ON SOME GINKGOALEAN LEAF IMPRESSIONS FROM THE RAJMAHAL HILLS

K. R. MEHTA & J. D. SUD Banaras Hindu University

INTRODUCTION two may perhaps be species of Gittkgoites. They are the first reports from the Rajmahal HE record of fossil Ginkgoalean leaves series of the Upper Gondwanas of India.l from India is comparatively meagre. The exact localities from which the specimens T The earliest reports were made by were collected are noted in the text. Feistmantel who described two species of Ginkgoites-G.lobata Feist. (1877a, 1877b, DESCRIPTION PL. I; FIG. 1) from Jabalpur; andG. crassipes Feist. (1879, p. 31, PLS. XV, XVI, FIG. 12) Specimen 1 (FIG. 1, locality Maharajpur)­ from the Madras coast. The genera Phoeni­ From an exposure nearly a quarter of a mile copsis and Czekanowskia are also mentioned to the south-west of the railway station near 2 (FEIST., 1877 , p. 19). Feistmantel ( 1881, the base of the second hillock. p. 121, PL. 46A ) also described Rhipidopsis The fossil is preserved as a leaf impression densinervis from the Damuda beds of Goda­ of a dark yellowish-brown colour on a soft vari district, although Sitholey (1943, p. shale. Lamina fan-shaped, deeply lobed 188) suspects that it may be a species of and attached to a petiole. The maximum Psygmophyllum, since the specimen differs length of the middle lobes is indeterminable from ' typical Rhipidopsis in the absence of due to the absence of the upper portion of small basal segments '. After a gap of over the impression. Overall breadth of lamina a quarter of a century Seward ( 1907) found 5·5 cm.; petiole 1·3 cm. long and 1 mm. Psygmophyllum H ollandi in Sir Henry Hay­ broad; lateral segments 2·8 and 3 cm. long. den's collection of Permo-Carboniferous Lamina divided into 6 primary segments, plants from Kashmir, although two years middle incision almost down to the petiole, earlier Seward and Woodward ( 1905, PL. 9, other incisions reach more than two-thirds FIG. 3 ) had described as Psygmophyllum sp. the depth of the lamina. Primary seg­ an incomplete leaf from a collection by ments in turn divided once, their incisions Noetling from the Vihi Valley in Kashmir. reaching almost to the middle. The entire Seward (1912, p. 7) also founded another lamina may thus have divided into a dozen species Psygmophyllum H aydeni from a linear segments with an average width of material collected by C. S. Middlemiss from 2·5 mm. per lobe. Venation faintly clear Golabgarh Pass in the Pir Panj al Range; only in the third primary segment from the and Seward and Sahni (1920, p. 37) referred right. There appear to be 5 veins which, in some detached leaf impressions of doubtful the ultimate dichotomy, will, therefore, num­ identity to the genus Phoenicopsis. After ber two or three per segment. another gap of over two decades Sitholey Due to unsatisfactory preservation all ( 1943) referred to Psygmophyllum H aydeni attempts to study the cuticular structure another specimen collected by Wadia from have failed. Dandlutar in Pir Panjal, and simultaneously The leaf is organically attached to an axis­ Ganju (1943) added yet another species like impression whose structure, however, Psygmophyllum Sahnii from Sirin spur, Vihi cannot be made out. A small swelling is district, Kashmir. associated with the base of the leaf which The present collection describes three leaf impressions from the Rajmahal series in 1. While this communication was practically Bihar. These specimens were collected by the ready for the press, the authors' attention was authors in 1949 and reported to the annual drawn to a note by Sah ( 1952) on the occurrence meeting of the Palaeobotanical Society at of Ginkgoales in the Rajmahal series of Bihar. The specimens reported by Sah are all from one of the Lucknow in 1950. One of these probably two localities from where the authors have mane represen ts the genus Baiera and the other their coHections. 51 52 THE PALAEOBOTANIST may have been the short shoot on which it B. Venation very dense, 20 veins or more was actually borne. per cm. in the apical part of the foliar Specimen 2 (FIG. 2, locality Sakrigali­ leaves ... Ginkgodium. ghat)-Whitish impression on a soft greyish II. Foliar leaves divided into primary shale. Lamina deeply incised into 4 linear lobes very deeply (up toi or more). segments. Probably only a part of the leaf. These primary lobes may be arranged in two In two of the segments thin streaks, six in groups and in turn are deeply divided at each, resembling veins, are faintly visible. least once, but mostly repeatedly. All the They seem to run parallel, but their bifurca­ lobes narrowly linear or almost so, including tion cannot be made out. The largest segment not more than 2 to 4 parallel veins in any is nearly 22 mm. long and 4 mm. broad. part ... Baiera. Specimen 3 (FIG. 3, locality Sakrigal:­ The remaining genera are so markedly ghat) - Three main segments, each dividing differen t from these four as well as from the into two(?) attached to a stout stalk. specimens described above that they need This specimen, too, is probably incomplete. not be considered here. It is indifferently preserved. The lobes are Specimen 1, according to this classification, deeply incised, their apices are either trun­ may perhaps belong to the genus Baiera for cate or got broken during manipulation. the following reasons: The primary incisions The deepest incision reaching almost the are very deep; the secondary incisions are apex of the stalk seems to suggest that also deep reaching almost the middle of the blade may have