The Pala-Sena and Others

History of Ancient India Volume V POLITICAL HISTORY AND ADMINISTRATION AD 750ñ1300) (Regional Powers and Their Interactions) (c. Editors Dilip K. Chakrabarti and Makkhan Lal Vive k anand a Int e r nat io nal Fo und at io n New Delhi Ar y an Bo o k s Int e r nat i o nal New Delhi Cataloging in Publication Data—DK [Courtesy: D.K. Agencies (P) Ltd. <[email protected]>] History of ancient India / editors, Dilip K. Chakrabarti and Makkhan Lal. v. 5 cm. Contributed articles. Includes index. Contents: v. 5. Political history and administration, c. AD 750–1300. ISBN 9788173054846 1. India—History. 2. India—Politics and government. I. Chakrabarti, Dilip K., 1941II. Makkhan Lal, 1954- III. Vivekananda International Foundation. DDC 954 23 ISBN: 978-81-7305-484-6 © Vivekananda International Foundation All rights reserved. No part of this book may be reproduced, utilised in any form or by any means, electronic and mechanical, including photocopying, recording or by any information storage and retrieval system without prior permission of the authors and the publishers. Responsibility for statements made and visuals provided in the various papers rest solely with the contributors. The views expressed by individual authors are not necessarily those of the editors or of publishers. First Published in 2014 by Vi v e k a n a n d a In t e r n a t i o n a l Fo u n d a t i o n 3, San Martin Marg, Chanakyapuri, New Delhi - 110 021 Tel.: 24121764, 24106698; Fax: 91-11-43115450 E-mail: [email protected] www.vifindia.org in association with Ar y a n B o o k s In t e r n a t i o n a l Pooja Apartments, 4B, Ansari Road, New Delhi - 110 002 Tel.: 23287589, 23255799; Fax: 91-11-23270385 E-mail: [email protected] www.aryanbooks.co.in Designed and Printed in India at ABI Prints & Publishing Co., New Delhi. Contents Foreword vii Editorsí Preface xv Part I NORTH INDIA I.1. The Gurjara Pratiharas ó Sima Yadav 3 I.2. The Paramaras ó Sima Yadav 28 I.3. The Chandellas ó Sima Yadav 74 I.4. The Kalachuris ó Sima Yadav 110 I.5. The Chahamanas ó Paras Nath Singh and Amit Upadhyay 133 I.6. The Gahadavalas ó Paras Nath Singh and Amit Upadhyay 142 I.7. The Guhilas ó Paras Nath Singh 157 I.8. The Pala-Sena and Others ó Rajat Sanyal 165 I.9. Assam from the Fifth to the Thirteenth Centuries ó Rajat Sanyal 214 I.10. Dynasties of Orissa ó Shailendra Kumar Swain 228 Part II THE DECCAN AND THE SOUTH II.1. The Pallavas ó G. Sethuraman 255 Political History and Administration vi II.2.1. The Cholas ó C. Santhalingam 272 II.2.2. Military Organization of the Cholas ó Soubhik Mukherji 288 II.3. The Pandyas ó V. Vedachalam 329 II.4. The Hoysalas ó M.S. Krishna Murthy 349 II.5. Eastern Chalukyas ó Rajat Sanyal 384 II.6. Notes on Some Dynasties of the Deccan and South India : The Rashtrakutas, Western Chalukyas, Cheras, Kakatiyas ,Yadavas and Silaharas ó Dilip K. Chakrabarti 392 426 III.3. Invasions of Ghaznavids ó Sima Yadav 441 III.4. Muhammad Ghori and the Establishment of Muslim rule ó Sima Yadav 460 Part IV Part III THE ROAD TO MUSLIM POLITICAL POWER III.1. The Arabs in Sindh, Kabul and Zabul ó Makkhan Lal III.2. The Shahis of Afghanistan and Punjab ó Sima Yadav 409 COINS, INSCRIPTIONS, ARCHAEOLOGY IV.1. Coins ó Devendra Handa 471 IV.2.1. North Indian Inscriptions ó T. P. Verma 500 IV.2.2. South Indian Inscriptions ó K. Rajan 525 IV.3. Inscriptions and Archaeology in the Mapping of Religious Settlements: A Case Study of the Surma Valley (Sylhet) ó Birendra Nath Prasad 532 Contributors 557 Index 559 I.8. The Pala-Sena and Others Editorial Note [The Pala-Sena imprint on the landscape of eastern India, especially Bengal and southern Bihar, is remarkable in the sense that the most visible range of sculptures and architectural specimens in this region belongs to the chronological span of this dynasty. The author of the following essay puts the basic history of this dynasty both in terms of the geopolitical divisions of its distribution zone and the complexities of its chronology. The author also outlines both the pre-Pala historical situation and the miscellaneous dynasties of southeast Bengal.] h THE ëPALA-SENAí ERA: PERSPECTIVES IN HISTORIES OF ART AND POLITY Even a cursory review of the available bulk of writings will enable one to realize that the essentially dynastic phraseologies ëPala and Senaí (Kramrisch 1929, 1994, Huntington 1985: 387-413) or ëPala-Senaí (Huntington 1984) have retained wider usages in the larger historiographical context of early South Asian art than that of political structures. Possibly, the earliest work highlighting the development of a distinct regional school of aesthetic expression under the Pala patronage, arising out of an earlier artistic tradition, was that of J.C. French (French 1928). The immediate response of Stella Kramrisch by the inclusion of the Sena period within this narrower dynastic frame not only broadened the scope of the subject underlined by French, but also ensured a platform from which the later art historians have tried continually to look into the patterns of genesis of this regional stylistic genre of sculpture in Political History and Administration 166 eastern India under the broader emblem ëPalaSenaí. In the domain of literature on the history of dynastic polity, however, the existence of ëPalaSenaí as a political era has been absent in major writings since the early twentieth century. The earliest account of a ëcompleteí history of early Bengal was by Rajanikanta Chakravarti in two volumes published in 1907 and 1909 by the Rangpur Sahitya Parishad (Chakravarti 1999). Of all the writings of the early twentieth century in vernacular, this still remains the pioneering and the most well-documented historical account of the Pala and the Sena dynasties on the basis of chiefly inscriptional records. Besides titling the book as The History of Gauda instead of that of ëBengalí, Chakravarti also published the readings of all the major epigraphs of the Pala and the Sena families known at that time. Among the other major attempts at reconstruction of regional histories, the narrative of dynastic polity of Bengal outlined by Ramaprasad Chanda composed in 1912 essentially sees the Pala and the Sena phases as the results of separate lines of development (Chanda 1975). The Bengali monograph of R.D. Banerji written in 1914 categorically emphasizes the ëfirstí and the ësecondí Pala empires under Gopala I and Mahipala I respectively, the intermediate phase being described as an era of tension between the Gurjara-Pratihara (GP hereinafter) and the Rastrakuta (RK hereinafter) powers on the face of which the Pala dominion of Gauda experienced a political massacre. In fact the significance of this intermediate phase was first brought into light by Banerji after his notice of some inscriptions of one Mahendrapala from northern Bengal and southern Bihar and Banerji believed that this king was the famous GP Mahendrapala, son of Bhoja alias Mihira. The Sena polity seems to have been visualized by Banerji as more closely aligned with the Islamic conquest of Bengal than with the rise of a distinct and powerful local polity (Banerji 1987). Pramode Lal Paulís pattern of narrative of the early historical mechanism of Bengal was also not a departure from the trend set by Banerji (Paul 1939). R.C. Majumdarís History of Bengal, Volume I published from Dhaka in 1943 and its later variants followed an almost identical line of reconstruction of these two polities, although for him, there were three Pala empires, the last centring round Ramapalaís recovery of northern Bengal (Majumdar 1971, 1998, 2005). N.R. Rayís emphasis on the Pala period rather than the succeeding Sena regime can be clearly shown by his use of the term ëPalayanaí, i.e. ëthe Pala formationí in the context of rise of the dynasty that he did not find suitable for the Sena phase which was, again, discussed more as a prelude to the Turkish conquest than as an important phase of Bengalís polity (Ray 2001). The only work of this phase that showed a departure from this general trend was the monograph of B.C. Sen (Sen 1942) on political administration and geopolity based almost exclusively on epigraphic records that practically laid the foundation of D.C. Sircarís coining of the phrase ëPala-Sena Periodí or ëPala-Sena Eraí in eastern Indian polity in the early 1980s. The work that looked at the importance of inscriptions in underlining the contours of political and cultural geography of early Bengal was undertaken by Barrie M. Morrison. Morrison concentrated on the quantitative profile of land transfer charters from Bengal and adjoining regions between the fifth and the thirteenth centuries in visualizing the development of political and cultural geography of Bengal under certain rigid ësubregionsí like the Bhagirathi delta, the northern Bengal plains, southeast Bengal and so on (Morrison 1970). But it has been shown The Pala-Sena and Others 167 elsewhere that the contours of polity under even the precise span of Sena rule in the twelfththirteenth centuries revolved round a continuous fluidity of geo-political identities with prominence of certain new tiers of administrative polity in certain geographical and topographical niches of the delta (Sanyal 2009b). Although not specifically argued, the implication of the phenomenon hints at the limitation of Morrisonís rigid and water tight geographically defined historic identities of ësub-regionsí in the spheres of polity and culture.1 D.C. Sircar precisely suggested that his ëPalaSenaí era can also be conveniently designated as either the ëPalaí era or the ëPala-Sena-Devaí era, for apart from these two major dynasties, a number of local dynasties (of which only the Deva dynasty that turned prominently only in the declining years of Sena supremacy appeared relevant to Sircar for reasons unexplainable) ruled in different parts of Bengal. He preferred the nomenclature in spite of this on three scores: i. The Pala family ruled for a period of no less than four hundred years in Bengal-Bihar and even parts of Uttar Pradesh, a span that remained unsurpassable for any other ruling dynasty of the region; ii. The history of the Sena family can be clubbed with that of the Pala, according to Sircar, apparently because the end of the Pala supremacy and the inception of the Sena dominance in Bengal is a chronologically overlapping set of phenomena as the first paramount Sena king Vijayasena was a contemporary and also probably a samanta of the last sovereign Pala king Madanapala; iii. The history of the Sena period was of further significance because of the annexation of some sectors of Bengal by the Turkish invader Bakhtiyar Khilji during the rule of Laksmanasena (Sircar 1982a: 4). However, notwithstanding the emphasis on the ëPalaí or ëPala-Senaí period of polity, the existence of such a distinctive ëPalaí period in Bengalís history is very difficult to justify. Firstly, a number of members of the Chandra, Varman and Kamboja families ruled in eastern and southwestern terrains of Bengal between the tenth and the thirteenth centuries when the Pala were emanating as the strongest polity of the delta and territories beyond it. Secondly, there is no definite proof that the entire delta was under Pala rule for even any considerable fragment of time, although there are occasional claims of Pala sovereignty in some of the areas beyond their political nuclei in north Bengal and south Bihar, an issue I shall return to in greater details later. The identity of a specifically ëSena periodí of history, on the other hand, is possibly more convenient to visualize on the basis of the geographical distribution of epigraphic records of the Sena family, as this chapter will reveal. PRELUDE TO THE RISE OF THE PALA DYNASTY: CHANGING LINES OF LOCAL POLITIES Political history of Bengal under Gupta domination in the fifth century is more or less coterminous with the political history of northern Bengal under the provincial administrative division called Pundravarddhana bhukti. All the rulers from Kumaragupta to Vishnugupta ruled in this sector of the delta with their administrative headquarters at the districts of either Kotivarsa (modern Rajibpur area of the South Dinajpur District of West Bengal) to the west or at Pancanagari (modern Panchbibi in the Jaypurhat district of Bangladesh) to the east. The existence of a lineage having Datta nameending is consistently supported by references to names like Chiratadatta, Brahmadatta and Jayadatta in four of the five Damodarpur copperplates. Quite interestingly all of them held the post of Uparika, i.e. provincial Political History and Administration 168 administrators and the latter two are also found to have assumd the epithet maharaja. D.C. Sircarís hunch that they might have been the descendants of Nagadatta, one of the rulers of Aryavartta defeated by Samudragupta, seems justifiable and relevant (Sircar 1985a: 141-2). However, this north Bengal- centric political structure soon assumes a highly fluid and fastchanging shape between the sixth and the mid eighth centuryñña phase distinctly marked by the rise and fall of a number of local polities not only within the historically defined delta consisting of modern West Bengal and Bangladesh, but also to the areas beyond it into parts of Bihar and Odisha. In terms of empirical records in the form of primary epigraphical texts, the fluid contours of the local geopolitical frames of Bengal before the rise of the Pala can be underlined prima facie on the basis of three separate groups of copperplate inscriptions dated between the early sixth and the early eighth centuries. These records bear testimony to political developments underlining the inception of the early medieval phase in Bengal and can be visualized from the distribution of copperplates within the historical boundaries of Bengal and the contiguous sectors of northern Odisha. Of these three groups, the first can be safely dated to the sixth century when Maharaja Vainyagupta was still having his administrative seat in AD 507/8 at Kripura in the Brahmanberia sector of southeastern Bengal.2 Till date, the historical identity of this king was known primarily on the basis of the isolated copperplate from Gunaighar. But the recent rediscovery and decipherment of another copperplate of the reign of the king by Ryosuke Furui tend to completely alter oneís understanding of the early history of Bengal. The plate is currently under the possession of a private collector in Dhaka and presumably hails from the Comilla region. The inscription is dated in the Gupta era 184 (= AD 503/4) belonging to the reign of pancadhikaranoparika-mahapratiharamaharaja Vainyagupta who is also described as paramabhattaraka-padaanudhyata.3 Here Vainyagupta copies and by implication reapproves a grant made nearly a century earlier in the Gupta year 91 (= 410/1) by Maharaja Mahesvara Nathachandra. ìLand plots are acquired through purchase, donation and so on at 17 settlements in Jayanatana of Purva mandala, while another plot is from a settlement in Daksina mandala.î While this record is of manifold significance for the study of early social-political and religious history of Bengal for which one should await the published version of the record by its decipherer, some of the readily understandable points that emerge from the inscription are: first and foremost, although the current form of the inscription is dated to early sixth century, it contains the earliest record of permanent land alienation in Bengal, as it predates the Dhanaidaha copperplate of Gupta era 113 (=c. AD 432/3), the earliest copperplate inscription so far reported, by more than two decades. Secondly, it proves that as early as the early fifth century a king named Nathachandra ruled in the Jayanatana area of the Comilla sector of southeastern Bengal where the districts of Peranatana and Guptinatana had developed by the seventh century as prominent regional administrative centres. Thirdly, this is the only inscriptional evidence recovered from Bengal where re-appropriation of a previously issued grant is recorded. Fourthly, this is again the sole charter from Bengal where, apart from land, a number of valuables in the form of ritual objects in metal and ivory are recorded to have been donated to a religious institution. It may be noted here that the place name Jayanatna ëappears as The Pala-Sena and Others 169 the name of an administrative unit belonging to Purva mandala on one hand and of a settlement where the shrine is located and some witnesses reside on the other.