/

'

PLATE IX.

Artiodactyla — S uidce, and Perissodactyla — Equidce.

Fig. 1. Sus cristatus, Wagner. The third left lower true molar; from the Cathedral cave (bed
Cc) No. F. 266. — Page 47.

„ 2. Sus cristatus, Wagner. The third left lower true molar ; from the Cathedral cave (bed

Cc). No. F. 266 a. — Page 47.

,, 3. Sus karnuliensis, Lyd. The third left lower true molar ; from the Cathedral cave (bed

Cd). No. F. 268. — Page 49.

,, 4. Sus cristatus, Wagner. The third left lower true molar; from the Cathedral cave (bed Cb).

No. F. 267. — Page 4 7.

,, 5. Sus cristatus, Wagner. Part of the right ramus of the mandible ; from a turbary in

Madras. No. F. 115. — Page 47.

,, t>. Sus karnuliensis, Lyd. The third right npper true molar; from the Cathedral cave (bed

Cd). No. F. 260. — Page 49.

,, 7. Sus cristatus, Wagner. The third right upper true molar; from the Cathedral cave

(bed Cf). No. F. 261 .—Page 48.

,, 8. Sus karnuliensis, Lyd. The third right lower true molar ; from the Cathedral cave (bed

Cb). No. F. 267a. — Page 49.

,, 9. Sus karnuliensis, Lyd. The third left upper true molar; from the Cathedral cave (bed

Cd). No. F. 260a. — Page 49.

,, 10. Sus cristatus, Wagner. Part of the left lower canine; from the Cathedral cave (bed Ca).

No. F. 271. — Page 48.

„ 11. Equus asinus, Linn. The third right lower true molar ; from the Cathedral cave (bed Cb).

No. F. 256. — Page 39.

,, 12. Equus asinus, Linn. The first left lower true molar; from the Cathedral cave (bed Ca).

No. F. 255. — Page 39.

,, 13. Equus, sp. The third right lower true molar ; from the Cathedral cave (bed Cb). No. F 248.

— Page 40.

,, 14. Equus, sp. A left upper cheek-tooth; from the Cathedral cave (bed Cc). No. F. 246. —

Page 39.

,, 15. Equus asinus, Linn. The third right upper true molar ; from the Cathedral cave (bed Cf).

No. F. 254. — Page 39.

,, 16. Equus, sp. The third left upper true molar; from the Cathedral cave (bed Cd). No. F.

253.— Page 39.

,, 17. Equus, sp. A left upper cheek-tooth ; from the Cathedral cave (bed Cc). No. F. 246a. —

Page 39.

* All the specimens are in the Indian Museum, and are represented of the natural size. Excepting
fig. 10, which is from the inner aspect, all the specimens are viewed from the oral aspect. In the molars
of Sus the letters a to f indicate the homologous elements of the upper and lower teeth ; in those of
Equus e indicates the antcro-internal pillar.

Geol. Surv. of India

TERTIARY VERTEBRATA.

Vol.IV PI. IX.

Berjeau. 8t Highley del etlith. We®1. Newman t, Co. imp .

.

■

■

.

-

'

PLATE X.

Perissodactyla. — Rhinocerotidce.

/

Figs. 1, la, 13. Rhinoceros karnuliensis, Lyd. The second and third left upper true molars, in a
middle stage of wear ; from the Cathedral cave (bed Cc). No. F. 233. Fig. 1 shows the
inner and grinding surfaces of the two teeth, fig. 1 a the outer surface of 2, and fig. 13
that of m- 3. — Page 41.

,, 2. Rhinoceros karnuliensis, Lyd. The inner half of the well-worn third right upper

premolar ; from the Cathedral cave (bed Cc). No. F. 236. — Page 42.

., 3, 3 a. Rhinoceros unicornis, Linn. The third right upper true molar, in an early stage of

wear; from a turbary in Madras. No. F. 114. Fig. 3 shows the inner and grinding

surfaces, and fig. 3 a the outer surface. — Page 42.

,, 4, 4<z. Rhinoceros karnuliensis, Lyd. Part of the symphysis and left ramus of the mandible ;

from the Cathedral cave (bed Cd ). No. F. 237. Fig. 4 is from the oral, and fig. 4a from
the external aspect. — Page 40.

* All the specimens are in the Indian Museum, and with the exception of figs. 4, 4a (which are £)
are represented of the natural size, a, anterior collis ; 3, posterior collis ; c, second costa ; d, first costa ;
e, crochet ; for, mentary foramen ; g, entrance of median valley ; i, posterior valley.

Geol. Surv. of India.

TERTIARY VERTEBRATA.

Vol. IV. PL X.

West, Newman &. C? imp

Berjeau. Sc.Higliley' del. etlrtK.

■ K%

V

'

«r •

m - • ■ •

PLATE XI.

Artiodactyla. — Bovidce and Cervidce.

Fig. 1, la. Cervus axis, Erxl. The third right upper true molar; from the Cathedral cave (bed Cc ).
No. F. 306. — Page 46.

,, 2, 2 a. Antelope, genus non. det. The second right upper true molar ; from the Cathedral cave

(bed Cb). No. F. 288. — Page 45.

,, 3, Cervus axis, Erxl. The third left upper true molar ; from the Cathedral cave (bed Cb).

No. F. 306a. — Page 46.

,, 4. Cervus axis, Erxl. The first or second right lower true molar ; from the Cathedral cave

(bed Cc). No. F. 3065. — Page 47.

,, 5,5 a. Cervus aristotelis, Cuv. The first right upper true molar ; from the Cathedral cave (bed

Cf). No. F. 302. — Page 46.

„ 6,6a. Cervus aristotelis, Cuv. The second left upper true molar; from the Cathedral cave

(bed Cd). No. F. 302a. Page 46.

„ 7. Boselaphus tragocamelus (Pall.). The first right upper true molar ; from the Cathedral

cave (bed Cd). No. F. 282. — Paqe 44.

., 8. Boselaphus tragocamelus (Pall.). The second right upper true molar ; from the Cathedral

cave (bed Ca). No. F. 280. Page 44.

,, 9. Boselaphus tragocamelus (Pall.). The first right upper true molar ; from the Cathedral cave

(bed Cb). No. F. 281. — Page 44.

,, 10. Boselaphus tragocamelus (Pall.). The second left lower true molar : from the Chapter-
House cave (bed Aa.). No. F. 284. — Page 44.

„ 11,11a. (?) Cervulus muntjac (Zimm.). The second right upper true molar ; from the Purgatory cave.
No. F. 308. — Page 47.

,, 12,12a. Tetraceros quadricornis (Blainv.). The left ramus of the mandible ; from the Cathedral
cave (bed Cd). No. F. 277. — Page 46.

„ 13, 13a. Antii.ope cervicapra (Linn.). The second left upper true molar; from the Cathedral
cave (bed Cb). No. F. 287. — Page 45.

,, 14, 14a. Boselaphus tragocamelus (Pall.). The right ramus of the mandible of a young individual ;
from the Karnul caves. No. F. 286. — Page 45,

,, 15, 15a. Gazella hennetti (Sykes). The left ramus of the mandible; from the Cathedral cave
(bed Ca). No. F. 278. — Page 45.

* All the specimens are in the Indian Museum, and are represented of the natural size. In the text
figs. 2, 5 and 6 are referred to the wrong side of the jaw.

Geol. Surv. of India.

TERTIARY

VERTEBRATA.

Vol. IV Pl. XI.

13a.

la..

14 a.

12 a.

15 a.

11a.

7/1 tn.4

3

ptn

try. 3

C.Berjeau del et lith.

West, Newman & C° imp .

* ■

MEMOIRS

OF

THE GEOLOGICAL SURVEY OF INDIA.