been equally divided primary segments; the ultimate lobes are into two halves as in the case of the first narrow and linear; and the veins not more specimen. Venation not clear. than 4 in each segment. The central secondary segments at the The generic and specific determination of point of bifurcation are 2 mm. and 2·5 mm. Ginkgoalean leaf impressions is not an easy broad. task. This is particularly the case with the genera Ginkgoites and Baiera which seem to IDENTIFICATION merge into one another at several points. While dealing with the subject Seward (1919, In his exhaustive treatise on the fossil p. 16) remarked that" as with the definition Ginkgoales Florin (1936, pp. 44-45) has of species within the genus Ginkgoites so also given a comprehensive key dealing with the the adoption of Ginkgoites or Baiera is to a classification of the genera known at the large extent, the result of individual pre­ present time. The relative part of the key ference and merely expresses an attempt is reproduced below: to classify in an arbitrary fashion the nume­ 1. Foliar leaves either undivided or divided rous types of leaves that in themselves afford more or less deeply into 2 to 8 primary lobes no sure guide as to precise affinity". which are tongue-shaped and wedge-shaped Another eminent worker, T. M. Harris towards the base. These are frequently in ( 1935, pp. 47-50 ), discussing the differences two groups, one on the right and the other between these two largest genera of the on the left. Normally they are further group has referred to the following schemes: deeply divided only once in the middle or 1. Leaf divided into several segments ... still deeper into secondary lobes which Baiera. include in the middle zone of the leaf 4 to Leaf merely indented ... Ginkgoites. 6 or more subparallel veins. II. Leaf with distinct petiole, outline of A. Venation, distant to moderately close, lamina as a whole being semi­ showing less than 20 veins per em. in the circular ... Ginkgoites. apical part. Leaf without a petiole ... Baiera. (a) Foliar leaves: in epidermal structure or Harris agreeing with Krausel (1917) other anatomical characters of systematic supports the second scheme, while Lundq­ importance closely resembling those of vist ( 1918 ) supporting the first one observes Ginkgo bUoba L. (recent) ... Ginkgo. that the question of the presence or absence (b) Foliar leaves: epidermal structure or of a petiole is " purely subjective". other anatomically importan t characters Reviewing the whole problem Seward either unknown or strongly different in one ( 1919, p. 34) has observed: " Leaves assign­ or more aspects from those of Ginkgo iiloba ed to Baiera vary within wide limits as ( recent) ... Gin/cgoites. regards the size, the number of linear seg- MEHTA & SUD - GINKGOALEAN LEAF IMPRESSIONS FROM RA]MAHAL HILLS 53 ments and their angle of divergence. In mens are from the same locality from which many cases the leaves are petiolate though Sah (1952) has reported his collection. in several instances the petiole is represented Our specimen 2 is perhaps comparable by a narrow basal region of the lamina as in to his Fig. 2, while his Fig. 1 represent­ Psygmophyllum. It is stated by some ing a very incomplete specimen roughly authors that the veins are subdivided but recalls our specimen I although the two though dichotomy is less frequent in Baiera are collected from places about 9 miles and may be absent in narrow parallel-sided apart. segments, it is by no means rare." Amongst the Mesozoic plants of Queens­ It is obvious that the schemes noted land described by Walkom (1917. 1918. above do not solve the problem, and the 1919 and 1924) the Ginkgoales, particularly confusion continues to exist. Nor does the genera Baiera and Ginkgoites, are promi­ epidermal structure help us to come to a nently represented. It is quite possible that definite understanding, since Harris ( 1935 ). our specimens are referable to the generaliz­ quoting Oishi ( 1933. p. 251 ). states that ed form Ginkgo ( Ginkgoites ?) digitata Brong­ "A comparison of cuticles of Ginkgoites and niart which exhibits as great a variation Baiera species shows that its structure can­ in its leaf form as the living representa­ not be used as a basis for separating the two tive G. biloba L. As Sah (loc. ciL) points genera ". out, the Rajmahal specimens are not com­ In view of the above discussion the parable to Feistmantel's Ginkgo lobata and assignment of specimen 1 to the genus G. crassipes because of their laminae being Baiera must be regarded as tentative and divided into distinct lobes. subject to revision when more details are The importance of this contribution lies available. For obvious reasons no attempt in the fact that Ginkgoalean leaves have is made for the time being to assign it to a been described for the first time from the definite species. Rajmahal series of the Indian Gondwanas The identity of the other two specimens from two new localities. is still more obscure. They are not only The authors express their gratitude to incomplete but are also very indifferently Professor ave Arbo Hoeg, Director, Birbal preserved. The smaller number of seg­ Sahni Institute of Palaeobotany, for his ments and a larger number of veins per lobe kind permission to use the library and for in the second specimen suggest a comparison providing other fasilities at the Institute. with Ginkgoites to which genus specimen They are also indebted to Dr. R. V. Si tholey 3 perhaps also belongs. These two speci- for many a helpful suggestion.

