,RECOGNIZABLE SPECIES OF TERTIARY PLANTS FROM

\

DAMALGIRI IN THE GARO HILLS, ASSAM

R. N. LAKHANPAL

Birbal Sahni Institute of Palaeobotany, LucknolV

INTRODUCTION

INNovember 1938 the Director, Geologi-cal Survey of India, sent ProfessorB. Sahni a small collection of fossilplants for examination. The collection con-sisting of 34 unregistered specimens, mostlyof leaf impressions, was made by Sir CyrilS. Fox "from Eocene beds at 500 yardssouth of Damalgiri Bungalow (25°32'N.:90°7'E.), Garo Hills". Incidentally, thereare at least three" Damalgiris " in the GaroHills district. The one from which thesefossils were collected lies eleven miles west ofTura, the district headquarter.

In 1945 Professor Sahni gave the collec-tion to Dr. G. S. Puri for investigation, butas the latter had to leave for studies abroadsoon after, the material was passed on tome. A short note about these fossils waspublished by me (1947) in Palaeobotany inIndia, VI.

Most of the impressions are too poorlypreserved to be identified specifically. How-ever, a few species can be recognized in thecollection and are described in this paper.The number of species is too small to betreated collectively as a flora and to drawconclusions about its palaeoecology andpalaeogeography. However, in the discus-sion of each species the habit and distribu-tion of its modern equivalent are given tosuggest the living conditions of the fossil.

MATERIAL AND METHOD

Dr. R. C. Misra of the Geology Department,University of Lucknow, has very kindlyexamined the rock specimens for me andreports as follows:

" In the hand specimen the rock is greyishwhite in colour, with small patches stainedyellow with secondary iron oxide. It ismoderately compact and foliated. Originally,most likely, a shale, now it is metamorphosedto a phyllite.

" Under the microscope a transverse sec-tion shows the foliated nature of the rock.

Much of the material is clayey and does notreact with polarized light. Very fine elon-gated laths of Sericite are, however, clearlyobserved. In a section parallel to thefoliation planes organic material of richbrown colour is prominent.

" The rock has suffered a moderate degreeof metamorphism."

The fossils are preserved as leaf impres-sions. The nature of preservation is ratherpoor and there is no cuticle on the surface ofthe leaves.

The identifications of the fossils are basedon their resemblances with leaves of themodern species and were carried out at theherbaria of the Indian Botanic Garden,Sibpur, and the Forest Research Institute,Dehra Dun.

SYSTEMATIC DESCRIPTION ANDDISCUSSION OF EACH SPECIES

All the species are angiospermic, belongingto the subclass Dicotyledoneae. They arearranged here according to the scheme fol-lowed by Willis ( 1951, pp. i-xlix) based onEngler's system of classification.

ORDER-URTICALES

FAMILy-ULMACEAE

Trema garoensis sp. novoPlate 1, Figs. 1, 2

Description - Leaves simple, ovate-oblong;average length 8·5 cm.; average width 4·2 cm.;apex? acute; base inequilateral, triplinerv-ed; petiole 1-2·5 cm. long, slightly curved,grooved; midrib slender, slightly curved;secondaries 9 pairs, lowest about 4·5 cm.long, coming off at the base at an angle of45°, angle of divergence slightly more in thesecond pair but gradually decreasing towardsthe apex to about 30°, individual veinscurving up along the margin to join withthe next upper ones in loops; tertiaries veryslender, running curved across the secondariesin broad cross-ties, some bifurcating and

27

28

,THE PALAEOBOTANIST

each limb joining the next secondary; marginthinly serrate; texture rather membranous.

Discussion - The shape and venation ofthe leaf with serrate margin and inequilateralbase strongly suggest affinities with Trema.The triplinerved base, very small serrations,grooved petiole and the texture and size ofthe leaves are mostly like those of the livingT. orientalis Bl. There is also considerableresemblance with T. amb01:nensis Bl., but inthe latter the leaves are larger.