1}

afeoutologiit Jrnfe,

BEING

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

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

Ser. X.

INDIAN TERTIARY AND POST-TERTIARY VERTEERATA.

Yol. IV.

Part III. EOCENE CHELONIA PROM THE SALT-RANGE
by R. LYDEKKER, B.A., F.G.

WITH 2 PLATES.

CALCUTTA:

BOLD AT THK

GEOLOGICAL SURVEY OFFICE, AND BY ALL BOGKSELLER8:
LONDON : TRUBNER & CO.

MDCCCLXXXVII.

BRIXTEB BT TUB SWBBKINIBKPJKT OB OOBBRKMsifl BBIKT1K6, 1KPH, 8, BiSTIMSt BTftKET, OAlCSTTi.

SS?*

• i

- X I

'

4 *

*

i

»

<

£.C>

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

EOCENE CHELONIANS FROM THE SALT-RANGE.

BY R. LYDEKKER, B.A., E.G.S., &c.

(WITH PLATES XII AND XIII.)

INTRODUCTORY.

The two specimens, forming the subject of this memoir, were obtained in 1886
by Dr. H. Warth, by whom they were presented to the Indian Museum. They were
both obtained at Nila (lat. 32° 29' 30"; long. 73°j in the Punjab Salt-Range; while
at least the example of Hemichelys (Plate XII) came from a bed situated 10 feet
below the coal-band of Nila.1 Erom the same bed Dr. Warth also obtained certain
other remains, which, according to a manuscript note by Dr. Waagen, with which the
writer has been favoured, comprise teeth of Lamina, Otodus , and Capitodus, together
with certain cephalopoda allied to Belemnites ; and Dr. Waagen is of opinion that
the bed in question belongs to the highest horizon of the “ Cardita beaumonti
group,” which is now regarded as a passage-bed between the cretaceous and the
eocene, although in being more nearly allied by its fauna with the latter than with
the former is provisionally classed as lowest eocene ; 2 and may not improbably
correspond homotaxially to the Cernayisian stage of Reims, and the Puerco group
of the United States.

The fossils mentioned by Dr. Waagen indicate that the bed in which they were
found is either of marine or estuarine origin ; and this is confirmed by the Chelonians,

1 See Wynne : ‘Mem. Geol. Surv. Ind.’ Vol. XIV, p. 168 (1878).

3 See Blanford : ‘ Mem. Geol. Surv. Ind.’ Vol. XX, pp. 4, 45, 46 <1883).

60

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

oneof which is covered with the ‘ spat ’ of oysters, while sharks’ teeth are embedded
in the matrix of the other. The affinities of the Chelonians show, however, that
their habitat could only have been estuarine, since they belong to essentially
fresh-water families. Both specimens when placed in the hands of Mr. Barlow, of
the British Museum, to be ‘ developed * were in a much broken condition, and from
being traversed by numerous veins of calcspar, along which they readily split, it
required all his skill to render them fit for description.

It should he mentioned that the only fossil Indian Chelonian from below the
level of the Siwaliks (where all the forms are more or less closely allied to those
now inhabiting India, and adjacent regions) which has been hitherto described is
Platemys leithi 1 (Carter) from the intertrappeans (lower eocene) of Bombay, which
is a small species with a shell of about five inches in length. Platemys, it need
hardly he observed, is a member of the sub-order Pleurodira, now characteristic of the
Southern Hemisphere, and is itself confined at the present day to South America. A
chelonian has been recorded by Sir R. Owen1 2 from the lower eocene of England under
the name of Platemys bowerbanki, hut Professor Rutimeyer3 has shown that this
generic determination is incorrect ; and that the presence of an incomplete meso-
plastral element in the plastron indicates affinity with the South American genera
Peltocephalus or Podocnemis, although the imperfect preservation of the type speci-
men renders its precise determination a matter of considerable difficulty. The same
authority also shows that the specimen figured by Owen 4 5 under the name of Emys
Icevis, Bell, is nothing more than the young of the above-mentioned species ; and it
may he added that the length of the band uniting the carapace with the plastron in
this specimen is indicative of affinity with Podocnemis rather, than Peltocephalus ,
and it has been provisionally referred to the former genus by the present writer and
Mr. G. A. Boulenger;6 that genus being represented by another species in the
same formation. A plastron, said to be from the London clay, was also referred by
Sir R. Owen8 to Platemys, under the name of P. bullocki ; hut this specimen is
really from the Purbeck, and belongs to the genus Pleurosternum.7

It will suffice to add in the way of introduction that in the memoir on the
Siwalik Chelonia published in Yol. Ill of the present work, the term * scute ’ is
applied to the bony framework of the Chelonian shell, and ‘ plate * to its horny
epidermal covering ; but, as this has been thought liable to lead to confusion the
terms * hone ’ and ‘ shield ’ are employed here in these senses.

1 Originally described as Testudo, but referred by Gray to Hydraspis, which is now generally included in
Platemys ; see ‘ Rec. Geol. Surv. Ind.’ Vol. XX, p. 66 (1887).

2 Monograph of Reptilia of London Clay — Chelonia I, PI. XXIII (1849).

* * Ver. Nat. Ges. Basel,’ Vol. VI, Art. 1, pp. 121, 128 (1878).

4 Op. cit., pi. XXII.

5 Geological Magazine, June 1887.

6 Ibid., pi. XXI.

7 See Lydekker and Boulenger, op. cit. Cope, ‘ Tertiary Vertebrata of the West ’ (Rep. U. S. Geol. Surv. Ters.
Vol. Ill) book 1, p. Ill (1884), makes Pleurosternum the type of a distinct family of Cryptodira, and erroneously

states that there is no intergular shield.

EOCENE CHELONIANS FROM THE SALT-RANGE.

61

Sub-order Pleurodira.

Characters. — The sub-order is characterised1 by having the carapace and
plastron fully ossified and united together ; by the anchy-
losis of the pelvis to both plastron and carapace ; by the
absence of at least some of the vertebral bones;2 and by
the presence of only a single heart-shaped supra-pygal
bone above the pygal, as is shown in the accompanying
woodcut. Other characters are found in respect of the
head and neck, which need not be mentioned on this
occasion. When horny epidermal shields are present, an
inter gular shield is developed on the plastron ; and a
mesoplastral bone may be present. As additional characters
of at least many genera, it may be observed that the nuchal
bone (in the Chelydidce) is smaller than the first vertebral
shield, and that the suture between the pygal and supra-
pygal bones is situated above the suture separating the
pygal shields from the last vertebral shield (as is shown in
the wood-cut) ; the reverse of these features obtaining very
generally in the Cryptodira. 3 The sub-order may be
divided into two families according to the presence or
absence of homy shields on the shell.

The right half of the carapace of
Sternothcerus sub nip er. Gray;
from Madagascar.

Eamily Carettochelydid^:.

History. — This family was founded by Mr. Boulenger 4 on Carettochelys
insculptus, Ramsay,6 from the Ely River, New Guinea.

Characters. — The distinctive feature of the family is the absence of epidermal
horny shields on both the carapace and plastron.

In Carettochelys the limbs are paddle-shaped, the anterior being much elongated;
only the first and second digits are furnished with claws ; the plastron is apparently
composed of only the normal nine elements, the mesoplastron (intraplastron) being
absent ; the vertebral bones are of very small size, and are separated from one another
by the meeting of the costals in the middle line, and the nuchal bone is greatly
expanded laterally.

Genus Hemichelys, nobis.*

Characters. — Vertebral bones in contact, and not separated by the costals ; a
small mesoplastron apparently present.