í Finally, it is the only primary source, so far available from Bengal, for the study of the Ajivika sect and the cult of Manibhadra. Publication of this inscription with detailed interpretative analysis of its decipherer is expected to mark a major shift in the currently prevailing paradigm of understanding the early history of Bengal. The four other rulers of this group, viz. Dvadasaditya, Dharmaditya, Gopachandra, and Samacharadeva, are known to have been operating with their power base principally at the Varaka-mandala visaya in the Vanga region comprising the Kotalipara sector of the present Gopalganj (former Faridpur) district.4 While the local developments in and around Faridpur in the eastern sector of the delta under all these four rulers is empirically convincing with consistent issuance of copperplate charters and coins, the more intriguing issue is the extension of Gopachandraís power in the central alluvial tracts of West Bengal and in the Odisha adjoining southwestern sectors which had already assumed distinct province (bhukti) level administrative identities as Varddhamana bhukti and Danda bhukti respectively, as revealed from the Malla Sarul and the Jayrampur plates. However, D.C. Sircarís observation that Samatata was also under the political dominance of Gopachandra demands further empirical support (Sircar 1985b: 111).5 The second line of development, which was aligned more closely with southeastern Bengal and areas beyond along the trans-Meghna routes rather than the mainstream political milieu of the Gangetic plains, took place under three major lineages like the Rata, the Khadga and the Deva in the Comilla-Noakhali-Chattagram regions across the trans-Meghna corridor. Samanta Lokanathaís isolated epigraphic evidence from the Tippera foothills shows that he was operating from the Suvvunga visaya (which is still unidentified on ground) and was a contemporary of Jivadharanarata, the predecessor of Sridharanarata who assumed the title of ëlord of Samatataí (Samatatesvara) and ruled from his headquarters at Devaparvvata in the southern foothills of Mainamati range in the Comilla region in the third quarter of the seventh century. The proper political and chronological status of the Samanta Murundanatha with one inscription from Kalapur also remains problematic. So far, the knowledge about Sridharanarata was based solely on the Kailan copperplate, but the recent discovery of a set of three 6 copperplates from the village of Uriswar in the southern Pararpur sector of the Comilla district records grant of land in the same Guptinatana district where lands will be again donated a couple of years later according to the Kailan plate. The Khadga rulers established power in Samatata with a power base at Karmmanta (i.e. modern Barkamta/Barkanta near Comilla proper) under Devakhadga. Our information on this lineage is based principally on the two plates of this king from Ashrafpur and one from the Comilla area, often named as the ëDevaparvataí plate. Out of the eight copperplates discovered from Salban Vihara, four are certainly known to belong to this Devakhadga and his sons Rajabhata and Balabhata. While the Ashrafpur plates were issued from his office at Karmmanta, all the later plates were issued from a place ëother than Karmanta or any other known administrative centre in the areaí, as Morrisonís notes confirm (Morrison 1970: 37).7 The only definite evidence on the date of this family, as recovered by Sircar, is supplied by the travelogue of the Chinese Political History and Administration 170 traveller Sheng-chi who found the devout Buddhist Rajabhata (alias Rajaraja) on the royal seat of Samatata in the third quarter of the seventh century (Sircar 1985b: 146-47).8 The issue of political dominance, chronology and interrelationship of the members of the Deva families of Samatata-Harikela having at least three different and apparently separate roots is a major problem. Previously, three kings between the late eighth and early ninth centuries having confirmed historicity were known to have had this Deva name-ending: Anandadeva, Bhavadeva and Kantideva, of whom the previous two were evidently stationed at Devaparvata while the third one issued land grant charter from Varddhamanapura in Harikela identified with modern Bara-Uthan village of the Patiya Upazila [a revenue unit] in Chittagong (Bhattacharya 2000: 473 and references therein).9 Sircar took them to belong to the two sections of the same lineage who ruled in the broader ëComilla-Noakhali-Chattagramí areas (Sircar 1982a: 182). With the recent discoveries of two copper vase inscriptions of Devatideva ( AD 715) and Attakaradeva (c. early tenth century) from the Chittagong area have led Gouriswar Bhattacharya to argue that Kantideva has to be placed between these two rulers of whom the first belonged to the well known Khasa tribe who in the early medieval period had settled in the Chittagong hill tracts having definite connections with Myanmar and Arakan, as the date seventy-seven recorded in the inscription has been corroborated in terms of the Burmese era commencing from AD 638 (Bhattacharya op. cit.). 10 However, the relationship between these kings is far from clear, more particularly because the Khasa identity of the first ruler of the line so proudly proclaimed is not reiterated by his successors ruling from the same geopolitical base. Suchandra Ghosh has recently shown with epigraphic and literary references that the Khasa tribe gained Ksatriya identity by virtue of their proficient military activities, finally finding place in the Pala military organization as indicated by their frequent mention in many Pala inscriptions (Ghosh 2006, 2011a; cf. Bhattacharya op. cit.). Another Deva family ruling from Devaparvata is known from inscriptions of two kings named Anandadeva and his son Bhavadeva donating land in the Peranatana district; however, this family seems to have hardly had any connection with the Chittagong Deva family. Further, Anandadeva of this line is found to have used the epithet Vangalamrganka, thereby probably implying an antecedent stage of development of Vangala as a separate geopolitical unit in southeastern Bengal (in Sylhet more precisely), finally finding an established identity in the early tenth century when during the rule of Srichandra the essentially ëlocalí Vangala matha is being differentiated from the Desantariya matha establishments. It is quite interesting to note that several administrative centres having a natana suffix like Jayanatana, Peranatana and Guptinatana either as a visaya or a major node of political administration continued in the Comilla sector from the time of Nathachandra to that of the Deva rulers of Comilla. Overall, the Samatata-Harikela areas of southeastern Bengal having their own distinctive and yet highly fluid geopolitical orbits, witnessed the rise and fall of a number of local dynastic polities that rendered the trans-Meghna region the identity of a major sub-regional node in the general canvas of geopolity in early medieval Bengal (Chattopadhyaya 1993-94, 2003: 87). The third and the final line of development of the pre-Pala phase in Bengal started with the accession of Sasanka on the throne of Gauda and the varying lines of continuous power conflicts The Pala-Sena and Others 171 of this phase in which Sasanka played a major role finally paving foundation to the circumstances leading to the rise of the imperial Pala lineage. Sasankaís name as a mahasamanta is found engraved on a seal matrix at Rohtasgarh in Bihar (Fleet 1863: 283-84); he is found to have had a sound administrative machinery from the level of the province to that of the village with elements of reciprocity at lower tiers in and around Danda bhukti attached to Utkala-desa according to the Antla (i.e. the so-called ëMidnaporeí) copperplates; and he is finally mentioned as the overlord of the Sailodbhava Madhavavarman II in the Ganjam area of Odisha in AD 619 (Hultzsch 1981). That Sasanka had his political-administrative headquarters at Karnasuvarna is known from Xuan-zangís travelogue (Beal 1981: 210) and that he was the paramount sovereign of Gauda (i.e. Gaudesvara) is known from the commentary of the text of Banabhatta called Harsacharita (Cowell and Thomas 1897: 275; cf. Hultzsch op. cit. 143 and Buhler 1892: 70). Considering the location of Karnnasuvarnna on the northern bank of the Bhagirathi in the Murshidabad district of West Bengal, as revealed from archaeological excavations (Das 1968), the political dominion of Sasanka has to be mapped over wider areas of Bihar, West Bengal and Odisha.11 Recently Dilip K. Chakrabarti has categorically underlined the geopolitical self-consistency of his kingdom in the context of ancient Indian routes of political and cultural linkages. His observation demands a thorough notice (Chakrabarti 2010: 73). In the political context of the time when the lines of domination were changing fast, Sasanka must be credited with the carving out of a kingdom which was reasonably coherent geographically. The alignment along the western bank of the Ganga/ Bhagirathi from Karnasuvarna to Medinipur and from Medinipur to Orissa is clear and has been a major line of movement in Indian history. The extension up to Rohtasgarh suggests that this kingdom also included the entire highland area from the western part of modern West Bengal to the Gaya-Sasaram sector of modern Bihar. Rohtasgarh is approachable from Sasaram, with the Son flowing around Rohtasgarh. This area gives direct access to the Palamau section of the Chhotanagpur plateau, and a major route from south Bihar to the Banares area via Bhabua and Chakia passes through this general area. The configuration of Sasankaís kingdom involves segments of two distinct routes of ancient India. Other than Sasanka, the only ruler whose name can be doubtlessly associated with this early Gauda sector is a Vaisnava king named Jayanaga known from a sole record from Maliadanga12 in the Murshidabad district of West Bengal. Although his administrative seat was located at Karnasuvarna, in the absence of any reasonable supplementary data, it is not possible either to ascertain his chronology or status in the contemporary political scenario. However, it has been surmised by Sircar that he might have ruled Gauda at the wake of the matsyanyaya immediately after the death of Sasanka and that his naga name-ending might indicate his use of the metronymic (Sircar 1985b: 112-13). BIRTH OF THE IMPERIAL PALA LINEAGE: ëELECTIONí OF GOPALA I AND THE ERA OF CRISIS It should be specified at the outset of this section that the rise of the Pala supremacy has to be looked at in the backdrop of the changing political complexes of eastern India discussed above and the numerous series of development within the Gangetic milieu of the north in general. After the shaking off of the Gupta yoke from the middle Ganga Valley orbit and the subsequent rise of five regional power Political History and Administration 172 structures, viz. the Maukhari of south Bihar (Magadha) and Kanauj (Kanyakubja), the Puspabhuti of Thaneswar (Sthanvisvara), the Later Gupta of Malwa, the unspecified lineage of Gauda and the Varman of Assam (Kamarupa), the earliest precisely datable record is the Haraha inscription Maukhari Isanavarma who claims to have subdued among others the king of Gauda (Shastri 1982, Thapliyal 1985: 23-24, 141-46). This must have happened before AD 554, the date of composition of the panegyric of Isana. struggle in the following manner (Chakrabarti 2010: 72): But the issue of the circumstances leading to the rise of the Pala king Dharmapala, son and successor of Gopala I, as the paramount ruler of Gauda almost unexceptionally called in contemporary literary and epigraphic texts as the Gaudesvaraññhas to be understood in the retrospect of the long-drawn rivalry between Gauda and Kanyakubja between the sixth and the eighth centuries. Sircar successfully explained the contours of this rivalry with reference to the ancient theory of ëthe circle of friends and foes of a kingí envisaged by early political theorists like Manu. It will be interesting cite in details the enumeration of Sircar of the political alignments involved in this struggle and then consider the same series of phenomena envisaged in one of the very recent studies. Sircar writes (Sircar 1985a: 23-24): According to the mitra-amitra-cakra (the circle of friends and foes of a king) conceived by the early Indian politicians, a kingís neighbouring ruler is his potential enemy while the neighbourís neighbor is a potential friend. This is exemplified by the Gaudas and the Later Guptas of East Malwa and their enmity with the Maukharis as well as the kings of Kamarupa. If one now juxtaposes the presentations of the same set of phenomena by the two scholars, the complicated structure of this struggle becomes fairly comprehensive. While in Sircar configuration, the Pusyabhuti of Kanauj are absent, in that of Chakrabarti the Kamarupa line does not figure. This is in fact because Sircar underlined the initial phase of the struggle in the sixth century when the Pusyabhuti were yet to appear prominently on the scene and Chakrabarti on the other hand was looking at the most heightened sub-phase of the conflict when the Maukhari-Pusyabhuti and Gauda-Later Gupta bilateral alliances and subsequent GaudaKanyakubja conflict assumed central focus where the Kamarupa line had nothing more to figure for. The appearance of the Pusyabhuti replacing the early Maukhari lineage in direct confrontation with Sasanka is then underlined by Chakrabarti: ë[t]he combined forces of Devagupta and Sasanka uprooted Maukhari Grahavarman of Kanauj who was married to the Thaneswar princess Rajyasri. This brought the Thaneswar forces in...í13 Very recently, Dilip K. Chakrabarti in his attempt to situate early Indian dynastic polities in contemporary geographical frames has delved into understanding the contours of this The conflict of the five regional lineages for political supremacy in northern India in the seventh century led to the rise of Adityasena as the ëlord of Magadhaí in the late seventh century In the seventh century Bengal, there was a kingdom in the old Gauda section, with its base on the northern bank of the Ganga/Bhagirathi in the Murshidabad district. Sasanka, the lone known major king of this kingdom with its capital at Karnasuvarna (Chhiruti in Murshidabad), figures in the political struggle between Pravakaravardhan, Rajyavardhan, and Harshavardhana of Thaneswar, Devagupta (presumably of the Later Gupta connection) of Malwa, and the Maukhari Grahavarman of Kanauj. The Pala-Sena and Others 173 as gleaned from spatial distribution of his inscriptions along the Bhagalpur-Gaya-Patna alignment having at least one secure date of AD 672 (Thapliyal 1985: 158-166). On the seat of Kanyakubja in the early eighth century, on the other hand, one finds Yasovarman who claims to have subdued the king of Gauda-Magadha according to the Prakrit text Gaudavaho and seems to have had considerable political control over southern Bihar (Sircar 1982a: 44). The recent publication of the inscription of Pahila, a minister of Devapala, has provided concrete epigraphical corroboration to the claim made by Vakpati in the Gaudavaho (Bhattacharya 1997, 2000: 481-87).14 The decline of Yasovarman in Kanauj in the middle of the eighth century paved the path for yet another lineage to flourish there under a king named Vajrayudha, while the fall of Sasanka as the paramount ruler of Gauda at the end of the first quarter of the seventh century is generally acceptedóin the extant historiography of political history of northern Indiaññas the age of inception of a long-ranging political crisis and alluded to as the matsyanyaya in some of the early Pala inscriptions and two texts of the early medieval-medieval periods. The event finally leads to the reappearance of the imperial polity of Gauda under one Gopala who was ëelectedí by his subjects who showed an august example of ësubordinating individual interests to a national causeí around the conjecturally framed date of AD 750, since the historicity and chronology of this king has to rely solely on the claims made by his successors (for early writings on this crisis and subsequent election, refer Banerji 1987, Chanda 1975: 16, Maitreya 1987: 12-7, Majumdar 1971: 82, Ray 2000: 49, Sircar 1982a: 52-57; for an estimate of the varying versions of historians on matsyanyaya and the election of ëthe Bengali kingí, refer Pal 2008). The issues of this anarchy and the actual significance of the story of election of Gopala may now be looked into at some details. The earliest epigraphic evidence on the anarchic situation of matsyanyaya (lit. ëíthe rule of fishí) where the weaker becomes an easy prey of the stronger is narrated in the fourth verse of the Indian Museum copperplate of Dharmapala dated to his twenty-sixth ruling year, falling in the third quarter of the eighth century, issued from Mudgagiri (i.e. Munger). The famous Kesava Prasasti was composed at Bodh Gaya in the same year (Chakravartti 1908, Maitreya 2004: 29). The Indian Museum plate records the installation of Gopala to put an end to anarchy by the prakriti (i.e. subjects). This Gopala is described as a maharajadhiraja and descendant of the progenitor Dayitavisnu. The same claim of election of Gopala has been repeated in the Khalimpur plate of Dharmapala dated to his thirty-second year issued from Pataliputra. The undated Nalanda Prasasti, however, is silent on the issue and the implication of this silence is also important as we shall shortly see. Apart from these two and some other epigraphic documents of the early Pala kings, two texts also present two different versions of this anarchy and subsequent election. The Buddhist Aryamanjusrimulakalpa of the eighth century records that after the death of Soma, i.e. Sasanka (Jayaswal 1934: 50-51): [T]he Gauda political system (Gaudatantra) was reduced to mutual distrust, raised weapons and mutual jealousyññone (king) for a week; another for a month; then a republican constitutionññsuch will be the daily (condition) of the country on the bank of the Ganges where houses were built on the ruins of monasteries. Thereafter Somaís son Manava will last for 8 months 5 days. Political History and Administration 174 The second text is a composition of AD 1608 by the Tibean saint Lama Taranatha who writes that the Chandra lineage ruled in Bhangala (i.e Vangala) from the time of Sri Harsa (i.e. Harsavardhana) to the rise of the Pala dynasty. Taranatha traces the beginning of the anarchy at the death of the last king Lalitachandra of this lineage in the beginning of the eighth century. According to the text (Chattopadhyay 1980: 251): After him though many ksatriyas were born in the Chandra line, none of them actually ruled the country... those who were born in the royal family lived a ministers, brahmanas, rich merchants etc. and were lords in their respective spheres. But there was no king as such ruling in the state. Then Taranatha presents the story of birth of Gopala by a tree-god on a ksatriya woman ënear Pundravarddhana.