ArticlePDF Available

Lower Tertiary Vertebrates from western India

January 1975

Authors:

Ashok Sahni

Panjab University

Vijay Prakash Mishra

Geological Survey of India, Lucknow

Fig~ i. FIGS. 1~3 TOMISTOMA TANDONI Sp. novo I a""-Dorsal view of skull, L.U.V.P. 11062a x 0.25 ..b-Ventral view x 0.25 2 a-Dprsal view of mandible, L.U.V.P. 11062 (Holotype) x 0.25 2 .b-;-Ventral view x 0.253-Dorsal view of left articular part of mandible, L. U. V.P. 1062 x 0.25

…

Figures - uploaded by Ashok Sahni

Content may be subject to copyright.

Content uploaded by Ashok Sahni

Content may be subject to copyright.

@191~

LOWER

TERTIARY VERTEBRATES FROM WESTERN INDIA

ASHOK SAHNI

VIJA Y PRAKASH MISHRA

DEPAI{TMENT OF GI!OLOGY, LUCKNO'W UNIVERSITY.. LUCKNOW,

AllS'l'l{AC l'.--Pal.lcogclle :uld J-,ower lvliocellC J'ock~ 01' Kutcllt We~terJl IJldiut IlavtJ yielded :l pl'ulilil; vcr-

tebrate fauna from three separate horizons: Middle Eocenc, Chattian (uppermost Oligocene) and Lower

Miocene. The Middle Eocelle fauna comprises of sililroid fislles, Ff!jllmia /I/tllOlli sp. nov.t I;'f!jllmia mijTfli sr.

nov., SOClIojJflca IIt/Tni sp. 11OV. and /I rill.\" kIlIChclIsi.r; trionycJlil1 tllrtl~s; a Il<'~W t()llli.~IOlllill, 'li'llli.rtl/lllcl lfmr!,mi

sr. novo alld crocodiliall reillains as well .11; tile Ill:unfn,us reprc1;elltcd by art:lla~()Cetcs, Otlolltocetcs, a 1;irclli:ul

and a probable terrestrial mammal. Of the mammalian genera and 1;peciest the following taxa are new: }->rolocetus

IlaTudiensis sp. nov., Indocelus ramanigen. et sp. nov.t Andrewsiphius kulcllensis gen. et sp. nov., and Andrew-

siP/lilts minor sr. novo hldocelus ramnni gen. et sp. novo suggests an intermediate stage between Mesonychidac and

Archaeoceti. The Chattian mammalian fauna is marine with the predominance of the dugongid Halililerium. The

Lower Miocene is a mixed facies fauna comprising of the sirenians, Irldosiren koeni,gswnldi sp. nov., as ,veIl as

definite terrestrial mammals represented by Deino.therium and the anthracotheriid Brac/!}'odus.

The lower Tertiary vertebrate fauna from Kutch closely resembles corresponding fauna8 from Egypt

and Nigeria.

INTRODUCTION

T ertiary vertebrates from KutC!1, western India,. were. first repor.ted by Gra~t (1840).

For well over a century untIl the present tIme, lIttle attentIon was paId to the

investigation of Kutch vertebrate,q except for a few report,q (Wynllc 1 ?'7?, Rao 1 ~).r;6, Tcwari

1959 and 'fall dOLI 1 ~71 ). ,!'he maill emphilsis was direcl(~d to tllc study of [oralnillifers alld

ostracodes (Sykes 1840, Sowerby 1840, Vredenburg 1906, Nuttall 1926, 'rewari and others

1956-1968, Tandon 1966, Singh 1967, Mohan alld others 1968-1970, SriVaS[clVa 1970). 'rhere arc

scvcral otller reports Oil invertebrates (j)' f\rclliilC illlCI 1.-lililn(: I f~:):1, OltllCilll i111(1 Slil<lell I f~f~:1,

Vredenburg 1925, 1923).

Systematic collection of '!'ertiary vertebrates of Kutch was undertaken by Lucknow

University field parties 'during the years 1970-1973. During six months of field prospecting,

a large and varied collection was made from a numb(~r or locC:llities (l"ig. I), .ll1(] froln Ilorizolis

ranging in age from the Lower ~~ocene to the Lower Miocene (Fig. 2). .

The Tertiaries of south-western Kutch are essentially a shallow water, coastal sequence,

exposed in numerous nala sections in the area (Fig. '2). r!'oe Palaeocene Madh Series over~ying

the Deccan plateau basalts, COIISists of red lateritic and tuffaceous, trap-wash material

containin,g plant fossils and representing non-marine deposition. These beds are overlain by

the Eocene Berwali Series attaining a maximum thickrless of 95 meters in soutll-western Kutch.

The series is divisible into two stages, the Kakdi Stage comprising Lower Eocene, and Babia

Stage comprising the Middle Eocene beds, and has been separated mainly on the basis of

ostracodal and foraminiferal assemblages (Biswas 1965, 1971 a). 1'he Oligocene Bermoti Series

are deposits of arenaceous marls, and are overlapped in several areas by a thick sequence

of Mio-Pliocene sandstones and carbonates (Kllari .and Kaukawi.lti Series) wllicll suggcst'thc

transition of marine conditions iu Lowcr Mioccne to terrestrial conditious in tile l>liocelle. 'fhe

Pleistocene is represented by a thin bed of Miliolitic Limestone. In south-western Kutch, Recent

and sub-Recent alluvium as well as wind-blown sand extensively blanket the Tertiary and

f\l1esozoic sequences. ,

The Eocene fauna includes fish, turtles, crocodiles, primitive cetaceans, sirenialls and a

doubtful moeritheriid. The fishes are dominated by large silurQids comprising of three genera

and five species. Elasmobranchs are represented by fivc species of sll.lrks illld oue ,qp(~cics of

rays. '!'he turtles are mostly cstuariue forms, comprising the genera Asperidites and l-riolryx.

The crocodilian fauna is diversified and consists of a new species, of' a.large tomistomiid, aud

Crocodylus. The mammals consist of archaeocete cetaceans as well as doubtful odontocetes.

The sirenians are represented by a single genus, Protosiren. rille presence of l,llld mammals

during Middle Eocene times in Kutch is recorded by a sacrum presumably belonging to a

moeritheriid.

A. SAHNI ,,1ND V. P. 1\IIISHl~A

The Chattian vertebrate fauna comprises mainly of sirenians represented by ribs, verte-

brae and mandibular fragments of Halitllerium.

The youngest vertebrate horizon has yielded a number of sharks, rays, batoids and other

teleosts; trionychid turtles, crocodilian vertebrae and scutes; sirenians are represented by a new

species of In(fosiren.. From the rocks of the same age (Burdigalian) but ,It a Iligllcr horizon, a'

femur of Deznorherzum has been recovered.

The types and figured specimens are deposited in tIle laboratory of V(~rlebratc Pc.lI,lCOll-

tology (L. U. V. P.), Geology Dep..trtment, T.uckllUW Ullivcrsily, I.ucknow, .llldi,l.

This study was undertaken as post-graduate research at University of Lucknow, India.

We wish to thank l">rof. R. C. Misra, Head of the Geology Department, Lucknow University for

providing laboratory facilities and encouragemcut. 'l'llanks arc also due to Shri Sushil K um,lr

Khare, research scholar in Geology Department, Lucknow University for help during the pre-

paration of the manuscript. We also wish to extend our thanks to Dr. Leigh Van Valen of

Chicago University and Dr. F. S. S2;alay of City University of New York for their comments and

suggestions on the archaeocete cetaceans in the fauna. The junior author is also thankful to

the University Grants Commission, and Council of Scientific and Industrial Research, Govern-

ment of India, for financial assistance.

Kutch Tertiary Vertebrate localities (Fig. l) yielding fossils described in this paper

E Longitude N Latitude

Locality Age

Locality No.

? Cilaltiall and Lower

Miocene 68"57' '2:3"32':17"

Matanomadll

L. U. 200l

68"1-7' '2:3nt}!)':iO'

LOWl~r Mioccnc

L. U. 2{)O2 L:lkllpilt

23"30'20"

68"4

~)'

Harudi

L. U. 2003 Middle Eocene

68°39' 23"31.'

Naredi (locally Lower Eocenc

knowr.) as

Narcda)

L. U. 2004

23°39'10"

68°41 '

L. U. 2005 Nareda (locally Middle Eocene

known as

Naredi)

23°39'

Middle Eocene 68°39'30"

U. 2006 Godhatad

23°31' 15"

68°55'30"

Denma

L. u. '2007 ? Chattian

23"26'35"

68°52'00"

L. U. 2008 Jangadia Lower Miocene

23°23'

68°51 '

L. U. 2009

Buta

Lower Miocene

23"2'

L. U. 2010 68"LiU'

Aida Lower Miocene

23"28'40"

Samcla Lower Miocene 68°52'

L. U. 201

23"28'15'

63°4.9'30'

J unagia Lower Mioccnc

L. U. 2012

23°29'42"

Middle EOCClle 68°45'30"

L. U. 2013 Jhulrai

68°31'45" 23"31'

Lower Miocene

L. U. 2014 Pipar

68"34'10" 23031'06'

L. U. 2015 }Zatipar ? Cllattiall

vVESTERN INDIA TERTIARY T/ERTEBRATES 3

.LA" HPAT

\~OL'V'2002 I 'I. ,', / i -"\ ,i ,{

, ./ \

,'-, ',I f.~ '

~ ' \" , ,

, I 1'

, , !

RANN OF

KUTCH

?,,;?I ""i"."e'~~-r'::~~~ '\ I

~Ol~ K U T C OJ j

I: I VI" .MO~O"O"Odh .ROPO~

.C!ouJ

-23°1' ':~=~=~

ARABIAI' ~

SEA ~-~~~~oi-J

1 ---r. ~ ~

--",'" c.

",'

, 'f-O

-..M.lo lIed "00 d

--UnmotOll'd ...od

LU V."ob..ol- 1..0111)' N..

/) L ."... I:oc_n-

f) Mlddl, ~OC."-

D ? Chott,on

0 low... "'oo..n.

't-/"

\ oi .I

>, ~. '-

, ,4\ LU2004 '-,-.-;

,,:!-' "' .'

-~ '.;1,;' "A~EOI '"

I \ " ~ ,

/ I ' "." \ t'.B"~ANO"

I" fI/lTIPAR ! \ I ,. r ~ ~

PIPA~.6"""'" "".U;2003 r ~

L,U20'14 IULU201~ ~'8HARUDI

, ,.' "1 1-',,-,"-"~

1/: I..,~J-" .f!

", . ,r

-'.' , ---

, I

,\ ,

, r'

,AVO R .', i

( .

, ,

\ r ,,- -~.

, .

I" '.'.

" 0 NALlS-

CtfAIIOPADI ,,!

~.~"-~ I

OENMA .'..,.O!il;f"

r- ,

I',, , , .

" .JHULRA' ,~ I,,' \,'

"u~OI' , I', i ,J'_.~

, \ ~ ,...

~AGHAPAOHAR '" .,SAMOA. 1 ,,' -"

JUNAGIA.O'.-!.O)LU~O"',I_' ,,"":

'11.2012 i.l ".'(,"-

r' II,I!' ~.-J.-'-.'"

(' "p.d.":lJ~~- ~

oj. ~ JANGAOIA .

:" ..1,

, ,

" "

"DA"~'dJ~ .~

...,2010'0!. ' ~ oil'); .."\\-:- { .

I \ " AV' k.

., " v / ---' ~

" -'-

"' ', / 0 ~ l.

'\t'/' 1/"

" ,", ,

/1'"

/ '...'"

.,,' ,..

,,' .."

~." ,,' ,,'

eEI/~p.\.'..."'"'.-'_."

-.,::"...'

SCALE

0 4 BKm.

I 1 I

ARABIAN SEA

./7

FIG. 1. Sketch map of western Kutch showing vcr IcOrate fossil localities.

KUTCH, vVE,S'1-1.i.'llN INDIA

TEl?TIAl?YVERTEBl?A'TE If A UNA 0/"

Collected 1970-1973 by V. P. Mishra

Lower Eocene Fauna

Elasmobranchii S coliodon sr.

Middle Eocene Fauna

I,'ishes

Elasmobranchii Notidanus primigenius Agassiz

Galaeocerdo cuvieri Le Sueur

G. adullcas Agassiz

Lamna sp.

Carcharia.S' tricu.rpidatus Day

Myliobatis sr.

Fajumia menoni sp, nov,

F, mi.rrai sr. nov.

SOCIl{)Pfl{~(l nnTfli ~l)' IIUV,

Arius kulchensis l{ao

Arius sr.

G)bium sr.

S phyraena sr.

Pycnodus sr.

Actinopterygii

Reptiles

Crocody lidae Tomistoma tandOlli Sp- novo

Crocodylus sr.

crocodilian coprolites

1-j-iol!)'_\, sr.

Chelonia

"'AT IW4 ' \ "

:-.-Al,.' .~\

->,'" \' "

" I" SASIAj. "-:;.-

, ".~ NARAyAN, '." PANANOMRO.l """")0,.:

.~ ~- --, '-- .,

, .; I ,--

' -'. .

, ,. '. LU 2005, \ ,

'.~ I"'. ';". \

r GOOHA TAp.' V NAREOA "

/' l) .,,;-., .',

/'L,U.2000\ .'

A. SAHNI AND .V. P. MISHRA

Mammals

Cetacea Protocetus sloani Sahni and Ivlishra

P. hQrudiensis sr. novo

cetacean assorted mandibul,lr fragments

lndocetus ram ani gen. et sr. novo

Andrewsiphius kutchensis gen. et sr. novo

Andrewsiphius minor sr. novo

Sirenia Protosirenfraasi Abel

Proboscidea (?) cf. Moeritheriid sacrum

Oligocene (Chattian) Fauna:

Mammals

Halitherium sp.

Sirenia

Lower Miocene Fauna:

Fishes

Elasmobranchii Galaeocerdo cuvieri Le Sueur

G. wynnei Mehrotra et al.

Hemipri.flis serra Agassiz

H. sureshi Mehrotra et al.

Hypoprion macloti Muller and Henle

J\I egaprion brevirostris Pocy

S coliod on .forrah:OlVall Cuvicr

IS'urus spallanzanii Bonaparte

Carcharodon megalodon robustus Davies

C. angustidens Agassiz

C. bigelowi Mehrotra et ;:II,

C. carcllarias Linna ellS

AloPias vulpe.\' Gmelin

.s'plryrna prisca Springer

Carcharias tricuspidatus Day

'. C. heptacusPidatus Mehrotra et 31.

Myliobatis curvipalatus Lydekkcr

Myliobatis sp.

Aetobatis sp.

Raja sp.

Actinopterygii Cybium sr.

S p!ryraena sr.

]-> seudoegertonia sr.

.s'parus (ChrysoPhlJ1S) sr.

Diodon sr.

Reptiles

Crocodylidae crocodilian scutcs & vertebrae

Chelonia T rionyx sr.

Mammals

Sirenia lndosiren koenigswaldi sr. novo

Proboscidcc:t Deinotherium pentapotallliac l,'a]collcr

~VESTERN INDIA TERTIARr VERTEl3RATES 5

A. SAHNI AND V. P. MISHRA

SYSTEMA'rIC D 1'"'. S CR J P T ION

EOCENE FAUNA FISHES

The present collection of Eocene fishes was obtained from the Uerwali Series of wcslcrll

Kutch from four principal localities: Harudi, Naredi, Godhatad and Nareda. 'rhe collection

contains both large specimens as well as micro-remains. Amongst the former are nine

fragmentary skulls, isolated cranial elements, jaw fragments, assorted teeth, dent.ll plJ.tes, SpillCS,

pectoral girdle elements and vertebrae. 'rhe specimens show a wide size variation ranging from

a large siluroid skull some 40 cm in length to a smaller one some 8 cm in length. Micro-fish

remains comprising of teeth, scales and vertebrae were obtained by the process of washing and

screening in the laboratory.

Of all the fish remains, the presence of siluroid cat-fish is particularly interesting, not

only because they are the single most well represented group of fishes, but also because of their

immense size, general similarity to African forms and palaeoecological parameters of the environ-

ment. 'rhe only other previously known siluroid from Kutch is a fragmentary skull of Arius

kutchensis (Rao 1956) from Lower Eocene beds of Naredi, 8°5 kms N 27°W of Harudi.

Class Chondrich thyes

Subclass Elasmobranchii

Order Selachii

An eroded tooth of "S'coliodon was collected from the [ossilircrcus sllalc :-1110 marl Ilnit of

Kakdi Stage (Lower Eocellc) exposed at Naredi.

From the Middle Eocene (Lutetian) horizon of Kutch only few fossil slla.rk teeth have

bcell collecled from Harudi and Nareda. III N,irt~da, thc sllarks arc relalivl:ly 11101.'(:. abulltlallt

than in Harudi. NotidallUs primi,f!;enius of tht~ !iuborder I-I(~xi.lnchoid{~a alld (.I'alaeocerdo cuvieri,

G. aduncas, Lamna sp., and CarchariaJ tricurpidatus of thc suborder G,llcoidca Ili.lVC becll

recognised (Mishra et al., 1973).

Order Batoidea

Suborder Myliobatoidea

}i'amily Myliobatidae

GENUS MYLIOHA TIS Cuvicr 1817

MYLIOBATIS sr.

Plate I, figs. 1-2

Material :-Dental plate--L. U. V. P. 11078 (Harudi); isolated teeth-L. U. V. P. 11079

(Harudi) and L. U. V. P. 11083 {Nareda) ; Spines-L. U. V. P. 11081 (I-Iarudi) and L. U. V. P.

11082 (Narcda). .

, Hori.~ons and localities: -Ossiferous gypseous sh;lles of Babia StaJ{c (M iddle I~ocenc) at

Harudi, and Foraminifcrallimestorlc of Babia Stage (Middle l~occne) at Nar('d;l.

Description :-Dental plate-L. U. V. P. 11078 (Pl. 1, fig. 1) is the only dental plate in the

collection comprising of the 8 rows of median teeth and two rows of lateral teeth of one side.

Only five teeth of median row are complete. I-he median teeth are hexagonal, seven times as

broad as long and antero-posteriorly arched. Inner lateral teeth are diamond-shaped, small and

nearly as long as broad. The outermost lateral teeth are the smallest and are longer than broad.

'I'he occlusal surface is fic1.t. The Kutch specimen resembles closely with the dental plate of

/Ii/. goniopleurus Agassiz from Lower-Middle Eocene of England (Woodward 1889) but has only

two lateral rows instead of three. Transverse sections of the dental plates of ~\1. ,g-oniopleuru.1' and

the Kutch specimen are nearly identical. The Kutch specimen also resembles the derltal plate

of j\'1. toliapicus [rom Middle Eocene of l~nglCJ.nd (Zittel 1932, p. 86, fig. 137) but has t1licker

WE's'TEl?N IND!.

TERTIA.RY'

v"ERTEBRATES '7

crowns. L. U. V. P. 11078 differs from the dental plate of Myliobatis curvipalatus from Eocene

of Kutch (Lydekker 1886 b) in being smaller in size and having two lateral rows instead of one

and possessing a flatetned plate.

'rhe other myliobatid teeth in our collection are isolated specimens, and belon~ to the

medial) row. These teeth are 1-7 times as broad ;1S IOllg' som(~ are sliglltly ;lr(;ll(~ll ;lllter()-

posteriorly, while others are nearly straight. '!'he coronal surface is smooth, while the basal

surfaces of the roots show distinct longitudinal ridges and grooves. The crowns are less thick than

in the median teeth of T.. tJ. V. P. 11078.

.S:pilles :-L. LT. V. P 11082 a (Pl. .1, fig. 2) is a c;:luoal spine whicll is llcarly 5 cm lOllg.

1'he spine is narrow and longitudinally striated, some of the striations being deep enough to form

grooves. The cross-section is similar to that of Myliobatis serratU.f from Oligocene of Alzey

(ZilteI1932, p. 86, fig. 138). The SpiliC willcllS llistcllly, tll(~ llel1licles prCS(~llt 011 lll(~ lat(~l'al c(l~cs

are directed distally making all angle of about 35u from the central axis. In the distal 1/3rd

region the denticles are absent, and only a fine cutting edge is present. The Kutch spines are

similar to that of Myliobatis sp. described from the Miocene of Balasore (Hora 1939), Lower

Miocene of Mayurbhanj, Orissa (Ghosh 1959) and M. serra/us from Oligocene of Alzey (Zittel

1932, p. 86, fig. 138).

Discussion :-Myliobatoid teeth alld spines are not common in the ]~ocene strata of Kutch

and dental plates are even rarer because the median alld lateral teeth in a dental plate get easily

detached, and are mostly found isolated. The few myliobatoid remains ill the present collection

are not sufficiently diagnostic to be referrable to any particular species of Myliobatis. A dental plate

of l\1ylioba.tis curviPalatus from the Eocene of Kutch described by Lydekker (1886 b) is the only

species of Myliobatis so far described from F.ocene horizon of India From Miocene, a median

tooth and a caudal spine from llalasol'c, Orissa (HOI"l l~)3~») and a spine from 7\1ayurbhanj, Orissa

(Ghosh 1952) have also been described. Presently in India, Myliobatis is found along the east

and west coasts, alld also in the mouth of G;ill,l{es and Chilka T,ake.

.<:lcl:\1; ()sll~jcllIJIY(':\

Subclass AClilJopterygii

Infraclass Teleostei

Superorder Ostariophysi

Order Siluriformes

Family Bagridae

GENUS F AJUMIA Stromer 1904

FAJUMIA MENON! sp. nov.

Plate I, fig. 3

Etymology :-For Dr. A. G. K. Menon, Superintending Zoologist, Zoological Survey of

T --:c1 ~ r, 1 ..

J.llQla, ualCutta.

Holotype :-1... U. V. P. 11140, a skull.

Horizoll alld locality :-llluisil-grey shales of l~abia Stage (Middle l':ocelle) at Godllatad.

Diagnosis :-A large siluroid with skull similar to that in Ariu.\' but dillcrs in neither

having the complex vertebra fused with the basioccipital nor in possessing the median supraoccipi-

tal depression; in these two characters it resembles F/!jumia .\'chiveinfurihi Stromer (i\ndrcws 1 ~)O6,

pp. 313-314) from Upper Eocene of I"ayum (Egypt). It differs from 1"- schweilljurtht in being wider

and having no rows of specially enlarged tubercles on the sides of the folltanelle ; also the anteri-

or border of the ethmoid is concave and not straight as in F. schu}eilifurthi. It diflcrs from the

skull of Socnopaea grandis Stromer from Upper l':ocene of Fayum (Egypt) in being smallcr in

length and lower in height; also the tubercles are big,ger and not arranged in radiating rows,

The median depression is smaller with only olle fontanelle instead of two as is ,S'oc/lopaea.

Description :-Skulllong and wide within the size variation of 11'ajumia schweinfurthi, the

type species. Only the following bones are preserved in the skull: the interneural plate, supra-

occipital, pteriotics, sphenotics, frontals, left prefrontal, left part of ethmoid, basioccipital,

paraspllenoid and complex vertebra.

A. ,~AHNI AND V. P. J.UISHI~A

The skull-roof is flat but elevated in the posterior part of supraoccipital as in the type

species. The interneural plate at the dorsal posterior extremity of the skull is large, its anterior

end being firmly attached in between the posterior extensions of supraoccipital j the length of

the plate is 9 cm. Anteriorly the plate is narrow but widens posteriorly. Skull-roof is ornamen-

ted with large, prominent, coalesced tubercles, without any regular pattern on any ccpll,uic plate.

The supraoccipital is small, sloping c.lnteriorly as well 'itS laterally j its posterior part c.llollg with

the interneural plate forms a median longitudinal ridge. Posteriorly the processes of the supra-

occipital enclose firmly the narrow anterior part of interneural plate j antcriorly tIle Rupraoccipi-

tal wedges in between the backwardly directed proccsseR of thc f(ontals j I,lt<~rally it i~ bOUlld<~d

by the sphcnotics, pteriotics and epiotics. Dorsal surf(l.cc of the supraoccipital is ornamentcd by

rounded tubercles. There is no median depression or posterior fontanelle on the supraoccipital.

Sensory grooves are shallow on the anterior side of the bone.

Pterlotics are small, sub triangular, attached internally with the supraoccipital, anteriorly

with sphenotics, and posteriorly with epiotics. 1'heir junction with sphenotics and supraoccipital is

convex, while that with epiotic is almost a straight line.

Sphenotics are elongated and about 9 cm.long, their posterior extremity makes the shape of

a V, one arm of which is in contact with the supraoccipital and the other with the pteriotic.