1 See Riitimeyer : ‘Verb. Nat. Ges. Basel,’ Vol. VI, Art. 1, pp. 21—33 (1878).

2 The hinder vertebral bones are absent among the Cryptodira in Cinosternum and Dermatemys.

3 Compare the figures in pis. XXI and XXIII of the preceding volume of this work.

4 Ann. Mag. Nat. Hist. 1887, pp. 170—172.

5 Proc. Linn. Soc., New South Wales, Vol. I, p. 158, pis. Ill — VI (1886).

6 Rec. Geol. Surv. Ind., Vol. XX, p. 66 (1887).

62

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA.

Hemichelys warthi, nobis.1

Characters. — Carapace depressed, expanded posteriorly, with five vertebral
bones, and its surface comparatively smooth ; surface of plastron pitted ; length of
carapace about 28 inches.

Description. — The type specimen is represented in plate XII, and comprises
the greater portion of the carapace and plastron. The former is tolerably perfect
posteriorly, but has lost the greater portion of the lateral marginals, and anteriorly
the nuchal bone, the marginals, and portions of the first costals; the figure is
restored from the recent Podocnemis pipiti (Gray).2 The plastron has lost both
extremities, and its junction with the carapace is also injured.

When entire, the carapace must have measured about 28 inches in length. The
Pleurodirian affinities of the specimen are conclusively shown by the general contour
of the carapace, which agrees very closely with that of Podocnemis pipiti, by the
presence of only five vertebral bones, and of a single supra-pygal, as well as by the
elongated form of the pygal and adjacent marginal bones. The surface of the
carapace is moderately smooth, and shows no trace of the marks of horny shields ;
the surface of the plastron (pi. XII, figs. 2, 3) is marked with small pits, and was
evidently covered merely by skin. The only one of the plastral sutures that can be
detected is that between the hyo- and hypo-plastron ; externally this suture divides,
apparently for the reception of a small incomplete mesoplastron.3 Whether the
extremity of the xiphiplastron was forked as in the restoration, or whether it was
entire as in Carettochelys ,4 cannot of course be determined.

Affinities. — The absence of horny shields 5 from this interesting form conclusively
separates it from all members of the Chelydidce, and affiliates it with Carettochelys.
The latter is, however, widely distinguished by the characters of the vertebral bones 6
which form minute elongated hexagons, separated from one another by portions of
the costals, which consequently meet in the middle line ; the present form is also
distinguished by the pitted sculpture on the plastron, and (apparently) the more deeply
cut axillary notch ; while it is highly probable that the nuchal bone was of the form
represented in the restoration, and was not of the curiously expanded shape
occurring in the New Guinea genus,7 which resembles that of the Trionychidce.

These differences leave no doubt as to the generic distinction of the present
form from Carettochelys , but the common feature of the absence of horny shields
apparently indicates such relationship as to justify at least its provisional reference
to the same family. While, however, Carettochelys is an extremely aberrant form

1 Loc. cit.

2 See ‘Proc. Zool. Soc.,’ 1871, pp. 747-48.

3 Compare Owen’s figure ( op. cit., pi. XXIII) of the so-called Platemys bowerbanki.

* Ramsay, op. cit., pi. III.

s In the preliminary notice, I thought it probable that these plates were present on the carapace.

6 Ramsay, op. cit., pi. IV.

7 Ibid.

EOCENE CHELONIANS FROM THE SALT-RANGE.

63

not showing any close relationship to the Chelydidce, Hemichelys appears to be
much more nearly allied to that family, and may be the representative of a still
earlier type from which the latter has been derived. The preservation of an appa-
rently still more primitive type in the Australasian region is an instance analogous
to that of the survival of Ceratodus and other primitive forms in the same area.

Family Chelydidce.

Characters. — Shell covered with horny epidermal shields, and an intergular
shield present on the plastron.

There is considerable variation as to the number of vertebral bones in the
carapace of the different genera. Thus, in the Australian genera Chelodina, Chelymys,
Euchelymys, and Elseya, vertebrals are entirely absent, the only azygos bones being
the nuchal, the supra-pygal, and the pygal.1 In Platemys there are normally six
such bones, but in P. raniceps the first three are wanting,2 and in a specimen of
P. planiceps in the British Museum there are none. In the eocene form figured by
Owen3 under the name of Platemys bowerbanki there are seven such bones ; the
same number being present in the existing genera Podocnemis, Peltacephalus, and
Sternothcerus 4 (woodcut). An incomplete mesoplastral element is present in Podo-
cnemis, Peltocephalus, Pelomedusa, Pentonyx, and Sternothcerus ,6 while in Pleuro -
sternum this element is complete. Peltocephalus is distinguished from all the other
genera by the fusion of the two pygals into a single shield, as in Testudo. The nuchal
shield is wanting in Sternothcerus, Pelomedusa, Podocnemis, and Peltocephalus.

Some writers regard the presence of a mesoplastral bone as a family character,
and accordingly apply the name Pelomedusidce to the group presenting this feature.

Genus Podocnemis, Wagler.6

Characters. — Shell thick, with a long plastro- costal symphysis ; nuchal shield
absent; plastron solid; vertebral bones seven in number, the first five being
elongated, and the seventh small, rhomboidal, and wedged in between the sixth
and seventh costals.7

Podocnemis indica, n. sp. nobis.

Characters. — Carapace oval, tectiform, not keeled, and narrowed posteriorly ;
the first vertebral shield wide, the second and third elongate, total length about
35 inches.

1 Riitimeyer, op. tit., p. 24.

5 Ibid.

3 Op. tit., pi. XXIII.

* Riitimeyer, op. tit., p. 24.

6 Riitimeyer, op. tit., p. 23.
6Syst. Amphibien, p. 135 (1830).

7 Riitimeyer, op. tit., p. 24.

64

INDIAN TERTIARY AND POST-TERTIARY YERTEBRATA.

Description. — The specimen on which this species is founded is represented in
plate XIII. The greater portion of the carapace remains, but a large part of the
right side and the whole of the posterior extremity are wanting ; the vertebral and
costal bones are distinctly defined, as are the boundaries of the anterior marginal and
first three vertebral horny shields. Only small fragments of the plastron are pre-
served, although sufficient remains to show that this was united by bone with the cara-
pace, and that the band of union was a comparatively long one. When complete the
total length of the carapace must have been about thirty-five inches.

In the epidermal shields the most noticeable point is the absence of the nuchal ;
the first vertebral is wider transversely than it is long ; while the second and third are
imperfectly hexagonal, and are elongated antero -posteriorly. In the underlying
bones the nuchal is of comparatively small size, and lies well within the limits of
the first vertebral shield. There are seven vertebral bones, of which the first five
are elongated, while the seventh is very small, and wedged in between the sixth and
seventh costals. Behind the last vertebral the seventh and eighth costals unite in
the median line; behind which there is the commencement of the supra -pygal
bone, of which the hinder half is wanting. The contour of the carapace is sub-oval,
the posterior portion being the narrowest; and the median line is much elevated,
with the costals sloping rapidly away on either side : there is no keel on any of the
vertebral bones.

Affinities. — It is quite clear from the general structure of the specimen that the
only families to which it could possibly belong are the typical Emydidce or the Chely-
didce ; and it is extremely unfortunate that the decisive evidence which would be
afforded by the plastron is not available. The characters of the carapace are, how-
ever, such as to indicate clearly that it is to the latter family to which it should be
referred. In the first place, the nuchal bone is of the relatively small size charac-
teristic of the Chelydidce ; and in the second, the vertebral bones agree precisely in
number and character with those of the Podocnemis group ; while there is the same
interval between the last vertebral and the supra-pygal bones, and the latter was
evidently a single one. If indeed the figure of the Indian specimen be compared
with that of the so-called Platemys bowerbanki given by Owen 1 (which, as we have
already mentioned, is referred to Podocnemis), it will be seen that the two evidently
belong to the same group, and might perfectly well be generically identical,
although their specific distinctness is indicated by the wider vertebral epidermal
shields of the English form. If again the figure of the present specimen be compared
with that of Podocnemis dumeriliana given by Gray,2 an equally striking resem-
blance will be observed in the general contour of the carapace, as well as in the
absence of a nuchal shield ; and there accordingly appears no doubt but that the
specimen belongs to the Chelydidce. With regard to its generic position, the presence
of seven vertebral bones differentiates it from all the existing Australian genera;

* Op. cit., pi. XXIII.