í Subsequently he is said to have come to the country of Bhangala where the common people elected him the king of the country already suffering from long-ranging anarchy. The inconsistencies in chronological and geographical configuration of the anarchy and election readily emerging from a review of epigraphic and literary evidences are many: Firstly, the statements on both the duration of anarchy and crisis and the location of Gopalaís election vary in Aryamanjusrimulakalpa and the travelogue of Taranatha. Sayantani Pal rightly detects that while for the former, the duration was more than a century and the location of the rise of Gopala was Gauda, for the latter this anarchy started only in the first quarter of the eighth century and that too in the ëBhangalaí region only (Pal 2008: 35). Secondly, Sircarís acceptance of Taranathaís statements on Gopalaís early occupancy in southeastern Bengal (Sircar 1982a: 51; also Chakrabarti 1991: 161-7) seems to contradict the statements of the Sian inscription of Nayapala (Sircar 1972-73, 1982b: 102-22, 1985c) who claims that Gopala I exerted a severe blow against the king ruling at Samatata, where only two image inscriptions from Baghaura (Bhattasali 1918, 1983a) and Narayanpur (Chakravarti 1956, Sircar 1942) in the Comilla reign of Mahipala I have been accepted as reliable source to advocate for Pala rule in the ëTipperaí region (Sircar 1952). If Gopala hailed originally from the southeast, his attack on Samatata cannot have any explanation (Chowdhury 1967: 18, Ghosh 2011: 222). It has been recovered from the statements of Aryamanjusrimulakalpa that Gauda was the most prominent part of Bengal to the author of this text and Gopala had acquired supremacy over this region (Pal 2008: 24). Thirdly, the distribution of epigraphic records of Dharmapala from Paharpur in the east to Nalanda in the west fairly shows his stronghold in north Bengal and south Bihar. His reference as a king of Vangala or Vanga in contemporary records outside Bengal was obviously because these fluid suborbits of Bengal like Vangala-Vanga-Gauda were possibly coterminous for the neighbouring polities for whom the only prominent figure in Bengal was ëDharmaí (i.e. Dharmapala). Finally, if one accepts Sircarís suggestion that the undated plate from Nalanda should predate the dated plates judging from the constitution of the charter, the issue of complete absence of the genealogy/ eulogy section in this plate remains unexplainable, particularly when the other two charters so charily record this election resulting from the chaotic situation of anarchy. May one think that the story of matsyanyaya and election became only relevant to Dharmapala at a later stage when he had to draw on legitimacy of his paramount status from his predecessor as he was aspiring to figure in the larger power-axis of The Pala-Sena and Others 175 northern India? Reference to the story of political unrest and election of Gopala and that to the capture of the seat of Mahodaya (i.e. Kanyakubja) in consecutive verses of both the Indian Museum and the Khalimpur plates at least tend to indicate this possibility. The most recent study aimed at understanding the historical geographical significance of the term rightly highlights the underlying factors responsible behind the currency of the story of matsyanyaya in the contexts of historical texts and historiographical models. It was quite natural for the court poets of Dharmapala to legitimize his supremacy by way of glorifying the ancestry of the lineage and the story of upheaval provided the most suitable plank for the supposed election of the ancestor of their patron. For the author of Aryamanjusrimulakalpa, on the other hand, the same story provided the ëideal atmosphere for the capture of power by a Buddhist line of kingsí and Taranatha finally composed his account nearly nine hundred years after the incidents involving obvious geochronological discrepancies mixed with legends preserved in Buddhist traditions, although his statement about an early association of Gopala with north Bengal seems correct. It is quite probable that Gopala achieved the throne of Gauda either with the help of a section of a preexisting bureaucracy or a small but powerful military set up that has been legitimized by the composers of Dharmapala as prakriti (Pal 2008). It is difficult to believe that in a state absolute lawlessness and acute chaotic political environment the common populace will unanimously ëelectí a person to gift him with the royal seat of Gauda. Sircar has shown more than one example of election of a king in ancient Indian tradition, but they probably form part of the larger process of legitimization of polity in early medieval India (Sircar 1982a: 56-57). PALA GENEALOGY BETWEEN DHARMAPALA AND MADANAPALA: NEW SOURCES A proper reconstruction of the genealogy and chronology of the Pala kings is primarily constrained by two factors: firstly, the Pala kings recorded dates in terms of their ruling years instead of that of any particular system of known reckoning like the Vikrama or the Saka eras, leaving a few crucially valuable exceptions though; secondly, the sources clearly show that in this family many of the kings used the names of predecessors, resulting into the appearance of several persons with one single name. The problem is further aggravated by references to Pala kings in the inscriptional records of other lineages like the GP whose records are precisely dated in terms of calendrical years and therefore the Pala dates are ought to be corroborated by the date ranges of kings of this and other contemporary families. Taranatha assigns Gopala I a reign of fortyfive years, whereas Aryamanjusrimulakalpa suggests that he ruled for twenty-seven years and passed away at the age of 80. Sircar discarded the mythical elements in these estimates and fixed the chronology of Gopala between AD 750 and 775 based on the conventional notion of one generation of twenty-five years (Sircar 1982a: 51). The Actual political history starts with the rise of Dharmapala as the paramount sovereign of Bengal in AD 775 according to Sircar. It is recovered from the epigraphic sources of Dharmapala that after the death of Gopala I, Dharmapala ascended the throne and two rulers following him, viz. Devapala, Surapala I, the son and grandson of Dharmapala respectively, ruled in Bihar-Bengal in the ninth century. After this, the line of Political History and Administration 176 Dharmapala closes and that of his younger brother Vakpala begins. There is so far no definite proof of rule of Vakpala and his followers till Vigrahapala I, 15 but epigraphic and literary evidences dated in the ruling years of his successors starting from Narayanapala to Madanapala testify to the rule of kings of the line of Vakpala from the late ninth to the mid twelfth centuries. Based on these collateral lines of succession, a considerable number of considerations and reconsiderations of the Pala genealogy and chronology have been attempted by a number of historians and epigraphists, of which a systematic summary till 1984 is available in Susan L. Huntington (Huntington 1984: 2938). The latest of the estimates available till the time of Huntington was that of Sircar envisaged in the 1975-76 (Sircar 1975-76), where he gave a detailed genealogy of all the members of the Pala family (including the kings who apparently did not rule and mentioned only in epigraphic and literary sources of ruling members of the family) and a through chronology of all the members of the lineage based on the then known latest dates form inscriptions and manuscripts. 16 Huntington offered her own proposal based on the compilation from all the then available sources (Huntington op. cit. 38), although she cautiously refrained from calculating the chronology for problems that will be discussed below.17 The complex issue of Pala genealogy and chronology and many related issue in northern Indian polity assumed an altogether new shape almost immediately following Huntingtonís publication in March 1987 when a copperplate inscription of a king named Mahendrapala, son and political successor of the Pala king Devapala, son of Dharmapala was discovered from the mound of Tulabhita or Salaidanga in the Habibpur Police Station of Malda district. Soon the record proved to be of paramount significance in the history of northern and eastern Indian polity because of the reason that as many as twelve dedicatory inscriptions on images, stupas and architectural members of a king named Mahendrapala were already known from Bengal-Bihar (Sanyal 2009a and references therein); this Mahendrapala was considered by R.D. Banerji to have been identical with the GP ruler of the same name. Thus, Banerji came to the conclusion that GP Mahendrapala had extended his political dominance in eastern India, subduing the contemporary Pala monarch in the first half of the ninth century (Banerji, 1918). The discovery of the Tulabhita plate completely disproved the above theory and made a number of fresh issues to emerge in the Pala political history as well as the history of the Pala-GP relations. Notwithstanding the significance of the discovery of the Tulabhita plate, one has to put on record that the report of this copperplate was but the initiation of a plethora of discoveries (or rediscoveries) of epigraphs and manuscripts that necessitated a thorough revision of the history of northern-eastern Indian polity of the Pala dynasty and their near contemporaries. Thus, the discovery of the Tulabhita plate was succeeded by the discovery of two plates of Madanapala of the same dynasty in 1990 from the mound called Sibbati in the Rajibpur village in South Dinajpur district, the publication of two hitherto unknown plates of the time of one Gopala, son and successor of Surapala I preserved in a private collection at Los Angeles, a third plate of the reign of the same king from Mohipur in the Bogra district of Bangladesh, the Biyala and the Rangpur plates of the time of Mahipala I from Jaypurhat and Rangpur districts respectively in Bangladesh. A unique stone inscription of the thirty-third ruling year of Mahipala I from The Pala-Sena and Others 177 Rajbhita in the Dinajpur district of Bangladesh also deserves notice (Haque and Kuddus 2005a, Roy 2009, for the reading of Gouriswar Bhattacharya).18 The discovery of the new copperplate inscriptions of the Pala family demanded, inter alia, a further reconsideration of the genealogy and chronology of the lineage particularly of the first group of kings of the line of Dharmapala. They prima facie showed the immediacy of three necessary amendments: 1. A king named Mahendrapala, the GP namesake of the Pala lineage, ruled between Devapala and Surapala I. 2. Another new king named Gopala, son and successor of Surapala I, ruled for at least four years; hence, this Gopala had to be numbered Gopala II as previously the son of Rajyapala was considered Gopala II and the grandson of Ramapala Gopala III (who are to be now called Gopala III and IV respectively) and 3. The discovery of the second Rajibpur plate proved that Madanapala ruled for at least twenty-two years and not eighteen as previously believed unanimously. Gouriswar Bhattacharya gave a tabular schem of the genealogy summarizing the entire database available till then (Bhattacharya 1998). S.C. Mukherjee attempted a reconstruction primarily based on his wrong reading of the digit 32 for 22 in the second Rajibpur plate along with the dates available in the records of other kings and suggested a daterange of 1143-1175/6; the reasons of assigning fifty-seven years by Mukherjee to Narayanapala remains equally questionable (Mukherjee 1999). S.C. Bhattacharyaís recent reconstruction on the basis of ëMKYRí (i.e. maximum known year of reign) available in the records of the kings of the Dharmapala line shows how the first six rulers of the dynasty, because of the incorporation of Mahendrapala and Gopala II, have to be accommodated in the pedigree within a span of about 123 years (Bhattacharya 2007).19 A review of the above schemes offered by epigraphists from D.C. Sircar to S.C. Bhattacharya would show one striking similarity in methodology: the inception of the lineage under Gopala I in AD 750, a datum line fixed by R.C. Majumdar and D.C. Sircar. It is curious to note that some of the other scholars, as Huntingtonís composite table would show, located this inception at dates which are closer to AD 750 but not exactly that date. But a detailed analysis of this specific difference will lead one to undertake issues that demand a separate focus of inquiry. A fresh attempt was made during the present work to re-address the issue of Pala genealogy in the light of all the available sources of which the most crucial sets of evidence surfaced in the last three decades, obviously because of a renewed interest of scholars in Pala history inspired by the discovery of the Tulabhita copperplate. A review of the available literature on Pala literary and epigraphic evidencesññ juxtaposed against the so far known estimates of the Pala genealogyññshowed that in spite of all the coherent attempts, a few crucial evidences have so far escaped the notice of scholars dealing in the problem. Considering the entire range of database in hand from the time of Dharmapala to Madanapala, the following are the chronologically arranged evidences that would lead one to a yet another reconsideration of the issue: 1. The discovery of the Tulabhita plate suggests that there was a king named Mahendrapala who ruled for fifteen years between Devapala and Surapala I, the latter being the younger brother of Mahedrapala as the Tulabhita plate suggests. Political History and Administration 178 2. A king named Gopala succeded Surapala I and ruled for at least four years, as evident from the two Los Angeles paltes; this has to result, as already stated, in renaming ëGopala IIí and ëGopala IIIí as Gopala III and IV, son of Rajyapala and Kumarapala respectively. 3. If the convention is to number kings of same name in terms of their chronological appearance, the reason behind not naming Surapala, son of Devapala, as Surapala I, is not understandable, for there was a second Surapala, son of Vigrahapala III who should be named Surapala II. Likewise, the reasons behind not naming Rajyapala, the second ruler of the line of Vakpala and the successor of Narayanapala, as Rajyapala II remains inexplicable, since there was a previous Rajyapala (who should be named Rajyapala I), the son of Devapala, who apparently did not rule as the available evidences suggest. 4. Sircar and all the scholars following him believed that out of the three Vigrahapala, only Vigrahapala III, successor of Nayapala, ascended the throne and ruled for twenty-six years. Huntington argued on the basis of her own stylistic analysis that an image of seated Buddha in the Indian Museum assigned unanimously to the twelfth or thirteenth year of Vigrahapala III of the mid eleventh century should in fact be placed in the reign of Vigrahapala I of ninth century, successor of Jayapala and nephew of Dharmapala (Huntington 1984: 47). However, there is more than one reason to argue that this image was carved during the reign of the intermediate Vigrahapala II, grandson of the renamed Rajyapala II in the tenth century. Although a detailed study of the image and the inscription is beyond the purview of this essay, a few points will be demanding to justify the argument. In the area of style, the nascent stage of development of the conical back-slab which is round in the ninth century and sharply conical in the eleventh, the crudely carved pancharatha pedestal with a beveled outer border showing an intermediate stage between the triratha pedestal of the ninth century and the sharp pancharatha of the eleventh and the high relief and thickset accommodation of the subsidiary figures on the back-slab markedly characterized by the absence of a kirttimukha, which becomes an essential component of the eleventh century idiomóall these stylistic features would put this image to the tenth century. In the realm of palaeographic features, even the lesser number of letters in the single lined inscription would explain that the style of writing shows a transitional phase of development from mature Siddhamatrika to Gaudiññthat initiated from the middle or second half of the tenth century. Since the only Vigrahapala who can be dated to the tenth century is Vigrahapala II, it is tempting to assign this image to his reign. The hypothesis gains a stronger plank with the recent publication of an image of the reign of one Vigrahapala where the inscription clearly suggests that the donorís father-in-law was a minister of Rajyapala (i.e. our Rajayapala II). Although this image does not provide the exact year of reign of Vigrahapala, it is clear from style and palaeography that it belonged to tenth century and hence to the reign of The Pala-Sena and Others 179 Vigrahapala II; thus this image clearly proves that Vigrahapala II ascended the Pala throne after Gopala III (Bhattacharya 2000: 281-91). If the present argument on the date of the Indian Museum Buddha is accepted, one has to again provide twelve years for this king demanding a second set of amendments in the line of Vakpala.20 5. It was unanimously believed that Vigrahapala III was succeeded on the Pala throne by his son Ramapala whose elder brothers Mahipala II and Surapala II did not have any chance to the throne and therefore in all the genealogies of previous scholars they do not figure as ruling kings. But S.K. Saraswati argued that a manuscript dated in the sixth year of a Mahipala should be dated to the reign of Mahipala II of the eleventh century on the basis of stylistic analysis of the illustrations (Saraswati 1978: 35, 38). However, his arguments were summarily dismissed by Sircar to whom this was an absurd proposition ëconsidering Mahipala I ruling in 1026 and the accession of Madanapala in 1143í, as he thought it would be impossible to accommodate this second Mahapala with six years of rule within this short span of time (Sircar 1982b: 87). But the claim of Saraswati has recently been categorically reiterated by Claudine Bautze-Picron in the light of stylization of the illustrations (Bautze-Picron 1999). The proposition of Saraswati and Bautze-Picron can be readily supported by an examination of writing style of the text also, bearing close resemblance to those found in the manuscripts of Ramapala than those of Mahipala I. One has to further remember that the rediscovery of Mahendrapala created a similar situation where one is compelled to accommodate three rulers within the time bracket between c. AD 833 and 860. Ramapalaís second brother Surapala II was also not known to have had any records dated in his own reign, allowing Sircar to surmise that he met an early death after a fratricidal conflict leading to the coronation of Ramapala. But a recent publication of a manuscript by Eva Allinger dated in the second year of one Surapala has proved beyond all doubts that this Surapala cannot be a ruler of ninth century as the stylistic features of the figurative representations as well as the palaeography of the manuscript clearly reveal that it has to be dated at a later date and hence should be assigned to Surapala II (Allinger 2006). Considering the authority of these scholars in the areas of eastern Indian iconography and art of the early medieval period, one has to accommodate, once more, now in the second line of the imperial Pala kings, the accession of three monarchs, viz. Mahipala II, Surapala II and Ramapala within the span from c. AD 1067 to 1075. 