The frontals are elongated with a mediall depression anteriorly, leading into a fontanelle.

Posteriorly the frontal is narrow and its junction with the supraoccipital is a simple zigzag line.

The tubercles on the frontal are small and isolated and are neither arranged in a linear manner

nor do they form two double rows of specially enlarged tubercles on either side of the fontanelle

as in F. schweiifurthi. Frontals are bounded posterolaterally by sphcnoti<:s and antcrolaterally hy

pre[rol1tals. Only tile left prcfrollt,ll i~ rr<~RI~I"VCO ill tll<~ RPI:cillll'll be,ll'ill~ I,u'gl', l:l('~NY ~pacl~o

tubercles. '!'he anteriormost bone is the ethmoid which is T -shaped, having a small alltcrior pro-

cess and a long posterior process enclosed in between the anteriorly directed pro(:c~S<~R (J the [ron-

tal. '-[he anterior border of the ctllmoid iR slightly COIICaV(~ ,lllO 11()t str,li~llt ,I.~ i,l /1, ,\(:h7(!(~il!lilrlhl.

Tllc dorsal sur[.lcc of tlll~ CtlllllOid i~ witlll)ut tllbl~I'cll'S.

Basioccipital is broad, thick and swollcn at its posterior side bearing two condyles j ante-

riorly this bone narrows and joins the parasphenoid with a collvex suture. 1'he "crucifix" is not

complete, as the post-temporal bars are 11Ot preserved. Anterior to tile position or tIle

right post-temporal bar, adjacent to the basioccipital condyle, a roughly circular but deeply

eroded otolith is present. The parasphenoid, situated anterior to basioccipital is c.llmost as broad

as basioccipital but more compressed dorsoventrally. The complex vertcbra is not fused with

the basioccipital, differentiating the skull from that of Arius. The complex vertebra is long and

narrow. The allterior face of the centrum is wide and concave, the ventral surface of the

vertebra is deeply grooved for the dorsal aorta ,as in genotype.

FAJUMIA MISltAI sr. novo

Plate I, figs. 4-5

Etymology :-l,'or Prof. R. C. Misra, Head of Geology Departmellt, J...ucl<now University.

Holotype :-L.U. V.l>. l1142,.l skull.

J'>aratype :-L.U.V.P. 11142 a, a part of the left pectoral girdle, found associ..ltcd with the

holotype.

Horizon and locality :-Bluish-grey shales of B;tbii.l St;'lge (Middle f~ocellC) at N;lreOi.l.

Diagnosi.f : -A large siluroid with skuJI similar tu Arius but' differs in not having the

complex vertebra fused with the basioccipital or a median depression arising from the supra-

occipital region. Skull nearly of the same size as in 1". menoni sp. novo but slightly shorter and

wider than in 1': schl.vei1ifnrlhi. Unlike in 1': sc!t,veiltjjlrlhi, 110 row of especially l;lrge tulJerclcs on

either side of median fontanelle as in J.: menoni sp. novo I". misrai sp. novo diflers from 1': menoni

sp. novo in having a lower supraoccipital, more compressed and wider basioccipital, and in the

presence of continuous longitudinal ridges of tubercles on the dorsal surface of the skull. It

differs from ~'ocnopaea grandis from Upper Eocene of Fayum (Egypt) in having ~l smaller and

lower skull, larger tubercles joined only along mid-dorsal line of skull, and in having only one

mcdian fon tanellc.

WESTERN INDIA TERTIARY VERTEBRATE,S' 9

L?"scriptiIJn :-Skl!ll is l<;,ng, similar in size. to th.at of the F. meno?i sp: nov., but slightly

shorter In length and a lIttle wider than F. ,I'chwetnfurtht. Only the folloWltlg bones are preserved

in the skull: the interneural plate, supraoccipital, post-temporal, pteriotics, sphenotics, frontals,

basioccipital and parasphenoid. The skull r,)of is almost flat except for being sli.(1;htly higher in

thc supraoccipital rc.(1;ion; it is lowcr ill this part than in /I', 1/li!/lllni sr. novo 'fIle intcrncural

plate is firlllly att:lchcd ill bCtwcCll tIle posl(~riol' processes or tllC supr:'loccipital, lllC suturc being

the same as in F. menoni sp. novo Supraoccipital n~arly as long as in F. menoni sp. nov., longer

than widc, slopin.(1; very .(1;cntly towards thc clntcrior and mar.(1;illal sides. '['ubcrcl<:s <Ill tliC supra-

occipit:ll :lrc promilient and arc joillcd OIl tIle mi<l-clorsal lillt: [C)('mill/!; COlitillll01lS Ic'llp;itu<linal

ridges running from the intcrncural plate through the supraoccipital towards the i'rontals. 'i'here

is no posterior fontanelle or median depression on the supraoccipital. A pair of sensory grooves

starts.from the supraoccipital and extends towards the frontal becoming deeper at the posterior

margIn of the frontal.

Only the right post-temporal is preserved, and consists of a small, nearly quadrangular

bone. The tubercles on its dorsal surface are prominent, rounded but unjoined. The post-

temporal slopes towards the lateral margin of skull, and at its outer extremity there is a distinct

tuberosity for the articulation of the upper part of supraclavicular. The parietal is a small bone

placed in betweell the supraoccipital alld the post-temporal, anterior to: which is the pteriotic

which is broken along the outer margins, the suture between the pte,riotic and supraoccipital

is a straight line. The tubercles on the p:el iotic are distinct and unjoined. The sphenotic which

is situated anterior to the pteriotic and external to the frontal, is an elongated bone, the posterior

extremity of which is 'V'-shaped as III F. menoni sp. novo

1'he frontals are elongated with a median depression at the juI)ction of their anterior

side which leads into a fontanelle. The median depression is slightly anterior in position than

in F. menoni sp. novo The tubercles on the dorsal surface of the frontal are slightly smaller than

on postcrior bones but distinct and form continuous ridges on the median side. On tile side or

the fontanelle coalesced tubercles also form continuous low ridges, but tubercles arc ncither

enlarged nor placed in two rows as in F. schweilifurthi. The prefrontal and ethmoid bones are

not preserved. .

On the ventral side of the skull, posterior to the basioccipital, the complex vertebra is

not fused with the basioccipital as in other species of Fajumia. 1'he basioccipital is more dorso-

ventrally compressed than in F. menoni sp. novo but wider. The basioccipital condyles, though

weathered, appear to be smaller than in F. meniJni sp. novo Tho post-temporal bar is not preserved,

thus the crucifix is not seen on the specimen. The otolith of right side is in its position in

the depres~ion formed bet,,\'een the basioccipital and the post-temporal bar. The otolith has a

diameter of about 2.5 cm but is quite weathered and no details can be made out. I he

parasphenoid, anterior to basioccjpital, is also wide and dorso-ventrally comprcs~ed but incomplete

on its anterior side~

Measurements in centimeters

29.5

24.5

23.5

Maximum length of the preserved skull

Length from basioccipital to the anterior part preserved

I..e!lgth from supraoccipital to the anterior part preserved

Pectoral girdle :-L.U.V.P. 11142 a is a part of the pectoral girdle of the left side with

the head of pectoral spine attached. It was also found associated with the skull (L.U.V.P.

11142), the holotypc. Of pectoral girdle, only the c1cithrum is preserved, the dorsal surface of

which is ornamented with pr.ominent ~ubercl.es, some ?f which are joihed but withouf',~;ny

regular pattern. The arm of clelthrum whIch artIculates wIth the post-temporal bone is ipcom-

plete and without. any ornamentation. The glenoid facet is weathered and partially contains

the head of the pectoral spine. The shaft of the spine is broken but appears to be strong and

stout, the articulating facet (head) of the spine bears two condyles. There is a small opening

on the ventral side of the head of the spine. "

10 A~ SAHNI AND V; P. MISHRA

Fa'mily Bagridae '

GENUS SOCNOPAEA Stromer 1904

,'SOCNOPAEA HOl~AI sr. novo

Plate II, fig. I

E:tymology :-For late Dr. S.L. Hora, a leading Indian ic~thyologist:

.fJolotype :~L.U.V.P. JII4.5, a skull, the only known specimen.

Horizon and locality, ;-Ossiferous gypseous shales of Babia Stage (Middle Eocene) 'at

H.arudi.

Diagnosis :-A large siluroid, resembles Socnopaea grandis Stromer from Upper 'Eocene of

Egypt. It. differs Fajumia in the same characters as S. grandis, i.e. in having a larger size, a

different type of ornamentation consisting of many small tubercles arranged in low and thin

radiating ridges; in having no rows ofspecially enlarged tubercles lateral to the median depres-

sion and haS" a posteriorly projecting supraoccipital. ~upraoccipital is high in the p~steromedial

part sloping steeply on the lateral sides; the basioccipital is wider than in FaJumia. The

skull differs from that of S. grandis in being slightly smaller in length and wider posteriorly,

also the median depression does not possess two fontanelles. The sculpturing on the cranial

plates also resembles Macrones aor (Bag~us aor) from Siwalik Formation. of India (Lyd.ekker

1886b) to some extent, but J.\1acrones aor IS more than 50 percent smaller m length havmg a

median depression extending the length of the frontals and posteriorly to the anteriormost part

of the surraoccipital, in contrast to the restriction of median depression: only to frontals in the

new specIes.

Description :- The skull is long and wide, larger than 11'ajumia but slightly smaller th..li

S. grandis. 'l'he skull is gypsified and the sutures between the cephalic plates are obscurc. rrhc

cephalic plates are thinncr than in tIle spccics of li'tljumia alld Ariu.r. rfllC skull roof slopcs aillc-

riorly and considerably on the lateral sides from the median part of the supraoccipital. The

tubercles are small, usually coalesced, forming continuous longitudinal and radiating ridges from

the supraoccipital towards the frontals, thus the ornamentation differs from that of the species

of the Fajumia and Arius in the collection. In ornamentation the skull resembles wilh that of

Macrones aor from the Pliocene Siwalik rocks of India (Lydekker 1886b), but the skull of S. horai

sp. novo is more than 50 percent longer and does not bear a median depression on supraoccipital.

The supraoccipital is elongated, longer than wide, sloping anteriorly and on the lateral

sides, having a high median ridge on the posterior side formed by an internal keel. The supra-

occipital projects posterior to the rest of the skull as in S. grandis. On the right side, the parts

of the post-temporal, parietal, pteriotic and sphenotic are preserved but gypsi{ication has obscured

the sutures and the details of the bones cannot be made out.

Anterior to the supraoccipital, the paired frontals are present, which are long and wide

but broken laterally in the specimen. The median depression in between the two frontals js

wide, long and deep and there is only one fontancll(~ 011 its autcrior sidc uuljk(~ two in tilC typc

species. The frontals are characteristically ornamented: on eac.h of thefrontals the tubercles

join to form low and fine radiating ridges somewhat assuming an acicular pattern on the

anterior side.

On ventral side, the skull is more gypsified than on the dorsal side. 'l'he basioccipital

and parasphenoid are wide, wider than,in }'ajumia and are dorso-ventrally compressed.

.-Remarks :-The familial status o[Fajumia and Socnopaea described originally by'stromcr

(1904) from Qasr-el-Sagha Formation (Upper Eocene) of}l'ayum (Egypt) is uncertain. Andrews

(1906, pp. 313-315) and Zittel (1932, p. 158) placed these genera in the family Siluridae. Later,

Peyer (1928) while discussing the validity of these genera included them in the family rrachy-

suridae, a suggestion which ,was later accepted by Jayaram (1955). Jordan (1923) and recently

Romer (1966) classified these genera in the family Bagridae, an assignation which has been

followed in the present work.

WESTERN IN.DIA TERTIARY VERTEBRATES 11

Family Ariidae

GENUS ARIUS Cuvier and Valenciennes 1840

Al{IUS KUTCHENSIS Rao 1956

PJ;ll(! J I, lil(R. 2-1

.lv/alerial :-L.U.V.P. IJ036, impression of the dorsal surface .of skull from Harudi;

L.U.V.P. 11047 and L.U.V.P. 11048 a and b, the sku]] and pectoral spille and ]>arl or pectoral

girdle from Nareda; and L.U. V .P. 11148, the l_'oslcrior I>art or skull from Godhatad.

Horizons and localities :-L.U.V.P. 11036 from shell limestone of Babia Stage (Middle

Eocene) exposed near Harudi; L.U.V.P. 11047 and L.U.V.P. 11048 from bluish-grey shales or

Babia Stage (Middle Eocene) exposed near Nareda;.and L.U.yT.P. 11148 from bluish-grey shales

of Babia Stage (Middle Eocene) exposed near Godhatad.

l~evised Diagnosis :-Skulilarge, much larger than of Arius crassus from Upper Eocene of

Barton, England (Newton 1889), A. iheringi from Upper Eocene of Tertiary lignite of Brazil

(Woodward 1901), A. egertoni from Mictdle Eocene of Bracklesham, Englaud (Woodward 1901),

as well as thc Recent species A. gagorides from India (Newton 1889), but nearly of Ule same

size as of the species of Arius from the Siwalik Formatian of lildia (Lydekker 1886b), from

which it differs in the pattern of ornamentation.

Skull flat but elevated in the supraoccipital region by a high triangular keel. Dorsal

surface ornamented with prominent and rounded t~1berc]cs, usu~11y so arranged as to form

radiating ridges. Supraoccipitallongcr tllall wide, l)rojcctiog 1;lr bcllillu tll(~ r(~Ht or tile skull.

Median depression quite prominent and deep, starling from the middle of the supraoccipital

region running towards tllC frontals. Orbits large; basioccipital robust; crucifix present.

De.\'criptioll :-I..V.V.P. 1101.7 is a partially preserved skulltlc:tr1y (,[tllc Satlli~ size as t1tc

ho10type' (Rao 1956) but slightly higher. Tlle following cephalic plates are preserved in the

specimen: supraoccipital, epiotics, pteriotics, sphenotics, frontal, basioccipital and part of para-

sphenoid bone. The ornamenation consists of small rounded tubercles, most of which have got

eroded. Supraoccipital is slightly smaller than o[ the type specimen (GSl tyP{~ 110 175:~9, stored in

Indian Museum, Calcutta). 'l'he anterior contact of supraoccipital with ti,e frontal is somewhat a

zigzag suture as in A. iheringi (Woqdward 1901). The median depression arises from nearly

the middle of supraoccipital be~ng narrow posteriorly, widening anteriorly, continl,ling towards

the frontals leading into a fontanelle anteriorly. A pair of sensory grooves for slime canal

extends on the side of median depression from suproccipital towards frontal having a divergent

angle of 35°. Epiotics and pteriotics arc smaller than other cephalic plates. rrlle sphenotics, situated

antero-Jateral to supraoccipital, are elongated enclosing the backwardly directed processes of

the frontals in between them. Frontals are slightly wider posteriorly tltan in holotype. Basi-

occipita] is robust, but condyles are broken. rl'he crucifix is not seen as the complex vertebra and

the post-temporal bars are not preserved.

Associated with L.U.V.P. 11047, the skull, were found a fragment of cleititrum of pcc-

tora1 girdle (L.V.V.P 11048 a), and an isolated pectoral spine (L.U.V.P. 11048 b), both of

right side: Pectoral spine is str:ong;-the head has two articulating condylar facets; the sh",tt is

laterally compressed with well-marked longitudinal striations. Botll tIle anterior cllld posterior

edges are with backwardly directed dcnticlcs. In tltc ltcad rcgiott, tile 1l1edi:ltl fUSSi.l is situated

on the posterior side.

L.U.V.P. 11036 is the impression of another skull of this species which is ]arger in size

than the holotype. The impression of the skull possesses supr.loccipital, epiotic, pteriotic, sphe-

notic and frontals; part of circum-orbital series on the outer side is also preserved showing the

position of a large 'orbit. The skull is ornamented with prominent rounded tubercles, arranged

in a radiating pattern. Supraoccipital is longer than wide, prujecting behind the rest of the

skull. The median depression arises from the middle of supraoccipital continuing towards the

frontals and is deeper than in holotype and L.U.V.P. 11047, described above. A sensory groove

on each side of the median depression arises from the supraoccipital diverging at all agle of 40".

12 A. SAHNI AND V. P. MISHRA

Still another specimen, L.U V.P. 11148 shows only the hinder part of the skull; the dorsal

surface of skull has weathered off, exposing the internal keel on the posterior si;:Ie jn between

the V-shaped ridge on the ventral side of supraoccipilal. The basioccipital is nearly of the same

size and shape as in L.U.V.P. 11047.

Discu.rsion :- The larger number of the specimens jn the prescn t colicci ion makcs it

possible to clarify certain points which were obscure in the holotype (Rao 1956). In Rao's

(1956, pl. 28) original descrjption of the holutype, the bone marked 'FR' is actually sphcllotic,

and anterior to the supraoccjpjtal, only a part of Ihc frontals is prcscnt wlljcll has l>('(~n Rupposcil

by Rao (1956, p. 181) as ethmoid; furthermore the ethmoid is never in contact with supraoccipital,

as supposed by Rao. Also, the bones marked 'SPRaT' and 'PTOT' are a single unit, the

pteriotic bone.

Arius is found both in fresh anti marine water conditions.

A. dussumierl) is more abundant on the Malabar coast. In India the living species

Infraclass Holostei

Order Pycnodontoformes

Family Pycnodontidae

GENUS PYCNODUS Agassiz 1833

PYCNODUS sp.

Plate II, fig. 5

Material :- Isolated splenial teeth (L. U. V .P. 11173).

Horizon and locality: -Shell limestone of Babia Stage (Middle F.ocene) at Haruui.

DescriPtion :-The splenial tooth of [:Jycnodus from Kutch is small and discoil.lal. Ils oral

surface is rugose and indented. '!'he basal surface is roun<l with concentric ~r()wth lines (lllU a

small pulp cavity. 1'he tooth resembles with that of Pycnodus toliapicus Agassiz (figured in

Woodward 1908, pl. 1, fig. 11), but is smaller in size. The vomerine teeth of Pycnodus lametae

described from Lameta beds (Woodward, op. cit.) are bigger in size than the splenial tooth from

Kutch. Remarks:-A skull anJ three vomerine teeth of Pycnodus lametae have been dcscriberl from

the Lameta beds (Cret;lceous) of Dongargaon in Madhya Pradesh (Woodward 1908). and teeth

of Pycnodlls sp. have been described from the Uttattur beds (Cretaceous) of South india (Stoli-

czka 1873). The report from Kutch extends the geological distribution of the genus into the

l':ocene in India. rl'he only other genus of Family Pycnodontidae described from India is

Coelodus, fro~ the Middle Eocene of Siju Limestone, Garo Hills, A~~am (Menon and Prasad

1958), and from Upper Cretaceous of Ranikot and Sind (Prasad and Rao 1958). It is of

interest to note that the pycnodont fishes are not found in the beds younger thaI! latest I~ocene

period. Class Reptilia

Order Crocodilia

Prior to current work, fragmentary reptilian remains were reported by Gran t (1857),

Wynne (1872), Feistmantel (1876), and Lydekker (1876b, 1877, 1879) from the Mesozoic and

Tertiary rocks of Kulch. During the present invcstigation, a large number of crocodilian

fossil remajns were recovered from the rocks of Babia Stage (Middle t:ocene) exposed near

Harudi and Nareda. Crocodilian remains which are gypsified but retain their original struc-

ture, are represented in the collection by skull, mandibles, vertebrae, and coproliles; Eocene

crocodiles in Kutch are referrable to the genera Tomistoma and Crocodylus. Previously, only a few

fossil tomistomin genera were known from Asia; Eotomistoma multidentata from Upper Cretaceous

of Chil1a (YOUllg 1964), Tomistoma petrolica from Upper t:ocene of China (Yeh 1958) and

l~hamphosucllus crassidens ii"om Siwaliks (Pliocene) of India (Lydekker 1886b). rIlle presence of

Tomij.tCJma in Middle Eocene of Kutch is interesting in as much as the Palaeogene species of the

genus were so {Oar known only from Egypt. T petrolica from Upper Eocene of China, though

originally described as a tomistol11in, is probably a juvenile gavial.

WESTERN INDIA TERTIAR11' VERTEBRATES 13

The following features have been taken as the basis for the generic and specific differen-

tiation of the Kutch crocodilian remains: size and shape of the cranium, shapes of premaxillae

and narial opening, arliculation of nas:lls Wilh premaxillae, number and structure of teeth, length

and the widlh of the mandibular symphysis, and the fuuction of the splenials in the formation of

the symphysis.

Suborder l~usuchia

li'amily Crocodylidtle Cuvicr 1807

Subfamily Tomistominac Woodward 1932

GENUS r!'OMIsr!'OMA Muller 1846

TOMISTOMA TANDONI sr. novo

Plate 111, figs. 1-3

Etymology :-For Dr. K. K. Tandon, Department of Geology, Lucknow University.

Holotype :-L.U. V.P. 11062, mandible with 14 alveoli, containing symphysis.

Para types :-L,U.V.P. 11062 a, the skull lacking the premaxillae and portion posterior to

orbits; and L.U.V.P. 11062b, ten vertebrae. 1'he halotype and the paratypes belong to the same

individual.

Horizon and locality: -Bluish-grey shales of Babia Stage (Lutetian) at Nareda.

Diagnosis :-Skull and the mandible tomistomid; skull elongated, longer than T: gavialoides

(Andrews 1906) and T schle,l!;eli (Konj uk<?va in Orloy's treatise 1964-, p; 522), but srlightly

smaller than T. eggenburgense (Toula and Kal11885), while the Portugese MIocene form T: lusz-

tanica (Antunes 1961) is 20 percent lonl?jer ~han Kutch spe.cies: The. rostrum gradu'.tlly tapers

anteriorly as in T: eggenburgense, T gavzalozdes and T: lu,\'ztanzca. WIdth of the cranIum at the

level of the anterior border of the orbits is greater than in 'T. kerunen,\'e mId T (Imericana,. but

slightly smaller. than in T eggenb~rgense, while in 'l~ lusitanica.it is or~e aud a half.times wider at

this level than In the Kutch specIes. Nasals deeply penetratIng as In T: champsozdes (Lydekker

1886 a) between the premaxjllae, and not reaching the exte~nal narial opening-. Estimated

number of the total teeth 20-21, U ppcr and lowl~r t<~(~th intcrlocking. M;llldil>lc ~llortcr ill

length thCl;n in T lusitallic.a' ~ americana (Scllar~s 1915),. T: africallum (An~rews 1906) an.d

T champsozdes. The mandIble IS nearly as wIde as In T lusztanlca. The mandIbular symphysIs

shorter in length than in T: lusitanica, T africanum, and T: americana, and extends posteriorly

up to 11 thJ12.th tooth as in ~ lusitanlca, unlike in T schlegeli a.nd T: afri~anum, in ~hich

the symphysIs extends posterIorly up to the 14th tooth, Splenlals enter I1lto mandIbular

symphysis exteuding anteriorly upto 8th tooth. 4th mandibular tooth is largest and is situated

at a higher level than the others.

Description :-L.U.V.P. 11062, the ru:'lndible, is complete on the alltcrior side, comprisiIlg

of both the rami along with the mandibular symphysis; the ri,ght ramus is preserved upto the

14th tooth, while the left ramus is preserved upto the 10th tooth. 'rhc articular region

comprising the right surangular and left sur;tn,l!;ular-;lrticular bones, LII()IIl{11 II"do'lbtcdly

of the same individual, was found dctachcd from LI1C dcntary alld i~ also prcscrved. ,!'hc

mandible is shorter in length than in T lusitanica, T. americana, T africanum and T champ-

so ides. The lateral margins of the mandible are nearly parallel from thc anterior end up to the

8th tooth, posterior to which bolll tile rami start (liv<:r.l!;illl{, lII<: a<:tll;II s<~I);ll'.lli<)ll l:lkcs pl.ll:C

posterior to 11 th tooth with a divergence angle of about 30) as in r: {tji'icallum. 'l'he auterior

extremity of the mandible is bluntly rounded. Sockets for the first mandibular teeth occupy

most of the area of the anterior end of the mandible and are large, uearly of the same size as

the second and third mandibular teeth, being smaller than the fourth manuibular teeth- PositiolIS

of the alveoli show that the anteriormost two teeth were large, projecting forward and outward

towards the upper jaw. Mandible widens slightly at the point of the second tooth which is

situated ,lrt<~r ;1, di:lst(~ma of 1.5 (:m from thl~ first tooLI1. '.l'lll~re i~ a promiuellt IIOtch

after the second tooth, which probably received the third premaxillary tooth. 'The socket for

the fourlh mandibular tOOtll is largest, circular in outline aud plilced at a higher level than the

otllcrs, i\lv<~oli for 5tll to 8th tl~ctll arc cquiuistalilly silllatl~d allu ilrl~ llc:lrly equal ill sizc. 'lnc

14 A. SAHNI AND V. P. AlJI,s'HRA

mandible starts widening and diverging pvsterior to the 8th tooth. As in T lusitanica, the

mandibular symphysis extends posteriorly up to the 11 th tooth, after which both the rami

separate out. 'fhe splenials form part of the symphysis, and extend more anteriorly, i.e. up to the

8th tooth than in T schlegeli and T lusitanica, where splenials extend anteriorly only up to 9th or

the diastema bctwecn the 8th and 9tll tooth; in T rifricullum the :oplcllial:o project alltcriorly Olily

up to the 10di tooth. r!'he mandibular symphysi:o is shorter in length than in 1: lUj'llauica,

T africanum and 7: americana. The mandibular alveoli posterior to lath tooth are not promi-

nently seen on the spccimcn.