It has not been possible to find any des-cribed fossil leaf resembling T. garoensis.

Trema is a tropical or subtropical genus.T. orientalis is a small, fast-growing ever-green tree. It has a remarkable capacity forspringing up on forest clearings, landslipsand wherever the ground is exposed. Insuch places it is generally the first tree toappear coming up gregariously. It is dis-tributed at the foot of the Nepal and SikkimHimalayas, in Bengal and Bihar, extendingsouthwards to Travancore. In Burma it isreported from Rangoon and Myitkyina dis-tricts. It is common in Ceylon and MalayanIslands. T. amboinensis occurs in Sikkim,Assam, Chittagong, Burma, Singapore and theAndaman Islands. It has the same habit asT. orientalis of which it is considered as alarger form.

Collection - Syntypes, Nos. GH 16 andGH 18.

ORDER-RAN ALES

FAMILy-LAURACEAE

Neolitsea sahnii sp. novoPlate 1, Figs. 3, 4

Description - Leaves simple, elliptical;length 9-10·5 cm.; width 3·2-3·8 cm., widestnear the middle; apex not preserved; basebroadly obtuse, almost rounded, slightlyinequilateral, triplinerved; petiole about2·7 cm. long, slightly curved, fairly stout;midrib prominent, straight; a pair of lateralprimaries coming off at base, each about6·5 cm. long, slightly curved near the baseand then running straight towards the mar-gin at an angle of about 20°; secondariesstart arising from the midrib at about one-third of its length from the base, approxi-mately 10 pairs, angle of divergence about60° decreasing towards the apex, each secon-dary curving up near the margin, runningconvex to it and then joining with the nextupper secondary in a loop; tertiaries very

slender, running as cross-ties between thesecondaries, enclosing rectangular areas, somearising directly from the midrib to join withthe lateral veins; margin entire; texturesmooth.

Discussion - The form and texture of theleaves are typically lauraceous. The tripli-nerved character in Lauraceae is met within N eolitsea and Cinnamomum. In Cinna-momum, however, the lateral primaries arevery prominent and extend almost to theapex. N eolitsea zeylanica Merr. presents avery close resemblance with our specimens inhaving the same shape and size, the tripli-nerved base, the lateral primaries running toabout two-thirds the length of the leaf, andthe tertiaries running parallel across thesecondaries forming rectangular meshes. Inthe allied species N. joliosa Gamble theleaves are longer and the base sub-tripli-nerved, i.e. the lateral primaries comingoff a little above the base. This tendencyof sub-triplinerved base is also found in someleaves of N. zeylanica, but far less in fre-quency than in N. joliosa and in N. sahniithis seems to be altogether absent.

There are two fossil species of N eolitseaknown to the author, N. Gardneri Bandulska,from the Upper Eocene of Bournemouth,England (BANDuLsKA, 1926), and N. lataMacGinitie from the Middle Eocene of Cen-tral Sierra Nevada, U.S.A. (MACGINITIE,1941). None of these is very closely com-parable with N. sahnii. N. lata is bigger insize, narrower at the base, has smaller numberof secondaries, and its lowest pair of second-aries does not come off from the base as inN. sahnii. In shape N. Gardneri seems morelike the present fossil, but the figure of N.Gardneri is not clear enough to make out thedetails of its structure for compartson. Itlooks slightly longer and narrower and hasfewer secondaries. The structure of itscuticle is given, but it cannot be com-pared as the cuticle of N. sahnii is notpreserved.

The living N eolitsea is a small to middle-sized evergreen tree. It is distributed throughthe Philippine Islands, Java, Indo-MalayanArchipelago to Burma and India. N. zey-lanica occurs in Mergui and Tavoy districtsof Burma; Assam in eastern India; Nilgiris,Madras and Coimbatore in southern India;and, as the name implies, in Ceylon.

I have named this species after my teacher,the late Professor B. Salmi, F.R.S., as atoken of respect and affection.

