PLANT ECONOMY AT CHIRAND,

YISHNU-MlTTRE

Birbal Salmi Institute of Palaeobotany, Lucknow

ABSTRACT 7·5 m. has been deposited in 705 years. Remains of food plants from the bottom layers Presuming that the rate of deposition has of Neolithic horizon at Chirand near , Bihar, been ± uniform, the bottom 2·7 m. (from comprise lentil, barley, rice, Pisum arvense, etc. 7·5 m. to 10·3 m.) of deposit should have These constitute the first record of Neolithic plant been laid in about 1900 years thus dating economy from North . The Neolithic plant remains known from BUl'zahom in the Kashmir the sample from layer 17 to 1755+1900 B.C. Yalley are of weeds associated with cultivation. = 3655 B.C. The rate of deposition of these From a series of C14 dates at this site, it appears sediments is variable as seen from the that the bottom layers are younger (405 B.C.• consecutive dates in relation to depth 1570 B.C.) than the overlying layers (1675-1750 B.C.). If no discrepancies arc involved, the bottom (0'4 m. = 530 years, 0·3 m. = 95 years, layers from which the plant remains have been 0·3 m. = 70 years), the calculated age discovered may be dated around 3000-4000 B.C. could be either between 2000 and 3000 B.C. or between 3000 and 4000 B.C. Narain (1970), INTRODUCTION however, believes that the lowest layer might be as old as C. 2500 B.C. The samples collected during recent ex• cavations of the site conducted in February remainsmaterialof foodof soilplantssamplesfrom andthe THEarchaeological site Chirand (Lat. 25° 1972 have net bf'en dated yet. The plant remains described here are the 45'N, Long. 84° 45'gE), District Saran, Bihar, first records of cereals from the Neolithic of a site about 8 Km. east of Chapra and North India. Records from the Neolithic situated near the confluence of Ghagra and the Ganga, was kindly sent for investigation of Kashmir (Burzahom) are of weeds indi• by Prcfessor B. P. Sinha, Professor and cating indirectly the cultivation of wheat or Head of the Department of Ancient Indian barley (Vishnu-Mittre, 1968, 1971). History and Archaeology, Patna University, Patna. The soil samples were found devoid DESCRIPTION OF PLANT REMAINS of pollen grains. The carbonized gr?ins of CEREALS cereals and legumes recovered from layer 17 of CRD-XI (Text-fig. 1) from the Neolithic WHEAT horizon extending from layer 12 to 17 are described here. The layers 10-17 have been PI. 1, Fig. 2 radiocarbon dated and surprisingly the top layers are older than the bottom layers The single kernel of wheat IS short, (d. Text-fig. 1). The discrepancy remains oblong, grooved and thick with the ends unresolved. Dr. D. P. Agarwal of Tata broad and blunt. It measures 3·75 mm. Iustt. of Fundamental Research, Bombay, long, 2·75 mm. broad and 1·75 mm. thick. who has dated the samples (Agarwal, 1971) The dorsal surface of the grain is domed, considers involvement of subsidence (perso• and the groove is deep. In these characters nal communication) whereas Prof. Sinha the grain of wheat approaches the grains of (personal communication) believes that the Triticum sphaerococcum Perc. In its lesser bottom layers probably got contaminated LIB, LIT, and higher BIL and TIL indices with water. The layers 10-13 are con• it further compares with the indices of sistently dated to 1050 B.C., 1580 B.C., grains of T. sphaerococcum as shown in 1675 B.C. and 1755 B.C. respectively and Table 1 below. layers 14-16 range in age from 405 B.C. to The differences may be due to carboniza• 1570 B.C. and layer 17 is dated to 605 B.C.• tion. The absence of transverse rippling 1370 B.C. The top one metre of deposit of the shell precludes the possibility of its comprising layers 10-13 from 6·5 m. to belonging to naked grains of barley. 18 VISHNU-MITTRE - NEOLITHIC PLANT ECONOMY AT CHIRAND, BIHAR 19

CRD- XI: SECTION FACING WEST ------~0~--~! T SHOWING THE POINT5 r~olo\ WHERE SM-IPlES I JmN rOR POll[N G) i

J G:AIN MIAmiS

CD IPEr~o~~I CD I mIOD-N t------__ -----JI. tl Atl I I

,

:-~_------lll PERIOD-ill

L...--- __ 0 - _iI I PERIOD-ITt B ®

TEXT-FIG.1 - Stratigraphy of verti• cal section facing west. The cultural sequence is shown on the right. Posi· II~' tion of pollen samples is indicated by 3000± 90 I PlIlJOD-lIA black dots. The C14 dates for the ® {H.\lOlllltlC. layers dated are inserted. Archaeo• botanical materials described in this 3530±100 ® paper were found in layer 17. @ :;:;~ If- 3625±140 ." I @ =< L 3705:±:155 ." -1 I 3465±155 ." I I 3490:::110 ®