6. Finally, the date twenty-two available in the second Rajibpur plate of Madanapala will necessitate an extension of the reign of this king by four years as he was previously known to have ruled for eighteen years. 7. In the area of dates of inscriptions, some crucially significant exceptions to the general practice of recording dates in terms of regnal years of concerned kings are found in some of the epigraphic records of the family. In altogether three inscriptions the dates are given in Saka and the Vikrama eras. Quite interestingly, all these dates come from donative records, Political History and Administration 180 presumably because of donorsí acquaintance with the then prevalent systems of reckoning in other parts of the country. Thus, an inscription of Mahipala I from Sarnath is dated in Vikrama Samvat 1083 equivalent to AD 1026 showing that Mahipala I was ruling in that particular year (Cunningham 1968: 182, Hultzsch 1885, Maitreya 2004: 104-9). A second pedestal inscription from Nongarh in Bihar dated in Vikarama Samvat 1201 (= AD 1144) similarly fell in the early part of Madanapalaís reign (Sircar 1967a). A third date of similar kind comes again from the reign of Madanapala who is found to rule in Saka era 1083, i.e. AD 1160/1 in his eighteenth year (Sircar 1951a, 1954a, 1958a). The fourth date has been recovered from a dedicatory record of the time of one Govindapala who is usually supposed to have succeeded Madanapala following the imperial Pala line in at least south Bihar (Banerji 1915: 109, Sircar 1965b); the problems of considering the issue can be discussed later. Thus, a reconsideration of the most current set of sources would lead one to amend a number of issues in the Pala genealogy: a. Apart from the inclusion of Mahendrapala and Gopala II in the first line of kings, Mahipala II and Surapala II have also to be included in the list of kings immediately preceding Ramapala. b. following the convention numbering homonymous rulers in terms of appearance, the well known Rajyapala, the son and successor of Narayanapala, has now to be called Rajyapala II for reasons explained above and c. Vigrahapala II has to be given the status of a ruling king for no less that 12 years in the light of the fresh dating of the Rohoi (Bihar) Buddha in the Indian Museum, though one should keep this issue open to further critic. It is found in all the relatively justifiable attempts at structuring chronology that the scholars have initiated with the preconceived conjecture that Gopala I started his reign in AD 750, and then they tried to match the lines of accession and to finally corroborate the absolute dates available in the inscriptions of Mahipala I and Madanapala. However, an easier option available after the discovery of the Balgudar inscription showing an absolute date was to go back from Madanapala, who must have initiated rule in 1143, tracking down to the reign of Dharmapala with a calculation of the maximum years of reign of each king available from a thorough sampling of their written records. Although the intricate calculations involved in this exercise is beyond the scope of this discourse, considering the acceptance of the new rulers like Vigrahapala II and others as well as the latest known dates preserved in epigraphs and manuscripts, the accession and decline of Dharmapalaís rule can be reasonably put around AD 766 and 798 respectively. This readily takes back the date of coronation of Gopala to AD 741. Note that the date of end of Dharmapalaís rule envisaged here is in sharp contrast to all the earlier estimates where this king is found on the Pala throne till the early ninth century. One has to mention with equal emphasis that because of the impediments caused by the nature of presentation of dates in most of the Pala inscriptions, a single new date in the reign period of any of the kings might entirely invalidate the present state of research and this is inevitable. FROM DHARMAPALA TO MAHENDRAPALA: ëKANYAKUBJA-GAUDAí CONFLICT A review of the major epigraphic database pertaining to the reign of Dharmapala clearly The Pala-Sena and Others 181 shows that the study of political history of the Pala lineage under his control is more or less coextensive with that of the struggle involving the Pala of Bengal, the GP of Rajasthan-Uttar Pradesh and the RK of Deccan centring round the capture of the seat of Kanyakubja in the late eighth-early ninth centuryññan issue that almost equivocally characterized the historiography on the theme till the later part of the twentieth century. The issue of this conflict of three was first underlined by H.C. Raychaudhuri who coined the term ëTripartite Struggleí, justifying his nomenclature by saying (Sen and Raychaudhuri 1934: 80): The tradition of empire attached to Kanauj from the time of Harsa to to the Muslim conquest. Rulers of the most distant corners of India counted it their proudest boast to have captured Mahodaya-Sri, i.e., the royal splendor of Kanauj. Bitter contests ensued for the possession of the imperial city. Soon the theory gained currency in the broader canvas of history of polity in early medieval northern India (Altekar 1934, Majumdar 1971, 2005, Mishra 1966, Puri 1957, Ray 1936, Sen 1942). A new dimension to this struggle was first visualized by D.C. Sircar has emphasized the role of a still inadequately known lineage called Ayudha having its root in Panchala, who captured Kanauj in the first half of the eighth century under one Vajrayudha, succeeded by Indrayudha alias Indraraja. On the basis of certain references to the Ayudha in contemporary literary and epigraphic records, Sircar concluded that immediately after rising to power as the lord of Gauda, Dharmapala resumed the old rivalry with Kanyakubja and attacked Indrayudha to enthrone his regent Chakrayudha. Sircar rightly detects that this marks a renewed phase of the old conflict that cenred round the Pala and the Ayudha families alternately supported by the GP and RK under fast-changing axis of power. Sircarís configuration of this conflict was based on the statements made in the Khalimpur plate of Dharmapala, the Bhagalpur plate of Narayanapala and some of the later records of the GP and the RK and their feudatories. But the time-frame he structured in relation to this struggle requires a revisit. Sircar believed following the claim made in the Khalimpur plate that initially Dharmapala was successful in subduing Indrayudha to install Chakrayudha at Kanauj. But this success was short-lived as Dharmapala was defeated by GP Vatsaraja (having his power base near Jodhpur in Rajasthan) when Indrayudha sought his help to regain Kanauj. But Dharmapala again attacked Kanauj (for reasons not supplied by Sircar). But now Indrayudha took refuge to RK Dhruva who had in the meantime subdued GP Vatsaraja; the result was a tie between Indrayudha and Dhruva, the latter defeating Dharmapala near Kanauj. But soon after his return to Deccan, the Gauda king again attacked Kanauj and usurped Indrayudha to install Chakrayudha. But then Dhruvaís son Govinda III, who had already defeated Vatsarajaís son Nagabhata II, came forward to help his fatherís ally Indrayudha in the beginning of the ninth century and forced Chakrayudha and Dharmapala to bow down before him. Then after his return to Deccan, Nagabhata again gained stronger political footage in northern India to the effect that his vassal Bahukadhavala defeated Dharmapala (Sircar 1982a: 60-1). Two things appear curious in the above scheme of reconstruction. Firstly, it is seen that a single Pala generation represented by Dharmapala is taken to be contemporaneous to two generations of all the other polities, viz. Dhruva and Govinda III representing the RK, Vatsaraja and Nagabhata II for the GP and Political History and Administration 182 Indrayudha and Chakrayudha for the Ayudha. The second issue that seems even more curious is that Dharmapala is found to have installed Chakrayudha twice, if Sircarís interpretation is accepted, on the seat of Kanauj. Let us now have a retrospect afresh at all the inscriptional sources currently available for reviewing the issue. Previously the only epigraphic source referring to Dharmapalaís win over Indrayudha was the Khalimpur plate of his thirty-second year. But now it is known from the date furnished by the Indian Museum plate that the installation of Chakrayudha by the Pala king took place at least before his twenty-sixth year of rule. The Pala sources that explicitly record this incident are: 1. The Indian Museum plate of year twentysix of Dharmapala (falling in c. AD 792 considering his accession in AD 766). 2. The Khalimpur plate of year thirty-two of the same king. 3. The Tulabhita grant of the fifteenth year of Mahendrapala, the elder son of Dharmapalaís successor Devapala, where an identical claim is repeated. 4. The Mohipur plate of Gopala II, son of Surapala I, where the composer clearly records the gift of villages including Kusasthala (i.e. Kanyakubja) by Dharmapala to his friend who was a descendant or grandson (naptr) of Yasovarman, the king of Kanyakubja. In all likelihood this anonymously refers to Chakrayudha of other inscriptions. I am fully in agreement with Ryosuke Furui that if this refers to Chakrayudha, he was probably Yasovarmanís daughterís son (Furui 2008: 67-70). 5. The Bhagalpur plate of year seventeen of Narayanapala (around c. AD 880/1). The records of the GP and the RK families or their feudatories precisely referring to the different junctures of this conflict, on the other hand, are all records of later rulers cherishing the memory of their ancestorsí political achievements, leaving the sole exception of the Nesarika grant of Govinda III who claims to have defeated the Vangala king Dharma before at least c. AD 805, the date of issuance of the Nesarika grant. The exact claims made in the records of these families regarding their achievements against the Pala are summarized chronologically in the following (for general outline of dates of RK sources, refer Reu 1933 and Altekar 1934 and for those of GP, Puri 1957 and Mishra 1966). “ Nesarika plate (c. AD 805) of RK Govinda III (c. AD 794ñ814) claims that he defeated Dharma, the Vangala king and snatched the royal banner depicting Bhagavati Tara (Gupta 1963, Sircar 1963e). “ Radhanpur plate (c. AD 808) of RK Govinda III claims that his father Dhruva (c. AD 785ñ 794) captured from GP Vatsaraja the two white umbrellas of Gauda (gaudiyam sarad-indu-pada-dhavalam chatradvayam) captured earlier by Vatsaraja from the king of Gauda (Keilhorn 1981a). “ Wani plate (c. AD 808) of RK Govinda III repeats the same claim (Fleet 1882: 15663). “ Baroda plate (c. AD 812) of RK Karkaraja informs that his uncle Govinda III put his younger brother Indraraja (i.e. the father of Karka) in charge of southern Gujarat (Lata) so that the gaudendra-vangapatinirjjaya-durvidagdha Gurjara king can be prevented from entering Malwa (Fleet 1883). The Pala-Sena and Others 183 “ Barah plate (c. AD 836) of GP Bhoja (c. AD 836ñ85) proclaims victory of his grandfather Nagabhata II (c. AD 795ñ833) who drove out Chakrayudha from Kanyakubja and granted land within Kalanjara mandala of the Kanyakubja bhukti (Shastri 1983a, Sircar 1983: 233-35). “ Jodhpur plate (c. AD 837) of the GP feudatory Bauka claims that the army of Nagabhata II under Baukaís father Kakka achieved fame in the struggle against the Gauda king at Mudgagiri (Majumdar 1983a, Sircar 1983: 236-41). “ Sanjan plate (c. AD 871) of RK Amoghavarsa (c. AD 814ñ78) claims that i. his grandfather Dhruva had defeated the king of Gauda in the land between Ganga and Yamuna ii. Dharma and Chakrayudha paid tributes to his father Govinda III (c. AD 794ñ814) when he reached Kanyakubja and iii. his father Govinda III ëcarried away in battles the fair and unmistakable fame of Nagabhata II. (Bhandarkar 1983b, Sircar 1983: 478-94). “ Gwalior Prasasti of GP Bhoja records i. the victory of his grandfather Nagabhata II (c. AD 795ñ833) over Chakrayudha who was of a ëlow demeanourí and dependent on others (parasraya-krta-sphuta-nicabhavam) as well as the Vangapati and ii. the victorious campaign of Bhoja against Dharmapalaís son (Majumdar 1983b, Sircar 1983: 242-46). “ Una plate (c. AD 899) of the GP feudatory Avanivarman suggests that his great grandfather Bahukadhavala defeated Dharma in war (Keilhorn 1981b). “ The Chatsu plate of early Kalachuri Baladitya (c. AD 900ñ20), possibly a subordinate of GP Nagabhata II, records that his matrimonial ally Guhila Sankaragana conquered ëBhataí, the king of Gauda, and made a gift of this kingdom to his overlord (Bhandarkar 1982, Sircar 1983: 363-71). “ The inscriptions of RK Krisna II (c. AD 880ñ 915) record that he forced the Gauda (king) to learn modesty and also subdued the Anga, Kalinga, Vanga and Magadha (Reu 1933: 75, cf. Sircar 1982a: 75). What is understandable prima facie from the above is, as already noted, the absence of dates of incidents in GP and RK records immediately after the time period concerned. Further, there is hardly any scope to situate Chakrayudha at Kanauj twice under the repeated attacks of Dharmapala who himself had to accept defeat at least from Dhruva, Govinda III and Vatsaraja. If one accepts the date of accession of Nagabhata II in c. AD 795 estimated by V.B. Mishra (Mishra 1966: 22-24) instead of c. AD 808 by B.N. Puri (Puri 1957: 42), one is compelled to either take the defeat of Chakrayudha and Dharmapala at the end of the latterís rule in the last decade of the eighth century or to argue that this second phase of GP-Ayudha struggle took place during the early years of Devapala who tried to help his fatherís young friend Chakrayudha but failed as suggested in the Jodhpur plate of Bauka.21 Further the repeated reference to ëDharmaí as the king of Gauda, Vanga and Vangala need not be necessarily taken in the physical sense of the terms of reference, for even in the Arab geographical sources, the farthest limits of southeastern Bengal is described to have been ruled by Dharmapala (Chakravarti 2002a: 165). It seems that the name Dharma became a generic representative for the polities of Bengal in the contemporary records beyond the region. Political History and Administration 184 It is surprising to note that while the GP king Nagabhata II occupied Kanyakubja and shifted his headquarters there for Rajasthan, neither the RK Govinda III nor Dharmapala ever aspired to annexe the city, although both of them had captured it. The claim of Devapala that his father was the lord of all the cardinal quarters is exaggerated and represents a stereotypical narrative of the Chakravarttiksetra. However, the areas described in the Indian Museum and the Khalimpur plates to have been occupied by himññconsidering his political base in the south Bihar-north Bengal plainsññis quite impressive. D.C. Sircar has placed Bhoja in the Berar region, Matsya around Jaypur, Madra in Sialkot, Kuru in Delhi and Meerut, Yadu in Gujarat, Yavana in the northwest, Avanti around the western Malwa region, Gandhara in the Peshwar region and Kira in the Kangra region (Sircar 1982a: 59).22 In any case, the claim of suzerainty along ëthe entire sweep of the Indo-Gangetic plain from the north-west to Bengal with extensions into the Malwa corridor and Vidarbhaí coupled with the ability to ëdonateí the chair of Kanyakubja to a ëperson of his choiceí will lead one to believe that Dharmapala was definitely a powerful ruler of northern India at least in the earlier half of his reign. Geopolitical horizon of the Pala dominion extended during the reign of Dharmapalaís son Devapala, while the fate of his other son Tribhuvanapala alias Haravarsa remains unknown. Devapala must have initially suffered from raids from the GP Nagabhata and the contemporary Tibetan monarch, but subsequently he regained form the queer situation as his own inscription and those of his successors till Narayanapala suggest. While the narratives of the ëpan-Indian empireí credited to him in the Badal Garuda Pillar inscription of Narayanapala (Keilhorn 1894a, Maitreya 2004: 70-85, Sircar 1982b: 194-95) are stereotyped, the claims of win over the Utkala, Kamarupa, Nepala and the Gurjara appear reasonable. But the identity of the Gurjara king is again problematic. He is not likely to have been Nagabhata IIís son Ramabhadra, as suggested by Mishra (Mishra 1966: 26), for this king ruled according to Mishra from c. AD 833 to 836 and was not powerful enough to have threatened the Pala monarch. The more reasonable probability is that of Nagabhata II in the later years of reign. In that case, the only way to explain the claim of Bhoja that he defeated Dharmaís son is to have a reinvestigation of the three years allotted to Ramabhadra when in fact Bhoja probably had made a sudden thrust against Devapala to retaliate the defeat of his grandfather at the hands of the latter. The Kamboja referred to in the inscription of this king would have come either from the northwestern (Pakistan-Afghanistan) or the northeastern (Tibetan) frontiers, though D.C. Sircar explicitly favours the latter assumption on linguistic, Tibetan legendary as well as later epigraphic sources. Although his conquests over the Huna and the Dravida are also not easily understandable, the massive ëaerial flightsí commissioned by his and his son Mahendrapalaís composers that Devapala was the ëstage director of the drama of great warsí (for e.g., maha-samara-nataka-sutradharah, found in the Tulabhita plate) do not appear to have been complete exaggerations. His direct rule over Bihar is clearly revealed by his Munger and Nalanda grants, while his sway up to the Varanasi region is clearly proved by the Lucknow Museum plate of his younger son Surapala I whose mother was stationed at the city on the other side of the river Kalmasanasa (i.e. modern Karamnasha forming the boundary between Bihar and Uttar Pradesh), while the executor of The Pala-Sena and Others 185 the plate was a provincial ruler of Vyaghratati on the Bhagirathi in the northern sector of Murshidabad. Devapalaís religious leanings and patronage towards Buddhism have strong empirical support that deserves note. The Ghoshrawa eulogy records Devapalaís favours to Viradeva, a Buddhist preceptor who came to Magadha and stayed there in the Yasovarmapura vihara (Keilhorn 1888, Maitreya 2004: 45-54). The Nalanda plate on the other hand makes reference to grant of five villages to a monastery at Nalanda constructed by the Southeast Asian Sailendra ruler Balaputradeva. The period between Devapala and Surapala I is of paramount significance and has been subjected to several discussions by a number of scholars, for it has permanently altered a major hypothesis on Pala-GP political relations. The entire bibliography of the Pala rule before the discovery of this plate suggest that after Surapala I, the rulers of Vakpalaís line came to dominate Pala polity starting with Narayanapala whose date was fixed by Sircar from the mid ninth to the early tenth century. As already noted, R.D. Banerjiís discovery of the inscriptions bearing the name of one Mahendrapala prompted him to invoke the theory of ëGurjara occupation of Magadha.