3 Cat. Sbield-Keptiles in British Museum, pt. I, pi. XXVIII (1855).

EOCENE CHELONIANS EROM THE SALT-RANGE.

65

while the absence of a nuchal shield affiliates it with the existing Sternothcerus,
Pelomedusa, Podocnemis, and Peltocephalus. From the first (woodcut, p. 61) it is
readily distinguished by the much longer vertebral hones. By the unfortunate
absence of the pygal region and the greater part of the plastron, the best distinctive
characters of several of these genera are not available. Podocnemis is, however,
distinguished from the others by the great length of the band uniting the plastron
with the carapace, and its thick shell ; in both of which respects the fossil agrees
with it. Podocnemis contains, moreover, the largest living member of the family ;
and both the young of P. expansa and the adult of P. dumeriliana are characterised
by their tectiform carapace; in which respect they also agree with the fossil.
The close resemblance in the general contour of the fossil carapace and that of the
latter species has been already alluded to, and since I can find no points of generic
distinction between the former and the existing species of Podocnemis, I think
the fossil may be at least provisionally referred to that genus 1 with the name of
P. indica.

Podocnemis, it need scarcely be observed, is at the present day confined to
South America ; but its occurrence, together with that of Platemys, in the lower
eocene of Western India, and also in the corresponding formation of England,
indicates that the original home of this group of fresh-water Chelonians was in the
northern hemisphere of the Old World, from which they have been driven in all
probability by the competition of the Bmydidce to find a home in the southern half
of America.

1 1 have not been able to compare the fossil with a recent Podornemis in which the horny shields have been stripped
from the carapace.

PLATE XII.

Chelonia.

Hemichelys Warthi, Lydekker ; fig. 1 the carapace, fig. 2 the left half of the plastron, fig. 3
sculpture of the plastron ; from the lower eocene of Nila, Salt-Range. Indian Museum.
Figs. 1, 2, i nat. size ; fig. 3 nat. size. N. nuchal bone. Page 62.

G M Woodward del. et lith

PLATE XIII.

Chelonia.

Podocnemis indica, Lydekker ; the carapace ; from the lower eocene of Nila, Salt-Range. Indian
Museum. ? nat. size. Page 63.

Ge ol . Surv. oi' India- .

TERTIARY

YERTEBBATA

Vol.IY.Pl.Xin.

G-.M .Woo<lwar*cL del. el lifch

West, Newmaji & C? imp.

.

*

Errata to Part II.

In list of Aves on pp. 51 and 58, before Ketupa , add Striges.

Page 57. — The Helix mentioned under the name of H. cysis turns out to be II. nicobarica ; while
that referred to as II. indica is more probably II. ligulata.

■

..

'

-

PALiEONTOLOGIA INDICA.

(In Royal 4 to.)

(See. I, III, V, VI, VIII.)— CRETACEOUS FAUNA OF SOUTHERN INDIA, by F. STOLICZKA, except
Vol I, Tt. I, by H. F. BLANFOIID.

Vol. I. The Cephalopoda (1861-65), pp. 216, pis. 94 (6 double).

II. The Gastropoda (1867-68), pp. xiii, 500, pis. 28.

III. The Peleeypoda (1870-71), pp. xxii, 537, pis. 50.

IV. The Brachiopoda, Ciliopoda, Eehinodermata, Corals, &c. (1872-73), pp. v, 202, pis. 29.

(See. II, XI, XII.)— THE FOSSIL FLORA OF THE GONDWANA SYSTEM, by O. FEISTMANTEL, except
Vol. I, Pt. 1, by T. OLDHAM and J. MORRIS.

Vol. I, pp. xviii, 233, pis. 72. 1863-79). Pt, l.R&jmahal Group, Rijmahal Hills. Pt. 2. Same continued. Pt. 3. Plants
from Golapilli. Pt. 4. Outliers on the Madras Coast.

II, pp. xli, 115, pis. 26. 1876-78. Pt. 1. Jurassic Flora of Kach. Pt. 2. Flora of the Jabalpur Group.

Ill, pp. xi. 64 x 149, pis. 80 (9 double) (I-XXXI x I A — XLVIIA). 1879-81. Pt. 1. The Flora of the Talchir-
Kurharbari beds. Pt. 2. The Flora of the Damtida and Panchet Divisions, Pt. 3. The same (concluded) .

I\, pt. I (1882). Fossil Flora of the South Rewah Gondwana basin, by O. Feistmaxtel, pp. 42, pis. 21.

(See. IX.)— JURASSIC FAUNA OF KACH.
ol. I, (1873-76). The Cephalopoda, by W. Waagen, pp. i, 247, pis. 60 (6 double).

(See. IV.)— INDIAN PRETERTIARY VERTEBRATA.

Vol. I, pp. vi, 137, pis. 26, 1865 — 85. Pt. 1 (1865). The Vertebrate Fossils from the Panchct Rocks, by T. H. Huxley.

Pt. 2. (1878). The Vertebrate Fossils of the Kota-Maleri Group, by Sib P. de M. Geey Egeeton and L. p.
Miall. Pt. 3 (1879). Reptilia and Batracliia by R. Lydekkee. Pt. 4 (1885). The Labyrintliodont from the
Bijori Group, by R. Lydekkee. Pt. 5 (1885). The Reptilia and Amphibia of the Maleri and Denwa groups,
by R. Lydekkee.

(See. X.)— INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA, by R. LYDEKKER, except Vol. I,
, Pt. 1, by R. B. FOOTE.

Vol. I, pp. xxx, 300, pis. 50. 1874-80. Pt. 1. Rhinoceros Deecanensis. Pt.2. Molar teeth and other Remains of

Mammalia. Pt. 3. Crania of Ruminants. Pt. 4. Supplement to pt. 3. Pt. 6. Siwalik and Narbada

Proboscidia.

II. pp. xv. 363, pis. 45. 1881-84. Pt. 1. Siwalik Rhinocerotkke. Pt. 2. Supplement to Siwalik and Narbada

Proboscidia. Pt. 3. Siwalik and Narbada Equidte. Pt. 4. Siwalik Camelopardalidse. Pt. 5. Siwalik
Selenodont Suina, etc. Pt. 6. Siwalik and Narbada Carnivora.

III. pp. xxiv, 264, plates 37. 1884-86. Pt. 1. Additional Siwalik Perissodactyla and Proboscidia. Pt. 2. Siwalik

and Narbada Bunodont Suina. Pt. 3. Rodents and New Ruminants from the Siwaliks, and Synopsis of

Mammalia. Pt. 4. Siwalik Birds. Pt. 5. Mastodon Teeth from Perim Island. Pt. 6. Siwalik and Narbada
Chelonia. Pt. 7. Siwalik Crocodilia, Lacertilia, and Ophidia. Pt. 8. Tertiary Fishes.

IV. pt. 1, pp. 18, pis. 6. Siwalik Mammalia— Supplement I.

(See. VII, XIV.) — TERTIARY AND UPPER CRETACEOUS FAUNA OF WESTERN INDIA.

