SOME REMAINS OF THE FROM THE KAREWA DEPOSITS OF THE KASHMIR VALLEY, !

G. S. PURl Forest Research Institute, Dehra Dun

ABSTRACT Key to the genera: Venation pinnate ...... The paper figures and describes plant remains belonging to Comus macrophy/la, C. capilala(?) and Venation pseudo-palmati-pin- Nlarlea bego1!iaefolia from the Karewa deposits of nate ... .., '" M arlea Kashmir. The present distribution of the fossil species is given in detail. A comparison of the past and present distribution of the fossil species on the GENUS Comus Linn. whole confirms earlier conclusions of the author regarding changes in the forest, climate and altitude In addition to leaf fragments one specimen in the Kashmir Valley since the Pleistocene. of fruit is also identified as belonging to Cornus capitata Wall.(?). The leaf impres­ INTRODUCTION sions seem to belong to at least two distinct species. HE specimens referred to the Coma­ Key to the species: ceae were collected from Ningal Nullah Venation pinnate-arcuate T and Liddarmarg on the Pir Panjal (a) leaf large, ovate-oblong, 6 to 7 pairs Range, in Kashmir. A detailed description of laterals arising at acute of the localities has already been given angles... • ... C. macrophylla ( PURl, 1948) and a brief description of the (b) leaf smal1, oblong, 4 to 5 pairs of fossils will now be taken up. laterals arising at somewhat obtuse angles ... ." ... Cornus sp. DESCRIPTION Venation strict-pinnate, leaf smal1, el1ip­ Except for one impression of a fruit, pro­ tic, 5 to 6 pairs of laterals arising at visionally identified with Cornus sp., all the acute angles ... C. capitata(?) specimens are leaf impressions, which are somewhat poorly preserved and fragmentary. Comus macrophyUa Wall. They are, in addition, poorly represented in the fossil beds, numerically. This may be Fig. 1 is a natural size photograph of two due to the thin texture of living leaves, which leaf fragments, one of which is lying on the feature helps in their quick decomposition apical part of the other. The leaf-lamina, and breaking up easily into fragments while which apparently has an ovate-Oblong shape, on their way to the site of decomposition by measures 2·2 in. long by 1·4 in. in the broad­ water or wind. est part in the larger leaf. The apex seems Living leaves of these fossil species were to have been probably acute. The margins, experimentally rotted in laboratory along though not well defined, seem to be entire. with leaves of other fossil species. The The base is rounded and bears 0·25 in. long leaves of Comus macrophylla and Marlea petiole in both leaves. begoniaefolia were completely decomposed The venation is pinnate-arcuate and reti­ in a two-week period, leaving a skeletal culate. A strong midrib runs straight in the network of larger veins. At this stage these lamina, slightly thinning out towards the api­ compared favourably with fossil impressions cal part. Six to seven pairs of secondaries, shown in phot?graphs. which are about half as thick as the midrib, arise from it at acute angles, in an opposite, or, in some cases, in an alternate manner. FAMILYCORNACEAE The lower pairs of secondaries arise closer The fossif fragments are referred to two than the upper pairs. These curve gradual­ different genera - Cornus and M arlea. ly and end in the margins without making