'l'he left surangular, angular and articular bOllt~':O arc preservcd (1>1. lIJ, Jig. 3), but tllc

fragment containing the external mandibular opening is wanting. rrhe surangular j:o long,

thin and high bone being about 12.5 cm high, a little more than in 1: africanum. r!'he angular,

lying postero-ventrally, forms a shelf-like border on the internal side of the mandible. rrlle

articular is situated internal to the surallgular bonc and dorsal to thc angular bolle. 'l'lle

glenoid fossa forming the articular surface for the quadrate is quite thickened, it is longer and

wider than in T aJrican.um. The articular surface is divided by a ridge into an inner and an

outer portions as in T. schlegeli (Andrews 1906, pp. 270-271) and differs from the condition in

7: ajricanum where the articular surface is simply concave. The articular is produced posteriorly

into a narrow process which is incomplete at its posterior extremity.

In ventral view the junction of the two rami is prominent. 'l'he lateral marg'jn of the

alveolar border of the mandible is sinuous, the concavities on the mandibular border are for the

reception of the upper teeth, The mandible is shallow anteriorly, becoming about 8.0 cm high at

the posterior extremity of the symphysis, while at the surangular region it is about 12.5 cm high.

Skull :-L.U.V.P. ll062a, the skull (Pl. JIl) fi1!;s. la, b) is partially prrs(~rve(I, r!'!IC

specimen consists of the posterior extensions of premaxillae, maxillae, nasals, prcli'olltals,

lachrymals, frontal and palatine bones. Anteriorly, the skull is broken from the poillt where the

notch or constriction in the rostrum occurs for thc reception of the fourth m;llldibular tooth;

and thus the alltcrior part of prcm;\xill;lri(~s ;lIOIJ~willi tll(~ (~xt(~rn;ll 11;lri:ll (11)c.~llill1!; is w;lJIlillg,

Posterior to the orbital region, the major part of the dorsal cranium is wanting, except [or the

median part of the pterygoid, the internal narial opening and the occipital condyle. '

The estimated length of the skull from its anterior extremity to the occipit~u condyle is

about 83 cm ; lilus thc skull of thc Kutch species is lollger tllC:lll thc skulls (I[ l{c(:l~llt '1: jchlegeli

(Konjukova 1964, p. 522; in Orlov's treatise) and of Egyptian Eocene T, gavillioides (Andrews

1906), but the skull of T, eggenburgensefrom the Miocene of Austria (Toula alld Kail1885 in

Zittel 1932, p. 365 and Antunes 1961, p. 41) is only slightly longer, while the Porlugcse Miocene

form 7: lusitanica (Antunes 1961) is 20 percent longer. The width of the cranium at the level

of the anterior border of the orbits is greater than in T kerunense and T gavialoidesJ. ol1ly slightly

more than in 7: schlegeli and T americana, while T. /usitanica is one and a half times wider at

the pre-orbital level than the Kutch species. ~rhc rostrum gradu:tlly t:ll)(~rs antc.'ri(lrly som(~wllat

as in T: schlege/i and is not suddenly constrictcd in front of tile orbits "lS in '1~ e,ggc/lburgellse,

7: gavialoides and T lusitanica. Between the 4th and 5th maxillary teeth tl'le margin is convex,

posterior to which there is a slight concavity in the border of the rostrum, 'I-he premaxillaries were

.long, ncarly o[ the same si~c as ill 1~ ,~avialoide,l", but :lrc sllOrtcr ill 1(~II,'-c1.'lll tll,lll ill 1-. 1IJ"littlni(;a,

1: chtlmpsoides, 1: america/la, 1-. eggenbulgfllse .\nd CVCll I: ,\chl(!geLi. 'J'JlC prc.~maxill.\rics

extend back to the level of the 2nd maxillary tooth. 1'he posterior junction of the premaxillaries

with the nasal s is of "champsoides" type (K alin 1955, p, 773), i.e. the nasals deeply ente r

in between the posteric)r processes of tile premaxillaries. Nasals arc IOllg, tllctr post(~rior juuclioll

with the fron tal bone is not clear, but they appear to be longer than in I: jchlegeli, 1 ~ ,ftavialoides

and T eggenbur.gense, but are shorter in length than in 7: lusitanica and T. americana (Mook 19~ 1).

The nasals arc nearly as wide' as in T. lusitanica, and are wider than in 1-. gavialoides, T. schlegeli

and 7: eggenburgense. The nasals reach anteriorly up to the 1st maxillary tooth, and do not

appear to be touching the external narial opening. Posteriorly tile sutures between the nasals,

lacllrym;lls, prcfrolltals, :.\nd frontal i\rC not clear. '.I.11e inter-orbital sp:.\ce of thc frontal is about

4.75 cm, and is wider than in 'r. schlegeli, 'r. gavialoide,l", and cven T. lusitanica, but it is sli~htly

narrower than in 7: americana and 'r. e,{!;genhurgense (the measurements given by Antunes 1961,

p. 42). The maxi1laries arc long and broad, their sutures with the nasals are nearly straight

lines.

TERTIARY

\f/ERTEBl~.. 15

rES

WESTE'RN INDI..

On the ventral side of the skull tile premaxillC1ries reach posteriorly upto the diastema

between 1st and 2nd maxillary teeth; the junction between the maxillary CI.nd premaxillary

i.lppCars to bc silnilar to the prcvi.lilillg colldilioll ill 'I: J'chlegeli, , GIlly 12 sockets of the maxillary

leelh are present, which are nearly equal in size. l\ssuming the presence of atleast 5 premaxillary

teeth, to.~cther with 12 cxistilll!; sockets of tIle m:'ixillnry lecUl, all<l al I(~asl:~ In<,rl~ lcclll Oil

tile broken posterior part of the llli.lXillae which is not fully preserved, tile lotCll Ilumber of tt'eth

estill\:'lted is 20; the number is identical to that found ill rr: l~,~,!!,enbur.1!,ell.re, but is less by one

maxillary tooth from the 1~ schlegeli alld 2 to 3 m;lxillary tt't~LlllcRs frol\1 'I: ,!,'{llliflloidl1s,

'The maxillt)-palatiue suture is at tile Icvcl of tll(~ IOLIl maxillary l('clil but its exact ()riell-

tatioll canuot be n1.ade out. The palatine forming tile inllcr m:lrgin or thc pterygoid fossae

is narrow. Posterior to the palatine, the median part of tile pterygoid bOllC is preserved i.tt the

posterior extremity of which is seen a semi-circular OP(~llillg, Ille illt(~rll;llll;l..i;ll OP(~llilll!;. Poste-

rior to tilis opening are the occipital coudyles, nearly 7 cm wide.

Verlebrae:-There are ten vertebrae (L.U.V.P, 11062b) iu collcctiou, which were found

in close associatiou with the mandible (L.U.V.P. 1 ]062) and the skull (L,U.V.P. ] 1062a) of

T, tandoni sr. novo Of the vertebrar, only the centra ~re well preserved except for three, in which

the region dorsal to centra is also preserved (PI, T I, fig. 10). l'he vertebrae are robust and belollg to

the cervical regiolll res(~mbling in size alld shape lO tIle cerevical vertebra of T africanum figured

by Andrews (1906, pl. XIII, figs.. 2, 2a) which i~ slightly larger and with a shorter celltra. The

centra are procoelous, longer than wide, spool-sllaped; aulcrior end is concave, wllile tile poste-

rior end is a convex projecting dome. In tile middle, tile centra are narrower tllan at their

extremities, the width on anterior side of the centrum being 4.5 to :>.1 cm. Ventrally tile ccrit-

rum bears a narrow and small keel-like strllcture, the hypapophysis, Th,rH;u to tll(~ (:(:lltrlllll is

the Ilcural callal, llearly (:ircular ill <.:ross-sccliou; llic crusg-s(~ctiull t)[ Ilcur:.u (:all;ll ill '1: {U,.icfJllulll

is triangular. 'l'he neural arches enclosiug the ueural canal are stout and projec t anteriorly

as pre-zygapophysis, posterior~y as r 1?ost-zygapopllysis, ~nd dorsa]ly as. a b~ckwardly directcd

strong but narrow neural spIlle. tile prc-zYI!;;lpophys's II;lS two artlclll;\I.II\~ r;l(:(:ls, III 011(;

vcrtebra, Lllc articulalillg facets for tile head atlll tubercle (,f tile rib arc preserved on the

lateral side of the centrum. Allcasuremcnts in ccntimeterJ'

l\;land ibLe

Length [rom the anterior extremity to the 14th tooth

Width at the anterior end of mandible

Width at the anterior extremity of the splenials

I Width at the posterior part of the symphysis

Depth at the anterior end of the mandible

Depth at the posterior end of symph)'sis

Length of the symphysis

Depth at surangular region

-.Angle between left and right rami after symphysis

4B.0

6.5

10.0

16.5

3.0

B.O

3B.0

12.5

300

5' kull 72.0

11.0

83.0

Le11gth of the skull from the basioccipital to the anterior part prc~erved

Estimatccl lcngtll or premaxillae broken .trltl~riorly

Estimated lellgth or the skull from the basioccipit..tl to the antt:rior extremity

of the rostrum

Length of the skull from the allterior extrt:mity of the orbit to tll<: tip or

ro~trum (cstim,ttcd)

Width of skull at the pre-orbital region

Width or skull at the anterior end of maxillae

Width of skull at the level of 10 tooth

Estimated length or the prcrnaxillarics (incl. posterior extensions)

Lengtll of tile nasals

Maximum width of the nasals

Widtll of the in ter-orbital bar

57.0

20.0

8.0

14.0

lB.O

38.0

3.5

4.75

T africanum

5.B

(2)

8.0

4.5

(3)

8.5

5.0

( 1 ,

7.2 approx

'ertebrae

Length of centrum

Widt)1 of CClltrum

Height of vertebrae from

tip of hypapophysis to

tip of neural spine 13.5 12.5

A. S..4HNI AND V. P. MISHRA

Discussion :-- The genus Tomistoma was restricted in its distribution duril1g the Eoccne,

being found only in Africa (Egypt), and now from Kutch in India. An Eocene species from

China (7 petrolica Yeh 1958) may well not be a tomistomin, mainly on the grounds that in this

species the nasals do not articulate with the premaxillaries, a character similar to the condition

found ill gavials alld mentioned by Yell (1958, p.212), AIs() tll(' sknl1 ()r tll(~ (:llill(~s(~ !l1)(~(:i(~s is

much smaller in length than all the otller known species of 'r omistoma al1d may belolll:,';

to a young individual of the gavials. rrhus the Kutch occurrCllce is the first report or

Tomistoma from the Palaeogene rocks outside Alrica,

The only other fossil tomistomin genus known from Neogene of Asian continent is

Rhamphosuchus crassidens (Lydekker l886b) from the Pliocene Siwalik beds of India. The present

distribution of Tomistoma is restricted only into the streams of Borneo, Sumatra and lVlalayan

Peninsula.

Antunes (1961) has discussed in detail the origin and evolution of the genera of

Tomistominae, and has also given their distribution in time and space. Recently, Sill (1968)

discussing the zoogeography of the Crocodilia, has shown the radiation pattern of the Tomisto-

minae. Sill (op. cit.) maintains "~urasia as the cent re of first radiation of l'omistominae, and

the Indian tomistmin confirms this hypothesis. From Eurasia the subfamily diversified from the

main crocodilian line during late Cretaceous and migrated to Europe. Asia, Afric(} and North

America. The second radiation of Tomistominae took place during Oligocene or early Miocene,

the centre of radiation this time being Africa, from where five species of T omistoma have been

found in Palaeogene rocks.

Family Crocodylidae

Subfamily Crocodilinae

GENUS Cl{OCODYLUS Laurcnli

CROCODYLUS sr.

Plate 11, fig. 6

7GU

ReJerred Specimen:-I...U.V.P. 11135, an isolated premaxilla with alveoli for three teeth.Horizon

and locality: ..:..Ossiferous gypseous shales of Babia Stage (Middle Eocene) at

Harudi.

Description:-Premaxilla is large. The anterior narialopening is large and triangular with

apex dircctcd postcriorly. Sol:kcts for three premaxillary teeth arc prcscl1t. Pl(Jxim.u Clld of pre-

maxilla is blunt and broad, and nearly J 1.5 cm wide. 'fhe socket~ of the teeth are small and

placed close'to each other on the sides of the mid-line of the prem().xilla. The posi tions or alveoli

show that the first premaxillary teeth were small. 1"he sockets of second premaxillary teeth are

situated after a diastema of 2.2 cm from the sockets of the first premaxillary teeth. These sockets

are larger and placed at a higher elevation than the first. Sockets of 3rd prcmaxillary tect]l

are very close to the secolld and are the largcst. 'fhe nasal does not appear to cntcr thc anterior

narial aperture, hut abuts at the posterior margin of this opening.

CI~OCODILIAN COPI~OLIT]~S

IJlate 11, figs. 7-8

Four crocodilian coprolites (LU. V.P. 11139) were collected from ossiferous gypseous

shales (Lutctian) expo~ed near Harudi. 'rhree coprolites are gypsiocd, while the fourth Olle is

impregnated with iron oxide. rrhe smallest coprolite is 8.5 cm long, while the largest one is

13.5 cm long. All are curved, three of them having distinct tapering ends. Surfaces of the

coprolites are smooth except that of ferruginous one which shows sphincter-pinched marks deve-

loped during excretion. Longitudinal striations are not seen on any of the specimen. The

coprolites from Kutch were identified as crocodilian from the illustrations and the description of

the crocodilian coprolites from the Palaeocene rocks of China (You ng 1964).

WESTERN INDlf

TERTIARYV

ERTEBl~ATES 7

Class Mammalia

Order Cetacea

Suborder Archaeoceti

'l'Ilc bulk of the collcctiOll OUl..lillcd from tIle Middlc Eocellc (Luleti..lll) rocks o[ WCSlCfll

Kutch collsi~ls of the remains of primitive cetacealls-the Archaeoceti and Odontoceli. Of

these groups-the Archaeoceti is a dominant and divcrsified f!,roup makil1g taxonomic idcntifica-

tion a difficult task. III all 16 skulls and mandibular fragmerlts (lS well ;lS rlumerous po~t-

crallial elements have been recovcred from the Babia Stage (Luletian) [rom tllrce sep..\.rate

localities -Harudi, Nareda and Godhatad (Fig. 1).

For the most part, the remains of cetaceans arc highly gypsified, so mucll so tllat irl many

cases the cranial sutures have been completely oblilcraled and teeth complctely replaced by

gypsum. However, a large number of specimens are still available particularly from the hard

fossiliferous shell limestone, which are well preserved. This is in sharp contrast to the specimens

recovered from the overlying gypsiferous shales which are more susceptible to the effects of

gypsification.

III gcllcral, the Kutch ccl:accallS resemble those from the Cairo and It'ayum region in

Egypt and from southern Nigeria. However, there are some important differences. Apart

from the presence of the common genus Pr%cetus, the Indian archaeocetes are characterised by

rather primitive characteristics in retention of the third molar, three-rooted posterior premolars

and molars, retention of inner cusp (protocone) and in possessing unserrated premolars and

molars One of the puzzling features of the Indian archaeocetcs is thc V,tri,ltioll ill tIle lcnglll of

the mandibular symphysis. In previously described archaeocete genera, nowhere does the mandi-

bular symphysis extends posterior to Pa: anterior root of Pa in Pappocetus lu,gardi, posterior root of

P2 in Prozeuglodon isis (Prozeuglodon alrox Andrews 1906) and Zeu.t:lodon o.firi.f Andrews 1906. In

Kutch archaeocctcs on the other IIClIIO, tIle sympllysis iN f()IIII(1 UJ I)l~ Ili!!;llly v,tri,u)lc ill

character. Only the m,\nuibular fragment, L. U. V. P. 1100'2, J.Jrolocelus .5loani (Sahni and

Mishra 1972) has a symphysis extending posteriorly to the second premolar. In the remaining

Kutch archaeocetes, the symphysis extends up to the posterior root of P:1 arId in some cases even

more posteriorly.

There are three mandibular fragments whose familial status is ullccrtain because of the

variable extensiOll of mandibular symphysis, extending posteriorly at least until P 4, after which

the right and left rami diverge and separate out. These specimens may indicate an intermecliate

position between Archaeoceti and Odontoceti. The symphysis extends posteriorly in L. U. V. P.

11061 (Pl. V, figs. 3 a, b) from bluish ,grey shales of Babia Stage at Nareda till anterior

border of P 4; in L. U. V. P. 11138 (Pl. V, figs. 4 a, b) from bluish-grey shales oj. Babia Stage at

Godhatad till anterior border ofP" as in I U. V. P. 11061, and in L.U.V.P. 1.1132 (Pl. V, figs.

5 a, b, c) from foraminiferal limestones at Nareda till the posterior root of P4 dorsally and

till posterior root of M1 ventrally. The extended symphysis is obviously a cetacean specialisa-

tion diverging sharply from the condition in the ancestral mesonychid stock. It is interesting

to note, however, that in t11OSl~ illstances, where tIle skull anci mmldibul;\r Ji'agml~llts llave been

found together, the dentition still retains a primitive character in tllat the protocolle is stili a

distinct cusp, the teeth do not possess secondary cusps and the molars still retain three roots,

with no proliferation in the number of the teeth.

'rhe following features are of the taxonomic value in archaeocetes: in skuLL, tile size, shape

and elongation of the rostrum, relative position of the external nares, relative hcights of the

sagittal and lambdoidal crests, width of the occipital region and shape and size of tympanic

bulla; in mandib~e, the size, degree of convexity of the rami, presence or absence of the groove

near the ventral border of ramus, length of the symphysis and diversion and separation of thc:

individual rami; and in dentition, length of dental series, number of the teeth, degree or molari-

sation of premolar, dimensions of teeth, diastem in between them, relative heights of different

cusps, length/width ratios of teeth, number of cusps and roots, presffice or' absence of serrations

and degree of heterodonty.

A. SAHNI AND V. P. MISHRA

Family Protocctidae Stromer 1908

GENUS INDOCETUS gen. novo

J:ype Species: -lndocetus ramani sp. nov.

Diagnosis:-Medium-sized primitive protocetid; frontal and parietal not in one platte as in

archaeocetes, but meet at an obtuse angle as in the mesonychids (e. g. Harpagolestes). Frontal

not as wide anteriorly as in Protocetus, maximum width of frontal at supraorbital process gr(~.ltcr

than that of Protocetus atavus atld Ilarpagoleste s orientalis but less so than in later zcuglouontids

and dorudontids. Post-orbital bar of frontal intermediate in prominence between mesonychids

and protocetids. Maxillo-palatine suture nearly at the same position as in P. atavus and

H. orientalis, but arched and not transverse.

'Upper dentition primitive, close to the condition in mesonychids; the known dentition

P3.M2 three-rooted, and with a rather' well-developed protocone in contrast to later archaeocetes

and even protocetids where it is not so prominent. Ml with a rather massive paracone, with a

slightly lower appressed metacone on its posterior slope.

Tympanic bulla large, massive and typically whale-like.

INDOCETUS RAMANI sp. nov.

Plate IV J figs. 1-3

/, Etymology;-After Mr. Raman V. Raste of Baranda (Kutch), who helped greatly during

field investigations.

IJolotypc:-L. U. V. P. 110:31-, U1C skull, ollly known spccimcll.

Horizon and locality: -Shell limestone of Babia Stage (I..utetian) at Harudi.

Diagnosis;-Only species of the genus.

Description:-Skull is well-prcscrved, large, wide and high, sloping anteriorly anti lacks

the portion anterior to the third premolars. Of the occipital region, only the portion or right

side with the tympanic bulla, occipital condyle and mastoid process is present. Parietals are

incomplete while the zygomatic arch is wanting.

Rostrum is broad and massive, diverging posterior to the third premolar. Posteriorly

the parietal is high corresponding to a high sagittal crest probably higher than in Protocctus atavus

and in Prozeu,glodon isis (I.Jrozeuglodon aLrox Andrews 1906). Frontals widen out suddenly to form

the broad and massive supraorbital process overhanging the orbits, at which point the skull is the

widest (20 cm), being broader than in P. atavus (16 cm) and Harpagolestes orientalis, but

narrower than in Zeuglodun osiris and Prozeuglodon isis. The skull roof is flat to gently convex.

The supraorbital process is more thickened, deflected downwards nearly at right angles to the

skull roof in contrast to the condition in P. isis, where the process is thinner and only slightly

deflected. This post~orbital process of frontal is not as prominent as in Protocetus atavus, but is

stronger than in mesonychids.

A small and sub triangular lachrymal is present antero-ventral to frontal, dorsal to the

anterior part of jugal, and posterior to maxilla, being concave at its posterior m;trh'in forming Ule

anterior border of orbits. Anterior part of jugal, which is narrow, is tbrust in between U1C

lachrymal and the maxilla. .

In the maxillae, only the roots for the third premolar to 2nd molar are present; the

lingual side of the crown of tile rigllt first molar is also preserved.

Lachrymal foramen opens posteriorly at the upper end of the maxilla at the level of the

anterior root of M2, having a diameter of 6 mm.

The maxillaries are large and posteriorly join the frontal and lachrymal; antcriorly

they are incomplete and the groove for the posteriorly directed facial processes of the pre-

maxillae cannot be determined, hence the premaxillae were restricted posteriorly to thc level

of the first or second premolar somewhat as in ProtoceLus atavus (posterior to P2), differing from

Prozeuglodon is is (Andrews) in which the premaxillae extend to the level of p4:

Thc vcntral sides of maxillac are nearly flat in the palatal region. Roots for the P3-M2

are present Of the crowns, only tile greater part or tile right lirst molar is prcserved. rrhe

WESTERN INDIA TERTIA)?Y VEllTEBRATES 19

maxilla of right side preserves the remanant of tIle bases of the crowns, hut the left maxilla is

more eroded exposing the roots. The maxillo-palatine suture is anteriorly convex rather than

transverse as in P. atavus and Ha rp ago le.\" le.\" orienlalis though occurring nearly at the same

position, i,e. at p4 in alllhree genera. The anterior border of the orbit is at the level of M2.

The occipital surf(ll:(~ is bl'o.ld, nc.l.rly 21-22 (~In .lcross ;1..1 I.II(~ In.lsloid pIO<':CSS, (lppl"oach-

ing the widlll at the same region in P. atavus ('24 cm) but narrower than in later basilo-

saurids. .Occipital surface nearly as high as in P. atavu.\". Supraoccipilal is broken. Tile

parietals slope laterally. Occipital conyles are nearly of tIle same size clS in J.), alavu.f but smallcr

than in Prozeuglodon isis, convex, wider at the upper end, narrowing towards the venrtal border

as in P. atavus. Foramen magnum with a diameter of about 3.5 cm is larger than in P. atavus

(2,6 cm), but smaller than in Prozeuglodon isis (Andrews) (5 cm).

The tympanic bulla is big, massive, bladder-like and typic.llly cctaccan. '1'he bulla

is attenuated anteriorly and widened towards the posterior side. The ventral surface of the

bulla is smooth and a broad shallow groove separates the smaller postero-internal and larger

postero-external lobes. The size of tympanic bulla of lndocetus ramani gen. et sp. novo is nearly

the same as in Protocetus atavus, having a maximum antero-posterior diameter of 7 cm and

maximum transverse diameter of 4.5 cm. The middle lacerate foramen 'lies on antero-internal

side of the bulla at thc posterior side of alisphenoid. The posterior lacerate foramen lies on

the postero-internal side of bulla and is deep.

Upper Dentition :- The dentition is primitive, close to the condition in the Mesonychidae.