Vol. I, pt. 1 (1871). Tertiary Crabs from Sind and Kutch, by F. Stoliczka, pp. 16, pis. 5. Pt. 1 (new 2) (1880). Sind
Fossil Corals and Alcyonaria, by P. Martin Duncan pp. 110, pis. 28. Pt. 3. The Fossil Echinoidea of Sind,
by P. Martin Duncan and W. Peecty Sladen. Fas. 1 (1882). The Cardita beaumonti beds, pp. 20, pis. 4.
Fas. 2 (1882). The Ranikot series, pp. 80, pis. 16. Fas. 3 (1884). The Khirtar series, pp. 146, pis. 18.
Fas. 4 (1884). The Nari series, pp. 26, pis. 5. Fas 5 (1885). The Gaj series, pp. 94, pis. 12. Pt. 4 (1883).
The Fossil Echinoidea of Kaehli and Kattywar, by P. Mabtin Duncan and W. Peecy Sladen, pp. 91, pis. 13.

(See. XIII.)— SALT-RANGE FOSSILS by WILLIAM WAAGEN.

Vol. I. Productus -Limestone Group : Pt. 1 (1879). Pisces, Cephalopoda, pp. 72, pis. 6. Pt. 2 (1880). Gastropoda and
supplement to pt. 1, pp. Ill (73-183), pis. 10 (vii-xvi). Pt. 3 (1881). Peleeypoda, pp. 144 (185-328) ; pis. 8
(xvii.-xxiv). Ft. 4 (1884). Brachiopoda, pp. 442 (329-770), pis. 62 (xxv-lxxxvi). Pt. 5 (1885). Bryozoa, Annelida,
Eehinodermata, pp. 64 (771-834), pis. 10 (lxxxvii-xcvi).

The price fixed for these publications is 4 annas (6 pence) per single plate.

To be had at the Geological Survey Office, Indian Museum, Calcutta ; or through any Bookseller. London : Triibner Sc Co.

GEOLOGICAL

OF INDIA

MEMOIRS

OF THE

SURVEY

(In Royal 8vo.)

Vol. I, pp. 309, 1859 [out of print).

,, II, pp. 341, 1859 [out of print).

,, III, pp. 438. Pt. I, (1803) [out of print). Pt. 2, 1864 (2 Rs.) : On tlie Sub -Himalayan Ranges between the

Ganges and Sutlej.

,, IV, pp. 450. Pt. 1, 1863 (2 Rs.) : Report on the Cretaceous Rocks of Trichinopoly District, Madras. Pt. 2, 1864

(2 Rs.) : On the Structure of the Districts of Trichinopoly, Salem; &c. Pt. 3, 1865 (1 Re.) : On the
Coal of Assam, &c.

>> V, pp. 354. Pt. 1. 1865 (3 Rs.) : Sections across N. W. Himalaya, from Sutlej to Indus.- — On the Gypsum of

Spiti. Pt. 2. 1866 (1 Re.) : on the Geology of Bombay. Pt. 3, 1866 (1 Re.) : On the Jherria Coal-field. —

Geological Observations in "Western Tibet.

,, VI, pp. 395. Pt. 1, 1867 (8 As.) : On the neighbourhood of Lynyon, &c., in Sind. — Geology of a Portion of
Cutch. Pt. 2, 1867 (2 Rs.) : Bokfiro Coal-field — R&mgarh Coal-field. — Traps of Western and Central
India. Pt. 3, 1869 (2 Rs. 8. As.) : Taptee and Nerbudda Valleys. — Frog beds in Bombay. — Oxyglussus
pusillus.

,, VII, pp. 342. Pt. 1, 1869 (3 Rs.) : Vindhyan Series. — Mineral Statistics. — Coal. — Shillong Plateau. Pt. 2, 1870

(1 Re.) : Kavharbiiri Coal-field. — Deoghar Coal-field. Pt. 3 1871 (1 Re.) : Aden Water-supply. —
Karanpura Coal-fields.

,, VIII, pp..353. Pt. 1 1872 (4 Rs.) : On the Kadapah and Karnul Formations in the Madras Presidency'. Pt. 2,
1872 (1 Re.) : Itkhuri Coal-field. — Daltonganj Coal-field. — Chobe Coal-field.

,, IX, pp. iv. 358. Pt. I, 1872 (4 Rs.) : Geology of Kutch. Pt. 2 1872 (1 Re.) : Geology of Nagpur. — Geology of
Sirban Hill. — Carboniferous Ammonites, pp. 65.

,, X. pp. 359. Pt. 1, 1873 (3 Rs.) : Geology of Madras. — Satpura Coal-basin. Pt. 2, 1874 (2 Rs.) ; Geology

of Pegu.

,, XI. pp. 338. Pt. 1, 1874 (2 Rs.) : Geology of Darjiling and Western Duars. Pt. 2, 1876 (3 Rs.) : Salt-region of
Koh at, Trans-Indus, pp. 330.

,, XII, pp. 363. Pt. 1, 1877 (3 Rs.) ; South Mahratta Country. Pt. 2, 1876 (3 Rs.) : Coal-fields of the Niiga
Hills, pp. 95.

„ XIII, pp. 248. Pt. 1, 1877 (2 Rs. 8 As.) : Wardlia Valley Coal-field. Pt. 2, 1877 (2 Rs. 8 As.) : Geology of the
Rajmahal Hills.

,, XIV, pp. 313. Geology of the Salt-range in the Punjab.

,, XV, pp. 192. Pt. 1, 1878 (2 Rs. 8 As.) : Geology of the Aurunga and Hntar Coal-fields (Palamow). Pt. 2,

1880 (2 Rs. 8 As.) : Ramkola and Tatapani Coal-fields (Sirguja).

,, XVI, pp. 264. Pt. 1, 1879 (1 Re. 8 As.) : Geology of Eastern Coast from Lat. 15° to Masulipitam. Pt. 2, 1880

(1 Re. 8 As.) ; The Nellore Portion of the Carnatic. Pt. 3, 1880 (2 Rs.) : Coastal Region of the Godavari
District.

,, XVII, pp. 305. Pt. 1, 1879 (3 Rs.) : Geology of Western Sind. Pt. 2 1880 (2 Rs.) : Trans-Indus extension of the
Punjab Salt-range.

,, XVIII, pp. 311. Pt. 1, 1881 (2 Rs.) : Southern Afghanistan. Pt. 2, 1881 (1 Re. 8 As.) : Manbhum and Singbhum.
Pt. 3, 1881 (2 Rs.) : Pranliita - Godaveri Valley.

,, XIX. pp. 242. Pt. 1, 1882 (2 Rs.) : The Cachar Earthquake of 1869. Pt. 2 1882 (1 Re.) : Thermal Springs of India.

Pt. 3, 1883 (1 Re.) : A Catalogue of Indian Earthquakes. Pt. 4, 1883 (1 Re.) : Geology of parts of
Manipur and the Naga Hills.

,, XX. pp. 240. Pt. 1, 1883 (2 Rs. 8 As.) : Geology of Madura and TinneVellv. Pt. 2, 1883 (2 Rs. 8 As.) : Geological
Notes on the Hills in the Neighbourhood of the Sind and Punjab Frontier between Quettah and Dera
Ghazi Khan.

,, XXI. Pt. I, 1884 (2 Rs.) : Geology of the Lower Narbada Valley between Nimawar and Kawant. Pt. 2, 1884 (1 Re.) :
Geology of the Kathiawar Peninsula in Guzerat.

,, XXII, pp. 344 (1883) : The Geology of Kashmir, Chamba, and Khagan.

The price fixed for these publications is 5 Its. (10s.) each volume.

Manual of the Geology of India, 2 Vols. and Map [out of print). Pt. III. (Economic Geology), 1881, price 5 Rs. (10s.).

To be had at the Geological Survey Office, Indian Museum, Calcutta ; or through any Bookseller. London : Triibner & Co.