LAKHANP'AL - RECOGNIZABLE SPECIES OF TERTIARY PLANTS

•29

Collection - Holotype, No. GH 2; para-type, No. GH 11.

ORDER - MALVALES

FAMILY- TILIACEAE

Grewia foxii sp. novoPlate 2, Fig. 5

Description - Leaf simple, ovate, asym-metrical, divided into two unequal halves;length about 5·7 cm.; width 4·3 cm.; apexbroadly acute; base broad, inequilateral,triplinerved; petiole about 1·25 cm. long,slender, twisted; midrib broad at the base(2 mm.) narrowing sharply towards the apex,slender, curved; secondaries 8 pairs, lowestarising from the base, running to abouttwo-thirds the length of the leaf; angle ofdivergence about 60° in the basal veins,gradually decreasing towards the apex, lessin the narrower half of the leaf than in thebroader; secondaries arcuate, curving up-wards near the margin and then runningalong it towards the apex; tertiaries fairlyprominent, thin, joining the secondaries inparallel cross-ties; the second pair of second-aries from the base short, forking into two,each merging into the tertiaries about halfway between the midrib and the margin;near the base tertiaries also coming off onboth sides of the midrib; tertiaries fromthe basal"pair oneconClaries on~the outer sideclearly visible going right up to the margin;margin mostly serrate, a little below theapex the serrations becoming blunt and atplaces rounded; texture thinly chartaceQus.

Discussion - The size, the division of thelamina into two unequal halves, the inequi-lateral base, the mixed serrate-crenate marginand the twisted petiole of the fossil presenta very close resemblance with the leaves ofthe living Grewia tiliifolia Vahl. The tex-ture and the shape of the lamina are verycharacteristic. The venation also is verysimilar, but in G. tiliifolia the basal second-aries are generally not much curved althoughin the variety G. tiliifolia vaL argentiaBurret they are almost as arched as in thefossil. No fossil record of a Grewia re-sembling the present specimen is known.

Saporta used the name Grewiopsis ferfossil leaves resembling those of Grewia.Plant remains belonging either to Grewia orto Grewiopsis have been described fromUpper Cretaceous to Upper Tertiary beds ofboth the Old as well as the New World.

The living Grewia is confined to the tropicaland subtropical regions of the Old World, i.e.Africa, Madagascar, Arabia, India, Burma,Ceylon, Andaman-Nicobar, Malay Peninsula,East Indies, Indo-China, extending to NorthAustralia. The genus is fairly representedin India. G. tiliifolia is distributed in eastTropical Africa and India. In India it occursin arid regions of Bengal, Bihar, central andsouthern India. G. tiliifolia vaL argentia isfound in Assam, Upper Gangetic Plain, cen-tral and southern India. Both G. tiliifoliaand its variety argentia are moderate-sizedtropical trees occurring in arid regions.

It is generally believed by the plant taxo-nomists that most probably the original homeof Grewia was Equatorial Africa from whereits line of distribution extended eastwardsto India, Burma, Siam and Indo-China(NARAYANASWAMI& RAO, 1950). If thisstatement is correct, the eastward march ofGrewia must have begun quite early becauseon the evidence of G. foxii it was present inthe Assam region in Eocene times. More-over, the fossil record shows that Grewia hada very wide distribution in the early Tertiary.In fact, it seems to have had a wider dis-tribution in the past than at present. Thepresent abundance of the genus in Africamay be due to the suitable environment whichthe plant got in that continent and may notbe an indication of its origin there. G.tiliifolia vaL argentia seems to be Indian inorigin as it is not distributed in Africa andits fossil equivalent G. foxii has been found inthe early Tertiary of Assam.

The species is named after the late SirCyril S. Fox who collected this material fromAssam.

Collection - Holotype, No. GH 17.