1. This work was done at Lucknow during 1940-42 under the guidance of the late Professor B. Sahni, F.R.S., to whom the author wishes to record his indebtedness. 107 108 THE PALAEOBOTANIST any loops. The upper pairs of laterals tend the leaves and the fruit were found in the to converge near the apical end. Tertiary same block of rock, they may belong to this veins, though not well preserved in the species. It is, however, difficult to identify leaves, can be seen at some places forming such fragmentary structures with certainty cross-ties. The finer reticulation is not pre­ with anyone species. served. Occurrence - Ningal Nullah at 9,500 ft. M arlea begoniaefolia Roxb. Number of Specimens - One only This species is based on two leaf fragments Type Specimen - Loc. 1 N. 490 one of which is more or less complete. The Collections - G. S. Puri, 1940 leaf-lamina is ovate or ovate-oblong with a The specimen is preserved in the Botany broad and cordate base. The margins are Museum, Khalsa College, Amritsar. wavy with broad teeth. The apex is acute. The fossil leaves resemble living leaves of The venation is pseudo-palmate and reti­ C01nus macrophylla Wall. with which they culate, consisting of one main rib and two are identified. basal laterals, which may give the venation a pseudo-palmati-pinnate appearance. The Cornus capitataU) Wall. primary rib gives off a number of secondary Fig. 2 is a natural size photograph of a veins and from some of them, especially the leaf which has been heavily coated with two at the base, a few smaller branches rubber solution in field. This has obscured arIse. The secondaries and the primaries the venation which has not been satisfac­ are almost of the same calibre. Tertiary torily brought out in the photograph. The ribs are well preserved and form rectangular leaf-lamina, which is elliptic in outline, is meshes. The finer reticulation is seen at slightly torn off on right margin near the some places in the leaves. basal region. It mea~res1·6 in. long by Occurrence - Ningal Nullah at 9,000 ft., O'73 in. in the broadest part. It gets narrow­ and 9,800 ft. ed towards the upper side into an acute apex Number of Specimens - Two only and into a rounded wedge-shaped base towards Type Specimens - N. 120 (FIG. 5) and the lower end. The margins are entire. N. 201 (FIG. 6) The venation is strict-pinnate and reti­ Collections - The specimens were col­ culate, consisting of a thin midrib and 5 to lected bv the author in 1940 and are 6 pairs of laterals, which are almost of the preserved in the Botany Museum, Khalsa same calibre as the midrib and arise from College, Amritsar it at acute angles. Second lateral from the The fossils resemble living leaves of base, on the right side in the photograph, M arlea begoniaefolia Roxb. in shape, mar­ gives off two thin branches near the margin. gins and all details of venation, and are Tertiary veins are not well preserved. The identified with this species. finer reticulations also are not preserved. Another leaf fragment shown in Figs. 7 Occurrence - Liddarmarg at 10,600 ft. and 8 was found bv R. R. Stewart from Number of Specimens - One only Ningal Nullah near one of the localities of Type Specimen - Loc. 3 L. 90 the author. This specimen shows some Collections - H. de Terra, 1932 resemblance to M arlea, but in the absence The specimen is preserved in the Botany of definite clear features it is not considered Museum, University of Lucknow. safe to identify it with this species. The fossil resembles living leaves of Cornus capitata Wall. in most of the features MODERN DISTRIBUTION OF THE FOSSIL that are visible, but on account of its rather SPECIES inconspicuous venation it is identified pro­ Cornus macrophylla Wall. grows in the visionally with this species. Himalayas between the Indus and Another leaf fragment shown in Fig. 3 at the altitudes of 4,000-8,00D ft. in conifer seems to show some resemblance to Cornus forests. It occurs on finely decomposed, capitata, and in the absence of any better rich and moist soils, along streams, or in comparison this is provisionally identified depressions, along with ] uglans regia, Aes­ as Cornus sp. culus indica, cherry, poplar, etc. Some­ The fruit specimen, shown in the natural times it is distributed in spruce and silver size photograph 4, seems to match favourably fir forests in gullies along with maples, with fruits of Cornus capitata, and as mostly Acer caesium. PURl - PLANT REMAINS OF CORNACEAE FROM THE KAREWA DEPOSITS 109