RECORDS OF THE GEOLOGICAL SURVEY OF INDIA.

(In Royal 8vo.)

The Records op the Geological Sttkvey op India will be issued at intervals of about three months (February, May
August, and November), and forwarded to subscribers.

The annual subscription for four numbers is 2 Rs. (4s.) Postage additional, if forwarded, for India, 4 As., for Great
Britain 8 As. (Is.)

Each number will contain the additions to the Libraiy and donations to the Museum up to the first of the next
preceding month, — that is, to the end of March, June, September, and December. The first volume (186S) contains only
three numbers.

Calcutta, October, 1886.

MEMOIRS OF THE GEOLOGICAL SURVEY OF INDIA

Vol.

I.

Vol.

II.

Vol.

111.

Vol.

IV.

Vol.

V.

Vol.

VI.

Vol.

VII.

Vol.

VIII.

Vol.

IX.

Vol.

X.

Vjl.

XI.

Vol.

XII.

Vol.

XIII.

Vol.

XIV.

Vol.

XV.

Vol.

XVI.

Vol.

XVII.

►4

o

>

XVIII.

Vol.

XIX.

Vol.

XX.

Vol.

XXI.

Vol.

XXII.

A

o

>

XXIII.

Vol.

XXIV.

Royal 8vo, pp. 309, 1859. Pt. 1 ( out of print): On the Coal and Iron of Cuttack. — Structure and Relatione
of the Talchir Coal-field. — Gold Deposits in Upper Assam. — Gold and Gold-dust from Shne-Gween. Pt. 2
(out of print) : Geology of the Kliasi Hills. — The Nilghiri Hills. Pt. 3 (out of print) : Geology of Ba»-
koornh, Miduapore, and Orissa. — Laterite of Orissa. — Fossil Teeth of Ceratodus.

Royal 8vo, pp. 341, 1859. Pt. 1 (out of print): Report ou the Viudhyan Rocks and their Associates in
Bundelkand. Pt. 2 ( out of print) : Geological Structure of the Central Portiou of the Nerhndda District.
— Tertiary and Alluvial deposits of the Ner'nudda Valley. — Geological Relations and probable Geological Age
of the several Groups of Rocks in Central India and Bengal.

Royal 8vo, pp. 438. Pt. 1, 1863 (out of print) : Report on the Ilaniganj Coal-field. — Additional Remarks ou
tne Geological Age of Indian Rock systems. Pt. 2, 1864 (out of priut): On the Sub-Hinialayau Ranges
between the Gauges nnd Sutlej.

Royal 8vo, pp. 450. Pt. 1, 1863 (price 2 Its.): Report on the Cretaceous Rocks of Trichinopoly District,
Madras. Pt. 2, 1864 (price 2 Rs.) : On the Structure of the Districts of Trichinopoly, Salem, Ac. Pt. 3,
1865 (price 1 Re.) : On the Coal of Assam, &c.

Royal 8vo, pp. 354. Pt. 1, 1865 (price 3 Its.): Sections across N. VV. Himalaya, from Sutlej to Indus. — On
the Gypsum of Spiti. Pt. 2, 1866 (price 1 Re.) : On the Geology of Bombay. Pt. 3, 1866 (price 1 Re.):
Ou the Jherria Coal-field. — Geological Observations on Western Tibet.

Royal 8vo, pp. 395. Pt. 1, 1867 (price 8 As.) : On the Neighbourhood of Lynyan, &c., in Sind. --‘-Geology o(
a Portion of Cutch. Pt. 2, 1867 (price 2 Its.) : Bokaro Coal-field. — Ramgarh Coal-field. — Traps of Western
and Central India. Pt. 3, 1869 (price 2 Rs. 8 As.) : Tantee and Ner'nudda Valleys. — Frog-beds in Bombay.
• — Oxyglossus pusillus.

Royal 8vo, pp. 342. Pt. 1, 1869 (price 3 Rs.) : Viudhyan Series.— Mineral Statistics — Coal. — Shillong Platean.
Pt. 2, 1870 (price 1 Re.): Kariiarbari Coal-field. — Deoghar Coal-field. Pt. 3, 1871 (price 1 Re.) : Aden
Water-supply. — Karaupura Coal-fields.

Royal 8vo, pp. 353. Pt. 1, 1872 (price 4 Rs.) : On the Kadapah and Karnul Formations in the Madras
Presidency. Pt. 2, 1872 (price 1 Re.) : Itkhuri Coal-field. — Daltougunj Coal-field. — Chope Coal-field.

Royal 8vo, pp. iv, 358. Pt. 1, 1872 (price 4 Rs.) : Geology of Kutch. Pt. 2, 1872 (price 1 Re.) : Geology
of Nagpur. — Geology of Sirban Hill. — Carboniferous Ammonites, pp. 65.

Royal 8vo, pp. 359. Pt. 1, 1873 (price 3 Rs.) Geology of Madras. — Sdtpura Coal-basin. Pt. 2, 1874
(price 2 Rs.) : Geology of Pegu.

Royal 8vo, pp. 338. Pt. 1, 1874 (price 2 Rs.) : Geology of Darjiling and Western Duars. Pt. 2, 1876
(price 3 Rs.) : Salt-region of Kohat, Trans-Indus, pp. 230.

Royal 8vo, pp. 363. Pt. 1, 1877 (priced lie.): South Mahratta Country. Pt. 2, 1876 (price 2 Rs.) : Coal-
fields of the Naga Hills, pp. 95.

Royal 8vo, pp. 248. Pt. 1, 1877 (price 2 Rs. 8 As.) : Wardha Valley Coal-field. Pt. 2, 1877 (price 2 Rs. 8 As.):
Geology of the Rajmahal Hills.

Royal 8vo, pp. 313, 1878. Geology of the Salt-range in the Punjab.

Royal 8vo, pp. 192. Pt. 1, 1878 (price 2 Rs. 8 As.) : Geology of the Aurunga and Hutar Coal-fields (Palamow).
Pt. 2, 1880 (price 2 Rs. 8 As.) : llamkola and Tatapani Coal-fields (Sirguja).

Royal 8vo, pp. 264. Pt. 1, 1879 (price 1 Re. 8 As.): Geology of Eastern Coast from Lat. 15° to Masulipatam.
Pt. 2, 1880 (price 1 Re. 8 As.) The Gneiss and Transition Rocks, and other Formations of the Nellore
Portion of the Carnatic. Pt. 3, 1880 (price 2 Rs.) : Coastal region of the Godavari District.

Royal 8vo, pp. 305. Pt. 1, 1879 (price 3 Rs.) : Geology of Western Sind. Pt. 2, 1880 (price 2 Rs.) : Trans-
Indus extension of the Punjab Salt-range.

Royal 8vo, pp. 300. Pt. 1, 1881 (price 2 11s.) : Southern Afghanistan. Pt. 2, 1881 ( price 1 Re. 8 As.):
Manblium and Singhbhum. Pt. 3, 1881 (price 2 11s.) : Pranliita Godavari Valley.

Royal 8vo, pp. 242. Pt. 1, 1882 (price 2 Rs.) : The Cacliar Earthquake of 1869. Pt. 2, 1 882 (prv e 1 Re.):
Thermal springs of India. Pt. 3, 1883 (price 1 Re.): Catalogue of Indian Earthquakes. Pt. 4, 1883 (price
1 Re.) : Geology of parts of Manipur and the Naga Hills.

Royal 8vo,pp.240, Pt. 1, 1883 (price Rs. 2-8): Geology of Madura and Tinnevelly. Pt. 2, 1883 (price R*. 2-8) .

Geological Notes on the Hills in the neighbourhood of the Sind and Punjab Frontier between Quetta an<i

Dera Ghazi Khan.