FAMILY- BOMBACACEAE

Bombacites orientalis sp. novoPI. 2, Figs. 6, 7

Description - Leaves large, digitately com-pound; each with five leaflets radiating fromthe tip of a long slightly curved petiole;preserved length of the petiole 12 cm., width3·5 mm., slightly swollen at its junction withthe leaflets, faint longitudinal grooves andridges visible on the surface of the petiole.Leaflets five, oblanceolate; length varyingfrom 10 to 14 em., width about 3· 2 cm.; apexnot preserved; base acute and decurrent withthe midrib to the point of junction with the

30 THE PALAEOBOTANIST

petiole; midrib stout, a slight longitudinalgroove visible on one side ( ? upper); second-aries about a dozen, very narrow, sub-oppo-site in the apical half, alternate below,distance between two adjacent veins about1 cm., decreasing towards the apex, angle ofdivergence about 55° in the secondaries onthe outer side and about 70° in those towardsthe middle of the leaf; secondaries curvingup near the margin but the formation ofloops not clear; smaller veinlets formingnetwork of irregular polygonal meshes; mar-gin entire; texture fair!y thick.

Discussion - In general form of the leavesthis species resembles Eriodendron anfruc-tuosum D.C. and Bombax spp. However, inone point the resemblance is more with E.anfructuosum. While in Bombax the leafletsare stalked, in Eriodendron they are sessileas in the fossil. This condition is almostapproached in B. insigne where the stalks ofthe leaflets are very small, but in this plantthe number of leaflets is generally 7-9 andtheir size is much bigger.

For digitately compound fossil leaves ofthe type of Bombax with large leaflets havingentire or toothed margins Berry ( 1916) gavethe name Bombacites. All the fossil recordsof Bombax and Bombacites are based on singleleaflets, while the present fossil shows leafletsjoined together on a petiole. From theavailable figures Bombax Sturtii Etting. des-cribed by Chapman and Crespin ( 1933-34)from Australia is the only species whichlooks somewhat like Bombacites orientalisalthough this comparison is not very reliable,as the single figured leaflet of B. Sturtii isincomplete. Most of the other fossil specieshave toothed margins.

There is no available record of fossil Erio-dendron. Some doubtful forms referred toBombax have been described from the Creta-ceous of America and Europe ( BERRY,1916 ).In the Tertiaries, Bombacites has been des-cribed from the Lower Eocene of Americaby Berry (op. cit.). In Europe, foliage ofthe type of Bombacites has been describedmainly from the Upper Eocene, but a fewspecies have also been reported from theother epochs of the Tertiary period. Bom-bax jiammeum has been recorded byMenzel (1920) from the Tertiary of Ca-meroons, Africa. Two species of Bombax,B. Sturtii Etting. and B. Mitchelli Etting.have been described from the Eocene ofwestern Australia by Chapman and Crespin( 1933-34 ).

Eriodendron occurs mostly in America.E. anfructuosum, a moderate-sized, soft-wood-ed, deciduous tree is indigenous to Burma,Andamans, Malay Peninsula and Archipelago,western parts of Indian Peninsula and inTropical America. Perhaps it also occurs inTropical Africa.

Bombax malabaricum is a large deciduoustree most commonly found on flat alluvialsoil near river-banks where it is often grega-rious. It is often found scattered in mixed,usually deciduous, forests and even in Sal(Shorea robusta Gaertn.) forests. It occursin regions showing a wide range of tem-perature and rainfall, but thrives best in acomparatively moist tropical climate. Itoccurs in tropical Eastern Himalayas andthroughout the hotter forests of India toBurma and Ceylon. It is also distributed inJava and Sumatra.

B. insigne is common in upper mixeddeciduous forests on well-drained often hillyground. It also extends into evergreen orsemi-evergreen forests on hill-slopes. It isessentially a tropical tree thriving best intemperatures with a maximum of 96°_110°F. and a minimum of 42°-60°F. andrainfall of about 40-150 in. Its natural homeis Burma, Andamans, Chittagong, MalayPeninsula and the Western Ghats of Indiafrom North Kanara southwards.