1~he known dentition P3-M2 is three-rooted possessing an inner root. Crowns of all the teeth

are broken except the lingual side of the right first molar.

p3 is triallgular a 11d witll tllree roots. LCIlgth or tIle l,-lbi:ll ,-llvl~oli is 3 <':111. r\ lltero-li.lbial

root is exposed, the depth of the root being 2.5 cm. Posterior roots arc in transverse plane,

having a width of 2 cm. Postero-lin~ual root has a little larger surfacG area tllan tile postero-

labial root, and is somewllat rounded. Th~ border of the postero-li'l~lIal root is '2.G!) cm from

the mid-line of tile skull. lletween the inner roots of the tllird and fourtll premolar, there is a

distinct pit for the reception of the protoconid of the lower premolar (probably P 4)'

p4 is situated after a diastema of 3 mm from P3. The tooth is tllrec-rooled, elongate

and fully molariform witll a l/w ratio of 1.16, being wider thall P:I, Ilaving an internal alveolar

border 2.45 cm from the mid-line of skull. Judging from the are.l or the Internal root exposed,

tllC protocone appears to have bcen a ratll~r prominent cusp occupying a ratller internal

position. As in p3, the rOot of the protocone is situated in the transverse plane of tIle metacone.

MI is ')ituated after a diastema of 4 mm from p4, The tOOtil is thrcc-rootcd, triangular, wider

tlli.ln long with a lengtll/widtll ratio of 0.85 and is smaller than P4. 'lile protocone is situated

lingually. Only the lingual side of the crown is preserved, the edges of the crown do not show

any serrations. As tile labial border of the crown is broken it is not pos."ible to statc with

certainty whether there were only one or two principal cusps on the labial side, but from the

general appearance of the broken apical part, it appears that both paracone and metacone were

present though closely appresscd to each other and had an undivided bi.lS(~, ProlOCOll(:: is lcss

posterolingually directed tlli.ln in pi, low., Ilcavily ~Iorn but dislill(:l i.Uld tll(~ (~ll;llncl ,ll tIle bi.lse

is slightly wrinkled. '-1"\\'0 pronounced smooth wear facets are present, one in the mid-line, and

the other on anterior side of the tooth, arising from the lingual slope of the paracone towards

the protocone.

A tllin serrated cingulum encircles the preserved crOWll, becomillg narrower lingually.

On the antero-lingual border towards base, the cingulum is also effected by the wear from MI,

A deep pit is situated in between protocones of MI and M2 for the reception of the protoconid

ofM2' M2 is situated after a diastema of 3 mm from M 1. 'l'he tooth is tringular, three-rooted

and smaller than Ml with a length/width ratio of 0.89. Base of the protocone is postero-

lingually directed. The postero-labial root is reduced.

20 P. MI.':J'HRA

A.SAlIN/AND V.

10.

o.20.

2~.

13.

Measurements in centimeters

Length of the P3-M2 $eries of the right side

Diastema between p3 and p4 (Right maxilla)

Diastema between p4 and Ml (" " )

Diastema between Ml and M2 (" " )

Width of the skull at the point of the supraorbital proces3

Width of the skull at the level of the anterior root of p3

Maximum width of the occipital surface (at the mastoid process)

Height of the skull at the anteriormost preserved part, i.e. at the level of p3

Height of the skull at the end of the frontals

Dimensions of the alveoli in centimeters

Length Width

p3 3.0 2.0

p4 2.85 2.45

Ml 1.7 2.2

M2 1.7 1.9

Discussion:--Indocetus ramani gen. et sp. novo is a rattler primitive archaeocetc retaining

some primitive characters related with supposed mesonychid ancestry. Unmistakable cetacean

modifications are found in the elongation of the skull, the structure of tympanic bulla, the

presence of an elongated nasio-turbinal passage, the shape of "Turk's Saddle" and the atrophy-

ing of the protocone in the upper dentition. Its affinity with the Mesonychidac particularly

the genera Harpagolestes arId Andrewsarchus lies mainly in the general pattern of the cheek teeth :

transverse teeth, three-rooted, unserrated crowns, absence of hypocone and a reduced metacone

on M2. The general shape of skull also resembles with that of Andrewsarchus and Harpagolestes.

There has been much controversy regarding the origin of the Cetacea with refcrence to

their monophyletic or polyphyletic origin. Also the ancestral stock is believed by some to be the

mesonychid condylarths (Van Valen 1966, 1968) and by others to be the creodonts or insecti-

vores (Dechaseaux 1961, pp. 856-857). The material from Kutch particularly Indocetus ramani

gen. et sp. novo \vould tend to support the mesonychid ,lncestry of the archaeoceles. 'l'hc Meso-

nychidae was fairly a well-entrenched group during Eocene in central, south and east Asia.

This group has been recorded from Middle Eocene Chharat Series (Oehm and Oettingen-

Spielberg 1958) and lower Kirthar bed (Pilgrim 194.0) in Pakistall; from basal M urree beds

(lower' Upper t~ocene) in Jammu and K.lshmir (Ranga Rao 1973); from Nanshung Basin

(Palaeocene) in South China (Young and; Chow 1963) and from Yuanchu Formation (upper

part of Upper Eocene) in Honan, China ~Chow 1965) as well as from a number of localities

and horizons in Mongolia (Szalay and Gould 1966): Narcln Bul;tk B(~ds of ]~;trly ]~O(:(~II(~, .lrdin

Manha Formation and Ulan Shireh Form,llion of Upper Eoccne and S,hara Murun It'ormation

and "'Ulan Gochu" Formation of latest Eocene.

The widespread occurrence of MesonYGhidae in the early Palaeogene stron~ly suggests

stable corridors for dispersal. '[he Indian and Pakistan occurrcnccs coult! only bc possiblc after

the abutment of Indian plate on Sino-Siberian platform sometimes in the early Lutetian. It is

also reasonable to surmise that terrestrial mammals reached Kutch from their northern dispersal

point at least by the end of Middle Eocene, thus providing a suitable stock for such a radiation

as represented by Indocetus ram ani gen. et sp. novo

f'amily Protocetidae Stromer 1908

GENUS PRorl'OCE':l'US Fraas 1904

PROTOCETUS SLOAN! Sahni and Mishra 1972

Plate V, figs. 1-2

Material :-L. U. V .P..IIO02 (holotype), the anterior mandibular fragment containing the

symphysis and the alveoli for C-P3; L.U. V.P. 11003, lcft mandibular fragment with alveoli for P4,

2.5

f;VESTEl~N INDIA TERTIARY VERTEBRATES 21

MI and anterior alveolus for M2; L.U.~.P. 11001, partial skull; L.U.V.P. 11043, anterior part of

the rostrum and L.U.V.P. 11146, posterIor part of skull.

Horizons and locali~y :-Shelilimestone and ossiferous gypseous shales of Babia Stage

(Middle Eocene) at Harudi.

Description :-The species has been described previously (Sahlli arid Mislua 1972)

including the. specimens L.U.V:P. ,11001, L.U.V.P: 1100~ and L..U.V.P. 11003. Additional

material obtamed after the pubhcatlon of the paper IS descrIbed Ilerem.

L. U. V.P. 11043 (Pl. V, fig. 1) is an anterior rostral fragment containing external narial

opening. The pI and p2 have a diastema of 2.2 cm in between t,hem. pI is single-roote? while

the p2 is two-rooted. p3 though not preserved, would have been situated at least after a dIastema

of about 2 cm from P2. Snout is constricted between pI and P2. On dorsal surface, the posterior

margins of the premaxillae extend to form the lateral sides of th~ posteriorly Yfjdening external

narial opening which extends backwards up to-the level of p2 unhke upto pI as III Protocetus atavus

(Fraas 1904) and Prozeuglodon isis (Prozeuglodon atrox, Andrews 1906). The length and the

breadth of the narial opening is greater than that of P. atavus (vide measurements given by

Andrews 1906, p. 256). The length of the preserved part ofL.U.V.P. 11043 is nearly equal to

the corresponding length in P. atavus but is wider.

L. U. V.P. 11146 (Pl. V, figs. 2 a, b, c) is a posterior portion or thc~ skull lacking tIle portion

anterior to parietals. The occipital surface is nearly as broad (26 cm) as in tile paratype of

P. sloani (L.U.V.P. 11001) and Zeuglodon osiris Andrews 1906 (28.8 cm), but is broader than in

Pro-tocetus atavus (22.5 -24 cm) and narrower than in ProzeuA1lodon i.ri.r (34 cm). In constrast

to L.U.V.P. 11001, the sagittal crest is lower, tIle occipital collclyll'!; ;lre largl~r wllile llll~ [oram~ll

magnum is nearly of the same size. Supraoccipital though broken uorsally, :'lppears to be

narrower than iri Prozeuglodon isis. Though the occipital condyles are weathered, it can be judged

that these were slightly larger than in L.U.V.P. 11001 and in Protocetus atavus and nearly as

large as in ProzeugLodon isis; the v~ntral border of the occipital. co?dyle .is narrow. l?xoccipitals

on the sides of the condyles run out mto a broadly expanded projectIon as m P. atavus (Dechaseaux

1961, p. 843, fig. 12), and in Prozeuglodon isis (Andrews 1906, p. 245, fig. 81), the upper edge on

outer side of the exoccipital unites with the posterior border of the squamosal. The basioccipital

is broad, the weathered tympani~ bullae are placed on its outer angles. 1'he mastoid process is

strong. Antero-dorsally the parI eta Is are narrow.

PROTOCETUS HARUDIENSIS sr. noy

Plate IV, figs. 4-7

Holotype :-L.U.V.P. 11037, incomplete skull with the roots for P:l-M2.

Paratypes :-L.U.V.P. 11037 a, left mandibular ramus with rools for P.,-M:J; J.. U. V. P.

11037 b, left !vI! and!vI2; I.-.U. V.P. 11037 c, isolated cusps of upper teetll. 't'ilc p,-lratypes were

found close to the holotype, and in all probability belong to tIle same individual.

Horizon and locality :-Sheillimestone of Babia Stage (Middle Eocene) at Harudj.

D.. D1f I I? Cl P4 M2 Sk 11 .dd I.

tagnOSts:- enta ormu a I? Cl P4 M3' u narrow, narrowlllg su en y anterIor

to p3 as in Protocetus atavus; nearly as high as in Prozeu,glodon isis but slightly higher than In

Protocetus atavus. The length of the P3-M2 series nearly equal to the corresponding distance in

lndocetus ramani gen. et sp. novo from Kutch but smaller than in Protocetus a.tavus, Prozeuglodon

isis and Basilosaurus. P3-M2 three-rooted, closely spaced and unserrated. Maxilla extending

3 cm posterior to the last molar for the reception of Ms.

Mandibular symphysis extending posteriorly to P2 as in Prqtocetus sloani. P 4 -M3 two

rooted; M! and M2 bicuspid.

SAIINI AND V. P. J\lIISHRA

Description and Comparisions :-rrhe skull (Pl. IV, figs. 4 a, b) is narrow, narrowing anterior

to p3 as in Protocetus atavus. The skull is nearly as high as in Prozeuglodon is is but slightly .higher

than in Proto cetus atavus. The length of the left pa -M2 series preserved (9.5 cm) is nearly the

same as the corresponding distance in Indocetus ramani gen. ct sp. novo (10 cm), bllt smaller thal1

in Protocetus atavus (12.5 cm), Prozeuglodoll isis (more tllall 15 cm) alld lJasilo.\'aurus cetoides

(21.6 cm). The skull slopes steeply on lateral sides in the preserved part. Dorsally the frontals.

nasals and upper part of left maxilla have been eroded off exposing the fillings of the narial canal.

Ventrally also, the long, tubular fillings of the narial canals are exposed as the ccntr:ll part of

maxillae and the palatine are completely eroded off. A longtitudinal groove on the external side

of the maxilla lodged the posteriorly directed facial processes of premaxillae which extended

posteriorly to the level of the posterior root of fourth premolar, a condition found in

Prozeuglodon isis.

The left maxilla w.ith the roots for P3-M2 is preserved. The maxilla extends posteriorly

beyond M2 for the receptIon of Ma. The teeth are three-rooted and closely spaced. The antero-

labial roots in P4.M2 have larger surface areas than of postero-labial and lingual roots. Isolated

cusps of upper teeth (L. U. V.P. 11037 c) show that the teeth are not serrated.

p3 is the longest in the series having a l{w ratio of 2.3. The lingual root is situated

nearer to thepostero-labial root, thus the protocone would have been posteriorly placed.

p4 is closely appressed to p3 having a ljw ratio of 1.8." -The lingual root is situated nearer

to the postero-labial root. .rrwo prominent and deep pits are present posterior ,to p4 and Ml on

lingual side for the reception of the protoconids of M1 and M2 which are preserved (L.U.V.P.

11037 b).

Ml is closely appressed to p4 having a ljw ratio of 1.9. In surface area, the postero-labial

root is small and reduced; antero-labial root is largest of all; the lingual root is smaller than the

antero-labial root and situatcd close to the poslero-labi.tl root.

M2 is similar to Ml but smaller in size. rrhe postero-labial root is reduced and close to

the lingual root which is 3.7 CIll from the mid-line of skull.

The maxilla extends posteriorly beyond M2 for the rcception of M3.

Left mandibular ramus (L.U.V.P. 11037 a) consists of the roots for P4~Ma as wcll as

crowns of first and second molars (L. U. V. P. 11037 b). P 4 -Ma are two rooted and closely

spaced. rl'he ramus is laterally compressed, narrow and deepens posteriorly. As in Pappocetus

lugardi, Prozeuglodon isis, and Zeuglodon osiris, there is no groove on external side towards the

ventral border. The left and right mandibular rami in collection are separate and it is reason-

able to suppose that the symphysis must have extended much further forward probably to the

posterior root of P2 as in j'>rotoct'tus sloani or what is more unlikely to the anterior root of Pa as in

Pappocetus lugardi (Andrews 1919).

The apical parts of the protoconids of M1 and M2 (I...V.V.P. 110:37 b) fit well into. the

pits situated posterior and lingual. to pi and Ml in the left m.lxi.l1a. M1 is antcro-post{~rlorly

elongated, two-rooted and bicuspid. The anterior cusp (protoconid) is largc and high, the

posterior cusp is smaller and lower by 50 percent than anterior cusp. f\pex of the anterior cusp

is worn, Edges of the cusp are not serrated; the anterior edge bears two shallow wear facets,

the one towards apical side is larger than the other ventral to it, .'I'he anterior edge is steeply slop-

ing from protoconid. One wear facet is present on antero-labial and the other on postero-labial

border of the protoconid. The broken basal part of metaconid bears a cingulum on lingual side

which continues anteriorly upto the protoconid and also on the posterior side of metaconid.

M2 is also antero-posteriorly elongated, two-rooted aud bicuspid. Apex of larger anterior

cusp is bluntly conical and unworn. There are two wear facets on the protoconid, the oue on

antero-labial border is more prominent than the other trending towards postero-labial border,

Posterior cusp is small; low and unworn. Apex of posterior cusp is linked with the posterior

margin of the protoconid by a ridge, The lingual cillgulum is prominent on the posterior cusp

continuing towards lingual side of protoconid; posteriorly on the hypoconid the cingulum

rVESTERN INDIA TERTIAl~lr VERTEBRATE~'

becomes thick and forms a shclf,-likc structure. The labial cingulum is .tbSl~nt, tllis is in contrast

to condition in Pappocetus lugardi where the cingulum encircles the M. on all sides.

f)ime",fio",\' t!f thl? fllu{'oli i" {;{~illilll{!ll'r.\"

Lellgtu

2.9

2.3

2.1

1.9

3.0

2.3

2.2

2.1

Widlll

1 ~3

1>.

1. &

1.0

Suborder Odontoceti

1'here are two mandibular rami in the collection which camlot be included in Archaco-

ceti and have been placed in the Eocene family Agorophiidae of the Odontoceti. AgoroPhius

Cope. and )."enoroPhus Kellogg, the previously known genera of the family Agorophiidae, are the

only odontocetes found in Upper Eocene rocks of North Aml~rica, and are represented by skulls.

rrhe Kutcll odontocct<'s are rcprcscIllcd by m.lndibular fl'agml'J.lts. 'f11(~S(~ I';lllii .ll'C distillct from

all other in the collection in being long and n.lrrow llaving extreme lateral compression, and

extremely long symphysis-the separation of the left and rigllt rami taking place at least as far

back as M3. Such a long symphysis is not found in archacocelcs and ig ch.,r.lctcristic of more

ac\vallccd cemCl~.lIIS. i\llolllcr characteristic is tllc: gll.lJ'P clivcrl!;<:II(;l: (I[ tIll.: r.llili J)ostl~rior to

Ma, in contrast to the gradual alld anterior divcrgcllcc in al'chacocctcs. 't'lle dC~Jltitioll, however,

retained its primitive number, the only specialisation being that the teeth have a tendency

to become homodont.

Family Agorophiidae Abel 1913

GENUS J:\NDREWSIPHIUS gen. noy.

Etymology :-After C. W. Andrews, the foremost worker on the Fayum vertebrate fauna.

'rypc spccic.!' :-.llndrewsiphius kutchen,fis sp. nov.

Diagnosis :-Mandible narrow, elongated and with a longer sympllysis tllan that prevailing

in more primitive archaeocetes, in which group the symphysis extcnds postcriorly tillI>a and

never to the molar series. Symphysis extending posteriorly at least up to the t\lJa, Depth of

ramus low as compared to archacocetes. rl'ecth tending toT,yards homodonty but number

remains the same as in Archaeoccti. Shallow longitudinal groove prescnt Ilcar the v(~ntral border;

in coss-section ventral border obtuse anteriorly, becomillg acute postcriurly.

ANDI{EWSIPHIUS

KUTCH1?,NSIS :;p. llOV

Pl.\tc V, jig. (j

Holotype :-L.U.V.P. 11060, malldible with alveoli for 12-M1.

Horizon and locality :-Bluish grey shales of Bahia Stage (Middlc Eocenc) at Narcda.

Diagnosis: -Mandible IOllg, compressed laterally, symphysis extends posteriorly at least

up to the l,l.st molar, i.e. MI. Depth of ramus low as compared to archaeocetes. 1"'he spacin.g

between anterior teeth nearly equal; posterior teeth closely spaced, teeth tending towards

homodonty but their number remains the same as in Archaeoccti. T...ike in most of archaeocetes

the ventra] border straight but with a slight concavity at the level of canine, c.f. Pappocetus

lugardi with a step-like appearance below M2.

24 A. SAHNI AND V. P. MISHRA

Description :-The mandible is gypsified consisting of both the rami with alveoli from the

second incisor t<:> the first 1!l°la; (Fig. 3). Thoug~ the. alveoli for ,M2 and Ms are not preserved,

the ventral portIon of the Jaw IS preserved. The Jaw IS smaller ill length than in Pappocetus

lugardi, Zeuglodon osiris Andrews 1906, Zygorhiza kochii, Prozeuglodon is is (Zeu,f?lodon is is Andrcws

1906), but is nearly of the same length as of the young individual of Prozeuglodon is is (l'>rozeuglodon

atrox Andrews 1906). The mandible gradually increases in depth, anteriorly it is 2.2 cm and

towards the molars it is 6.5 cm.

The most important feature of the mandible is its long symphysis, extending posteriorly

at least up to Ms. Among the archaeocetes, nowhere does the symphysis extends posterior to Ps,

and is up to the anterior root of Ps in Pappocetus lugardi (Andrews 1919). The symphysial junction

is well defined by a shallow groove from the .anterior end of the mandible up to the middle of the

P 4, after which the symphysial region widens and the right and left rami diverge dorsally, but

remain ventrally united at least up to the Ms.

QQ~

QQ-

~~:t::~::~i::'=~~-~

l C I 2. P3 ~ M, ~ T3

Fig. 3. Andrewsiphius kutchen.sis gen. et sr. nov., Holotype, the mandible

(L.U.V.P. 11060), occlusal view x 1/3'

The lateral sides of the rami, are slightly convex. On the lateral side near the ventral

border of each ramus, there is a shallow, wide groove running from the anterior end, posteriorly

up to M3' This ventral groove is absent in most of the archaeocetesexcept for in Protocetus sloani

from Kutch, where the groove fadcs out posteriorly at P2' III later cctaccans (e.g. 11e.fperoinea

dalpiazi Moncharmont Zei 1956) this groove is also present. The ventral border of the mandible

is rounded anteriorly up to the first premolar, posterior to which the border is acute. The cross-

sectional profile of the mandible anteriorly is nearly oval while at M1 it is triangular. Like

most of archaeocetes the ventral border is straight but with a slight concavity at the level of

canine. In Pappocetus lugardi also there is a step-like appearance below M2 (Andrews

1919). .

The crowns of none of the teeth are preserved. The alveolus of first incisor is also

wanting. 12 to P1 are single-rooted, the teeth posterior to PI are two-rooted. l'he alveoli of

anterior teeth from second incisor to second premolar are separated from each other by nearly

equal diastema of 3.5 cm. The alveoli of P2 from P3 and P3 from P 4 are separated from

each other by the diastema of 1.8 cm. P 4 and M2 are closely spaced. The pits caused by the

occlusion of cusps of upper teeth are seen only posterior to P 2 and P 3' The alveolar border deepens

posteriorly with a concavity between C and P1 which received the upper canine. The alveoli

are smaller than those of archaeocetes.

2.5.

6.6.

Ji~4

¥~

!:",~

Meaj'urements in centimeters

Total measured length of the mandible

Length from anterior border of 12 to the middle of M1

Depth of jaw beneath 12

" " C

II " PI

" " Pa

" " P 4

" " M 1

WESTERN INDIA TERTIA1?T VERTEB1~ATES 25

Dimensions of tht alveoli ill cell/i'l/fler.\'

~ . Width

0.7

0.8

O.l~

0.9

0.8

1.0

O.!)

0.85

12 (alveolus DIlly)

J.3 ( " ,,)

C ( " ,,)

PI ( " ,,)

P 2 (alveoli only)

P 3 ( " ,,)

P 4 ( " ,,)

M 1 ( " ,,)

ANUl~E\\'SIPHlUS MINOR sp. novo

Plate V, fig. 7

Holotype :-1... U. V. P. 1] ]65, incomplete mandibular j-

I-MI, Horizon and locality :-Ossiferous gyp3eous shales of Babia

with the alveoli of

fragment

pStage Midd.Je Eocene) at

Harudi.

Diagnosis :-Mandible small, smaller than Al/drezt'.\'iflhius I..utchell.\'i.\' gen. ct sp. IIOV.,

narrow anteriorly but differing from A. kutchen.\'is gen. ~t sp. novo in diverging abruptly posterior

to P3. Ventrally, the rami remain united at..least up to M2 and hence the mandibular symph)'si'i

extends up to M2. P3 -M2 closely spaced. D~pth or the r;lmu~ leR~ tllan in Clr(~II:I.(~()crlcs.

Thc teeth are trausverscly compre.'isl~d.

Description :-L. U. \-'.P. 11165 is an anterior mandibular fragment of a smCl.!1 species (If

Andrewsiphius, consisting both the right and left rami. Mandible is l1arrow anteriorly till P 3,

posterior to which it abruptly broadens. The mandible differs from .4. Icutrhen.\'i,\' ~cn. ct sr. novo

in the sudden divergence and widening of rami posterior to third premolar. 'l'\IC two rami

diverge posterior to third premolar but do not separat.~ a\v.ay from each other at least up to M2

and are united ventrally.

The mandible is broken anterior to Pl. Cro\\'tls or PI and P2 arc not prc.,crvccl. Pi is

single rooted and appear.., to be a sma.l1 tooth. P 2 is situated after a diastema of 4 mm from PI.

P 2 is two rooted, the length of alveoli being 2 cm. Anterior alvcolus i5 antero-postcriorly clonga-

ted and compressed; the posterior alveolus is circular and external with respect to tIle anterior

one. Diastema between P 2 and P 3 i~ 1.2 cm. 'l'he mandible starts wiclelling all(l cle(~pcllillg

posterior to P3. P3 js two rooted and 1.8 cm long. Posterior to P:1, the teeth are situated clost'.

to each other without any diastema. P 4 is two rooted, clongatecl, laterally compr(~sscd and

\vider posteriorly, the length or tOOtll being].n CIllo M1 i~ t~o rool~(l ;I.O(t I.r;tllsv('r~(~]y (:()m-

pressed, the posterior root b<.'il1g muctl wider than alltcri()1.' oO(~; b;lS(~ ()f' I.II(~ (~I.'(IWII is ;u.'i() pl.'C-

served; the length of the tOOtll is about 2.0 cm. At M2, the jaw i~ (Iorsally l~l.'o(I('(1 ;lllcl 1.11<.'

boundaries of the alveoli are not clear. Posterior to M2 the jaw is broken. J~oth lin,gually and

labially the roots of P3 -Ml arc cxposed due to the erosion of the r;lmi. 1'II(~ deptll of r,lmus

is less than in archaeoceles, .1Ild even slightly less tllan in A. kutch~II.\'i.\' g(;u. ct sr. IIOV. j\ gro()VC

is present near the ventral border on the external sides of the mandible which f~ldc." posteriorly

at the level of anterior border of P 4. Anteriorly the cross-section of mandible is nearly oval,

but posteriorly it becomes triangular.