Royal 8vo, pp. 286. Pt. I, 1884 (price Its. 2-: Geology of the Lower Narbada Valley. Pt. 2, 1884 (price

lie. 1): Geology of Kathiawar. Pt. 3, 1885 (price Rs. 2) : Coal-field of South Rcwah. Pt. 4, 1885 (price

Re. 1) : Barren Island.

Royal 8vo, pp. 344, 1883. The Geology of Kashmir, Cliamba, and Khagan.

In preparation.

Pt 1, 1887. The Chhindwara Coal-fields, (price l Re. 8 As.).

The price fixed for these publications is 5 Its. (ICG.) each volume.

Manual of the Geology of India, 2 Vols. and Map, 1879, (out of print). Vol. Ill (Economic Geology), 1881, priced Rs. (10s.)

Vol. IV (Mineralogy), 1887, price 2 Rs..

To be had at the Geologicul Survey Office, Indian Museum, Calcutta, or through any Bookseller. London : Triibner & Co.

RECORDS OF THE GEOLOGICAL SURVEY OF INDIA.

Vols. I to XIX, 1868 to 1886.

Rkcohds of the Geological Survey are issued quarterly, — in February, May, August, and November. They contai:.
brief reports and papers; abstracts of more detailed work ; notices of recent discoveries; donations to Museum, and addition!
to Library, &c. They are of the same size as the * Memoirs,' but are separately paged. Begun in June 1868.

The annual subscription fot four numbers or parts is 2 Its. (4*.). Postage additional j if for India, 4 As., foi Great
Britain, 8 As. (1*.).

Calcutta, September 7, 1887.

PALiEONTOLOGIA indica

(Series I, III, V, VI, VIII.)

CRETACEOUS FAUNA OF SOUTHERN INDIA, by F. STOLICZKA, except Vol. I, Pt. I, by H. F. BLANFORD.

Vol. I. The Cephalopoda (1861 -65), pp. 216, pis. 94 (6 double).

„ II. The Gastropoda (1867-68), pp. xiii, 500, pis. 28 (10 parts).

„ III. The Pelecypoda (1870-71), pp. xxii, 537, pis. 50 (13 parts).1

„ IV. The Brnchiopoda, Echinodermata, Corals, &c. (1872-73), pp. v, 202, pis. 29.

(Series II, XI, XII.)

THE FOSSIL1 FLORA OF THE GONDWANA SYSTEM, bv O. FEISTMANTEL, except Vol. I, Pt. I, by T.

OLDHAM and J. MORRIS.

Vol. I, pp. xviii, 233, pis. 72. 1863-79. Pt. 1: Rajmahdl Group, Rajmahal Hills. Pt. 2: Same, continued. Pt. 3 :

Plants from Golapilli. Pt. 4: Outliers on the Madras Coast.

„ II, pp. xli, 115, pis. 26. 1876-78. Pt. 1 : Jurassic Flora of Kach. Pt. 2: Flora of the Jabalpur Group.

„ III, pp. xi, 64 + 149, pis. 80 (9 double), (I-X XXI + 1.4-XLV1 IA. 1879-81. Pt. 1 : The Flora of the Talchir-Karharbari
beds. Pt. 1 : (Suppl.). Pt. 2 : The Flora of the Damuda and Panchet Divisions. Pt. 3 : The same
( concluded ).

„ IV, pp. xxvi, 52 + 71, pis. 35 (2 double), I— XXI -t- I^t— XIV4). 1882-86 Pt. 1: The Flora of the South Rewah
Gondwana basin. Pt. 2 : The Flora of some of the Coal-fields in Western Bengal.

(Series IX.)

JURASSIC FAUNA OF KACH.

Vol. I, pp. i, 247, pis. 60 (6 double). (1873-76). The Cephalopoda, by W. Waagen.

(Series IV.)

INDIAN PRETERTIARY VERTEBRATA.

Vol, I, pp. vi, 137, pis. 26. 1865 — 1885. Pt. 1 (1865): The Vertebrate Fossils from the Panchet Rocks, by T. H, Huxley. — -

Ft. 2 (1878) : The Vertebrate Fossils of the Kota-Maleri Group, by Sir P. de M. Grey Egerton and
L. C. Miall. Pt. 3 (1879) : Reptilia and Batrachia, by R. Lydekker. Pt. 4 (1885): The Labyrinthodont
from the Bijori Group, by R. Lydekker. Pt. 5 (1885) : The Reptilia and Amphibia of the Maleri and
Denwa groups, by R. Lydekker.

(Series X.)

INDIAN TERTIARY AND POST-TERTIARY VERTEBRATA by R. LYDEKKER, except Vol. I, Pt. I, by

R. B. FOOTE.

Vol. I, pp. xxx, 300, pis. 50. 1874-80. Pt. 1 : Rhinoceros deccanensis. Pt. 2 : Molar teeth and other remains of Mam-
malia. Pt. 3 : Crania of Ruminants. Pt. 4 : Supplement to pt. 3. Pt. 5 : Siwalik and Narbada
Proboscidia.

„ II, pp. xv, 363, pis. 45. 1881-84. Ft. 1 : Siwalik Rliinocerotidse. Pt. 2 : Supplement to Siwalik and Narbada Pro-

boscidia. Pt. 3: Siwalik and Narbada Equidse. Pt. 4 Siwalik Camelopardalid®. Pt. 5 : Siwalik Selenodont
Suina. Pt. 6 : Siwalik and Narbada Carnivora.

„ III, pp. xxiv, 264, pis. 37 (1884-86). Pt. 1 : additional Siwalik Perissodactyla and Proboscidia. Pt. 2 : Siwalik and
Narbada Bunodont Suina. Pt. 3 : Rodeuts and new Ruminants from the Siwaliks. Pt. 4 : Siwalik Bird*.
Pt. 5 : Mastodon teeth from Perim Island. Pt. 6 : Siwalik and Narbada Chelonia. Pt. 7 : Siwalik Crocodilia,
Lacertilia and Ophidia. Pt. 8 : Tertiary Fishes.

„ IV, Ft. 1 (1886). Siwalik Mammalia (Supplement 1), pp. 18, pis 6. Pt. 2: (1886). The Fauna of the Kamul caves :

(and addendum to pt. 1) ; pp. 40 (19-58), pis. 5 (vii-xi). Pt. 3 : (1887). Eocene Chelonia from the Salt
Range : pp. 8 (59-66), pis. 2 (xii-xiii).

(Series VII, XIV.)

TERTIARY AND UPPER CRETACEOUS FAUNA OF WESTERN INDIA.

Vol. I, pp. 16 + 110 + 382 + 91, pis. 5 + 28 + 58 + 13. Pt. 1 (1871): Tertiary Crabs from Sind and Kach, by F. Stoliczka.

Pt. 1 (new 2) (1880): Sind Fossil Corals aud Alcyonaria, by P. Martin Duncan. Pt. 3 (1882-86) : The Fossil
Ecliinoidea of Sind, by P. Martin Duncan and W. Percy Sladen. Fas. 1 .- The Curdita beaumonti beds.
Fas. 2. The Ranikot series in Western Sind. Fas. 3 : The Kirthar Series. Fas. 4 : The Nari Series (oligocetu).
Fas. 5 : The Gaj Series (miocene). Fas. 6 : The Makran Series (pliocene). Pt. 4 (1883): The Fossil
Ecliinoidea of Kachh and Katty war, by P. Martin Duncan and W. Percy Sladen.

(Series XIII.)

SALT-RANGE FOSSILS, by WILLIAM WAAGEN, Ph.D.

Productus- Limestone Group: 1 (1879). Pisces, Cephalopoda, pp. 72, pis. 6-

2 (1880). Gasteropoda and supplement to pt. I, pp. Ill (73-183), pis. 10 (vii-xvi) ;
(1 double).