•· l6" -1. r 2355::':125 I I 3520±1I5 @ I ~ /" ." I 3220±\.O5 ." I- I ® ." . PEIlJOD-l l ______• Ia ! ~w mut

•• 1 I 2555:t125 .• I I 3325±IOO I I ® •• I I I .• I I ·.•. I . , ! 'i I I N ATUP-AL 50lL I ~ 20 THE PALAEOBOTANIST

TABLE 1- SHOWING THE VARIOUS INDICES OF MODERN AND CHIRAND WHEATS

(The dimensions of gnins of various species of Triticum are based upon hundred grains each. The com• parative discussion of statistical data of the modern species is discussed in Vishnu-Mittre, 1974)

0·560·730·320·790·40LITB/LT/L0·872·141-620·611·090·460·63T/B3·070·37 2·68LfT32·441·761-36 ChirandT. compactumsphaerococc1lmwheat T. aestivum

BARLEY produced by the wild species and the PI. 1, Fig. 2 broader ones by the cultivated species. There are no other external morphological All the thirteen grains referred to barley characters to distinguish the grains of indivi• are very much worn out measuring about dual species. The statistical evaluation of 4-5 X 2·25-3·25 X 1·50-2·25 mm. Most of the various dimensions and the calculation them are naked since the pales are absent of LIB X T indices of the wild and cultivated and the surface bears rippling marks. The species of Oryza in India (d. Vishnu-Mittre, grains are evenly rounded in cross-section 1971) has given reliable information to apart from the ventral furrow which is distinguish various species as shown in shallow and a soft longitudinal depression Table 4. The LIB xT indices of the carbo• on the dorsal side. Some grains show slight nized grains are calculated and compared deviation from the straight shape suggesting with those of the modern species (Tables that the spike was six-row and probably 3 and 4). of the nodding type, a Hordeum sp. In LIB X T index of the large carbonized three of the grains the surface is charac• grains approaches that of O. sativa var. terized by a faint longitudinal stripping spontanea and O. sativa (cultivated) and arising from the remains of the pales or that of the small grains is not different from the longitudinal surface cells of the seed coat. The cross-section of these grains is not angular. These might have belonged TABLEL/BxT1·672·83B2·05T2'502·251·251·001·851·801'503 - DIMENSIONS5·004·25 OF CHIRAND to the hulledSmall barley.grains The4'505'25 measurements of index GRAINS OF RICE six randomLarge'grainsgrainsare given below (Table 2). L

TABLE 2 -" MEASUREMENT OF RANDOM CHI~~ND GRAINS OF BARLEY 2·25B3'252·803·002·752·501·75T1'502·251·702·004·17L 5·254·804·255·00 Average 4·00

TABLE 4 - LIB X T INDICES OF C ULTI• VATED AND WILD SPECIES OF ORYZA BASED UPON HUNDRED GRAINS EACH. THE TABLE ALSO INCLUDES INDICES OF CARBONIZED GRAINS FROM CHIRAND

Species of Oryza L/BXT index RICE Chirand large grains 1'76 PI. 1, Fig. Oryza sativa var. indica 1·71 O. sativa var. japonica 1·70 The kernels of rice, ten in number, are O. sativa var. spontanea Koenig 1'77, 1·79 all five-ribbed. Of these five are slender O. perC11nis Moench 2·20 O. 0fficinalis W aJ I. 2·36 and five broader. The examination of O. ruftpogon Griffiths 2'64 kernels of wild and cultivated species of Chirand slender grains 2'45 Oryza reveals that slender grains are usually VISH~U-l\nTTRE - NEOLITHIC PLANT ECONOMY AT CHIRAND, BIHAR 21 from that of the wild species O. perennis scar is present in most of them and measures and O. rufipogon. The differences could be about 1-1·5xO·50 mm. due to carbonization. These statistical comparisons reveal that the carbonized OTHER REMAINS grains were derived from both the cultivated UNIDENTIFIED FRUIT and wild species of Oryza. PI. 1. Fig. 7 LEGUMES Two halves of a fruit measuring about 12 m. long, 6·7 mm. broad and 2·50 m. The identification of leguminous seeds is thick remain unidentified. They are oblong based upon the shape of the grain and the with a wrinkled surface. Inside there is a position of the hilum scar. There is reduc• broad oval area surrounded by a tissue in tion in dimensions of both the grains and of vertical alignment. A distinct micropylar hilum scar owing largely to carbonization and region is present. The rind outside seems being uninfluenced by repeated cultivation. to be less than half a millimetre thick.