í This hypothesis provided a plank for the coining of the second set of hypothesis of tripartite and subsequently quadripartite struggle involving these two polities. The date ranges of both Narayanapala and GP Mahendrapala I (the latter consecrated around c. AD 890 to rule till c. AD 907) fit so well into the then available sources on the Pala-GP history of the period that the proponent of and subscribers to this theory readily argued that the Mahendrapala of the dedicatory records of eastern India was actually the GP king of that name who gained a strong footage in Gauda-Magadha in the ninth-tenth century subduing the contemporary Pala monarch Narayanapala. The complete absence of Narayanapalaís record between his seventeenth and fifty-fourth years also made sufficient room to insert GP Mahendrapala I and surmise that after hiding for this intervening phase Narayanapala regained his throne at Magadha sometime before his fifty-fourth ruling year. But the discovery of this new copperplate has permanently discarded this myth of GP Mahendrapala I ruling in Bengal-Bihar. While he had apparently no role to play in the combat, the situation gets further complicated, after the inclusion of the two new rulers in the Pala genealogy, as Bhoja is now found have been a contemporary of at least four Pala rulers from Mahendrpala to Narayanapala, an issue to which I shall return. Some preliminary observations on this phase of struggle between Bengal and Kanyakunbja may be relevant. First of all, it is clear from extraneous epigraphic evidences that though the role of the Ayudha dynasty in this conflict can hardly be denied, the adjectives applied to Chakrayudha in the Gwalior inscription in no sense portrays the image of an imperial political stature. It is also quite curious that a lineage which had at least three rulers in succession on the throne of Kanauj never issued a single copperplate proclaiming their presence there. Thus, the ëpowerfulí and ëmightyí status ascribed to the Ayudha kings, on the basis of two references in literary texts, seems to have been an overemphasis (Sircar 1982a: 58, 1985a: 6, 16).23 Secondly, it is difficult, as it is illustrated above, to understand how Dharmapala played the role of the sole representative of Gauda in this struggle the eighth and early ninth centuries balancing between two generations of all the other three power bases. Thirdly, the Tulabhita plate proves that GP Mahendrapala I ruling in Political History and Administration 186 the late ninth-early tenth centuries never had anything to do with Bengal-Bihar and therefore, the phase visualized as the era of exclusive GP dominance in eastern India is now to be found as possibly the only phase of a sustained break in the entire contour of Pala-GP conflict when Narayanapala ruled peacefully for more than half a century without any raids. Finally, the removal of Bhojaís son from the scene and the discovery of new Pala rulers between Devapala and Narayanapala will show that Bhoja was on the GP throne during the reigns of all the four kings between the two mentioned above. And this chronological frame is further significant, for there is no claim in the GP records of the period, other than Devapalaís defeat in the hands of Bhoja which must have been short-lived for reasons irrelevant here, of tassel with the Pala. Thus the entire span of Bhojaís rule in Kanauj saw a phase of stable extension of the Pala empire in eastern-northern India, excepting the early phase of Bhojaís reign involving his victorious campaign against Devapala. On the whole, the new discoveries in the east and the reconsideration of the comparative chronology show a thorough and revealing shift in the contours of regional power structures of north India on a broader canvas. The Tulabhita plate does not furnish any specific data regarding the political profile of this Pala monarch, who is surprisingly not mentioned in any of the subsequent Pala records except the Siyan inscription of Nayapala, 24 from the testimony of Surapala Iís inscriptions noted above coupled with the evidence of the still unpublished Sarnath image dated in his reign (AR-ASI: 1907-08: 75). But it is quite convenient to believe that Surapala I had held sway over considerably larger territories in Bengal-BiharUttar Pradesh along the mainstream Gangetic route. He possibly inherited this tract from his elder brother Mahendrapala who in turn had received it from Devapala. The evidences that necessitated the second major alteration of the Pala genealogy was a set of three copperplates from the reign of Surapal Iís son Gopala who is now numbered Gopala II. Nothing specifically is so far known about this king excepting that he had his capitals at Sohana and Mahasala and that he had control over Pundravarddhana province with districts of which at least one, viz. the Sthalikata visaya (evidently the Sthalikkata visaya of Dharmapalaís inscription), was inherited from his predecessorsí territories. It is interesting to note that some administrative officers of the rank of mostly army chiefs and subordinates are regularly found to have patronized Buddhist religious institutions and in all the cases such activities are being endorsed by the central authorities by way of permanent endowments made in favour of such institutions. The remarks on the issue by the decipherer of the Indian Museum and the Mohipur plates in the context of the former deserve citation (Furui 2011b: 150-51): The acts of mahasamanta Bhadranaga recorded in this inscription conform to the pattern of donation witnessed in some of the early Pala grants, namely, the establishment of religious institutions by subordinate rulers and their petition for land or village donations to the king. In the Khalimpur plate of Dharmapala, mahasamantadhipati Narayanavarman asked the king to donate four villages to the deity Nunnanarayana established by him. Similarly, the Jagajjibanpur plate of Mahendrapala and the Mohipur plate of Gopala II record the construction of Buddhist viharas and the petitions for donation by mahasenapati Vajradeva abd mahasainyapati Kokkaka respectively. The present case is comparable to the last two also in terms of the format of the inscription, which contains The Pala-Sena and Others 187 the genealogy of the subordinate ruler, though its location is different from that of these inscriptions. The composite set of evidence found in these Pala inscriptions is expected to provoke explorations into patterns of royal patronage under local autonomy vis-‡-vis the role of the central authority on the broader processes of legitimization of polity. THE NEW BEGINNING: FROM VIGRAHAPALA I TO VIGRAHAPALA III The first two members of the second line named Vakpala and Jayapala are known only from references in other membersí inscriptions and literary accounts. Similarly, Huntingtonís arguments for more-than-a-decade long rule of Vigrahapala I do not seem to have sufficient grounds. Thus, the rule of the second line has to initiate with Narayanapala, who was previously supposed to have spent more than two-third of his political life in exile. Although now it can be conveniently argued, on the basis of his copperplate recording grant in north Bihar (Tira bhukti) and an image carved in Bihar Sharif showing his latest date so far known (Banerji 1915:), that he enjoyed dominance over entire Bihar and passed it on to his son Rajyapala II. The sources dated in Narayanapalaís time do not supply any significant information about his own rule. However, both the Bhagalpur plate and the Badal stone eulogy are reliable evidences for understanding of the political profile of his predecessors; the former also proves that this ruler was the first in the family to embrace Saiva faith shifting from the Buddhist doctrine of his ancestors. Rajyapala II was earlier believed to have ruled only in Bihar and the non-availability of major epigraphic evidences from Bengal from the time of Devapala onwards had led scholars to believe that the Pala dominion came to be confined to the southern Bihar region and was gradually being divided probably into smaller principalities (Majumdar 1971: 134-35, Chakrabarti 2010: 99). But the Bhaturiya inscription from the Rajshahi district offers some crucially significant information. First of all, this is a unique epigraph representing a hybrid composition involving characteristics of a karasasana and a Prasasti (for kara-sasana, refer Sircar 1965a: 111-14), where a village was donated to a Siva temple built by a minister of the king against a nominal tax of 100 purana/ annum. But the high-sounding claims of the political exploits of the king demand examination. It proclaims that he subdued the Mleccha, Anga, Kalinga, Vanga, Odra, Pandya, Karnnata, Lata, Suhma, Gurjara, Krita and the Cina. D.C. Sircar rightly remarks that the reasons behind conflict with the Pandya, Karnnata and the Lata are difficult to understand, while the Krita was a Muslim principality of northwest and the Mleccha were probably Arab Muslims. Note that Kalinga had a different geographical connotation from Odra and the connections of Suhma with the latter needs reinvestigation, particularly because a segment of Radha encompassed Suhma adjacent to Utkala/Odra for a considerably long span of history and this tract formed a major route of communication from western Bengal through the northern and western Odishan corridor. The march against Anga would have resulted presumably from a local upsurge, while Vanga is known to have been occupied by the Chandra kings with its epicentre at Vikramapura. On account of literary reference to Tibet as China, D.C. Sircar equates Rajyapalaís China adversaries with the Kamboja enemies of Devapala, although it is prima facie impossible to explain the renaming of the group within a span of less than a century. Considering Political History and Administration 188 his reference in the an inscription of Gopalavarman of Kamarupa as Gaudaraja (Sarma 2003: II/113, Sharma 1978: 214, 218) and repeated claims of Devapalaís campaigns against Kamarupa further reflected in Narayanapalaís Bhagalpur plate, it appears that Pala-Kamarupa conflict was in fact in smouldering continuity of the broader Kanyakubja-Gauda polity (for details, Choudhury 1988: 108-33). 25 A second Kamboja lineage headed by Saiva Kunjaraghatavarsa (c. AD 91525, according to Sircar 1982b: 181) having his solitary inscription at Bangarh claiming his identity as Gaudapati (Chanda 1975: 41, Westmacott 1872 for reading by Rajendralal Mitra), probably snatched north Bengal from Rajyapala II, who possibly regained the territory sometime before the issuance of the Bhaturiya inscription. That the Pala-Kamboja relation remained cordial afterwards is shown by the homogeneity of names of the Kamboja Rajyapala and his wife Bhagyadevi with Pala king and his consort of the same name. The two rulers succeeding Rajyapala II were Gopala III and Vigrahapala II. While this third Gopala has no important epigraphic record after the sixth year found from Jajilpara in Malda, this inscription is a fair indicator of his reign in north Bengal and south Bihar, for the record was issued from south Bihar at Vataparvatika (which is one of the two properly identifiable administrative headquarters of the Pala kings) and land was granted in northern Bengal. It is further important in the area of administration that while Kuddalakhataka here appears as a visaya level administrative centre, it was the administrative camp from where the Tulabhita plate of Mahendrapala was issued. One major political issue of this phase was the Pala-Chandra conflict in the eastern sector of the delta. The discovery of the famous Ganesa image from Mandhuk has been accepted as an indicator of sustained Pala supremacy within Chandra dominion towards southeast (Sanyal 1950). The claims in the plates of Srichandra on the other hand that he reinstalled (samropana) Gopala and returned his beseiged consort to the Pala king leaves no room to doubt that at the later part of Gopala IIIís reign the Chandra lineage again regained independence in eastern-southeastern Bengal. Although recent discoveries and new approaches to existing ones will lead one to assign more-than-a-decade of kingship to Vigrahapala II, the polity of the period remains unknown until new epigraphic material surface. As noted in the beginning, the inception of Mahipala Iís rule has been viewed by earlier scholars as the establishment of the ësecondí Pala empire. D.C. Sircar had rightly repudiated this theory of revival since there was no acute crisis in the immediately succeeding phases as already discussed in the foregoing. But the reign of Mahipala I certainly witnessed events that set it apart from that of his predecessors in the preceding century. The issue of the retrieval of pitrya rajya claimed in his inscriptions and that of janaka-bhu mentioned in Kamauli plate (Maitra 2004: 127-46, Venis 1894) and the Ramacharitam (Sastri 1969: 26) remains problematic, for there is no evidence that the possession of north Bengal was at stake unless new evidences prove that Kunjaraghtavarsaís successors continued to battle against the precursors of Mahipala I; the Nayapala of the second Kamboja line possibly took his name from the king of the same name who succeeded Mahipala I on the Pala throne. Another important issue is the discovery of dedicatory records recording the name of this king from two sites in southeastern Bengal. While such discoveries might apparently seem to be stray or results of locally transported evidences, if one considers The Pala-Sena and Others 189 the report of the Kamarupa ruler regarding the defeat of the Gauda king in the hands of Tailokyachandra, a counter-attack in the peripheral Chandra dominion might not be altogether devoid of substance, particularly considering the aforementioned claim made by Srichandra. The politically more significant developments of the reign of this king had greater concern with contemporary south Indian politics. The temporary defeat inflicted on the Pala by Kalachuri Gangeyadeva (c. AD 1010ñ41) was quickly retaliated by Mahipala I as suggested by the discovery of his inscriptions in the Banares region which his predecessors must have won in an opportune drawn from the Kalachuri-GP conflicts, though the identity of the Vangala king defeated by Tripuri Kalachuri Laksmanaraja in the second half of the tenth century is difficult to realize (Ganguly 1966: 61, Sircar 1982b: 82; for Kalachuri-GP quarrel, Mishra 1966: 45). However, D.C. Sircar recovers from Arab sources that Gangeya made a sharp comeback in the Banares area within c. AD 1034. This clearly reflects the Kalachuri ambition in extending power base from central India to the middle Ganga Valley into the Varanasi area via the Satna-Rewa-Mirzapur route (Chakrabarti 2010:119). The inscription of Rajendrachola dated AD 1021 makes claims of winning victory over a number of local rulers of Tandabhukti, Takkanaladam, Uttiraladam and Vangala-desa. The list of kings is: 1. Dharmapala of Dandabhukti, definitely a predecessor of Kamboja Nayapala, 2. Ranasura of Daksina radha, possibly ruling in the Garh Mandaran area on the Bankura-Hooghly border, 3. Sangu and his overlord Mahipala in the Uttara radha and 4. Govindachandra, the last member of the Chandra dynasty of Vangala (Hultzsch 1981, Shastri 1955: 251-52). Sircar makes emphasis on the defeat of