3 (1881). Pelecypoda, pp. 144 (185-328), pis. 8(xvii-xxiv),

4 (1882-85). Brachiopoda, pp. 442 (329-770), pis. 62 (xxv-lxxxvi).

5 (1885). Bryozoa — Annelida — Echinodermata, pp. 64 (771-834), pis. 10 (Ixxxvli-
xcvi).

6 (1886). Coelenterata, pp. 90 (835 — 924), pis. 20 (xcvii-cxvi).

7 (1887). Ccelenterata, Protozoa, pp. 70 (925— 994), pis. 12 (cxvii-cxxviii).

The price fixed for these publications is 4 annas (6 pence) per single plate.

To qe had at the Geological Survey Office, Iad'an Museim, Calcutta, or through any Bookseller. London : Triibner & Co.

^ 'K<

o -y

o'

"b 0X

r

Jr v ' ✓ >y - - + y. y.> * ~ _ , vl ^ > , v *>. * ^ . /v A

'‘\/^^ 3 r^%y *> »0> v\^Y*o^> *“'*\/V < '*% * 9 N° / *' *«, %** " '*'** s'** '*%

A*' ' U V. ’V^ v ^ e*V*V5fe'. V & S \%$ - ~ A*' * ^Sfe'* <

$ >■ m® : ,^V */ ..:, \ -. y ** -■ .- lJ"b : *V* A- ,. . ■•; ; y <£

o 1. ^<0f * A# - ^liwilr « «A <V 0 <SW " V1 ^ - pB p% * 'V O Rf# ‘

\0ox

« 7 ° VM v- - 0

> “. ■ > V < oV

<5. ' , , \ *

C> ^

V‘. Vf 0

.y >.

Ap -J.V,

•V <p

^ ^ A r >

., N 1O-1 ~o§ *

' s' V '* O

’■>' V

© ox

<°' v, * a r. s > \ na

,0y (. 0 ' * <s> V& < •

K eC Tl^ * '?r .A' x Jt

^ Vs ;

v0c>

\ <-<• -r

*>* .^r*/y

o5 ^ " 7;

V Vl\W\s* > \v- *

✓

* 5 N 0 v .

” 0

* *^V. r-

* . II

.v * v//k <» »

ov - „ wu r.J ■’> '^. oX’

aa ^...4 V y°'^ >'

n o v

’o 0N

^ ,* x<* ■’•*,. „ '.'0CP‘ ° ^ V '•

' \/7'i&VW -^ir, %/ •^'•-■^ > .'A'. ■

> <p.

V

, \ '3 cv y >
yo0^ =' •: . -^/;

A oV ^ V N*

r 0 N c. . <b ' ■> <* S A
c «. '1' .

* 0 V 0 5 ^

y a

r ' ‘ C

A' oJ co ''*,To:,,’ ^

\ vN , ' ",

■e,. <y * V a, .w

° y ^ -• ^ ® :*i

•V ./>

0 y = ■

w> v » -V '- v ■<> >1, •-*

,00

>u c;’
^ V

0 o5 ^
v ^ ^

* y A'f

</* 6 V

^ V

•^ '/\ •••. rv ’ -■•/: *•*. v • ■ ■ ' >°j, - - "c

> ^ ^ a'

^ 'i.

.v.,, %y°^v° «“"**%v :'VA •‘,*
r^/%oo^ iiflf

* ^ ^ *

> ^ Ke cfV *

U “

-> *

aV <P

* 'f,> j -i - • y- •

^•.-.y-v\- ■ * rv* • ' V% •• • 1 • • ' >'

'-A-^W-. V A ..">^ * j.„ , AA- A V A ♦

,0 O,

r o N 0 ' A^

\> a ^ ° * /

0 N C

/- v

V c ^ y, ; ••'-^/ * ■ x00^ *

e, ■ * , , ,*'V /' \ °V 0 '

<^- 8 * ' *N s 0 * / V . V v * n 8 I ' s 9 ^ /

✓> . cy s ^ r o O v ^ w / ^ ,v v a

•P. vN ^ 0 <\. Ar ff \ h ^ ^ a <r? v\.

^ ^ ^ Lf ' a : * \* - Ji; 1 ® ^ ^ ^ ^ I

* A^' % ' llyp^ ^ \ : ^ %• '•

C^- * ^ ^ ay - V, ^ ^ ^ aV

V ■'•*♦.'' X\A 1 If ^ yo»».'‘ A0 0 w c , ^ ' “ “ s'' a\A Ul/i ^0.^ o N c <, 'J

^ °° C°’A5^P^- C°'^W' ’*■- aX-

V

Jr . . * \v

8 I '

/’ s S * * , ,' V * » ~ 0 ’* v\^‘ , . « , % * 9 ' ' 4 V° S ' : / 1%

^ A* -
tP ^ v

y/ * « > ^ \ s . , , » >* ° 5

/■ ^ .9 •» s „ 7 ft o

t* •_ - ,\^ '

^'':- \ = y, ^ ^ ^

■f1 . *s

00'

^ ^ y

* •> N 0 ' ' t . 0 » » I ' 3 ^s»0 x t « , -A

, . jS- > A% fc“. A .»'*• * • •*

' > A • 9

br, . \\' ^ fA r A -

yp \V « : \Mj ] - <v

• / -M-; ' ■ n ; '

.) N (,

* -'■*- ,^N vo ' \ [ *_, r'b.

» t "bo'' T 'ixc

; x0^, a ^

♦a- b ^ %_ %

y

’ ■ o*\'sy '+

,XA ^ ^ ^ -

* ~ n A ^ i0

V ' V - V 1 * * '^Pv y 0 * ^ *

',\ .A b ^,.% 0 o°

^ V

O oO

c= A 'i ? „.

+ 9 1 ’7, s- ;^7y* aH°V\ ^ * * >; * • ' ^ 1 n N0 v> % J i,

^ ^ ■- f yOy '- ^*7 % A ::

^ * 0

O 0

^ a0

.A -y -

* ^ ^ y -1 o

O- 'o» ,* .,CA

o5

> Or V

* a* c^

> a,*

ts. . * Q f '/> ' N v \' N 4 f.

cy s ^ y/ cy N* ^ A

A ' ^

\ ^

; 7 \ \W

© «/

0n°

^ 0^ :

■ : 4 ^ ,

©c-

' *P t'

V> ^ ^ >

«

,Jb- ^

^0^

^ « W.

A.= . ^ -/■

»* y.\

‘‘ ‘ *• ' " ,AV v. 1 . u 0 v Au t o *■ t . V

r4> # '* ll*r °0 (-O’

^ *££%>>+ .r <, -\^B2|'' ^

' ^ %

r*- ^

v0 C>

\ <^»

, . . . ••'•;V^'v:ry*“**v'\-*, •>

* ^ , O'

,^y ^ o

- ^ '>■= «■ ^ a

2> ^ o . , -* /

- . ^ rP- c

0 9 X

§V ^ '*

1 'OO

.\^ v ^er*'*w ^ ^ rf.5» • *V > -*> x ., ... o

a\* J2>5 0» r- ^ * <f*. A*, o. J\S5»6T /V, n >►. .A <*- * V

*

*- Vg®§f- ^ *

* T i># v ’V.-*^k\Jw “® V/v AT 9 *

* * v ' • • -;>\ ■%• ; v ■ • ' *>. tr *-V • ■ '>'.. - '.iL- /%

,V ^

•v -' ' 'Sk. ■*

,0’

A s

£ ^ A *

> 'V O

'' ^ vl, \''"X+ 4
V • , % ,0*

X. -a\ ^ , . .

^ V - ‘W

Cl