Pisum arvense L. U~IDENTIFIED INSECT PI. 1, Fig. 4 PI. 1, Fig. 6 Mouth parts, the only preserved remall1S The spherical seeds measuring about 2-4 are shown in Plate 1, Fig. 6. mm. in size are compressed and provided with an oblong hilum scar about 1xO·50 CONCLUSION mm. There are smaller grains also, about 2 X 1-1,25 mm., which are perhaps the The Neolithic plant economy at Chirand abortive grains of Pisum arvense or of some consisted of barley, wheat, rice, lentil and other legume. Pisum arvense. Some leguminous weeds have also been recognized. The rice grains Lathynts sativus L. belong to both the cultivated and wild Oryza sativa, and probably also to O. PI. 1, Fig. 4, top right perennis and O. rufipogon. Wheat, barley Three compressed and wedge shaped seeds and Pisum arvense as identified belong measuring 4·50 ~3·50·x3 mm. Hilum scar to the same species as recognized from the is 1-1·5xO·50-0·70 mm. Harappan and Chalcolithic sites in western India. The records of lentil, wheat and Lens culinaris Maedik Pi sum at this site predate those from the Chalcolithic of Navadatoli Maheshwar dated PI. 1, Fig. 3 to 1660 B.C. (Vishnu-Mittre, 1968, 1971). The site Chirand is estimated to date from There are eleven lenticular seeds with C. 2500 B.C. and the evidence of wheat, keeled edge and measuring about 2·25-3 X barley, lentil and PisU1n arvense is suggestive 0·80-1·25 mm. Hilum scar is 1-1·5X of contemporaneity of Neolithic Chirand 0·25-0·50 mm. Compares with small sized with the Harappan cultures in western grains of Lens culinaris. India. Should the site prove earlier in age to the Harappan cultures, it would be an Other Legumes interesting though solitary example of the PI. 1, Fig. 5 diffusion of wheat, barley and lentil from western Asia predating the Harappans. There are a few extremely small seeds Chirand is the first archaeological site in about 2 mm. in diameter. The spherical India which has provided substantial details and flattened among them measure about of plant economy of the Neolithic period in 2 X 1 mm. and compare with small lentil India. Among the other Neolithic sites grains; a rounded one about 2 X 2 mm. looks in the north of India, Burzahom (Kashmir like that of Pis~tm and one 2 X 1·25 mm., Valley) has yielded indirect evidence of square to oblong seeds, seems to be that of wheat or barley cultivation, and rice from black gram (Phaseolus mungo). In view of the eolithic of Orissa has been found to their extremely small size they may belong belong to O. perennis (Vishnu-;\1ittre, 1971), to weeds rather than to cultivars. Hilum a wild progenitor of O. sativa. 22 THE PALAEOBOTANIST

REFERENCES

AGARWAL,D. P. (1971). Tata Inst. Radiocarbon VISHNU-MITTRE (1974). The Beginnings of Agri• Date List. Mimeogr. Circular. culture: Palaeobotanical Evidence from India. NARAIN, LALA ADITYA (1970). The Ne01ithic In ••Evolutionary Studies in World Crops: Diver• settlement at Chirand. Jour. Bihar Res. Soc. sity and Change in the Indian Sub-Continent, 46: 1-35. Ed. Prof. Sir Joseph Hutchinson, Cambridge, VISHNU-MITTRE (1968). Protohistoric Records of V.l(,: 3-33. Agriculture in India. Trans. Bose Res. Inst. 31: 87-106.

EXPLANATION OF PLATE

1. Carbonized grains of rice. The slender 4. Pisum arvense L. A single seed of Lalhyrus grains are on the extreme left of the figure. sativHs is seen on extreme right. 2. 'Wheat grain (Triticum sphaerococcum) on 5. Other leguminous seeds the extreme right marked as ,"V. Barley grains 6. Mouth parts of an unidentified insect. are seen on the left of the figure. 7. Two halves of an unidentified fruit. 3. Lens culinaris Maed. THE PALAEOBOtANlSt, VOL. 21 VISHNU-MITTRE - PLATE 1

I 5

3 2 w

4