Judging from the development of the molar series L.U. V .P. 11165 does not appeclr to be

a juvenile individual of Andrezv,\'iphius kutchensis gen. et sr. novo

lvleasurements in centimeters'

3.B

Length of preserved specimen

Depth of 1nandible at P 2

" "P 3

" "P 4

Width of mandible at PO)

" "M;

6 -6.5

2.5

J-.lcn,gth

1..'2

].3

l.~)

1.5

2.0

2.25

2.0

1.75

26 A. SAHNI AND V. P. MISHRA

Dimensions of the alveoli in centimeters

Length Width

P1 (alveolus ~nly) 0.7 0.6

P2 (alveoli only) ~.O O.G (post.)

P 3 (" ,,) 1.8

P 4 ( " " ) 1.8

M1 (" ,,) 2.0

Discussion; -Primitive odontocetes are represented by the Family Agorophiidac, consistilJg

of two genera; AgoroPhius and XenoroPhus from the Upper Eocene of North America. Both the

genera were described on the basis of skulls (Leidy 1869, Cope 1895, True 1907, Kellogg 19~3)

and hence there is no data on the morphology of the mandibles. However, judging from the

rather long and narrow rostra of both these genera, it can he reasonably presumed that the corres-

ponding mandibles would be narrow, elongated and with a longer symphysis than that prevailing

in more primitive archaeocetes, in which group the symphysis extends posteriorly till tlle P 3 and

never to the molar series. These Kutch mandibles have been placed in li'amily Agorophiidae

in view of the obvious specialisations in comparision to the archaeocete mandibles.

There is as much controversy as to the ancestral stock of odontocetes as there is to that

of archaeocetes. .Some workers maintain at present (Gidley 1913, Miller 1925, Kellogg 1928,

Kleinenberg 1958, 1959, Dechaseaux 1961, Yabalakov 1964) that the odontocetes did not

originate from archaeocetes but from a closely related group. Romer (1966, p.299) is of the

vie\v that,

"although the odontocetes presumably derive from a stock related to the

archaeocetes, the t,vo groups must have diverged at an early stage".

but some authors favour the. monophyletic origin of thc c(:taceallS. rl'hi~ COIl(:(~pt Il,lS b(;ell

aptly stated by Gregory (1951, p.450) that,

."...the skulls of the archaeocetes form an ideal structural ground plan for the

divergent paths leading perhaps through ArchaeodelPhis on the one' hand to the

mysticetes and on the other to odontocetes".

Gregory (op. cit., p.443) also remarks that,

"...after long study, the evidence still seems to me to support the view of C. W.

Andrews (1906) and O. Abel (1902), ,,-'ho pointed out essential features in which

.the skull of Zeuglodon fore-shadowed those of the odontocetes (toothed ,.,,-hales) ".

.G~egory (op. cit., pp.443-444) further expresses that the agorophiid skull is more or less

IntermedIate bet\veen the archaeocete and odontocete stages and even some authors include

agorophiids in the archaeocctes. '

However, ~ an \I? alen (1 ?68) in his recent discussion on the origin of the whales has revie-

wed the data and IS of the Op11110n that among the archaeocetes, the Protocetidae may have given

rise to recent whales. Van Valen (op. cit.) mentions following points which are common to the

archaeocetes and odontocetes ;-

1. Teeth present.

2. Lower jaws meet at symphysis.

3. Cervical vertebrae somewhat movable.

4.. Manus is usually pentadactyl.

5. Sternum has several bones.

6. Three or more pairs of ribs attached to the sternum.

7. In Agorophius the external nares in a position intermediate to that found in recent

genera and archaeocetes.

8. Fusiform body in Protocetidae, Dorudontidae and Odontoceti.

9. Tympanic bulla with same structure in all the three suborders.

r'fESTEI~N INDI..

TERTIAI~r

'ERTEBIlATES 27

The position of .4ndrewsiphius kutchensis gen. et sp. nov. may be significant in the context

of odontocete evolution and may form an intermediate stag-e between the archaeocetes and

odontocetes.

'"!'he Kutch odontocete mandibles have followillg Cll,1f,lctel's Wllicll HIIO\V lll(~ir .ldv;ulced

spccialisation over archaeocetes: narrowness of the mandible, lollger sympllysiH, lle.trly regular

spacing of the teeth except in molar series .tnd teeth tendiog towards homoclonty stl'ol1j!;ly f.lvour

their agorophiid affinities. but tilC llumber or teeth remaills tile !!alTle ,lS ill '\I'(;II:ll'o(:cli.

rrhe age of Kutch odontocetcs also raises some interesting questions. Previously described

earliest kno\\'11 specimens are Upper Eocene in age, no odontocete up to the present was

reported from the Middle Eocene (Lutetian). The recovery of Lutetian odontocele,; from

Kutch provides evidence that primitive cetaceans h;td st,irtcd diversifying by lltc Middle L~o<;:cnc

rather than the Upper Eocene.

Order Sirenia

The shell limestone of Babia Stage (Lutetian) expJsed near Harudi is the onLy horizon

to provide ~:ocene sirenian remains which are interesting inasmuch as they show close similarity

with the pelvis of .l~olherium ae,gypticum described by AbeL (1904) from Middl(~ Eor.cn(: or l~gypt.

Later Sickenberg put the remains described by Abel (op. cit.) under the synonymy of Protosiren

fraasi on the basis of associated vertebrae. There appears to be a discrepancy regarding the

locality of the skull and pelvis of the Proto.rirenftaa.ri Abel 1904. Arldr(~ws (1906, pr. II ~), 197,

204 215) mentions that Abcl (1901) described l~'olherillm, ae.!!yjJli(;/lm from Wllitc I..imcstonc of

I..o~er Mokattam Series exposed in Mokattam Hills near Cairo, corresponding to the Wadi

Rayan Formation (lower part of Middle Eocene) of Fayum. Ho\vever, Dechaseaux (in Piveteau

1958, pp. 344:-34:?) states th~t the occurrence of ~he cranium and. the }?elvis?f Proto.riren jraa,fi

(Eotherillm aegYPllcum) .d.esc.nbc~ by A~l (19?4) IsfrC?l;t the, I..~tetlan ot,tllc fayum (Egypt).. In

view of Andrews' familIarIty wIth the Egyptian localities, It IS more likely that the pelvIs of

Protosiren was recovered from Mokattam Hills, Cairo and not from Fayum.

9:J4

amily Protosircnidac Sickenbcrg

GENUS PROTOSIRF.N Abel 1904

PROTOSIREN FRAASI Abel 1904

Plate VI, IiI:!;. I

i\laterial :-L.U.V.P. 11038, left fragmentary pelvis alld L.U.V.P.Jund

very close to each other and probably belong the same individual. 1039, the ilium '~.ere

Middle Eocenc at Harudi.

Horizon and locality Shclilimcstonc of Babia Stage

Description:- The left pelvis (Fig. 4) is one of the few sirenian skeletal elements in the

collection from the Eocene of Kutch. The specimen resembles very closely witl1 the correspond-

ing bone in Protosiren fraasi described under Eotherium ae,g.ypticum hy Ab{~l (1~)O1) from the

Middle Eocene (Lutetian) of Mokattam Hills, Cairo (Egypt).

1'he specimen is incomplete as the ischium and pubis are broken distally; the ilium is ,uso

not complete distally. The estimated length of pelvis of Kutch sircnian is nearly twice than

that of the same elements in Eotheroide.f li/ryca, Prototherium verO/lense and Halitherium schinzi

figured by Dechaseaux (1958, pp. 348, 350-), but judging from the pelvis ligured by Andrews

(1906, p. 214), the Kutch specimen is only lon.ger by 50 percent than of liotlleroide.f (/~'().\"irell)

libyca and nearly by 30 percent than of Halitherium schinzi. 1'hc Kutch specimen is longer

than the same element in P.fraasi from Egypt by 33 percent and approaches the len,!?,"th of the

pelvis of .t\1oeritherium (Andrews 1906, p.214).

A. .s'AHNI AND V. P. MISHR.ll

",-,

,'" "\

-\-p~ '.

""--'.

I '

I f

Of' ,

.-i !-

I '

, \

, '.

I\\!!

,if.,..11.

---

~---~~

-~---~..

--,'

_f """-

,.,~

.. --.

,--

-~ ~ is.

~ ,

\. ,

'w"""

Fig. 4. Proto.!ircnfraasi Abel, left pelvis (L.U. V.P. 11038 and 11039), x 1/2 ac.-acetabulum, il.-ilium, ip.-ilio-

pectinal tubercle, is.-ischium, o. f.-obturator foramen, pu.-pubis.

The acetabulum formed by the fusion of ilium, pubis arId cotyloid bone i.1; nearly oval and

elollgated antero-posteriorly having a longer diameter of 4.6 cm and smaller diameter of 3.5 cm.

It is larger than in the Eotheroides libyca, Prototherium veronense, Halitherium schil'zi, as well as

P. fraasi from ~~gypt, approaching in size to the acetabulum of the Moerilherium from which it

differs in antero-posterior elongatic)11 of tile a(~etabulum aR OppOSl~U to a U()J'so-vl'lltr:ll ()n(: ill

lvloeritlterium. rl'he acetabular border particularly on the dorsal side is prominent and elevated

and has a greater curvature than on ventral side. Postero-dorsally, the acetabular border bears

a prominence providing ventrally a notch, the acetabular notch, situated postero-medially as in

pelvis of P.fraasi from Egypt. The acetabular notch is the point of tll(' origin of reclus fermoris

muscle. The acetabular region of the pelvis is quite broad, having a width of 5.2 cm, being

broader than in the known pelvic girdles of other sirenians, but is 20 percent narrowt"r than

that of Moeritherium. The ventral side of the acetabular region bears a wide shallow groove

arising from the obturator foramen fading away towards tIle ilium after a short distarlce ; tllis

groove apparently divides the acetabular region into a wider dorsal and a narrower ventral part.

Antl'rior to acetabulum on tile vl~ntro-in{crnal side or tllC ilillm, there is a triangular projcclioll,

the ilio-pectinal tubercle, 7 mm high and 1.7 cm long. Tile ilio-pectinal tubercle is better

developed than in Prototheriunz veronense, l-Ialither ium schinzi as well as in P. fraasi from l~gypt

but is less developed than in Moeritherium; in Eotheroides the ilio-pectinal tubercle is absent.

There lies a shallow fossa ill between the ilio-pectinal tubercle and the acetabulum.

Cl(.

Anterior to the acetabular region, thc iliac region suddenly becomes more slcllder ilS in

the Egyptian specimen. 1'he ilium (L.U.V.P. 11039) was found less than a metre away from

the L.U.V.P. 11038 and in all probability is a part of the pelvis de:;cribed. It is a narrow and

slender bone longer than the Egyptian specimen. Distally it is trihedral in ~ection, 1{r:lc1ually

bccomillg oval towoarus its proximal side. 't'I1C isciliulll, thou.!!,"ll, Ilot I)rc",(~rvc(l :It its })()illt (If

attachment with acetabular part, shows that it was quilc broad, broader lhall ill t':gyptiall spcci':'

men, but slightly less so than in Moeritherium. The width of the pubis is similar to that of P. fraasi.

The obturator foramen though not preserved, appears to be complete and was almo:;t of thc

same size as in the P. ftaasi from Egypt. In E'otheroides, Prototherium and llalitherium the

ob1moator foramen is partially or completely absent.

.Discussion :-The remains of Eocene sirenian are fairly cosmopoliton in distribution:

Eotheroides (Eosiren, Archaeosiren, Eotherium) is known by skulls, mandibles, vertebrae, and pelvic

girdle clements from the Middle to Upper Eocene of Korth Africa; Prorastomu.\' is known by

skull and mandible from Jamaica (West Indies); Prototherium by skull and pelvic elements from

Upper ~-:ocene of Italy and Protosiren by skulls, a pelvis and vertebrae from North Africa

and France and now from the Middle Eocene of Kutch. 'fhis is tbe first record .of a sirenian

from Eocene horizon in India and also from the Asian sub-continent. The material from Kutch

is rather scanty and consists of girdle elements and a fevv fragemcntary vertebrae. As already

WESTERN INDIA ,.TERTIA:J:lY"VJj;f;lTEfJ RA TES

mentioned, the pelvic girdle is specifically itldistinguishable from that of Protosiren fraasi. It

possesses a number of primitive features common to terrestrial mammals but later atrophying or

becoming absent in later sirenians. Primitively the pelvis is elongated with a long ilium and

having a well-developed acetabulum indicating the presence of a large functional femur

with a complete obturator forame,n. -"As pointed out by Andrews (1906) the pelvic girdle

of Protosiren is morphologically indistinguishable from that of Moeritherium except on being

smaller in length, possessing a less distinctly developed sacral surface on ilium and slightly

thicker pubis and suggested that early.proboscideans and'sirenians are closcly related and had

a common ancestor in early '-rertiary times, probably in the Lower Eocene. The primitive

sirenian-moeritheriid stock was co~stal, marsh-dwelling, equally adaptable to the life in

coastal marshes a s well as near ~hore conditions. The material from Kutch at this stage cannot

throw much light on the origin of sirenians. The presence of a well-developed sacrum described

hereafter and pr:obably a moeritheriid, supports the close similarity of Eocene sirenians and

moeritheriids.

Order Proboscidea

Suborder Moeritheroidea

Family Moeritheriidae

The existence of terrestrial mammal in the Middle Eocene fauna of Kutch is suggested by

the presence of a moderately large and robust sacrum, L. U. V.P. 11069. The development of a

sacrum, in which sacral vertebrae ankylose to form a single composite bone, has been attributed

by many authors (Weichert 1958, pp. 393-394, Goodrich 1958, p. 200, Kent 1965, pp. 141, 191)

as a necessary modification for a life on the land. rrhere is certainly a marked tendency amongst

aquatic mammals to lose this structure, which is primarily useful in forming a -firm skeletal

support in the movement of land mammals. Modern whales lack sacral vertebrae (Zittel 1925,

p. 80, Parker and Haswell J.897, p. 499) and in archaeocetes also only one or two sacral verteb-

rae were present (D~chaseaux 1961, p. 843). Similarily, sirenians also do not possess a massive

sacrum and the element is represented by one or two sacral vertebrae (Zittel 1925, p. 264).

In early sirenians also sacrum is lacking as Eotheroides (Eosiren Andre\v-s 1906, pp. 212-213).

possesses a sacral vertebra and in Protosiren (Eotherium aegypticum Andrews 1906, p. 119) the ilium

shows a less distinctly developed (more developed in Moeritherium) sacral surface, indicating

only a single sacral vertebra rather than a sacrum. In both archaeocetes and sirenians, the

structure of the pelvic girdle excludes the possibility that the ilium was firmly attached

to the sacrum. In other mammalian orders, such as the proboscideans the sacrum was robust

and firmly ankylosed for stability.

if. MOERITHERIID SACRUM

Plate VI. fig. 2

Material:-L.U.V.P. 11069, a sacrum.

.if.

SAHNI A"NDV. 'Po MISHRA

0 , ,~, ' ..

.., "

" :

'. ;'1-.

, , "

../" _.~-,',-,

~' .

' -\,~

, '\

, ,

, a

Fig. 5. if. Moeritheriid sacrum (L. U. V. P. 11069), anterior view x 1/2-

:broken, and so it is not possible to conjecture whether the width decrease.d posteriorly, but the

'thinness of the bone posteriorly, 'however, seems to' suggest that the width did decrease. The

articular surface of the anterior centrum is a transversely elongated oval as in Moeritherium,

while that of the posterior centrum is considerably less so. The anterior zygapophyses are well-

developed, but the anteriorly directed metapophyses are broken. 1'he neural canal is shaped

like an isosceles triangle as in Moeritherium but of smaller dimensions. The walls of the neural

canal are sloping somewhat posteriorly and bear a low but ma~isive neural spine as in Moerithe-

rium, the anterior neural spine is more prominent than the posterior one. Anteriorly, the

transverse processes (pleurapophyses) of the fused first sacral vertebra project beyond tIle ante-

rior surface of centrum while in Moeritherium the pleurapophyses do not project anterior to

centrum. Pleurapophyses possess an obliquely trending ridge, lateral to which the bone is

deflected downwards, this forms a roughened surface for attachment with the ilium. rl'he iliac

surface is borken laterally on both sides. It was well-developed and extended at least up to the

middle of second sacral vertebra. There is a pair of nearly circular intervertebral foramina

(sacral foramina) for sacral nerves in between the junctions of the broad pleurapophyses of the

first and second and second and third vertebrae. The intervertebral foramina in Moeritherium are

larger and antero-posteriorly elongated. The second and third vertebrae are elongated with

low neural spines and small post-zygapophyses (except the posteriormost in \vhich the structure

is broken but appears to be better developed).

Ventrally the surface is not flat as in Moeritherium, but in the centre there is a raised

region formed by the three fused centra which are very strongly convex transversely and concave

antero-posteriorly. The lateral margins, particularly those bearing the anterior iliac surface, are

prominently raised.

The sacrum also resembles the same element found in carnivores in general shape, in

.having a wide lateral expanse anteriorly and narrowing posteriorly.

Measurements in centimeters

Mocrithcrium

21.4

20.4

7.2

6.2

3.8

1.8

2.5

B.8

L.U.V.P.11069

Length 14.8 (measured)

Length from anterior to posterior faces of centrum 12.2 (measured) ,

Width of the right half of sacrum anteriorly 5.5

Length/Width Ratio. 1.35 1..

Width of centrum (anteriorly) 5.0 r

Height of centrum (anteriorly) 2.7

Width of neural canal (anteriorly) 2.5

Height of Neural canal (anteriorly) 1.6

Diameter of intervertebral foramina 1.2

Height of sacrum 7.0 (estirpated) :

Discussion :-With the possible exception of lndocetus ramani gen. et sp. nov., no other

Kutch marine mammal known at present would have possessed such a well-develeped sacrum.

The sacrum strongly suggests the presence of a terrestrial mammal. In view of its morphologi-

'cal similarity with the sacrum of Moeritherium, the authors feel that the sa,crum is more

converiient'ly placed in the Moeritheriidae than in any other systematic position. The sacrum

'possibly represensts a primitive mammal differentiating from the ancestral condylarth stock.

WESTERN INDIA TERTIARr VERTEBRATES 31'

MIOCENE FAUNA Class Chondrichthyes

Subclass Elasmobranchii

Ordcr Sclachii

Suborder Galeoidea

The selachian teeth belonging to the suborder Galeoidea are very abundant in the grey

'{;oloured gypseous shales of Khari Series (Lower Miocene) in the vicinity of Matanomadh and

Lakhpat, while in Pipar and Jangadia, the teeth are not so abundant. Prior to present investi.

-gation shark teeth from Kutch have been described by Tewari (1959) and 'Tewari et at. (1964)

from the Burdigalian horizon. Following species of Miocene sharks from Kutch have been

,already described by V.P. Mishra in a separate paper (Mehrotra, Mishra and Srivastava 1973) :-

Family Carcharinidae

Galaeocerdo cuvieri Le Sueur

G. wynnei Mehrotra et ale

HemiPristis serra Agassiz

H. sureshi Mehrotra et al.

Hypoprion macloti Muller and Henle

Negaprion brevirostris Poey

S coliodon sorrakou1ah Cuvicr

Family Isuridae

Isurus spallanzanii Bonaparte

.-' Carcharodon megalodon robust us Davies

C. angustidens Agassiz

C. bigelowi Mehrotra et at.

C. carcharias Linnaeus

Family Alopiidae

Alopias vulpes Gmelin

Family Sphyrnidae

Sphyrna diplana Rafinesque

Family Carchariidae

Carcharias tricuspidatus Day

C. heptacuspidatus Mchrotra et al.

Among the shark genera found in the Miocene of Kutch, Galaeocerdo, Hemipristis and

Negaprion of the family Carcharinidae, and Isurus and Carcharodon of the family Isuridae were

:£ound to predominate.

Order Batoidea

Suborder Rajoidea

Family Rajidae

GENUS RAJA Linnaeus 1758

RAJA sr.

Plate VI, fig. 3

Material :-Isolated teeth, L.U.V.P. 11099. .

Horizon and locality :-Grey coloured arenaceous shales of Khari Series (Lower Miocene)

.at Lakhpat.

Description :- The tooth is small with a height/width ratio of nearly one. The crown

,is rhombic in outline in oral view with feeble rugose ornamentation. A shallo,v and wide

32 A. S AHJVl AND J(. P. MISHRA "

groove is present towards the anterior side of crown, the border of the groove is slightly raised..

On the postero-ventral side, a small vertical groove divides the crown into two. A triangular.

enamelled lip extends downwardly frpm the crown on the external side.

The root is small, vertical to the crown and birid, the two halves being separated by a root.

canal. Two small foramen are present in the root canal, one going into each half of the root.

In India, the family Rajidae (skates) is presently represented by a single genus Raja, the

species of which (R. andamanica, R. johannis-davisi, R. mamillidens,. R. rever sa, R. powelli) are.

found in Bay of Bengal, Andaman Sea, Arabian sea, Travancore coast and Gulf of Manaar..

Prior to present work, there has been no fossil record of Raja from India.

Suborder Myliobatoidea

Family Myliobatidae

GENUS My'LIOBATIS Cuvier 1817

MYLIOBA TIS sp.

Plate VI, figs. 4-6

Isolated median teeth of Myliobatis have been reported from Miocene of 'Balasore

1939) which are shorter in length than the Kutch specimen. (Hora

L. U. V.P. 11097 (Pl. VI, fig. 6) is a fragment of spine of Myliobatis. The spine is longi- .

tudinally striated. The edges of spine bear fine denticles directed distally:; the spine widens.

distally, the denticles being absent in the distal region. The spine 'resembles the caudal spine

of M. meridionalis (Leriche 1957). Myliobatid spines have been described from, Miocene of"

Orissa (Hora 1939, Ghosh 1959) in India. .

Presently, three species of Myliobatis (M. nichofi, M. milvus andM. maculatus) are found..

in India along the east and west coasts, and also at the mouth of Gang~s and Chilka Lake.

WESTERN INDIA TERTIARY VERTEBRATES

GENUS AETOBATUS Blainvil1e 1816

AETOBATUS sp.

Plate VI, fig. 7

lvlaterial :-Isolated teeth, L.U.V.P. 11169, L.U.V.P. 11170 and L.U.V.P. 11171.

Horizon and locality :-Grey coloured gypseous shales of Khari Series (Lower Miocene) at

Matanomadh.

DescriPtion :-1'here are few isolated teeth of Aetobatus in collection from Lower Miocene

beds of Matanomadh. L. U. V.P. 11170 (Pl. VI, fig. 7) is a small, hexagonal tooth, nearly:

twice as broad as long. The occlusal surface is smooth. The root is divided longitudinally

into ridges and grooves, the number of ridges being 13. The root is nearly as high as the crown.

The ledge between the junction of the crown and root is prominent, and fits into the corres-

ponding groove 011 the posterior side of next anterior tooth. The tooth is slightly curved, the

curvature is well-exhibited in the basal view.

A tooth of A. arcuatus bariPadensis has been described from the Miocene of Mayurbhanj,

Orissa in eastern India (Ghosh 1959) from which the Kutch specimen differs in being narrower,

less curved and in possessing a smaller number of unbifurcated root ridges not continuing to the

posterior border of enamelled surface. The Kutch specimen of Aetobatus differs from the tooth

of A. arcuatus (Leriche 1957) in being smaller in size and less arched.

At present in India, the family Myliobatidae (Eagle Rays) is represented by only two

genera, Ae/obatus and Myliobati,)'. Only two species (A. flagellum and A. oceLLatus) of Aetobatus

(Duck-Billed Ray) are presently found in Indian region.

Class Osteichthyes

Subclass Actinopterygii

Infraclass Teleostei

Superorder Acanthopterygii

Order Tetraodontiformes

Suborder Tetraodontoidei

Family Diodontidae

GENUS DIODON Linnaeus 1758

DIODON sp.

Plate VI, figs. /].9

Material :-A ]arge number of dental plates: L.U.V.P. 11095 (Lakhpat) and L.U.V.P.

11087, L.U.V.P. 11088 and L.V.V.P. 11089 (Matanomadh).