Collection - Syntypes, Nos. GH 3 andGH 5.

AGE OF THE FOSSILS

As already mentioned, the present fossilswere collected at 500 yards south of Damal-giri Inspection Bungalow in the Garo Hillsdistrict of Assam. According to furtherdetails supplied at my request by the Geologi-cal Survey of India, the collection comesfrom the Kopili stage. Kopili is the upperof the two stages into which J aintia seriesof Assam is divided.

About the age of the Kopili stage Evanswrites (1932, p. 176):

" The Kopili stage includes a number offossil beds. From the Burmah Oil Company'scollections, Mr. Eames has identified a num-ber of forms suggestive of Middle Khirtharbeds ... it seems safe to assume that theKopili stage, as a whole, is approximatelyequivalent to the Middle Khirthar, i.e. UpperLutetian. "

According to this statement the stage isMiddle Eocene in age. The Geological Sur-

LAKHANPAL - RECOGNIZABLE SPECIES OF TERTIARY PLANTS 31

vey has, however, suggested that the Kopilibeds may be regarded as uppermost Eoceneto Oligocene (HERON, 1939).

Un ortunately, there has not been enoughprevious work on the Tertiary floras of this

part of the world with which comparisonscould be made for the correlation of strata.For the present there is no evidence againstthe early Tertiary age of the beds in whichthese fossils were collected.

REFERENCES

BANDULSKA,H. (1926). On the Cuticles of someFossil and Recent Lauraceae. Jour. Linn. Soc.( Botany). 47: 383-425.

BERRY, E. W. (1916). The Lower Eocene Florasof Southeastern North America. U.S. Geol. Surv.Prof. Paper. 91.

CHAPMAN.F. & CRESPIN, I. ( 1933-34). The Palae-ontology of the Plantagenet Beds of WesternAustralia. Jour. Roy. Soc. West. Australia. 22:103-136.

EVANS, P. (1932). Explanatory Notes to accom-pany a Table showing the Tertiary Succession inAssam. Trans. Min. Geol. Inst. I11dia. 27:155-260.

HERON, A. M. (1939). General Report of the Geo-logical Survey of India for the year 1938. Rec.Geol. Surv. India. 74(1): 1-169.

LAKHANPAL, R. N. (1947). Tertiary Leavesfrom the Garo Hills, Assam. Palaeobotanyin India - VI. Jour. Ind. Bot. Soc. 26(4): 261-262.

MACGINITIE, H. D. (1941). A l\Iiddle EoceneFlora from the Central Sierra Xe\'ada. CarnegieInst. Wash. Pub/. 534: 1-178.

MENZEL, P. (1920 ). Dber Pftanzenreste ausBasalttuffen des Kamerungebietes. Beil.Geol. Erforsch. Deulsch Sc/mtzgebiele. 18:17-32.

NARAYANASWAMI,V. & RAO, R. S. (1950). Apreliminary Note on the Indo-Burmese Speciesof Grewia Linn. Jour. Ind. Bal. Soc. 29(4):177-190.

WILLIS, J. C. (1951). A Dictionary of the Flower-ing Plants and Ferns. Cambl'idge.

PLATE 1

EXPLANATION OF PLATES

PLATE 2

1, 2. Trema garoensis sp. novo Leaves. Nos.GH 16 and GH 18 respectively. x 1.

3. Neolitsea sahnii sp. novo Leaf. No. GH 2. X 1.4. N. sahnii sp. novo Incomplete counterpart of

the same showing venation more clearly. No.GH 11. X 1.

5. Grewia foxii sp. novo Leaf. No. GH 17.xl.

6. Bcmtaciles orienlalis sp. novo Leaf. No. GH 3.xl.

7. B. orinlla/is sp. novo Leaf. No. GH 5.xl..

THE PALAEO BOTANIST, VOL. 3LAKHANPAL- PLATE 1

LAKHANPAL - PLATE 2 THE PALAEOBOTAN[ST, VOL. 3