Mahipala Iís subordinate Sangu rather than the Pala king himself in the hands of the Chola army. While Sircarís remark that the Chola expedition indirectly helped Mahipala I to conquer southwestern Bengal suffers from selfcontradiction, 26 K.A.N. Shastri rightly identifies him with the Pala monarch (Shastri 1937).27 One intriguing point to note however is that if Mahipala I was physically involved in the struggle in northern radhaññand this seems quite plausible because he is exclusively mentioned in that contextññhis sway over that region in the early eleventh century has also to be accepted. In spite of the rapid sequence of upheavals, this king ruled for no less than forty-eight years and donated lands in the two major subdivisions of northern Bengal. Of these the more significant one was the Panchanagari visaya that reappears as an administrative division after nearly six hundred years (cf. Appendix 1, nos. 5 and 68). Of equal interest are his donations towards Pasupata Saiva preceptors at the cloister at Bangarh, as revealed from an inscription of his son (Sircar 1973-74, 1982b: 85-101). Mahipala Iís son Nayapala ruled for at least fifteen years and the spatial distribution of five epigraphic documents 28 of his time from northern Bengal through western Bengal to south Bihar is quite impressive. That the ëprotractedí encounters and negotiations between the Pala and the Kalachuri polities continued till the time of Nayapala are known from Kalachuri inscriptions, where Gangeyaís son Karna is described to have been the conqueror of Gauda and Vanga. The subjugation of Karna by Nayapala on the other is indicated by Tibetan legendary accounts referring to a treaty between the two as well as a further clash during the reign of Vigrahapala III in which he came out victoriously resulting in matrimonial tie with Karnaís daughter Yauvanasri as alluded to by Political History and Administration 190 Sandhyakaranandin (Sircar 1982a: 85, 1982b: 115). Interestingly enough, there is concrete inscriptional corroboration of claims of both the parties. The defeat of Gauda (Nayapala) is recorded in an inscription from Paikore in the Birbhum district (AR-ASI 1921-22: 115, Majumdar 1971: 146), 29 while the victory of Nayapala over the Kalachuri is mentioned in the Siyan inscription of Nayapala. The recent notice of a dedicatory inscription during the reign of this king at a site called Mandalgram in Bardhaman is also interesting (Sanyal 2006). The three inscriptions from Gaya allude to the existence of a subordinate Brahamna family: 1. Sudraka, son of Paritosa, 2. Visvarupa alias Visvaditya, son of Sudraka and 3. Yaksapala, son of Visvarupa. Their consistent recognition of the sovereignty of Nayapala in their records will lead one to exclude any possibility of an internal struggle. The significance of the Saiva donation by Nayapala has already been underlined. One has to note that a second line of struggle was aligned with south Kosala under Somavamsi Mahasivagupta Yayati, presumably during the reign of Nayapala and again the western Chalukya Vikramaditya, successor of Somesvara I Avahamalla (c. AD 1044-68) in the second half of the eleventh century. However, it seems chronologically more justifiable to place the last battle in the reign of one of the two elder sons of Vigrahapala III than himself, as generally believed (Sircar 1982b: 86, Chakrabarti 2010: 119). While no complete land transfer record of the time of Nayapala has so far surfaced, the three records of the time of his son from north Bengal and south Bihar again makes it difficult to think of any crisis due to the Chalukya attacks. THE LAST CENTURY OF PALA RULE: SURAPALA II TO MADANAPALA AND BEYOND The reigns of Vigrahapala IIIís sons Mahipala II and Surapala II do not offer any specific information, but the manuscripts cited above will clearly indicate that they ruled for nearly a decade and only this considerable span falling roughly between c. AD 1067 and 1075 can allow one to incorporate the Chalukya raid which is too early to be contemporary with the reign of Vigarahapala III.30 It is difficult to believe that the Pithi lineage of Gaya revolted against any of these rulers, as the Pithipati Bhimayasah, the lord of Magadha, helped Ramapala to regain parts of north Bengal The identity and geographical affiliation of maha-mandalika Isvaraghosa who ruled for no less than thirtyfive years is however intriguing. While his power base is unanimously accepted to have been located at Dhekkari in the northwestern Bardhaman where one also comes across considerable archaeological evidence of early medieval settlements around the Gaurangapur area, the land donated in his inscription was located in the northern Bengal plains. The more convincing issue is that of the Kaivartta (fishermen) revolt that brought Ramapala onto the Pala throne in connection with the much discussed reoccupation of north Bengal by this Pala ruler. According to Ramacharitam, the Varendra sector of north Bengal was besieged by the Kaivartta chief Divya, followed by his brother Rudoka and Rudokaís son Bhima. That this revolt had a connetion with a larger scale of polity is indicated by the Ramacharitam that refers to raids in Varendri by the RK Sivaraja simultaneously with the revolt of Bhima, further intensified by the attacks of tha Vanga king Hari who has to be identified with Harivarma of eastern Bengal. In his expedition to wrest Varendra Ramapala was assisted by his RK The Pala-Sena and Others 191 maternal uncle Mathana and his son and nephews, besides a number of local chiefs. The list of the feudatories includes: 1. Pithipati, the lord of Magadha and conqueror of Kanyakubja, named Bhimayasah. 2. The daksina-simhasana-chakravartti Viraguna of Kotatavi. 3. Jayasimha of Dandabhukti who had defeated the Somavamsi Utkala king Karnakesari. 4. Vikramaraja of Balavalabhi located near Deva grama. 5. Laksmisura of Aparamandara styled as atavikasamantachakra-chudamani. He must have been a descendant of Ranasura of northern radha, defeated by Rajendrachola. 6. Surapala of Kujavati. 7. Rudrasikhara of Tailakampa. 8. Bhaskara alias Mayamallasimha of Ucchala. 9. Pratapasimha of Dhekkari whose relation with Isvaraghosa remains uncertain. 10. Narasimharjuna of Kayangala mandala. 11. Chandarjuna of Samkata grama. 12. Vijayaraja of Nidravali. 13. Dvorapavarddhana of Kausambi and 14. Soma of Paduvanba mandala (Sastri 1969: 38, cited in Sircar 1982a: 89-90).31 Many of these peripheral polities can be identified. The pithi were located in Bodh Gaya; Kotatavi (lit. ëthe forest fortí) is more likely to have been the forested Dumnigarh area to the east of Bishnupur in Bankura extending into the present Garh Mandaran (i.e. Aparamandara) forest in Hooghly which was the core territory of the ëcircle of feudatories of the forestí, for the chief of this principality has been called the ëcrest jewelí of the circle; 32 Dandabhukti was modern Dantan in West Medinipur; Kujavati has been identified by Sircar with the area to the north of Dumka in Santhal Parganas; Dhekkari in this case was definitely Dhekur in Bardhaman; Tailakampa is the submerged temple site of Telkupi in Puruliya; Kayangala is modern Kankjol that formed a separate provincial division under Laksmanasena with the name Kankagrama; Vijayaraja is usually identified with Vijayasena of the Sena lineage; identification of Kausambi with Kusumbi in north Bengal will imply a new line of force geographically widely separated from all the other chiefs of western Bengal; the case of Paduvanva identified with Pabna suffers from the same problem. However, the overall configuration of the growth of these local principalities during the time of Ramapala, coupled with a few contemporary to his predecessor Mahipala I, would lead one to surmise that a steady and extensive intraregional process of political formation characterized the polity in this sector of early medieval Bengal from the early eleventh century under certain mandala or chakra (both signifying a confederate ëcircleí) that evolved in less than one century into prominent foci of microscopic power structures in the eastern fringe areas of the Chhotranagpur plateau. It will be rewarding to inquire if the genesis of a large number of medieval bhuma territories in this precise belt ultimately culminating in the growth of the Mallabhuma in the Bankura-PuruliyaWest Medinipur regions, encompassing at least one of the principalities involved in Ramapalaís endeavour, had its genesis in the broader context of Pala polity in the eleventh century. The allusion in the Ramacharitam to the struggle between Ramapala and the rulers of Utkala and Kalinga has possibly a reasonable corroboration in Anantavarman Chodagangaís inscriptions. An almost identical claim by Kulottungachola however demands further evidence (Sircar 1982a: 91).33 It is unfortunate to note that though he ruled for fifty-three years with his base at Ramavati (identifiable with the extensive site of Amati in Malda) in north Bengal, all the records of the time of his reign are Political History and Administration 192 personal donations and are concentrated in Bihar only. Ramapala had four sons of whom only Kumarapala definitely had a share of the throne before Madanapala, as his minister Vaidyadeva ruled for at least four years according to his Kamauli inscription which shows that Kamarupa continued to form a part of the Pala dominion annexed earlier. But Sircarís allotment of two years for him based on the above theoretical assumption lacks any direct evidence and at the present state of information there is no justification in including this king in the list of members who actually enjoyed the throne. He was succeeded on throne by his son Gopala who is now Gopala IV. He ruled for fifteen years as so far known, but no copperplate of his reign is yet reported. However, the contents of a commemorative inscription from Nimdighi in the Rajshahi district recording his name has led scholars like R.C. Majumdar and D.C. Sircar to permanently discard some unjust scars of cowardly attributes attached to his personality by R.D. Banerji earlier. The Nimdighi inscription refers to his death in the battlefield, although both the epigraph and the epigraphists are silent about the identity of this enemy (Sircar 1965b). Gopala IV was succeeded on throne by Madanapala, the last known paramount ruler of the dynasty. The reign period of this king is of momentous significance as it provides the only telltale evidence of a corroborative date in a dedicatory inscription from southern Bihar on which the whole issue of reconstruction of the Pala chronology has to rest. Three years after the coronation of Madanapala, Gahadavala Govindachandra issued his Lar plate from his base at Mudgagiri, having already his presence felt by the end of Ramapalaís reign as shown by his Maner copperplate (Sircar 1982b: 95). On the other hand, Vijayasena, the historical founder of the Sena family also probably exerted periodical blows to the Pala supremacy in Bengal-Bihar having his strong base initially in the radha sector, if his identity as Vijayaraja of Nidravali is accepted. The Antichak inscription of Masanikasa alludes to the conflict between the Gauda and the Vanga kings in which the former succeeded with helps from Sahura, father of Masanikesa (Sircar 1972-73, 1979a: 238). The Rajghat inscription on the other hand records the construction of a Saiva temple at Varanasi by one Bhimadeva who is convincingly believed to be identical with Madanapalaís minister of the same name recorded in the Manahali copperplate. The second major event was the aggression of the Ganga king of Kalinga named Anantavarman Chodaganga who is stated to have vanquished Mandara (certainly the Aparamandara of Ramacharitam) besieging the Aramya city (modern Arambagh in the Hooghly disrict) and in this onslaught he was assisted by his friend (sakha) Vijayasena as suggested in the Vallalacharita (Dhar 1904: 24). The route of Chodaganga from the southwestern district of West Medinipur through the ChandrakonaGhatal sector straight into Garh Mandaran via Bandar and finally to Arambagh in Hooghlyññ would firstly strengthen the probability of Vijayasenaís early connections with this region and will further vindicate that the local polity of this territory under the Sura lineage was still quite vibrant in the middle of the twelfth century. D.C. Sircar believes that Madanapala must have had formed an alliance with Vijayasena during his raid into the Gahadavala heartland in Uttar Pradesh. But this friendship acted as an interval in the general sequence of hatred, renewed by the raids from the Kalinga direction. That the Pala polity was in decline in the second half of the twelfth century is clearly reflected in the Rajghat inscription where Bhimadeva boasts of The Pala-Sena and Others 193 acting as the saviour of the ësinkingí (varidhimadhya-guptam) glory of the Gauda dominion from the attacks of the king of Kalinga and the ruler of the Rayari lineage, the latter identified with the contemporary local polity of the Sylhet region, though this is difficult to prove (Sircar 1982b: 136-39). It appears from a review of the geochronology of the so far known inscriptions of Madanapala that the Pala dynasty had lost political control over northern Bengal in the later years of his reign and the strong sweep of friendhip and enmity with Vijayasena finally acted in the favour of this Sena king who soon held sway over Bengal with his seat at Vikramapura. However, the Pala kings retained their control over Bihar-Bengal at least till the early years of Madanapalaís reign is clear from the distribution of their epigraphic documents in the region. Most of the historians dealing in the polity of the Pala dynasty, most significantly including D.C. Siracr, believed that the dedicatory inscriptions and manuscripts dated in the ruling as well as the ëlostí years of two kings named Govindapala and Palapala are records of Madanapalaís successors. While the latest known date of Govindapala comes from a manuscript dated in his thirty-ninth year, D.C. Sircar allows four years to him without any explanation. The latest known date of Palapala on the other hand is thirty-five and this is readily accepted by Sircar on equally unreasonable grounds. So far there is no reason to take these rulers as members of the imperial Pala line, since their relationship with Madanapala is nowhere indicated in any source. Further their chronological status is also absolutely unknown.34 However, one has to note that oneís knowledgebase on Pala polity is fast-changing with the discovery of every next inscription and this is equally applicable for these rulers, since their consistence reference as ruling kings obviously suggests that they managed safe niches in the polity in southern Bihar in the twelfth-thirteenth centuries. The issue of much debated Pala rule beyond north Bengal should be briefly addressed. The discover of two sets of dedicatory records of the times of Gopala III and Mahipala I have been acceped as convenient evidences for advocating in the favour of Pala rule in the Comilla-Tippera region. Notice has also been made to the recent report of an inscription of the time of Nayapala from a region close to the site of the PalaKalachuri battle in western Bengal in the eleventh century. While these records hardly denote any permanent political possession, 35 the context of the discovery at Mandalgram and the precise time frame covered by these evidences from the second half of the tenth to the first half of the eleventh centuries will compel one to believe that at least at different points of this time bracket the common populace in these respective regions temporarily recognized the authority of these Pala kings. THE SENA PERIOD: POLITICAL ORGANIZATION The fall of the longest polity in Bihar-Bengal paved way for a new Saiva-Vaisnava lineage in the same region with Sena name-ending. Twelve copperplate inscriptions of the Sena kings so far discovered from the different geographicalcultural terrains of West Bengal and Bangladesh indicate that this lineage politically consolidated the largest part of the delta within a very short frame extending a little over a century from the late eleventh to the early thirteenth centuries. Besides, two major panegyrics from Bangladesh and three dedicatory records from Paikore in West Bengal (AR-ASI 1921-22: 78-79, Majumdar 1929: 168), Dhaka in Bangladesh (Majumdar 1929: 116-17) and Sanokhar near Bhagalpur in Political History and Administration 194 Bihar (Sircar 1958f: 78-82, 1982b: 154-57, 1983: 123) help trace the geographical horizon of their kingdom. The most dependable source for the study of political history of the Sena, however, remains the famous Prasasti of Vijayasena from Deopara in Rajshahi, coupled with the recent discovery of a stone panegyric dated in the sixth ruling year of Laksmanasena from the village of Bagbari in the Nababganj (former Rajshahi) district of Bangladesh (Haque and Kuddus 2005b). The inscriptions of Vijayasena and his successors report that the Sena lineage started with Virasena, followed by Samantasena and Hemantasena, whose historicity is not directly proved. The first historical monarch of the family was Vijayasena who was a contemporary of the last sovereign Pala king Madanapala.Vijayaís four followers ruled from the middle of the twelfth to the first quarter of the thirteenth centuries (Figure 10). The lineage expressly claims its southern origin and further proclaims its Brahma-Ksatriya identity, suggesting an admixture of Brahmana and Ksatriya blood (Sircar 1982a: 113-15). Vijayasena ruled for at least 62 years and the early years of his reign were spent warring either with the Pala kings or their enemies. The Deopara inscription narrates his campaigns in different directions. The kings specifically named are Nanya, Vira, Raghava and Vardhana, besides the rulers of Gauda, Kamarupa and Kalinga. The Gauda king was obviously