Horizons and localities :-Grey coloured gypseous shales at Matanomadh, and grey

coloured arenaceous shales at Lakhpat; Khari Series (Lower Miocene).

Descriptions :-A large number of dental plates of Diodon were collected from Lo",'er-

Miocene of Kutch. Dental plates in collection range in size from 0.8 to 4 cm in their longer-

diameter. The dental plates consist of a number of sub-triangular to oval lamellar plates,.

piled one over the other. In L. U. V. P. 11095 (Pl. VI, figs. 8 a, b) there are seven thick

lamellae, while in L. U. V. P. 11089 (Fl. VI, figs. 9 a, b) there are four thin lamellae. A

vertical plane divides every plate into symmetrical halves. The plates on the apical side are smaller-

while those in the mid-region are bigger. The edges of the dental lamellae are not crenula-

ted as in D. scillae (Woodward 1901, pp. 572-573, fig. 20). Every dental plate possesses 9 to 10

teeth on the superior side which is concave as in D.foleyi (Lydekker 1880).

In upper view, the teeth of all the lamellae ,are seen arranged ill transverse rOW5.. The

median teeth are larger than the marginal ones.

A. SAHNI AND V:P. 'MISHRA

Discussion :- The record of the fossil diodontids from India is meagre. Dental plate of

Diodonfoleyi has been described from the Eocene of Ramri Island, Burma (Lydekker 1880, 1886 b)

and dental plates of Diodon sp. from Nicobar Island though the horizon is uncertain (Nair

1.945). The Kutch species i~ slightly smaller in ~ize and ha~ Ic~~ numb~r of lamcllac tlliln

in D. foleyi and dillcrs from thc specics of Diodon frOln Nicobar Isl:lnG in 11ilving Icss numbcr

of lamellae with a concave upper surface. 'fhe Recent D. Irystrix inhabiting Indian Ocean is

smaller in size with a fiat upper surface. The presence of fossil Diodon in Ramri and Nicobar

Islands and now from Kutch, and the living species in Bay of Bengal indicates that the genus

inhabited the Indian Ocean throughout the ,!'ertiary.

Order Perciformes "

Suborder M ugiloidei

Family Sphyraenidae

GENUS SPHYRAENA Bloch and Schneider 1801

SPHYRAENA sp.

Plate VI, figs. 10.1 r .

Alaterial:-A large number of isolated teeth: anterior teeth (L.U.V.P. 11174 and

L.U.V.P. 11175), lateral teeth (L.U.V.P. 11178 and L.U.V.P. 11179).

Horizon and locality: -Grey coloured gypseous shales of Khari Series (Lower Miocene) at

Matanomadh.

De.rcrijllioll : -'I'llc Clllt(';rior tcctll (1")1. VI, fig. 10) ,lrc !;m,lll :lll<ll,lt<:r:llly (:ollll}rc~~cu; tll<:i.'

anterior edges are sharp with worn serrations, 9-11 per mm. 'l'he posterior edge is straight

forming a hook-like structure towards the apical part of crown. 'l'he cross-section of tooth is

asymmetrical, the anterior side is acute, while posterior is rounded. Both external and internal

faces of the crown possess longitudinal striations, 2-3 per mm, up to the middle length of tooth.

The anterior side is convex with a marked sygmoidal curve as in S. olisiPonensis (Jonet 1967, pp.

189-190, pl. 1) described from the Miocene of Portugal. There is no pulp cavity.

Measurements in millimeters

Height Width

L.U.V.P.11174 11 5

L.U.V.P.11175 15 7

Lateral teeth :-L. U. V.P. 11179 (Pl. VI, figs. 11 a, b) is compressed and straight with

sharp edges having 8-9 serrations per mm. The cross-section of tooth is sub-elliptical, the internal

face is convex, while the external is less so. External face is smooth, the internal face is with

finer striations running up to the middle of tooth. There is no pulp cavity.

As far as fossil record of Splryraena in India is concerned, Hora (1939) described two

isolated teeth as Splzyraena incertae sedis from Miocene of Balasore but Hora himself \...as doubtful

about the identification.

Foraminiferal limestone (Middle Eocene) at Nareda in Kutch has also produced teeth of

Splryraena, which are slightly smaller in size.

Suborder Scombroidei

Family Scombridae

GENUS CYBIUM Cuvier 1829

CYBIUM sp.

Plate VI, fig. 12

lliaterial :-lsolated teeth, L.U.V.P. .11177, L.U.V.P. 11178 and L.U.V.P. 11180.

.Horizons and localities :-Grey co loured gypseolls shales at Matanomadh, and

coloured arenaceous'shales at Lakhpat ; Khari S~ries (Lower Miocene). grey

WESTERNJND.(A TERTIARY' VERTEBRATES 35

.Description:- Teeth are small, symmetrical and laterally compressed with acute apices.

The height/width ratio is 1.8 to 2.0; in C. serralheiroi (Jonet 1967) from Miocene of Portugal the

h/w ratio is slightly more than one, being equilateral in shape. The Kutch specimens have

sharp cutting edges with serratiolls, 10-11 per mm. The pulp cavity is ab~cnt. Doth tnc faccs.possess

10n.g"itudil1al striations, 10-12 per mm, limited to crown's basc. Kut<:ll SpCCil11CI1S of

(:ybium diilcr in shape from those of (:. angu.\"tidell.r from Eocene of I\frica in wl1ic\l tIle sides

are parallel curving sharply at apex. Cross-section of tooth is biconvex in contrast to hexagonal

in C. bottei from Miocene of Italy and plana-convex in C. serralhciroi from Miocene of Portugal.

Though presently Cybium or seer-fiffies are common in Bay of Bengal. and Arabian Sea

{C. commersonii, C. guttatus, C. kuhli and C. interruptus), there has been no record of their fossil

forms in India prior to the present investigation.

The ~iddle Eocene foraminiferal limestone at Nareda in Kutch has also yielded smaller

; teeth of Cyhzum.

Class Reptilia

Order Crocodilia

Crocodilian scutes (Pl. II, fig. 9) :-Three crocodilian scutes (L.U.V.P. 11161 a, b, c) from

Pipar and one scute (L.U. V.P. 11166) from Matanomadh were collected from the rocks of the

Khari Series (Miocene). The specimens are small in size, the biggest being 7 cm long.

Maximum thickness of the scutes is 1Jearly 1 cm. l'he upper surfaces are deeply pitted, only

at a few places the pits are expanded into elongated grooves.

Crocodilian vertebrae :-Five crocodilian vertebrae have been recovered from Lower

Miocene (Aida Stage) at Samda (L.U.V.P. 111.12, 11153), Buta (I.-.U.V.P. 11154-), Junagia

(L.U.V.P. 11155) and Pipar (L.U.V.P. 11158). Only the centra of vertebrae are preserved

which have characteristic crocodilian procoelous, spool-shaped structure.

Class Mammalia

Order Sirenia

The first record of fossil sirenian from India was by Grant (1840) who recorded the

'occurrence of rib bones from the Nummulitic limestone and marl unit exposed near Aida in

Kutch; the rib bones were believed hy Cleft and Owen in Grant (1840) similar to those of

.iUanatus but flatter. The Nummulitic limestone and marl unit of Grant is equivalent to Wynne's

Nummulitic Group which in Biswas' (1965, 1971 a) classjfication is represented by Babia Stage

(Lutetian) and Berm(~ti Series (Oligocene). nuring present investigation in Ajda, the sircnian rib

bones were collected from the Aida Stage (Chattian-Aquitanian), and probably Grant's collection

was also from the same beds: ' .

36 A. SAHNI AND V. P. MISHRA

Wynne (1872) reported the occurrence of vertebrae and large rib bones which he believed'

to be mammalian from the beds he had designated as the Arenaceous and Argillaceous Groups,

which are now regarded as Aida and Vinjhan Stages respectively, of Miocene age (Biswas 1965).

Wynne collected these vertebrae and rib bones from Buta, Joonagia, Jangadia, Lakhpat, Pipar

and Samda. Our investigations of these localities have also led to the recovery of vertebrae and'

rib bones which are definitely sirenian. Biswas (197Ib) also collected a large number af rib

bones from the upper part of rocks of Waior Stage (Chattian) exposed near Dedhapar and'

Walasara, which he believes to be of some reptile, but our examination of these bones revealed'

them to be sirenian. ..

Prior to present investigation, only a few fossil sirenian genera were known from Asia ~

Anomotherium from Upper Oligocene of Eastern Asia; lndosirenjavanense from Upper Miocene-

of.Java (Koenigswald 1952) and Miodugong brevicraniusfrom the Miocene of Ceylon (Deraniyagala

1969).

Family Dugongidae

Subfamily Halitheriinae

GENUS HALITHERIUM Kaup 1838

HALITHERIUM sr.

1>1ate VI, fig. 13

Jyfaterial :-L. U. V.P. 11122, left and right rami of the mandible with the alveoli for"

11 to Ms.

Horizon and locality :-Bioclastic limestone of Chattian (upper part of Waior Stage) at.

Matanomadh.

Description :-The mandible is robust with a length of 19 cm; the rami are slightly'

incomplete both on their anterior and. posterior sides. The estimated length of the K utcll.

specimen is about the same as in Dugong. Halianassa, Protosiren fraasi (Eotherium aegypticum),

and Eotheroides (Eosiren) but less than in Manatus, Rlrytina, Felsinotherium, Halitherium and.

Desmostylus.

The rami are long, high and laterally compressed, being narrowest at the position of"

P1-PZ' dorsal to mental foramen, after which the rami gradually widen posteriorly. The sym-

physial region is decurved to the same degree as in Halitherium schinzi, Eotheroides (Eosiren) and!

Protosirenfraasi, more decurved than in Manatus, Prorastomus and Desmostylus, but less so than in.

Dugong, Rliytina and Felsinotherium. The symphysial region is thickened, each ramus beco'ming~"

somewhat swollen at its antero-ventral border. The symphysis is shorter and narrower than in,

Manatus and Eotheroides (Eosiren) but nearly of the same length and width as. in Protosiren fraasi

(E'otherium aegypticum). The mandible of Prorastomu.f .firenoides from the Eocenc of Jamaica

(Owen 1855) with a very long symphysis is narrower than the Kutch specimen. At the posterior

end of the symphysis, the rami are raised both dorsally and ventrally.

The mental foramen opening anteriorly, lies nearly in the mid-line of the ramus and is.

oval in shape, 4 cm long, antero-posteriorly elongated and narrowing posteriorly. Its anterior

border is 2.5 to 3 cm from the anterior margin of the mandible. In the Kutch specimen the"

anterior margin of the mental foramen is entire with no grooved an!terCi>-vental extension which

is seen in Dugong, Manatus, Prorastomus, Felsinotherium, Halianassa, Eotheroides (Eosiren), and

Protosiren fraasi. In Halitherium schinzi there are 6 to 7 small mental foramina and\ in Rlrytina

the mental foramen is absent.

The alveolar border. is slightly convex as in Halitherium schinzi,. Protosiren fraasi and!

Eotheroides, less convex than in Dugong and Rhytina but more clDnvex than in Prorastomus and

Manatus, the last two belonging to the family Trichechidae. The occlusal border of mandible.

posterior to Psis wider than in Manatus and Protosiren, but ventral border is. narrower than in!

Prorastomus and Eotheroides. The ventral border is concave forming an arch in between the

bulbous symphysial region and enlarged angular region. The Kutch specimen is..less arched than

in Dugong, Felsinotherium and Halitherium, more arched than in. Manatus. and Prarastomus and!

similar to the condition found in Protosiren and Eotheroides.

WESTERN INDIA TERTIARf VERTEBRATES 37.

The coronoid process arises externally and sharply from the ramus, posterior to the last

molar as in H. schinzi. 1'he upper portion of the coronoid process, the articular region and

posterior portion of the angular region are broken in both the rami.

Crowns of none of the teeth are preserved. Alveoli for single-rooted three incisors and a

canine occur on the decurved symphysial region. The alveolus for first incisor is constricted

laterally, the tooth was Jirected forward. The alveolus for second incisor is after a diastema

of 5 mm from 11 and is nearly circular in outline; 12 was also directed forward. The alveolus

for third incisor lies close to 12, and is nearly of the same size and shape as 12, but the tooth was

directed obliquely upward. 1'he alveolus for canine is situated after a diastema of 3 mm from

Is and is smaller than the alveoli for all the three incisors. Canine was also directed obliquely

upward.

Posterior to the canines the dorsal margins of the mandibular rami are heavily worn and

the presence of P1 and P2 cannot be determined. rrhe rami broaden posteriorly from the

position of Ps. Ps is two-rooted, the anterior root is smaller and narrower. Posterior to third

premolar the alveoli for two-rooted P 4 and M1 -Ma are closely spaced, having nearly square

outlines and are of the same size except for Ma which is smaller by nearly 33 percent than M2"

Measurements in centimeters

19.0

25.0

7.5

6.0

8.0

4.0

Measured length or the mandible

Estimated length of the mandible

Depth of the mandible at alP 1

" "" M1

" "" Ma

L'ength of the me~tal foramen

Dimensions of the alveoli in millimeters

Length

Width

13 (post.)

Subfamily Halitheriinae

GENUS INDOSIREN Koenigswald 1952 .

INDOSIREN KOENIGSWALDI sr. novo

Plate VI, figs. 14-15

Etymology :-After Prof. G.H R. von Koenigswald, Geological Institute, Rijks University

of Utrecht, Netherlands. -

Holotype : L. U. V.P. 11149, part of the left maxilla having last three molars.

Paratypes :-L.U.V.P. 11l49a, incomplete premaxillae, L.U.V.P. 11149 b, fragments of"

~ku!l ,bones, and isolated teeth. Holotype and the paratypes belong to the one and same

IndIvIdual.

Referred specimen :-L. U. V.P. 11150, premaxillae from Samda.

.Horizons and localities .:-The holotype and paratypes are from the grey coloured. shal~s

of AIda Stage (Lower MIocene) at Matanomadh; L.U.V.P. 11150, the referred specImen IS

from the ferruginous ossiferous conglomerate of Khari Series (Lower Miocene-probabl);'

Burdigalian) at Samda. "

38 A. SAHNI AND V.' P. MISHRA

Diagnosis :-Snout moderately decurved; a pair of enlarged incisors forming tusks

present ; inci~ors cylindrical in olltline anteriorly becoming laterally compressed posteriorly.

Narial openin~ large, nearly as wide as in Recent Dugan.!?, but narrower than in Protosiren ftaasi

from Upper Eocene of Egypt, Prototherium veronense from Auversian of Italy and Halitherium

schinzi from Oligocene of Europe. The anterior height of skull ~r(~atcr tll~ll in i\;[(lllntus and

Protosirenftaasi, but less so than in Dugong, Rlrytina, Halitherium and felsinotherium.

.The cheek teeth (Ml_M3) resemble in shape \'lith the molar of lndosiren.javanense described

from the Upper Miocene of Java (Kocnigswald 195'2), but are 50 to 75 percl~nt larger ill size.

Molars are elongated in shape alld lonser than in 1. javanense ; labial cusps higher tllall lingual

cusps and also higher than the labial cusps of I. javanense ; bilophodont, anterior loph higher

than posterior loph in contrast to I. java'lense where the lophs are of equal heights; eacll loph has

three cusps; transverse valley dividing protoloph from metaloph shallower than in I. javanen,fe

and divides M2 into anterior larger and posterior smaller subdivisions in comparision to nearly

equal areas in Ijavanense. Anterior and posterior cingula of I. javanense represented by high

ramparts (ridges) in the Kutch species, separatel from protoloph and metaloph respectively by

sharper valleys.

Description: -Although the major portion of the skull is present in the collection, it is too

poorly preserved to be completely reconstructed.

L. U. V.P. 11149, the holotype, is the posterior part of left maxilla containing thrcc molars,

which are interesting inasmuch as they show certain similarities with lndosiren javanense described

from the Upper Miocene of Java (Koenigswald 1952). The maxilla is robust with a convex

buccal edge. The palatal surface of the maxilla is eroded. The premolar dentition is not

preservcd.

-In general, the molars are longer than wide being 50-75 percent larger than l.javanen.s-e,

and bilophodont, tl1e anterior loph is higher than posterior loph ; anterior and posterior lophs in

I. javanense are nearly equal in height. Molars are three-rooted; roots are decp, being nearly

23 cm deep in an isolated molar (L.IJ.V.P. 11149 b). The anteriormost molar is a rather

transverse tooth with l/w ratio of about one. The crown of the fir:ot n-:olar is broken. 1'he

roots are thick, stout and deep. The lingual root is slightly shifted internally.

left Ml M3 (L.U. V

H\llolypc

49'

Fig. 6, lndosiren koenigswaldi sr. llOV

The penultimate molar (M2) is elongated, oval in shape and constricted in the middle.

What remains of the labial ~usps suggest that as ill J. ,javancn.rc, tile I,l!>iall:usps, p;l""l!.:()11(~ ,lll<1

metacone were higher than their lIngual counterparts, protocone and hypocone. 'luc paracone

was the highest cusp and was linked to the protocone by a high ridge, except for a short valley

separating protocone from protoconule. Another ri(Jge arises from the protocone descendin<Y

anterolabially and forming a high rampart on the anterior side of the crown. 1'he protoloph i~

separated from the metaloph by a shallower valley than in I. javal/ense. This valley in M2

di,;,ides the tooth into two un~q~al portions, the. allterior being larger than the posterior; in

I. Javanense the transverse \'alley divide" tooth (M2) Into two nearly equal halves. 'l'his is in

contrast to condition in M3, where the median valley divides the crown into two nearly equal

halves. The metaloph is transversely oriented and not obliquely as in M3. l\nother rido"e

arises from hypocone trending postero-labially forming a ramp:lrt on the posterior side of the

crown.. Both the anterior and poste~ior r:amparts are sep3;rated from pr.otoloph and metaloph

re~pectlvely by sharper valleys than ill I. .lavanellse. l\ntcrlor and posterior I amparts rcprcscnt

the cingula in lndosiren javanense.

ftVESTERN INDIA TERTIARY VERTEBRATES 39

Based on the morpholo,sical characters outlined above, the holotype of Indo.\'iren javanense

:£rom the Upper Miocene of Java (Kocnigswald 195'2) is believed to be a second left molar.

The last molar is elongated with rounded anterior and posterior margins and constricted

,in the middle. The paracone ~'hich is a high and conical cusp, i.q joincd to a .qmallcr protoconllie

by a descending ridge (protoloph). Lingually the protoloph does not connect with the proto-

cone, which though higher than protoconule,. is smaller than paracone, being separated from the

protoloph by a deep sharp valley. In the metaloph as well, the metacone is the Ilighest cusp and

in turn joins "he metaconule, being separated from the hypocone by an internal valley. 'l'here

arc two sets of ridges representing cingula arising from the protocone and hypoconc and trend-

ing anteriorly and posteriorly respectively. rrhe anterior ridge is well-developed, being more

prominent than the posterior rampart, which terminates in a smaller accessory cusp.

The cheek teeth of I. koenigsltJaldi sp. nov. resemble those of living Manatus but have

better developed lophs with distinct cones, separated by valleys. Manatus also lacks the upper

tusks. In Dugong (Recent), Felsinotherium (~)liocene) and Halitherium schinzi (Oligocene) the'molars

are bunodont. In Halitherium schinzi, the internal cusps are prominent in contrast to

the external ones in I. koenigswaldi; also the third molar possesses 3-4 accessory cusps on

posterior side, In Miosiren (Pliocene) and Metaxytherium (Miocene) the molars are less bilo-

phodont, while in Desmostylus (Miocene of California ancl.J ava) the cusps in the molars are in the

form or tIle par.tllel cylindl'ical columns arrallgl~d in tllrcc sIIc<.:cssivc pairs allu a Incuiall posterior

one. In primitive sirenians ( Protosiren, Eotlzeroides, Prototherium) the molars arc also bilopho-

dont as in lndosiren from Miocene beds.

An incomplete premaxillae (T..,U.V,P. 1111,9 a) [ormed tllc greater p;lrt ()r tIle sli,r,rlltly

downwardly curved rostrum having an angle or 1350 as in 1\1lanatus. The rostrum in Dugo/lg

(115°) is more downwardly turned, while in Halitherium (125°),-Halianassa (130°), Felsinotherium

(130°), Miosiren (130°), Rlzytina (125°) and Desmostylus (125°) the rostrum is markedly decurved,

more so than in I. koenigswaldi sp, nov., but less so than in Dugong. In Eocene forms Proto,S'iren

.(145°) and Eotheroides (140") the rostrum is less dccurved than in I. koenigswaldi sp. novo The

narial opening is large, elong:lted and widened posteriorly. The anterior portion of the narial

opening preserved is nearly as wide as in Dugon,g, but less so than in Protosiren, Prototherium

and Halitherium. I n front of the narial opening, the snout i~ q uitc broad, broader than

,in Halitherium, Miosiren, Eotheroides and Prototherium, but narrower than in Dugong and

Felsinotherium and rounded from side to side as in Protosiren ftaasi; the lateral sides of

snout are steeply sloping. 'l'he rostrum is higilcr at tile anterior end of narial opening in the

new species than in Manatus and Protosirenfraasi, but lower than in Dugong, Eotheroides libyca,

Rlrytina, Felsinotherium and Halitherium schin;,:i. Anteriorly the premaxillae become laterally,'compressed

enclosing the enlarged first incisors. The first incisor is large, elongated, laterally

compressed and downwardly directed, assuming the shape of a tusk having a cylindric,al,anterior

cross-s(~ction. In sulvivilll!; ~irl~Jli.lIJ~, ()Jlly lJu.I.:.()llg rO~HC~Hl~~ tuskH wllilc MIlII,(llu,I', ;llIU [llry/lll(l (Suu-

l"{,ccent-cxtinct by 1768 A.D.) lack tusks.

Discussion :-Halitherium was fairly a cosmopoliton genus found during the I~owcr

Oligocene to T...ower i\![ioccnc from I~llropc, Olig()CCll(~ ()[ Mall,lg,l~c,lr ;lllU i'Mio!:cllc of North

.."-merica, and now from Kutch. Indo,S'irell on the othcr Iland appears to be confined to souUlern

Asia having been reported previously only from Java, and now from western India. In follow-

ing essential characters lndosiren differs from Halilherium : molars are bilophodont with external

.cusps larger than internal counterparts,' rostrum is less dccurvcd, snout broader and narial open-

mg narrower.

Although lndosiren has been recovered from a relatively younger horizon, it bcars a number

of features more primitive than those of f-/alitherium. The cheek teeth in contrast to llalitherium

and later sirenians arc devoid or a(:cessory CUgpS and crcstlets, In the bilophodont nature, they

are remarkably similar to Eocene Sirenians such as Protosiren, and Eotheroides and Prototherium ; in

Protosiren the molars are square with larger inner cusps, while in Eotheroides there are only two'cusps

on proto-and mctC1;lophs wi~ho~t any ac~essory tubercles. rfhe d(~(:urva~llrc ,of the, rostrum

:is also less marked than In later Slremans but IS nevertheless more so than In Eotherolde.s. and

.Protosiren. "'c

40 A. SAHNI AND V. P. MISHRA

The distribution of fossil sirenians follows closely the present day distribution of the

order which comprises of two families Dugongidae and 'rrichechidae represented respectively by'

Dugong (Sub-family Dugonginae) found in West Pacific, Red Sea and Indian Ocean (as well as

in Gulf of Kutch) and Manatus (Trichechus) inhabiting Atlantic shores of Africa and America.

Another genus Rhytina (Hydrodamites) constituting the sub-family Hydrodamalinae under Family

Dugongidae, inhabiting North Pacific became extinct by 1768 (Zittel 1925, p. 266). As far as,

fossil forms are concerned, the Eocene genera Protosiren from Egypt and France, and Prorastomus

from Jamaica have been given separate family status namely Protosirenidae and Prorastomidae

respectively. Dugongidae comprises of three more sub-families: R ytiodinae Abel 1928 repre-

sented by Rytiodusfrom Upper Oligocene in Europe; Miosireninae Abel 1919 represented by'

Miosiren from Lower Pliocene of Europe; the third sub-family Halitheriinae comprises of fossil

genera indicating the main evolutionary trends leading their present day descendants Rhytina

and Dugong. Eotheroides from Middle-Upper Eocene of Egypt and Prototherium from the Upper

Eocene of Italy are the earliest known representatives of the sub-family Halitheriinae. Late,r

halitheriids are represented by Halitherium from Lower Oligocene to Lower Miocene of Europe,

Oligocene of Madagascar, Miocene of North America and now from upper Oligocene of Kutch;

Halianassa from Miocene of Europe, Upper Miocene and Lower Pliocene of Western North

America, Thalattosiren from Middle Miocene of Europe; Hesperosiren from Middle Miocene

of North America; lndosiren from Upper Miocene of Java and now from Lower Miocene of

Kutch, and by Felsinotherium from Pliocene of Europc. and Lowcr Plioccne of Norlll Africa

and North Amcrica.