REFERENCES

FEISTMANTEL. O. ( 1877a). Notes on fossil flora of HARRIS, T. M. (1935). The fossil flora of Scoresby India. Rec. Geo!. Surv. India. 10(4). Sound. East Greenland. Medd. am Gronland. Idem (1877b). Fossil flora of the Gondwana 112(1 ). system. Flora of the ]abalpur group (upper KRAUSEL, R. (1917). Ober die Variation der Gondwanas) in the Sone Narbada region. l\lem. Blattform von Ginkgo biloba L. und ihre Bedeu­ Geo!. Surv. India. Pal. indica. 2(2). tung fur die PaUiobotanik. Centralbl. f. Min. Idem ( 1879). Ibid. Upper Gondwana flora of the No. 63. outliers on Madras coast. Ibid. 1(4). *LUNDQVIST, G. (1918). Variations typen von Idem (1881). Ibid. The flora of Damuda and Baiera minuta Nathorst. Stockholm Geo!. Form Panchet divisions. Ibid. 3(2). Forhandl. 40. FLORIN, R. (1936). Die fossilen Ginkgophyten OISHI, S. ( 1933). A study of the cuticles of some von Franz] osephland Nebst Erbrterungen Ober Mesozoic gymnospermous plants from China and Vermeintliche Cordaitales Mesozoischen Alters. Manchuria. Sci. Rep. Tohuku Imp. Univ. 2 Ser. II. Pal. Beit. Zur. Natur. Der Vorzeit. 81 (B). (Geo!.), 12B. 82 (B). SAH, S. C. D. ( 1952 ). Occurrence of Ginkgoales in GANJU. P. N. (1943). On a new species of Psygmo­ the Rajmahal series of Bihar. Curro Sci. May phyllum (P. Sahnii sp. nov.) from the lower 1952. 21: 129-130. Gondwana beds of Kashmir. jour. Ind. Bot. Soc. 22 ( 2-4 ). *Not available to the authors. 54 'THE PALAEOBOTANIST

SEWARD, A. C. (1907). Permo-Carboniferous plants Kashmir. Mem. Ceol. SUyv. India, Pal. Indica. from Kashmir. Rec. CeDI. Surv. India. 36. N.S.2(2). Idem (1912). Lower Gondwana plants from SITHOLEY, R. V. (1943). On Psygmophyllum Golabgarh Pass, Kashmir. Jl1em. CeDI. Surv. Haydeni Seward. Jour. Ind. Bot. Soc. 22 India. Pal. Indica. N.S. 4(3). (2-4). Idem ( 1919 ). Fossil plants. 4. Cambridge. 'WALKOM,A. B. ( 1917). Mesozoic floras of Queens- SEWARD, A. C. & SAHNI, B. (1920). Indian land. Part 1. Queens. Ceol. Surv. Pub. 259. Gondwana Plants: a revision. Jl1em. Ceol. Surv. Idem (1918). Ibid. Part II. Pub. 262. India. Pal. Indica. 7: 1. Idem (1919). Ibid. Parts III & IV. Pub. 263. SEWARD, A. C. & WOODWARD, A. S. (1905). Per­ Idem ( 1924 ). On fossil plants from Bellevue near mo-Carboniferous plants and vertebrates from Esk. Mem. Queens. Mus. 8: 1.

EXPLANATION OF PLATE 1

1. ?Baiera sp. Leaf impression attached to an 2. ? Cinkgoites sp. x 2·5. axis. X ca, 1'5. 3. ? Cinkgoites sp. X ca, 2'5. ~ C r

:r:

'J.