Madanapala. If Raghava was son of Chodaganga Anantavarman, the separate mention of the king of Kalinga cannot be explicated. Identity of the king of Kamarupa is problematic. Nanya was the founder of the Mithila branch of the Karnata power. The logical identification of Vira and Vardhana with Viraguna and Dvorapavarddhana of Ramacharitam will again confirm the powerful status of these local polities. The location of his only copperplate signifies that he extended the limits of Pundravarddhana till the extreme southern borders of Western Bengal along the east of the Bhagirathi. His son Vallalasena had acquired some parts of southern Bihar as evident from the discovery of the Sanokhar inscription. The legendary classification of his dominion according to Vallalacharita into Vanga, Varendra, Radha, Baghadi and Mithila has been rejected by Sircar on account of obvious anachronism, but the inclusion of the Baghri region of Murshidabad may not be altogether an interpolated claim, since in one of the inscriptions of his son Laksmanasena (Appendix 1, no. 96), the donated land was located still in the Vyaghratati [mandala]. His son Laksmanasena has a new evidence from Bagbari as already stated, where he is found to have consecrated a tank in Abhira palli or the hamlet of the milkmen and caused the construction of a temple of Visnu, besides recording his victory over the city named Karnapura and a village called Vakasiya, though I have failed to understand the context of the latter. Equally interesting is the mention of Tira bhukti which was a major administrative division in north Bihar under the Pala rule. The inscription was engraved by Sulapani, son of the great artist (silpip-kalpa-vitapi) Brihaspati, implying his identification with the officer (ranaka) of Vijayasena who engraved the Deopara inscription. However, the complete absence of any section dedicated to genealogy of the family is quite curious and surprising. Laksmanasenaís latest known date is the year twenty-seven found in the Bahwal copperplate. His assumption of the epithet Gaudesvara has supplementary support from the claim of erecting victory pillar within the Gahadavala dominion at Banares and also at Puri The Pala-Sena and Others 195 after his victory over Kalinga. The claims of his court poet Umapatidhara on his victory over Pragjyotisha, Kainga, Gauda, Kasi, Magadha, Cedi and the Mleccha seems stereotyped, although the last of these was probably not altoghether out of substance. That their conflict with the Gahadavala kings was heightened during his fatherís reign is shown by a copperplate of Gahadavala Jayachandra dated in c. AD 1175 recording land grant around Patna. The use of the Laksmanasena era of reckoning in private records within the Gahadavala territory and the contiguous Bihar region also lends support to the Sena-Gahadavala struggle in the twelfth century. Laksmanasena was survived by Visvarupasena, after he fled to eastern Bengal having been succumbed to Islamic invasion of AD 1205 and started issuing settlement records from headquarters in villages such as Dharyagrama and Phasphagrama. Visvarupasena issued three copperplates in two of which the name his son Suryasena is found. He was previously mistaken as Kesavasena due to wrong reading of these plates that showed signs of corrections during the later years of his fatherís rule. (Sircar 1982b:). The other son Purusottamasena has no direct record to his credit. These later Sena rulers continued to control eastern Bengal till the first quarter of the thirteenth century, while their fatherís subordinate in southern Bengal named Dommapala has his only record precisely datable in AD 1196 from southwestern West Bengal. ADMINISTRATION IN THE PALA-SENA ERA Any discussion on Pala and Sena administration has to start with the expression Jayaskandhavara found without any exception in all the land grant charters distributed over West Bengal- Bangladesh as well as Bihar. The copperplate inscriptions of the Pala kings refer to the existence of as many as thirteen such ëcamps of victoryí from where the various charters were issued. The meaning of this experession has been explained as ëthe royal camp or capital; epithet of royal camps or residencesí (Sircar 1966: 134). The continuously changing names with a few reappearing in irregular intervals indicate that these were shifting establishments. The explanation behind not locating at a fixed base can only be explained in terms of the continuous battles characterizing their polity right from the beginning till the twelfth century. And this was the root behind their formation of solid military machinery involving the employment of a mixed milieu of ëregular and irregular troops of various localitiesí like Gauda, Malava, Khasa, Kulika and Huna, besides also sometimes Lata and Karnata (Ghosh 2006). At an upper level were the samanta, already discussed in details, who not only served the kings in military raids and occasions of crisis, but also engaged themselves in regularly patronizing religious establishments. In the area of civil administration under a complex design based on local and extra-local levels of administrative tiers, the general scheme of orientation from province (bhukti) to the village (grama) via the various intermediate tiers (like visaya, mandala, khandala or vithi) the hierarchy of whch are often confusing, unlike the Gupta frame of administration which was almost unilateral with elements of reciprocation al lower levels (Chattopadhyaya 1990, for a detailed disacussion). However, the gradual prominence of certain mandala category divisions of administration having essentially toponymic identities or showing floral or faunal characters of the landscape demands a separate focus of study. Political History and Administration 196 The stereotyped list of administrative officres addressed in the copperplates on the occasions of land transfer offers a kaleidoscopic profile of the nature and ramifications of the different administrative departments. Apart from the ranaka (minister), dutaka (executor of a grant) and the rajaputra (prince) who are almost omnipresent in inscriptions, one comes across posts such as maha-kumaramatya (chief minister of a province?), maha-sandhivigrahika (minister of defence), angaraksaka (royal body guards), mahadandanayaka and dharmadhikara (officers of the judiciary), mahapratihara, dandika, dandopasika and dandasakti (broader class of officers of the police department), and rajasthaniya (the broader class of high official). In the area of territorial administration, the names of visayapati, dasagramika, and sasthadhikrta are frequently referred to (for a broader discussion, Majumdar 1971: 273-80, Ghosh 2006). The administrative polity under Sena rule demands a closer examination than they have so far been subjected to. The copperplate inscriptions of the Sena family stand out in more than one significant respect when compared to their preceding or near-contemporary counterparts. A careful scrutiny of the physical constitution of these plates will readily bring home the uniqueness of some of their contents and the patterns of representation of such features. Firstly, for reasons unknown, the composers of the Sena plates traced the genealogy of the ruling king invariably down to the three previous generations and not beyond that in any case. Secondly, the address of the king to the royal officers and other dignitaries include the rajni (i.e. the queen). Finally, the following three points in the grant portion of these documents are noteworthy: i. these inscriptions attract immediate attention because of their ëunprecedented precisioní and contentment about the measurement system of land and the usage of numerous measuring rods for the purpose (Gupta 1996). ii. a section of this set of inscriptions refers to some administrative divisions that either appear for the first time in Bengal inscriptions or conveyó when interpreted in terms of contemporary perceptions of representing politicoadministrative localesññaltogether new and quite significant connotations. iii. in some of the later Sena inscriptions, names of certain jayaskandhavara (i.e. the administrative headquarters) having grama name-ending appear, for the first time in the history of the delta and here again, the pattern of representation of the names of these places is essentially exceptional. If one takes a cursory look at the copperplate inscriptions of the Sena period, out of the twelve charters one each were drafted during the reign of Vijayasena and Vallalasena, seven during the reign of Laksmanasena and the remaining three in that of his son Visvarupasena and his son Suryyasena. A re-examination of some of the details recorded in the grant portion of the Sena copperplates may be undertaken in order of revealing certain aspects of contemporary local political administration and economy. The appearance of new administrative divisions in Sena records was first noticed by D.C. Ganguly, the editor of the Saktipur copperplate, although he did not try to underline the probable nature of such administrative divisions. As D.C. Ganguly rightly observes (Ganguly 1984):36 The inscription furnishes us with the names of some hitherto unknown territorial divisions in Bengal... The relation between khandala and vithi is not known. The division of vithi seems to have been a The Pala-Sena and Others 197 khatika...Between khatika and chaturaka was a unit called vritta. Vritta was divided into chaturaka, chaturaka into grama and grama into pataka. Among all the above-mentioned types of territorial divisions of administration, the three most commonly encountered names found on records belonging to varying geographical niches are those of avritti, chaturaka and pataka, although the last one was not unprecedented in Bengal inscriptions, but assumed an altogether different connotation in later Sena charters of land transfer. Let us start with avritti. Altogether three avritti are known from Sena inscriptions. Arranged in their chronological order of occurrence, these are: the Kantallapura avritti of the Varendra tract in the Paundravarddhana bhukti mentioned in the Madhainagar plate, the Bandana avritti forming directly a part of the Paundravarddhana bhukti mentioned in the Rajabadi plate and the Madhuksiraka avritti within the same bhukti recorded in the Madhyapada plate.37 Available literature dealing in the administrative history of Bengal in the early medieval period has taken avritti as ëan administrative unit (emphasis added) like a Parganaí (Sircar 1965a: 380, 1966: 41) or simply ësub-divisions of a kingdomí, (Chowdhary 1971: 373), though no satisfactory explanation of such a comparative derivation has been offered. Unfortunately, the inscriptions are simply silent about the nature and characters of them. Then what could have been the nature of a avritti level centre of administration? A somewhat logical clue is provided in a work of lexicography called Amarakosa. The text refers, in the section on adjectives, to the term avrita as essentially a general place-name and further suggests that an avrita should be morphologically described as an ëarea covered (or secured) with a moat or ditchesí (Bhattacharya 2001: 313). The only logical inference one can drawóthough tentatively and at the risk of taking the term in its physical senseóis that an avritti (undoubtedly in an extended sense of the term avrita) was plausibly a fortified (or moat protected) nucleus/ core settlement area of a larger administrative division labeled with the same name. Out of the three avritti category administrative centres of the Sena dominion, one was located within the geo-political boundary of Varendra in Paundravardhana province, while the rest were located in the Vanga region. Some of the avritti mentioned above incorporated within their spatial limits, another newly introduced politico-administrative centre called chaturaka. The Vasusri and the Navasamgraha chaturaka were parts respectively of the Bandana and the Madhuksiraka avritti and were located presumably between Sabhar and DhakaBikrampur regions in Bangladesh. 38 This statement does not imply that a chaturaka was exclusively a subdivision of avritti. There are as many as five more instances in the Sena copperplates where a chaturaka is found to form a part of some other category of administrative centre; a different dimension of this fact will concern us later. Thus, the Vetadda, Kantallapura, Kumarapura, Lauhanda, and the Ura chaturaka were not only located in different geographical sectors, but were also not part of any avritti. Discussions chaturaka by some of the ranking authorities have been more than that on avritti. A number of suggestions on the character of this administrative division in Bengal are available in the extant literature. Depending on its occurrence in a Deccanese compendium of the medieval period named Lekhapaddhati, a group of scholars took it as a ëpolice stationí or ëtax for the maintenance of a police station.í39 A Political History and Administration 198 second group of scholars intended to derive the term from the composition of four, depending on the root word chatuh (i.e. ëfourí) from which the term originates. 40 While both of these suggestions are partially correct, a more intensive study of the etymology and the spatial distribution of chaturaka will lead one to reveal some crucially significant attributes of this geographical-administrative centre. It can be reasonably ascertained from a study of etymology of the term that chaturaka possessed a dual identity as an administrative centre. Initially emanating out of the root ëfourí, the term in its extended sense meant the nucleus of a larger spatial entity connected with the provision of security to its residents. Secondly, such chaturaka also acted as tax collectorates (attention may be drawn here to the name Navasamgraha chaturaka, lit. ënew collectorateí of the Madhyapada plate). In order of underlining the genesis of functional characters of the chaturakas, the most crucial clue seems to be hidden in the pattern of their geographical distribution. Firstly, even a casual review of their spatial distribution clearly reveals that each of them was located on or along the actual coastal uplands of deltaic Bengal, where landscape is dominated by the sweeping deposition of secondary alluvium. Govindapur, Bakultala and Saktipur plates record donation of lands both to the east and the west of the river Bhagirathi; the Rajabadi/Bhawal plate refers to land grant in areas around the confluence of the Padma and the Yamuna rivers; all the three chaturaka mentioned in the Madhyapada plate may be located around the Vikramapura bhaga (ësectioní) of Vanga. Secondly, this administrative centre is located at consistently shifting levels in the general ladder of administration, suggesting changes in the spatial limits of this tier at different regional and sub- regional levels. Finally, one of the interesting accounts on the character of landholding elites at chaturaka level is delineated in the Bakultala plate of the time of Laksmanasena. It refers to the creation of a plot of agrahara for a santyagarika category of official in the Mandala grama within the Kantallapura chaturaka and further refers, in relation to the specification of boundaries of the granted land, to the preexisting four agrahara settlements held by four other santyagarika officers (Sanyal 2005). The third tier of regional administration that finds place in some of the later Sena inscriptions is pataka. The earliest epigraphic reference to the word as part of a village settlement appears in the Damodarpur copperplate of Gupta year 224 (AD 543ñ44), which records transfer of 2 kulyavapa of land in three localities of which one was Svacchanda-pataka, the two others having been Arddhati and Lavangasika. Now, if one takes the term pravesya in the phrase svacchandapatake-arddhatipravesyalavangasikayanca, to refer, following D.C. Sircar, to ëthe revenue assessment of a village with anotherí (Sircar 1966: 263), it may be taken to imply that Svacchanda-pataka did not form a village per se and this was possibly the precise reason behind the commission of joint revenue assessment for two rural localities spatially smaller than a grama, by the authorities of the Kotivarsa visaya. Sircarís suggestion following Keilhorn, in that case, that a pataka was actually ëthe outlying portions of a village...which had a name of its own, but really belonged to a larger villageí (Sircar 1966: 242), seems quite reasonable and justified. Pataka, in the above sense of the term, continued to exist in parts of eastern and western Bengal throughout the sixth-seventh centuries. But it is quite curious to note that the practice of using pataka as a specific unit of land measurement, having The Pala-Sena and Others 199 varying denominations with changes in time and space, also continued simultaneously. 