It occurs in the Samana Range, North­ recorded from Sikkim at 9,000 ft. It grows West Frontier; Malkandi, Kagan Valley, in Kagan, Hazara, between 3,800 and 8,000 Hazara, at 6,000 ft.; Bagnotar, Hazara; and ft.; and is recorded between Ramsu and in the Murree Hills commonly at 6,000­ Banihal, Chenab Valley, Jammu. In the 7,000 ft. In Kashmir, it occurs in the Tawi Kashmir Himalayas it grows in Chineni, Valley, Jammu, at 6,000 ft. in 'ban' oak Keran, Kishtwar, Marwa Dachhan, Muzaf­ ( QuerC"Usincana) forests; and in the forests farabad, Ramban and Udhampur. It is of Keran, Kishtwar, Marwa Dacchan, Ram­ present in the Kashmir Valley proper. ban and Udhampur. It is present today in Eastwards, it occurs in the Beas Valley, the Kashmir Valley near the fossiliferous Kulu, at 4,000 ft.; Tons Valley, Jaunsar, at localities in gullies and depressions. East­ 3,000 ft.; Bashahr, at 7,000 ft.; the Glen, wards, it occurs in the Parbatti Valley, Kulu, Simla, at 6,000 ft. In the Mussoorie Hills and is recorded from Salori to Bhandal, it grows at 5,000 ft. and extends further Chamba, at 5,000-7,000 ft.; below Sungri, eastwards to Garhwal; Malia Nagpur, Naini Bashahr, at 8,300 ft.; Teklech, Bashahr, at Tal, at 5,500 ft.; and East Almora, Kumaon 4,500 ft.; Mushobra, Simla; and the Glen, division. Simla, at 6,000 ft. It occurs at Landour, This species grows in moist shady ravines Mussoorie, at 7,000 ft. commonly. Further and in dense shady forests. The fossil eastwards, it grows at Range, Garhwal, leaves of this species have been found from at 6,000 ft.; Tala Kaliphat, Garhwal, at Ningal Nullah, the place from where a rich 4,500-8,000 ft.; and in Doti district, . mesophytic flora consisting of an abundance Cornus cap£tata Wall. grows in the Hima­ of willows and poplars, which now grow on layas, between the Beas and Bhutan, at river banks or moist ravines, is discovered. altitudes of 3,500-8,000 ft. This species, This species seems to fit in well with the unlike the first, occurs mostly in forests of ecological picture given by the Ningal Nullah Quercus l:ncana and is found even in Chir flora. Ordinarily parts of trees occurring pine community. It affects outer sunny in moist ravines should be better fitted to slopes and occurs on hard dry soils. reach the site of deposition, if they are It does not grow anywhere in Hazara or sufficiently resistant to decay. But the Murree Hills, but is recorded from Muzaffara­ leaves of this species have been found to be bad in Kashmir. It is absent from the unsuitable on account of their thin texture Kashmir Valley proper. Eastwards, it occurs and easy decay in water. the leaves com­ in the Parbatti Valley, Kulu, at 6,000 ft. pletely rotted after two weeks in the labora­ It grows at Chor-Pubaha Road, Simla; and tory experiments conducted by the author in the Bashahr State, at 6,000 ft. In the during the months of November-December Mussoorie Hills, this species is found com­ 1941 at Lucknow. monly in the ban oak forests on the outer slopes between 3,500 and 7,000 ft. above CONCLUSIONS sea-level. It extends to Garhwal, at 4,000­ 6,000 ft.; Gangolihat, Almora, in Kumaon; A comparison of the past and present West Almora, Kumaon; Deota, Tehri-Garh­ distribution of the fossil species seems to wal, at 8,000 ft.; and in Phallaloong Forest, confirm the author's earlier conclusions Darjeeling, at 8,000 ft. (see PURl, 1945-1951a) regarding the condi­ M arlea begom:aefolia Roxb. grows in the tions of forest, climate and altitude during outer Himalayas from the Indus to Bhutan, the Pleistocene in the Kashmir Himalayas. ascending to the altitude of 6,000 ft., but is These conclusions may not be repeated here.

REFERENCES PURl, G. S. (1945). Some fossil leaves of the forests of the Kashmir valley during the Pleisto­ Salicaceae from Ningal Nullah and Laredura, cene. Ibid. 22: 232-256. Pir Panjal, with a note on the significance of Idem (1947). Fossil and the Himalayan temperate species in the Pleistocene flora of uplift. Jour. Ind. Bot. Soc. 25: 167-184. Kashmir. Proc. Ind. Acad. Sci. 22 : 87-112. Idem (1948). The flora of the Karewa series of Idem ( 1945a). The genus Quercus in the Karewa Kashmir and its phyto-geographical affinities, • deposits of Kashmir, with remarks on the oak with chapters on the methods used in identifica- 110 THE PALAEOBOTANIST

tion. Ind. For. 74 ( 3, 4: 5, 6): 105-122; 152-164; cene of India. Quart. Jour. Geol. Min. &- Met. 210-225; 240-244. Soc. Ind. 23: 25-41. Idem ( 1949). The historical factor and its appli­ Idem (1951a). Fossil fruits of Trapa and re­ cation to forest ecology. J our. Ind. Bot. Soc. mains of other fresh-water plants from the 27: 63-68. Pleistocene of Kashmir. Jour. Ind. Bot. Soc. Idem (1951). The family Rosaceae in the Pleisto- 30:113-121.

EXPLANATION OF PLATE 1

1. Leaf of Comus macrophyUa. X 1. 6. Leaf of Marlea begoniaefolia. x 1. 2. Leaf of Cornt~Scapitata(?). x 1. 7. Marlea, leaf fragment. x I. 3. Leaf of Cornus sp. xI. 4. Cornus sp. A fruit. xl. 8. M arlea, part of leaf fragment in Fig. 7. 5. Leaf of Marlea begoniaefolia. x 1. enlarged. x ca. 5. -

...., :e t'l 'i:l >... > t'1 0 til ~ 0... > Z If> ,-l < ...0

IV