The diet of present day Kutch dugongs are mainly angiospermic plants growing in shallow

waters on the basis of which it can be inferred that the fossil sirenians also subsisted on sea

grasses. Crabs and calms (pelecypods) have also been found in Dugong's stomachs (Kingdon

1971, p. 395) ; this is of interest in context that in the Miocene be~ of Matanomadh and

Lakhpat in which sirenian remains are in abundance, t,he fossil crabs have also been recovered,

in large numbers. Kutch fossil sirenians probably occupied the same coastal marshland that

they occupy today indicating the relative stability of the environments.

Order Proboscidea

Suborder Deinotheroidea ,

Family Deinotheriidae

GENUS DEINOTHERIUM

DEINOTHERIUM PENTAPOTAMlAE Falconer

Plate VI, fig. 18

Material:-L.U.V.P. 11130, head of right humerus.

..Horizon and locality: -Ferruginous ossiferous conglomerates of Vinjhan S,tage (Burdigalian) I

atSamda,

Description :-L.U.V.P. 11130 represents a massive and rounded head of the right

humerus of a large proboscid, nearly that of the size of the present day Indian elephant. The

greater tuberosity present laterally on the external side of head is broken. The lesser tuberosity

situated medially on outer (anterior) side of the head is separated from greater tuberosity by a

bicipital groove at the proximal end of the bone. rrhe head is robust, rounded, larger posteri-

orly and sloping anteriorly towards the lesser tuberosity. The deltoid ridge is broken.

Measurements in centimeters

Estimated length of humerus 80.0

Width of the head of the humerus 111 r;

WESTERN INDIA TERTIARY VERTEBRATES

tllc ~uid DicoryjJnocnuerusjale.!,'ar!lll,\'i,\' [roI1l1i'at(~.l\',ldll (70"51/ : '2;~"1"2') ill (~,l~t(~),11 Kult:JI (J))',l~;lU

1964, 1967) ii1 grits i.ll1d cOllglolncralCs, cq uivalcllts of the horizon providilJg Deillotheriul/l

from Samda. This indicates that for the greater part of Kutch, land conditions prevailed during

Lower Miocene, as expected, this coincided with a period of marine regression. In the I.ower-

Middle Miocene of Indian sub-continent, tJl~re occur a number of simil;lr f,lullas (~om[Jri~illg of

anthracotheriids, suids and proboscidcans in Hugti Beds and L;'atehjang zone in Pakistan, Vinjhan

Stage in Kutch, upper Boka Bil Stage in Assam, and Kama Sta~e)hear Prome in Burma.

SUMMARY AND CONCLUSIONS

6.7.

9.10.

12.

13.

15.

Palaeogene and Lower ivliocene deposits of western Kutch are mainly nearshore

marine deposits with minor non-marine intercalations.

Vertebrate faunas have been collected from three horizons: the oldest of Lower

Eocene and Middle Eocene (Lutetian) age, the intermediate of Upper Oligocene

(Chattian) age and the youngest of Lower Miocene (Aquitanian~Burdigalian) age.

The Eocene vertebrate fauna is essentially marine comprising of large siluroids,

trionychid turtles, tomistomid crocodiles, as well as Crocot{ylu.\'. MClmmals arc

represented by primiL.ivc ;l,rclla(~()Ccu's, rrol);lbl(~ od()nt()c(:t(:~, ~ir(~IJi;lll~ :I~ w(~ll ~lS ~l

land mammal represented by a mocrilhcriid S;lcrum.

The siluroids are generically indistinguishable from corresponding forms found in

Egypt and Nigeria.

Although tomistomills are fairly wide-spre~ld in the Eocene, tIle gCllUS 1-0111isloll1a s. 's.

is restricted to Africa and now western India.

Indocetus ramani gen. ~t sp. novo is a primitive archaeocete possessing a number of

features in common with mcsonychid condylarths.

Additional material of Protocetus sloani, and P. harudiensi.\' sp. novo reconfirm our

earlier conclusion (Sahni and Mishra 1972) that the genera Protocetus and P appoce-

ius are not congeneric but distinct.

Andrelu.\'iphiu.\' kutchensis gen. et sp. novo and Andrewsiphius minor sp. novo are

characterised by an extremely long symphysis extending up to the molars and

probably represent the oldest record of odontocetes.

Thc Eocene sirenian from Kutch are represented by a p(~lvic girdle which is

specifically indistinguishable from that of Protosiren fraasi from Middle Eocene of

Cairo (Egypt). '.

The existence of terrestrial mammal during Middle Eocene in Kutch is suggested

by the presence of a moderately large and robust sacrum presumably belonging to

a moeritheriid.

Chattian vertebrate assemblage is also marine and consists exclusively of sirenian.

remains of the genus Halitherium.

Lower Miocene horizon has yielded lndosiren koenigswaldi sp. nov., closely allied to

lndosirenjavanense from Upper Miocene of Java.

1'he youngest vertebrate producing unit (Burdigalian) is non-marine and has

yielded terrestrial mammal represented by Deinotherium.

The early Palaeogene faunas of western India show a remarkable similarity to those

of Africa, lasting at least until Lower Miocene. Unrestricted migration of Palaeo-

gene marine mammals was possible across the sea extending from western India

to Africa. The more northerly, Lutetian mammal faunas of Kashmir are terrestrial

representing a different facies and are quite distinct fro~ the Kutch Lutetian

mammals. .

In the Lower Miocene, there occur a number of similar faunas comprising of

anthracotheriids, suids and proboscideans in Egypt (Africa), Bugti-I,'atehjang

(Pakistan), Kutch and Assam (India) and Prome (Burma). I.,

,;I'

'.~

42 A. SAHN/ AND V. P. Af/SHRA

I{Eli'ERENCES

ABEL, 0., 1904

ANDRF.WS, C. W. 1906

ANTUNES, M. C. F. TELLES 1961

BIsWAs, S. K, 1965

1971 a

1971 b

Gnow J M. IQ{;,\

COPE, E. D., 1895

D' ARCHIAC, A., ANU I-lAIMR, 1853

DECHASEAUX, C., 1958

1961

DEHM, R., AND OETTINGEN-

SPIELBERG, T. ZU., 1958

DERANIYAGALA, P. E. P., 1969

1883

DUNCAN, P. M., AND

SLADEN, W. P.,

l"EISTMANTEL, 0., 1876

FRAAS, E., 1904

GID~EY,J. W.,

t913

GnqSl-r, n. K., 1959

GOODRICH, E., 1958

G~ANT, C. W., 1840

1857

GREGORY

W.K.1951

J-JOR~, S. L 1939

AYAI{AM, K. C., 1955

JORDAN, D. S., 1923

.JONBT, S., 1967

;.. ~.

...

.Or ."

"f '".'

.:~:'

WESTERN INDIA TERT IARr VERTEBl~ATES 43

1955

KALIN,]., Crocodilia. In Traite fIe Pnleontolo,l,'ie, .J. l'ivctcau, cd., Masson

ct Cic, l~ditcllrs, Paris, V, 695-784.

Description of an apparently' ncW toothed cetacean from South

Carolina, Smith. Inset. Miscl. Coll., 76 (2723), 1-7.

The history of \vhales: Tllcir adaplation to life in thc waler,

Ollart. Rev. BioI. 3 (I), 29-76 ; 3 (2), ]74-208.

Compa';.'ative anatomy of the vertebrates, C.U. Mosby Co., Saint Louis,

14], 199.

East Afi.ican Mammal.l", An Atlal" of Evolution in Africa, I; Academic

Press, IJondon & New York, 1-4'16.

The origin of the Cetacea, Doklat{v Akad. "iVauk SSSR., 122,

950-952.

On the origin of Cetacea, Proc. XV Int. Congo ;:'ool., 445-447.

Fos~il Sirenian from Java, Proc. R. Neth. Acad. Sc.,Amsterdam,'.'

B 55 ~5), 610-612.'

Crocodilia, In Osnovy Paleontologii, J. A. arlo\', ed., "Nauka",

Moscow, 506-523.

Thc extinct mamlilaliall fauna of Dakota alld Nebraska, including

an accolmt of some allied form!! from other localities, together

with a synopsis of the m.lmmalian rcmains of North America,

Jour. Acad. Nat. Sci., Philadelphia, Ser. 2, 7, 1-472.

Les Poissons Neogenes de la Bretagne dc I' Anjou et de la Tollraine,

.Wem. Soc. Grol, de Franre, N.~., MCIII. 11,36 (1), I-fiG, j>ls. XLIV

to XI.Vll.

Indian Tertiary and I'ost-Tertiary \'ertebrata : Molar Teeth and

other remain!! of Mammalia, Pal. Indica, Ser. X, 1 (2), 1-69.

Occurrencc of Plrsiosa7lrll.1" in IJldia, Rec. Grol. .S'llrv. luflia, 9 (4.),

]54.

Notices of New anu oliler Vcrteurala frmn Indian 'l'crliary.and

Secondary rocks, Rec. Geol. Surv. India, 10 (I), 35, 43.

Indian Pretertiary vertebrata. Fossil l{eptilia and Batrachia:

VII-Plesiosaurlll" from the Umia group of Kutch, Pal. Indica;

Ser. IV, I (3),28-3].

Teeth of fossil fishes from ]{amri Island and Pul1jab, Rec. Geol. Surv.

India, 13 (I), 59-61.

On Ule occurrence of the crocodilian genus T omistoma in th~

Miocene of thc Maltese Islands, (b/art. Jollr. GeQl. Soc. Londo.n,

42, 20-22.

Indian Tcrtiary and Post-Tcrliary Vcrtcurata: Tertiary f'ishes,

P,ll. Indica, Ser X, 3 (8),24.1-261.

Miocene Sharks from India. Recent /lesearches in GeQlogy (Jhingran

(~ommemorative Volume), Geology Deparlment, Delhi Uni-

versity, 191-200.

Coelodus jacobi, a new Pycnodont fish from Eocene IJeds of the Garo..

Hills, Assam, Rec. Geol. Surv. India, 85 (4), 563-567, pl. 21..

Eocene Sharks frpm India, Proc. GOt/I III fl. Sci. Con.lIr. Assoc., Abst-

racts,4, 31. "

Burdigalian microfauna from Kutch, India, Proc. Nfl!. Inset. Sci.

India, B 34 (4), 159-190. '.

Microfauna and age of the "GypscoliS Shale~" Western Kutch,

India, Contr. Cushman Fo7lnfl. Foram. /les., 19 (I), 21-2G.

Middle Eocenc Plmlklonic foraminiferal 7:on:ltion of Kutch, India,

Micropal" 16 (I), 37-46.'

Hesperoinia delpiazi N.G. et N.Sp. (Platanistidae, Cetacea) Della

Pietra Leccese, lv[em. 1st. Geol. ,Win. Univ. Padova, 19, 1-10.

Skull characters and affinities of the extinct l'~lorida gavial Gizvialo-'

suchus americana (Sellards), Bull. Amer. l\IUS. Nat. Hist., 44 (5). ;:',

On fossil fish-teeth from the Nicobar Islands, Proc. Natl. It/St. St~;

India, 11, 149-151. ;';. ~

A contribution to the history of I~ocene Siluroid {ishc!!, Proc. Zoot::

Soc. London, 201-207. .

The zonal distribution and description of larger Foraminifera of

the Middle and Lower Kirthar Series (Middle Eocene) of parts

of western India, Rec. Geol. Surv. India, 59 (1),115-164. ,

On the fossil skull of a Mammal (]C>rorastomus sirenoides) from the ~

Island of Jamaica, OJ/art. Jour. Geol. 'Soc. London, II, 541-543.

A Text BoQk of Zoolo,!,'y, II, Vertebratel", MacMillan .and .Q~mpany,

Ltd., London, 1-683. .:;',io' .: .

..,"""" ...

KELLOGG, R., 1923

1928

KENT JR., G. C., 1965

1971

KINGDON, J.,

KLEINENBERG, S. E., 1958

1959

1952

---,

KOENIGSWALD, G. H. R. von,

1964

KONjUKOVA, E. D.,

Lt-:IDY, J., 1860

1957

LERICHE, M.,

lU76 a

LVDEKKER, It.,

1876 b

1877

1879

1880

1886 a

1886 b

MEHROTRA, D.K., MISHRA, V.P., 1973

AND SRIVASTAVA, S.,

MENON, A. G. K., 1958

AND PR'SAD, K. N.,

MISHRA, V.P., CHOUDHARY,N.K., 1973

AND KHARE, S. K.,

MOHAN, M., AND BHATT, D.K., 1968

AND GUPTA, S.C., 1968

" AND SOODAN, K. S., 1970

MONCHARMONT ZEI, M., 1956

MOOK, C.,

1921

NAIR, K. K., 1945

T.,

1889

NEWTON,

NUTTALL, W.L.F., 1926

OWEN, R., 1855

1897

PARKER, T.]., AND

HASWELL, W.A.,

.~~.~.~~~:

SAHNI AND V. P. MISHRA

PEYER, B., 1928

PILGRIM, G.E., 1940

~:':~;':I

"..iJ' ...:.,r".'[~:,:~:;.:i~

PRASAD, K.N., 1964

1967

--,

AND RAO, V.R. 1958

RANOA RAO, A., 1973

RAO, V.I{, 1956

I{oMER, A,S. 1966

SAHNI, A., AND MISHRA, V.P., 1972

SELLARDS, E.H., 1915

SILL, W.D.,

SINGH, M.P. 1968

1967

SOWERBY~J.De C., 1840

SRIVASTAVA, I.P., 1970

STOLICZKA,.F., 1873

STROMER; E., 1904

SYK~S, W.H., 1840

...""'~

~~~t!"".1';-':

f..,,~ "'"

",;, ~'- "

~(.,' .. SZALAY, F.E., AND GOULD, S.J.,

TANDO~, K.K'.,

1966

1971

TEWARI, B.S.~, 1956 a

~.7 1956 b

1957

1959

AND llHAROAVA, O.N., 1959

1964

1967

'~.N'; '.i't.,.',W'~~.

" .' 1968

,.AND

AND TA~~I;>Ql'ft:~~'~;' 1960

Die WeIse des agyptischen Alttertiars nebst einer Kritischen Uber-

sicht uber aile fossilen WeIse, Abhandl. Bay. Akad. U'iss., math.

-nat. Abt., n.s., 32 (3), 1-61.

Middle Eocene mammals from north-west India, Proc. Zoot. Soc.

London, B 110, 127-152.

Miocene vertebrates from Cutch district, Gujrat, India, LJull. Ceo/.

Soc. India, 1 (2), 9-12.

Fossil mamma.1s of Cutch district, Gujrat, India, Quart. Jour. Ceo/.

Min. Met. Soc. India, 39 (3), 187-192.

Fossil Pycnodont fish teeth from I{anikot, Sind, llec. Ceo/. Surv.

India,85 (4), 557-562.

Notice of two new mammals from the Upper Eocene Kalakot Beds,

India, Direct. Ceol., Oil and Nat. Gas Comm., Dehradun, spl.

paper 2,1-6.

The sku!! or an Eocene Siluroid fish from western Kutch, India,

Jour. Pal. Soc. India, 1 (I), 18.1-185.

Vertebrate Paleontology, The University of Chicago Press, Chicago

and London, 1-468.

A new species of Protocetus (Cetacea) from the Middle Eocene of

Kutch, western India, Palaeontol., 15 (3),490-495, pl. 97.

A new Gavial from the late Tertiary of Florida, Amer. Jour. Sci.,

Ser. 4, 40 (236), 135-138.

The zoogeography of the Crocqdilia, CojJeia, 1, 76-313.

Study of Lower Tertiary rocks of Kanoj, Kutch (with special

reference to planktonic Foraminifera), Acad. Thesis, Lucknow

University (unpublished), 1-220.

Systematic description of organic remains from Cutch, Trans. Ceol.

Soc. London, Ser. 2, 5, 327-329.

Geology and Micropalaeon~ology of the Miani-Vinjhan arCfis of

south-western Kutch, Gujarat (\vith special reference to the

study of Ostracoda) Acad. Thesis, Lucknow Uhiversity (unpubli-

shed), 1-269.

The Cretaceous fauna of southern India: The Brachiopoda,

Ciliopoda, Echinodermata, Corals, etc., Pal. Indica, Ser. 8,4,

1-202.

Nematognathi aus dem Fajum und dem Natronthale in Aegypten,

Neues Jahrbuch, 1, 1-6.

A notice respecting some fossils collected in Cutch by Capt.

Walter Smee of the Bombay Army, Trans. Geol. Soc. London,

Ser. 2, 5, 715- 718.

Asiatic Mesonychidae (Mammalia, Condylartha) , Bull. Amer. A/us.

Nat. /list., 132, (2), 127-174.

Stratigraphy and Micropalaeontology of 'fertiary rocks of south-

western Kutch, Acad. TheJis, Lucknow University (unpublished),

1, 1-439; 2, pis. 1-52.

On the discovery of mammalian and reptilian r(~lrlaiJls from lilt:

Middle Eocene rocks of S. W. Kutch, india, Curro Sci., 1O,

(16),436-437.

The genus /lalkyardia from Kutch, Western India, Jour. Pal.

Soc. India, 1 tl), 172-175.

The genus SpiroclyjJeus from Kutch, Western 111dia, Curro .s'ci.,

25,319-320.

Geology and stratigraphy of the area between Waghopadar and

Cheropadi, Kutch, Western India, Jour. Pal. Soc. India, 2,

136-148.

On a few fossil shark teeth from Miocene beds of Kutch, Western

India, Proto Nat. Instt. Sci. India, B 25 (5), 230-236.

Kutch Microfauna: Aquitanian Foraminifera from W aior, S. W.

Kutch, Jour. Pal. Soc. India, 4, 6-11.

1'wo new species of shark teeth from Gaj beds of Matanumarh,

Kutch, Jour. Pal. Soc. India, 9, 74-76.

Tertiary beds of Kanoj-sehe area, Western Kutch, Curro Sci., 36,

(9),237-238.

On the presence of Auversian beds in the Kanoj-Sehe section,

Kutch, Jour. Pal. Soc. India, 13, 48-52.

Kutch Microfauna: Lower Tertiary Ostracoda, Proc. Nat. Instt.

Sci. India, B 26 (4), 147-167.

J;VESTERN IN.DIA TERTIARY VERTEBRATES 45

TOULA, F., AND KAIL, J., 1885

1907

TRUE, F.W.,)

VAN V ALEN, L., 1966

1968

VREDENBURG, E.W., 1936

1925

1928

WF:ICHF,RT, C.K.,

WOODWARD, A.S., 1 !)5B

1889

1901

1908

WYNNE, A.B. 1872

YABALAKOV, A.V., 1964

YEH, HSIANG-K'UEI, 1958

1964

1963

YOUNG, C.C.,---,

AND CHOW, M.,

ZITTEL, KARL, A. V ON, 1925

U'ebercincnkrokodilschadel aus den lcl'liarablagcl'Imgcl1 , von

Eggerlburg. Nieder OeJiterrcich, Denks. K. K. Akad. 11IiJs.,

lv/ath. Naturw, Kl., Bd. L., 1 (2),299.

Remarks on the type of the fossil cC'tacean .lJgorophiuJ' pygmaeus

(Iv[uller), Smith. Instt. .~'l)l. l-'ubl., 1694, I-B.

Deltalheridia, a ncw order of mammals, Bull. Amer. lVUS. Nat.

HiJt., 132 (I), 1-126.

Monophyly or Diphyly ill lite origin of whales, Euolution, 22 (I),

3741. '

Nummulites douuilli, an undescribed species from Cutch with remarks

on zollal di:otribution of Indian Nummulites, Rec. Geol. Suru.

India,34 (2), 79-94.

Dcscriptiuns of Mollusca from the post-Eocene Tcrtiary formation

of north-wc:;lcrn India: Cephalopoda, OpiJithobrallchiata;

Siphono~tomata, ,VC,'11. Ceol. ,S'/IYlJ. Inr/in, 50 (I), 1-:150.

Descriptions of Mollu:oca from the post-F:ocene Tertiary formation of

north-\\'est~rn .Jnrliil: (:~stror()rl~ {in p~rtf ~nd Lame1/ibranch-

iata, 1\.1,:/1/. &1'111. Srllil. iIIliil" ~O (2), :{:,I-:,OI;.

Ana/o7r~v of/he Ghord{/tl~f, Mc(;raw Ilill Hook (~o., Inc., 393-3~)ll..

Catalogue,if thejo,fJiljifhes in /hl' Bri/ish lvfu.fl'/im (.Natural His/ory) ,

1, 1~lasmobraIlcllii, llritisll Mu:oellm, Londoll, 1-474, pis. 17.

Cata/o,gue of the .fos.ril ./i.rlll'.f in Ihe Hri/i.fh ,~fnse/lm (N(llnral History),

4, AI;lillor'l~rygj;ln 'li~I"oslrlllli. ISrilisl1 :\11I!1('llln. I,OIldoJl,

1.631;, pis. 19.

On some fossil remains from the Lamcta beds at Dongargaon,

Central Province, Pnl. Indica, N. S., 3 (3), 1.6, pl. I.

Memoir on till' Geology of Kllcth, to accOlllpany a map compiled

by A. IS. WYJ11IC alld I". rcddcJl, duriJlg lll(~ scaSOl1S of 1867-68

and 1868-69, Mem. Geol. Suru. India, 9 (1). 1-293.

Convergenee or parallelism in the evoliltio!l of cetaceans, l.Jalaeont.

Zhur., 97-106 (In I~ussian), translated in Int. Geol Reu., 7

(1965), 1461-I4G8.

A new (~rocodile from Maoming, Kwangt'mg, Vert. Palaj'iqt., 2 (4),

237-240 (Chinese), 241-242 (in English).

New fossil crocodiles from China, Vert. Palasiat., 8 (2) ,:189-208.

Cretaceous and Palaeocene vertebrate horizons of north ~wang-

tung, .s'c:. Sinica, 12 (9),1411.

Text-Book of PalaeoltlQlo,~y, III, Mamlnalia (ill English,' ed.

A.S.. Woodward), Macmillall and Co., Ltd., Londo!l, 1-3IG.

Text-Book of PalaeontolfJ.lf.y, II, Ii'ishes to Birds (in Englisll, ed.

A.S. Woodward), l\-[acmillan and Co., Ltd.., Lonr!im, I.'l64.

1932

Manuscript received May 10, 1973.

Revised manuscript received April 5. 1974.

46 A. SAHNI AND V. P. MISHRA

EXPLANATION 01" PLATE

Fig.

FIGs. 1-2 MVLIOBATIS Sr.

I-Basal view of dental plate, L.U.V.P. 11078 x 1.2

2-Caudal spine, L U. V.P. 11082 x 1.55

1o'IG. 3 FAJUMIA MENONI sr. novo

3 a-Dorsal view of skull, L.U.V.P. 11140 (Holotype) x 0.35

3 b-Ventral view x 0.35

3 c-Left lateral view x 0.35

FIGs. 4.5 FAJUMIA MISRAI sr. novo

4 a-Dorsal view of skull, L.U.V.P. 11142 (Holotype) x 0.35

4 b- Ventral view x 0.35

4 c-Right Jateral view x 0.35

5 a-Dorsal view ofcleithralpart of pectoral girdle, L.U.V..P. IJ142 ax 0.5

5 b-Ventral view x 0.5

Fig.

EXPLANATION aIi' PLATE II

Fig,

~I!'("

f.~.~.I Fig.

FIG. 1 SOCNOPAEA HORAI Sp. novo

I-Dorsal view of skull, L.U.V.P. 11145 (Holotypc) ,x 0.3

FIGs. 2-4 ARIUS KUTCHENSIS Rao2-Dorsa1

view of skull, L. U. V.P. 11036 x 0.3

3 a-Dorsal view of skull, L.U.V.P. 11047 x 0.4

3 b-Ventral view x 0.4

4 a-Dorsal view of pectoral spine, L.U.V.P. 11048 a x 0.5

4 b- Ventral view x 0.5

4 c-Anterior view x 0.5

Fig.

FIG. 5 PYCNODUS Sr.5

a-Oral view of splenial tooth,L.U.V.P. 11173 x 10

5 b-Basal view x 10

Fig.

FIG. 6 CROCODYLUS sp.

6 a-Dorsal view of premaxillae, L.U.V.P. 11135 x 0.33

6 b-Ventral view x 0.33

FIGS. 7-8 CROCODILIAN COPROLITES

7---Coprolite, L. U. V .P. 11139 a x 0.5

8::-Coprolite, L.U.V.P. 11139 b x 0.4

FIG. 9 CROCODILIAN SCUTE

9-Crocodilian scute, L. U. V.P. 11161 x 0.5

FIG. 10 TOMISTOMA TANDONI sp. novo

10 a-Side view of cervical vertebra, L.U.V.P. 11062 b x 0.55

10 b-Posterior view x 0.55

Fig.

1~':;1!.:id'~"; "f..'~~~~l;~

:~~,,:~~,.~,;

Fig.'.'\~'~:;;!

Fig.

~"','!;;t;o"C

.~i~",~;;ii!;':~~.f

.~.~~,,'~'

.~r

...

e ! , ,'~;);~:\""J~.

;"'N~"'I'-

r'tc-~""";'

{..~g'..J'

I,,;

"~"'i ,~. 1\':'.:

..,,!i ) "

...'t. .

EXPLANATION OF PLATE

Fig~

FIGS. 1~3 TOMISTOMA TANDONI Sp. novo

I a""--Dorsal view of skull, L.U.V.P. 11062a x 0.25

..b-Ventral view x 0.25

2 a-Dprsal view of mandible, L.U.V.P. 11062 (Holotype) x 0.25

2 .b-;-Ventral view x 0.253-Dorsal

view of left articular part of mandible, L. U. V.P . 1062 x 0.25

b'..'iI,. ,.

~~'.I;r; .;:,::.,;~~

~; ::

ft~~ ".,~f~i

"~",i~~~::

£:'~'1.,.\,.

-!,,~~;jc'i:;.\!,~itf!\",c.,.

WESTERN INDIA TERTIARY VERTEBRATES

Fig. FIGS. 4-6 TRIONYX Sr.

4 a-Dorsal view of carapace, L.U.V.P. 11077 x 0.5 (From Khari Series at Matanomadh).

4 b-Ventral view x 0.5

5-Dorsal view of carapace, L.U.V.P. 11182 x 1.2 (l"rom Babia 1:)tage at l-larudi).

6-Dorsal view of carapace, L.U.V.P. 11172 x 0.33 (From Khari Series at Aida).

EXPLANATION OF PLATE IV

FIGS. -3 INDOCETUS RAMANI gen. et sp. novo

I a-Dorsal view of skull, L. U. V. P. 11034 (Holotype) x 0.33

I b-Ventral view x 0.33

I c-Left lateral view x 0.33

2 a-Dorsal view of the right posterior side of skull x 0.33

2 b-Ventral view x 0.33

2 c-Posterior view x 0.33

3- Lingual view of right MI, L. U. V.P. 11034 b x I

FIGS. 4-7 PitOTOCETUS HARUDIENSIS Sp. nov.

Fig. 4 a-Left lateral view of skull, L.U.V.P. 11037 (Holotype) x 0.5

4 b-Ventral view x 0.5

4 c-Occlusal view of left maxilla x 1

5 a-Occlusal view of left mandibular ramus, L. U. V.P.. 11037 a x 0.5

5 b-Left lateral view x 0.5

6 a-Labial view of left MI-M2, L.U. V.P. 11037 b x 1.2

6 b-Occlusal view of left M2 x 1.5

6 c-Lingual view of left M2 x 1.5

6 d-Labial view of left M2 x 1.5

6 e-Occlusal view of left Ml x 1.2

7 a-Lingual view of isolated upper tooth, L.U.V.P. 11037 c x 1.25

7 b-Occlusal view x 1.25

EXPLANATION OF PLATE V

FIOS. -2 PROTOCETUS SLOANI Sahni and Mishra

I-Occlusal view of anterior rostral fra,gment, L. U. V.P.

2 a-Dorsal view of skull, L.U.V.P. 11146 x 0.3

2 b-Ventral view x 0.3

2 c-Posterior view x 0.3

1043 x 0.5

FIGS. 3-5 ASSORTED CETACEAN MANDIBULAR FRAGMENTS

Fig, 3 a-Occlusal view of mandible, L.U.V.P. 11061 x 0.33

3 b -Left lateral view x 0.33

4 a-Occlusal view of mandible, L.U.V.P. 1113Bx 0.33

4 b-Left lateral view x 0.33

5 a-Occlusal view of mandible, L. U. V.P. 11132 x 0.33

5 b-Left lateral view x 0.33

5 c-,-Ventral view x 0.33 l'

FIG. 6 ANDREWSIPHIUS KUTCHENSIS gell. et sr. novo j

..~..

~,": .{})

.~i;!

;Ji'~:~,~~'"",{

, -f' .i'

L ,"

6 a -Occlusal view of mandible, L. U. V.P. 11060 (Holotype) x 0:3;:J

6 b-Left lateral view x 0.33

SAHNI AND V. P. MlSHRA

FIG. 7 ANDREWSIPHIUS MINOR Sp. novo

7 a --'"Occlusal view of mandible; L. U. V.P. .11165 (Holotype) x 0,5

7 b-Right ]ateral view x 0.5

EXPLANATION OF PLATE VI

~~~

.'" .:(N:"

:.~ljJ

I+': ~.'i.~~

"':',.~,1.:

" ,

Fig.

FIG. I PROTOSIREN FRAASI Abel

External view of left pelvi'~, L.U.V.P. 11038 and, L.U.V.P. 11039 x 0.5

Internal view x 0.5 .

J.'IG. 2 if. MOERITHERIID SACRUM

2 a-Dorsal view of sacrum, L.U. V.P. 11069 x 0.4

2 b-Ventral view x 0.4 .

2 c-Left lateral view x 0.4

Fig.

FIG. 3 RA,JA Sr.

3 a-Dorsal view of tuoth, L.U.V.P. 11099 x 4.2

3 b-Ventral view x 4.2

Fig.

ti!;,t.

FIGS. 4-6 MVLIOBATI5 Sr.

4 a-Basal view of median tooth, L. U. V. P. 11096 x

4 b -Oral view x 1.2

4 c-Lateral view x 1.2

5 a-Basal view of lateral tooth, L.U. V.P. 11102 x 2

5 b -Oral view x 2

6-Spine, L.U.V.P. 11097 x 1.3

FIG. 7 ~ETOBATUS sp.

7 a-Basal view of tooth, L.U.V.P. 11170 x 0.75

7 b-Oral view x 0.75

Fig. FIGS. 8-9 lliODON sp.

8 a-Posterior view of dental plate, L.U.V.P. 11095 x I.]

8 b--'-Anterior view x 1.1

9 a-Posterior view of dental plate, L.U.V.P. 11089 x 2

9 b-Anterior view x 2

fi~.

10.

FIG~. 10-11 SP~IYRAENA sr.

External view of anterior tooth, L.U.V.P. 11175 x 1.5

-Internal view of lateral tooth, L. U. V.P. 11179 x 1.3

-Side view x 1.3

.:Fig.

'..Fig.

}.'IG. 12 CVBIUM sp.

12-:-IRternal view of tooth, L.U.V.P. 11177 x 1.3 "

FIG. 13 HALITHERIUM

13 a-Occlusal view of mandible, L.U. V.P. 11122 x 0.313

b -Left lateralview x 0.3

FIGS. 14-15 INDOSIREN KOENIGSWALDI sp. novo

14 a- Dorsal view of premaxillae, L. U. V.P. 11149 a x 0.51

l4 b-Latera1 view x 0.5115a-Occlusa1

view of left ma,x,iI1a with Ml.M3, L.U.V.P. 11149 (Holotype) x 1

15 b-'"-Lingua1 view x 1

.1"IGS. 16-17 SIRENIAN VJo:;RTEBRA AND ltm

16-Vertebra, L.U.V.P. 11128 x 0.33

,17.,--Rib, ~U; V.P. 11121 x 0.5

" 'FIG. 18 DEINOTHERIUM PENTAPOTAMIAE Falconer '...18-':-H;ead

ofrigltt humerus, L.U.V.P. 11130 x 0.27

Fig.

~ig,

;' ,

~'FI,g.

~~~~

SAHNI AND MlSHRA: WKS'I'KRN INDIA

MONOGR. PAL. SOC. INDIA, No.3, 1975 PLATE III

;",.;;.t;: $t\F{NI AND MISHRA: WESTERN n

MONOGR. PAL. SOC.. INDIA, No.3, 1975 PLA TP- V

"Discovery of Miocene Whale Fall Fauna from India: Taphonomy and Palaeoecology of a Vertebrate-Invertebrate Assemblage."

Article

Oct 2024
Hist Biol

This research explores the fascinating ecosystem surrounding whale falls during the Miocene Epoch (more than 20 million years ago) in India, shedding light on the intricate relationships between vertebrates and invertebrates in these unique habitats.

Taxonomic revision of the giant marine snake genus Pterosphenus Lucas, 1898, based on new fossil material from the middle Eocene (Bartonian) Harudi Formation of Kachchh (Kutch) Basin, India

Article

Jun 2024
J Vertebr Paleontol

Fossil snakes of the family Palaeophiidae are primarily known from rare, isolated vertebrae, leading to a poor understanding of the intra- and inter-columnar variation within a palaeophiid species. This leads to ambiguity in the taxonomic identity of these snakes, as in the case of the Eocene-age marine snakes of the genus Pterosphenus, in which the majority of species have been described by characters of problematic taxonomic value. Here, we propose a taxonomic revision of the genus based on an articulated to semi-articulated section of the vertebral column at the anterior to mid trunk-transition of Pterosphenus schucherti discovered from the middle Eocene (Bartonian) Harudi Formation, Kachchh (Kutch) Basin. The characters that separate Pterosphenus schucherti from ‘Pterosphenus schweinfurthi’ and ‘Pterosphenus biswasi,’ viz., the height of the pterapophysis and the anterior concavity of the zygosphene, have been found to be non-diagnostic, and the latter two species have been synonymized with the type species. Additionally, we utilized 2D and 3D landmark-based geometric morphometry to determine the morphometric affinity of palaeophiids amongst the alethinophidian snakes. The body size of the snake was estimated based on extant snakes with a morphometric affinity to Palaeophiidae. The cosmopolitan distribution of Pterosphenus schucherti is modelled based on the sea surface temperature (SST) constraints of the modern cosmopolitan snake Hydrophis platurus, and the known fossil localities of the species. The present findings provide crucial insights into the global paleoecological landscape of the Eocene, and the interactions of Pterosphenus schucherti with other palaeophiid snakes as well as the other Eocene fauna.

Late Cretaceous-Early Palaeocene Lepisosteiform and Siluriform fish remains from Central India: palaeoecological, palaeoenvironmental and palaeobiogeographical implications

Article

May 2024
CRETACEOUS RES

Late Cretaceous-Early Palaeocene geo-climatic events played an important role in the diversification of the modern ichthyofauna. Lepisosteiformes and Siluriformes are two diverse clades of freshwater fishes, poorly known from India in this time interval. Their fossil record documents their early diversification and can be used to reconstruct palaeobiogeographic relationships among the continental masses during the Late Mesozoic. Indeed, the Cretaceous-Palaeocene is an exceptional time span for the Indian plate as it underwent a primary spatial reorganisation and remarkable geologic and climatic changes with extensive outpouring of the Deccan basaltic magma. Field investigation in a Deccan sedimentary sequence at Kisalpuri, Central India has yielded new, rich, and taxonomically important fossil material of Lepisosteiformes (Lepisosteus indicus, Lepisosteidae) and Siluriformes (Siluriformes indet.), which significantly improves their Cretaceous-Palaeocene fossil record from the Indian subcontinent. These fish lived in freshwater environments such as large rivers, lakes, and coastal areas. Their presence along with other aquatic biotic elements suggests that the Kisalpuri was an abundant aquatic ecosystem rich in nutrients, which attracted a diverse range of organisms to live together with fish. Biostratigraphically, these fishes first emerged in the Late Cretaceous-Early Palaeocene on the Indian subcontinent, and later, they spread to other parts of subcontinent throughout the Cenozoic. Their earliest intra-continental distribution was controlled by the palaeodrainage network influenced by the volcanic flows of the Deccan traps. Their inter-continental palaeobiogeographic distribution was broadly controlled by the sequential break-up of Pangaea. There appear to be close palaeobiogeographic linkages between South America and India during the Cretaceous-Palaeocene era based on the similarity between the Lepisosteiformes and Siluriformes remnants from India and those reported from the Upper Cretaceous of the Bauru Group, Brazil.

First documentation of Miocene shark teeth from South Garo Hills District of Meghalaya with a synopsis on the Miocene Chondrichthyes of India and palaeoenvironmental interpretation

Article

Dec 2024

Shark teeth from representatives of four families, including Lamnidae, Hemigaleidae, Carcharhinidae and Sphyrnidae belonging to five genera, namely Isurus, Lamna, Hemipristis, Carcharhinus and Sphyrna, were recorded and documented for the first time from the South Garo Hills District of Meghalaya from the Baghmara and Chengapara formations of the Garo Group of rocks of Oligo-Miocene age. They show close affinity and similarities to the Miocene shark tooth assemblages recorded from the Baripada beds of Orissa and the Surma basin of Mizoram. These selachian assemblages from South Garo Hills point to a shallow marine, nearshore coastal environment well connected to the open sea. Thus, the present study provides new insights on the palaeodiversity and palaeoenvironmental setup of the Miocene sediments of South Garo Hills of Meghalaya and also enhances the existing knowledge of chondrichthyan diversity in India during the Miocene time.

Miocene cartilaginous fishes (Chondrichthyes, Elasmobranchii) from India: A review on global palaeobiogeography

Article

Full-text available

Feb 2025

The Miocene deposits of India are well known for yielding a diverse assemblage of invertebrate and vertebrate fossils. However, research on Miocene elasmobranch fauna of India is limited to a few Miocene sites and the previously reported species also need to be reviewed, taking into consideration their significance on global Miocene palaeobiogeography. In this work, we compile the Miocene elasmobranch fauna of India and also update the status of accepted and unaccepted species reported earlier. It is found that the Miocene sediments of India specially from Kutch, Gogha coast, Piram Island, Quilon Bed, Pondicherry, Mizoram, Baripada, etc. have yielded various chondrichthyans remains, comprising of 17 families, 29 genera and 36 species from the Early Miocene. There are 5 families, 10 genera and 7 species from the Middle Miocene and 17 families, 26 genera and 49 species from the Late Miocene. Here, a correlation of the Miocene elasmobranchs of India and their global counterparts has been done using the beta diversity (Sørensen-Dice coefficient) data to understand the biogeographic dynamics. The high similarity index of elasmobranch fauna of the Early Miocene of India to the Mediterranean region suggests the presence of the Tethyan seaway passages till the early Aquitanian and the existence of the remnant water body that has a connection between the Indo-Pacific and the Mediterranean Sea. In the Middle Miocene, the similarity index of the Indian to the Mediterranean region has decreased from the Early Miocene which suggests a short-term reopening of the marine pathway during Langhian. The similarity index of elasmobranch fauna during the Late Miocene time, is more toward the South Pacific Ocean and Indo-Pacific region, suggesting a shift in migration seaway passages after the permanent closure of the Mediterranean. The majority of extant species of elasmobranch fauna of India are presently found in the Indian Ocean.

Miocene cartilaginous fishes (Chondrichthyes, Elasmobranchii) from India: A review on global palaeobiogeography

Preprint

Full-text available

Dec 2024

The Miocene deposits of India are well known for yielding a diverse assemblage of invertebrate and vertebrate fossils. However, research on Miocene elasmobranch fauna of India is limited to a few Miocene sites and the previously reported species also need to be reviewed, taking into consideration their significance on global Miocene palaeobiogeography. In this work, we compile the Miocene elasmobranch fauna of India and also update the status of accepted and unaccepted species reported earlier. It is found that the Miocene sediments of India specially from Kutch, Gogha coast, Piram Island, Quilon Bed, Pondicherry, Mizoram, Baripada, etc. have yielded various Chondrichthyes remains, comprising of 17 families, 29 genera and 36 species from the Early Miocene. There are 5 families, 10 genera and 7 species from the Middle Miocene and 17 families, 26 genera and 49 species from the Late Miocene. Here, a correlation of the Miocene elasmobranchs of India and their global counterparts has been done using the beta diversity (Sørensen-Dice coefficient) data to understand the biogeographic dynamics. The high similarity index of elasmobranch fauna of the Early Miocene of India to the Mediterranean region suggests the presence of the Tethyan seaway passages till the early Aquitanian and the existence of the remnant water body that has a connection between the Indo-Pacific and the Mediterranean Sea. In the Middle Miocene, the similarity index of the Indian to the Mediterranean region has decreased from the Early Miocene which suggests a short-term reopening of the marine pathway during Langhian. The similarity index of elasmobranch fauna during the Late Miocene time, is more toward the South Pacific Ocean and Indo-Pacific region, suggesting a shift in migration seaway passages after the permanent closure of the Mediterranean. The majority of extant species of elasmobranch fauna of India are presently found in the Indian Ocean.

New Record of Selachian (shark) fauna from the early Miocene (Aquitanian) of Kachchh Region (western India): Palaeoecological and Palaeobiogeographic significance

Preprint

Full-text available

Dec 2024

The past two decades have witnessed a renewed interest in the Neogene (mainly Miocene) interval of the Kachchh (= Kutch) region owing to its potential to yield numerous records on fishes, reptiles, birds, and marine and terrestrial mammals. However, the Piscean (mainly selachian) faunal records from the Neogene (Miocene) interval have been sparsely emphasized. The current investigation reveals the most diverse assemblage of selachian (shark) fauna, to date, from the Aquitanian (~ 21 Ma) Khari Nadi Formation at Kotada, Kutch Basin, western India. The recovered selachian (shark) fauna comprises five genera, including Galeocerdo , Physogaleus , Carcharhinus , Negaprion and Hemipristis representing seven species. The presence of Galeocerdo mayumbensis , Carcharhinus falsiformis , and C. brevirostris comprise the oldest Neogene records from the Kutch Basin, western India. The diversity similarity coefficient data suggests that the Aquitanian selachian assemblage is quite dissimilar compared to the previously known Langhian-Serravallian assemblage from the Kutch Basin, western India. Global distribution of the selachian fauna reveals intermittent biogeographic connectivity between the Atlantic, the Mediterranean, and the Indian Oceanic realms during the Aquitanian to Tortonian. However, complete stratigraphic records across various Oceanic realms are warranted to fully understand the diversity change(s) and the direction(s) of migration(s) of the selachian fauna. Considering the data on Modern relatives of the recovered selachian fauna, it is most likely that the recorded selachian fauna dwelled in warm-temperate to tropical shallow coastal waters in the Neritic to Intertidal zones, i.e., coastal-reef to mesopelagic environment(s).

Current developments in Paleogene vertebrate palaeontology in view of India's final drift phase and India-Eurasia docking: an appraisal

Article

Mar 2024

Vivesh V Kapur

This article aims to present an overview of the Paleogene vertebrate faunal records, especially mammals, with an emphasis on research conducted within the last 5 years i.e., from 2019 to 2023. Recent developments highlight the significance of vertebrate fossil records (particularly that of mammals) in our understanding of the origin, evolution, and historical distributions close to the India-Eurasia docking phase. Overall, the fossil records of early Paleogene vertebrates indicate that India was a hotspot for biodiversity for several orders of terrestrial (such perissodactyls, primates, and artiodactyls) and marine mammals (like cetaceans and sirenians). The directions of biotic migrations between India and its surrounding continents between the ~ 66 to ~ 50 Ma time slice remains to be fully understood; however, the sweepstakes and/or island-hopping dispersal mechanisms were the most likely mode of migration for the small-and medium-sized mammals. Additionally, according to a recent study, plausible existence of corridor(s) for secular migration allowed for faunal exchanges between India and Eurasia close to ~ 56-57 Ma, particularly for the large-sized mammals (> 35 kg) (e.g., cambaythere perissodac-tyls). Further, there is now a revived interest in the early Paleogene vertebrate palaeontology of India due to recent data on biostratigraphically constrained Paleocene and early-middle Eocene vertebrate records (such as fish, reptiles, and amphibians along with associated coprolite ichnofossils) from the western and northeastern regions of India.

Response of tropical rainforest to warming during Middle Eocene Climate Optimum (MECO): Evidence from palynological record from the Bartonian deposits of Kutch Basin, Western India

Article

Apr 2025

First record of Chelonian coprolites from the Early-Middle Miocene Kutch Basin, western India, and their palaeodietary and palaeobiological implications

Article

Apr 2024
Geobios

10 (I), 35, 43. Indian Pretertiary vertebrata. Fossil l{eptilia and Batrachia: VII-Plesiosaurlll" from the

S'llrv
Luflia

S'llrv. luflia, 9 (4.), ]54. Notices of New anu oliler Vcrteurala frmn Indian 'l'crliary.and Secondary rocks, Rec. Geol. Surv. India, 10 (I), 35, 43. Indian Pretertiary vertebrata. Fossil l{eptilia and Batrachia: VII-Plesiosaurlll" from the Umia group of Kutch, Pal. Indica;

Ã contribution to the history of I~ocene Siluroid {ishc!!

201-207

India

India, 11, 149-151. ;';. Ã contribution to the history of I~ocene Siluroid {ishc!!, Proc. Zoot:: Soc. London, 201-207. .

AND AND TA~~I;>Ql'ft:~~'~

Jan 1960

O N And Llharoava

AND llHAROAVA, O.N., 1959 1964 1967 , '~.N'; '.i't.,.',W'~~. ".' 1968 ,.AND AND TA~~I;>Ql'ft:~~'~;' 1960

The Cretaceous fauna of southern India: The Brachiopoda

1-202

Echinodermata Ciliopoda
Corals

The Cretaceous fauna of southern India: The Brachiopoda, Ciliopoda, Echinodermata, Corals, etc., Pal. Indica, Ser. 8,4, 1-202.

On a few fossil shark teeth from Miocene beds of Kutch

230-236

Geology
Cheropadi
Western Kutch
India

Geology and stratigraphy of the area between Waghopadar and Cheropadi, Kutch, Western India, Jour. Pal. Soc. India, 2, 136-148. On a few fossil shark teeth from Miocene beds of Kutch, Western India, Proto Nat. Instt. Sci. India, B 25 (5), 230-236.

Lower Tertiary Ostracoda

147-167

Kutch Microfauna

Kutch Microfauna: Lower Tertiary Ostracoda, Proc. Nat. Instt. Sci. India, B 26 (4), 147-167.

132 (I), 1-126. Monophyly or Diphyly ill lite origin of whales, Euolution, 22 (I), 3741. ' Nummulites douuilli, an undescribed species from Cutch with remarks on zollal di:otribution of Indian Nummulites

Deltalheridia

Deltalheridia, a ncw order of mammals, Bull. Amer. lVUS. Nat. HiJt., 132 (I), 1-126. Monophyly or Diphyly ill lite origin of whales, Euolution, 22 (I), 3741. ' Nummulites douuilli, an undescribed species from Cutch with remarks on zollal di:otribution of Indian Nummulites, Rec. Geol. Suru.

Convergenee or parallelism in the evoliltio!l of cetaceans, l

Jan 1965
1461-1465

Convergenee or parallelism in the evoliltio!l of cetaceans, l.Jalaeont. Zhur., 97-106 (In I~ussian), translated in Int. Geol Reu., 7 (1965), 1461-I4G8.

11061 x 0.33 3 b -Left lateral view x 0.33 4 a-Occlusal view of mandible, L.U.V.P. 1113Bx 0.33 4 b-Left lateral view x 0.33 5 a-Occlusal view of mandible, L. U. V.P. 11132 x 0.33 5 b-Left lateral view x 0

L U V P Of Mandible

a-Occlusal view of mandible, L.U.V.P. 11061 x 0.33 3 b -Left lateral view x 0.33 4 a-Occlusal view of mandible, L.U.V.P. 1113Bx 0.33 4 b-Left lateral view x 0.33 5 a-Occlusal view of mandible, L. U. V.P. 11132 x 0.33 5 b-Left lateral view x 0.33 5 c-,-Ventral view x 0.33

Lower Tertiary Vertebrates from western India

Figures

Recommended publications

Lignite deposits of the Kutch Basin, western India: Carbon isotopic and palynological signatures of...

A new species of Protocetus (Cetacea) from the Middle Eocene of Kutch, Western India

A new Species of Protocetus ( Cetacea) from the Middle Eocene of Kutch, Western India

On some Paleogene and Neogene crabs of Kachchh, Western India.

Upper Cretaceous-Early Palaeogene palaeobiogeography of India based on terrestrial vertebrate faunas