41 However, it is very difficult at the present state of information to underline the exact relationship between pataka as a rural settlement locate and that as a specific measure of land, particularly because there are epigraphic evidences to the reference of pataka having both the connotations within the limits of a specific geographical area at a given point of time. Precisely from the second half of the seventh century pataka as a rural settlement unit smaller than grama ceased to exist in epigraphic recordsóthe last known example being the Ashrafpur copperplate of Devakhadga datable to c. AD 670-71 where Tala pataka, Markatasi pataka and Dara pataka are mentioned as parts of the granted land42ññto reappear after more than five hundred years in the Saktipur copperplate of Laksmanasena, where Raghavahatta, Varahakona, Vallihita, Vijaharupura, Damaravada and Nima pataka of the Kumarapura chaturaka are said to have been granted in the Uttara Radha territory. Three other inscriptions record the donation of pataka. These are: the Madhainagar plate that mentions Dapaniya pataka having as its southern and western boundaries two other pataka named Cadaspasa and Gundisthira, ëin the direction ofí Kantapura avritti, the Idilpur plate recording grant of land in the Talapada pataka in Vanga and the Madhyapada plate where one comes across the existence of Ramasiddhi, Ajikula and Ghagharakatti pataka, all possibly located within the periphery of the Navya sector of Vanga. It is interesting to note that by the time these inscriptions were drafted, early medieval texts of lexicography had already precisely defined a pataka in spatial terms. Thus, the Abhidhanachintamani of Hemachandra prescribes, in connection with the spatial features of a grama, that a pataka is patastutadarddhe-syat, i.e. ëhalf of a villageí (Chowdhary 1971: 48, 80). Therefore, it can be fairly surmised that unlike pataka of sixthseventh century, which necessarily formed a hamlet or an ëoutlying territoryí of a larger village, the pataka recorded in the Sena inscriptions were gradually being recognized as prominent and independent rural settlement units replacing, in select areas, grama that formed so far the smallest territorial unit in the general hierarchy of administration. Possibly in compatibility to this new connotation of a pataka, a whole new and complicated system of land measurement involving various denominations of smaller units was evolved by the Sena rulers in areas where the existence of pataka as a settlement unit was in vogue. The final manifestation of of this development can possibly be viewed in the Bhatera copperplate of Govinda-Kesavadeva that refers to Bhatta pataka as an independent settlement locality of Srihatta, where 375 hala of land was donated; this Bhatta pataka is unquestionably identifiable with modern Bhtapara, still a major settlement locality in Sylhet. Some of the pataka freshly emerging in the Sena inscriptions deserve special attention, as their names would demand. Such a name as Ramasiddhi pataka, implying evidently a personal name-prefix, might suggest that the person after whom the pataka was named must have been a man of prominence in the concerned locality. Further, the name Raghavahatta pataka (hatta meaning a ëmarketí) located in close proximity to five other pataka one of which is specified as a pura (viz. Vijaharupura), is obviously indicative of the fact that this pataka category settlement had a domineering role to play in the development of local commercial set up within the Madhugiri Political History and Administration 200 mandala. These occurrences are admittedly inconsistent patterns, but when and where they are found to exist, their impact on the neighbouring rural settlement structure is undeniable. Apart from the interesting accounts on some so far less discussed names of administrative and geo-political divisions, a second category of evidence from the same grant portion of all the later Sena inscriptions may be highlighted through a review of the pattern in which the administrative headquarter or the more commonly known ëcamp of victoryí (jayaskandhavara) is represented in epigraphic sources. From the time of the Pala rule onwards the names of royal administrative headquarters, officially responsible for the issuance of legal documents, are represented in the grant portion of the copperplates in the form of a conventional sentence that starts with phrase Iha/Sa kahlu, followed by the name of the concerned headquarter (which is invariably a pura or a nagara), again succeeded by the phrase samavasita srimaj=jayashakndhavarat. The only exception to this is found in the three so far discovered copperplates of the Pala monarch Madanapala, issued from the jayaskandhavara located at Ramavati and hailing from Manahali and Rajibpur. I shall return to the exact point of exception later. All the early copperplate inscriptions of the Sena rulersññchronologically ranging between the Barrackpur plate of Vijayasenaís sixtysecond year and the Saktipur plate of Laksmanasenaís sixth yearññwere issued form their office at Vikramapura; the jayaskandhavara in all these inscriptions is, as usual, represented through the stereotypical sentence noted earlier. But it is quite intriguing to notice that in four out of the five Sena copperplates datable to the first quarter of the thirteenth century, a crucially significant variation in the pattern of representation of the jayaskandhavara is found, although apparently they seem to belong to the larger homogeneous group of Sena epigraphs. Thus, a review of the sentence in the Madhainagar, Rajabadi, Idilpur and the Madanpada plates referring to the jayaskandhavara bears the following morphology: Iha/Sa khalu phasphagrama/ dharyyagrama parisara samavasita srimajjayaskandhavarat. One can readily bring home two points from the above citation: firstly, instead of earlier pura or nagara types, the administrative headquarters mentioned in all of these inscriptions are grama; secondly, the names of royal offices in these plates are invariably suffixed with the unusual expression parisara. It is evident, from the regular appearance of the term in a particular set of records within a given time span, that its reference was neither normative nor arbitrary. The phrase in question has been translated as ëhere from the camp[s] of victory situated in Phasphagrama/ Dharyyagrama.í The apparently conventional sentence that contains this termññ if one contextualizes the word with the manner of its expressionññmay be better translated as ëhere from the camp[s] of victory situated within [the confined limits of] Phasphagrama/ Dharyyagrama.í43 The question that immediately strikes attention is obviously related to the contention behind such an unusual specification in the representation of the administrative headquarters. In order of having a justifiable answer, one may now recall the exceptional representation of the jayaskandhavara in the copperplates of Madanapala. Apart from the Sena inscriptions referred to above, the land grant charters of this last paramount Pala king are the only epigraphic documents of early The Pala-Sena and Others 201 medieval Bengal that referred to the ëcapitalí as parisara samavasita. These are also the only known Pala inscriptions issued from the well known city of Ramavati. If one takes a look at the description of the city during the reign of Ramapala, as gauged from the Ramacharitam of Sandhyakaranandin, the narrative is clearly that of a flourishing urban centre.44 But since we do not have a single copperplate inscription of this ruler so far known, it is difficult to speculate about its status as a ëcamp of victoryí in the eleventh century. And when it finally figures in the inscriptional corpus around the middle of the twelfth century, it is precisely and repeatedly described as an essentially ëenclosedí place. In the first place, it is fascinating to notice that the largest concentration of these regional and sub-regional centres of political administration like avrittis and chaturakas is found in the Navya (lit. ënavigableí) sector of Vanga and the Khadi areas contiguous to the estuarine mouth of the Bay of Bengal. Thus, they were mostly located in regions that formed a potential buffer zone in the context of wider Indian Ocean exchange networks along the littoral areas, evidently through the Samatatabased linkages of trade. 45 Besides, sporadic concentration of such territorial divisions is also occasionally found in the central Bhagirathi delta having sweeping distribution of older alluvium deposits fit for expansion of sedentary agrarian settlements. It is interesting to note that altogether eight agrahara plots in two presumably contiguous sets within the spatial orbits of four chaturakas in the Khadi and the Navya territories, find mention in the Bakultala and the Madhyapada plates. Now, the composite landmass around such hydrogaphically and geologically dynamic spaces (viz. Navya and Khadi) is highly suitable for the expansion of agrarian settlements. 46 Therefore, increasing concentration of newly created ërent-freeí settlements in these active alluvial plains must have had a close bearing on the broader rural agrarian set up of the coastal uplands. Further, it has been suggested quite rightly that such concentrations as reflected in the case of the Bakultala plate are indicativeññthough not in any spatio-temporally uniform mannerññof certain interactive relationships between the secular landholders at the village level and the agrahara holding religious elements at the immediately superior (i.e. chaturaka) level (Chattopadhyaya 1990: 36-7). Secondly, gradual prominence of certain pataka as administrative divisions in select geographical terrains more suitable for spread of agriculture and consolidation of economic affairs through trade linkagesó sometimes having close bearing on the rural settlement structures or on organization of local economy (as evident from names like Ramasiddhi pataka and Raghavahatta pataka)ó obviously reflects an intense trend in regional and sub-regional geo-political growth. This development of pataka as a settlement locale vis-‡-vis revenue unit spatially distinguishable from a grama was in compatibility to a highly advanced and intricate system of landmeasurement and revenue assessment in which several newer and smaller units of measurement bhu-khadi, khatika, kakini, and udana (i.e. unmana) were freshly implemented in areas where such geopolitical growths can be effectively implemented. Furthermore, spatial distribution of these geopolitical centres of local administration would reveal that they were created in some deliberately chosen geographical niches either highly fertile for agriculture and sedentary growth or essentially vibrant in the context of local and extra-local levels of trading linkages. Thus, it will be Political History and Administration 202 difficult to visualize these geo-political centres as omnipresent ëunitsí like visaya and mandala; rather they acted as sub-regional organizations, inter alia, of economic and political administration in select geo-hydrographic terrains of early mediaeval Bengal. It is again interesting to notice that such a purposefully geared regionalization in the geopolity was getting strengthened in a period that had just witnessed the attacks in and subsequent loss of the northern parts of the Sena dominion under the Islamic powers in AD 1205, as already suggested. Possibly such regional and local political-administrative and economic developments coupled with a complex and efficient machinery for revenue administration in the eastern segmentññthat became the centre of Sena polity in the first quarter of the thirteenth centuryññhad contributed to the fact that this ruling family continued to survive for more than two decades in spite of the sudden and powerful thrust from alien groups. Finally, it is quite crucially significant to find that at the later years of both Pala and the Sena dominions their administrative headquarters are described in their legal documents in a completely different and unusual manner. Deliberate use of the term parisara in expressing the physical character of the offices, which were essentially villages during the later Senas rule, is possibly indicative of a conscious proclamation of the local administrative authorities, as one finds an identical precedence in the later years of Pala rule in north Bengal, that the most important administrative offices responsible for the final endorsement and issuance of legal records are still located with a ëconfinedí area under the direct intervention of the central political authorities. CONTEMPORARY LOCAL LINEAGES: POLITIES BEYOND THE PALA-SENA ORBIT Apart from the two principal dynastic frames that formed the principal subject of inquiry of this discussion, a number of other independent local lineages figured in the broader political atlas of early medieval Bengal contemporary to the Pala and the Sena. The most prominent and politically sustained existence was that of the Chandra dynasty whose origin is unknown except for an interpolated statement by Taranatha that one Lalitachandra ruled in Bhangala before the risae of the Pala dynastyññinformation that can prove nothing on their origin. The genealogy preserved in the copperplates of rulers having unimpeachable historicity draws therir descent from one Purnachandra followed by Suvarnachandra and Trailokyachandra. The historical lineage starts with Srichandra followed by his descendants Kalyanachandra, Ladahachandra and Govindachandra, who ruled from the the early tenth to the mid eleventh century in unilineal succession. The family was initially Buddhist in affiliation while Ladahachandra was a Vaisnava. While their political relations with the Pala has been discussed, it will be relevant to reiterate the record that though their original base was at Vikramapura, Srichandra during his half-acentury rule made his strong presence felt in the southeastern borders. Two inscriptions of the time of Srichandra have recently been published, but no new information is contained in them. The line developed by Kunjaraghatavarsa probably died out without any following, while a second Kamboja line is known form epigraphic records of southwestern Bengal. This second ine had its base at a place called Priyangu in Odisha. Two inscriptions from Irda and Kalanda of the time of Nayapala in the eleventh century are The Pala-Sena and Others 203 known and they record donation within Dandabhuki, but it had by then lost its provincial identity and got transformed into a mandala. A number of local lineages across the geopolitical units of Vanga and Samatata ruled in the concerned period. The Varma family staing with Harivarma via his brother Samalavarma to Bhojavarma ruled for century in the eleventh-twelfth centuries. The original progenitor Vajravarma and his successor Jatavarma have no records. An inscription of a minister of Harivarma named Bhatta Bhavadeva was found from Odisha, though its original locale was probably the Dhaka region (Sircar 1983: 105-11). The statement that Bhavadevaís grandfather was a minister of Vangesvara will necessitate the inclusion of the the preecessors in the list of historical kings of the lineage.A second Deva family of Samatata-Harikela regions having a number of branches ruled between the late eleventh and the mid twelfth centuries. Inscriptional evidences of a number of these kings of this lineage like Ranavankamalla Harikaladeva, Damodaradeva, Viradharadeva, Dasarathadeva, GovindaKesavadeva and Isanadeva are known. That this lineage could retian its political identity long after the decline of the Sena polity is reflected in the precise dates available in their records. The many subordinate lineages in southern Bihar have already been taken notice of. The brief outline of a chronologically wider horizon covering nearly a millennium will demand one final observation. The Pala ruled for more that four hundred year radiating enormous influence on the norethern Indian polity as the principal power of Bengal. All the rulers starting from Dharmapala to Madanapala enjoyed the epithet Gaudesvara in its connotation per se, but the sources so far revisited have shown amply that at not a single point of their entire political existence, thay were able to bring the entire delta under one umbrella. A broad review of the spatial distribution of the Sena inscriptions shows a range, on the other hand, from Sundarban in the south to Madhainagar in the north and from Bhawal in the east to Paikore in the west, even after keeping aside the Sanokhar inscription of Vallalasena discovered from Bhagalpur in Bihar. This is a geographical stretch of eastern India that not a single Pala king has ever claimed to have held sway over in the entire span of their political career. NOTES 1. Dilip K. Chakrabarti rightly advocates for this inherent fluidity of Bengalís geographical and historical ësub-regionalí identities and explains this mechanism with the example of the northern province of Pundravarddhana bhukti. It will not be out of place here to put on record his remarks on the problem of Morrisonís methodology (Chakrabarti 2001: 18): ì[H]is readiness to accept the four sub-regions of the delta as rigid historical entities regrettably lessens the value of this otherwise useful study. If one looks beyond property transfers, one realises that the history of the Bengal delta has not evolved within the framework of rigid geographical units defined by administrative terms in inscriptions. Historicallyññ and probably geographically tooññthe lines of these subregions were always fluid. Let it not be forgotten that the Bengal delta itself was a unit; one cannot fix properly the date when the present Padma mouth came into existence. Further, the Brahmaputra once joined the Meghna, the literary antiquity of the latter being uncertain. The shape of the present Bhagirathi mouth also was not the same in the ancient context. The present Hooghly channel is comparatively recent and thus the present configuration of the Rupnarayan flow was also different in the ancient context.î Political History and Administration 204 2. Since the provenance of the plate, located about 30km northwest of present Comilla town, itself was an ancient agrahara (Gunikagrahara, i.e. modern Gunaighar), it can be safely surmised that this Vainyagupta exercised some political dominance in southeastern Bengal. And such a direct line of political control is not altogether unlikely as the territory of Samatata formed one of the frontier principalities of his predecessor Samudragupta as recorded in the Allahabad eulogy, although one can hardly be certain about the date of annexation of this sector by Vainyagupta (Sircar 1985b: 105). 3. It may be relevant here to note that a recent interpretation of the term padanudhyata in the light of exhaustive epigraphic and literary sources shows that it was more an honorific than an indicator of subordination, at least before the tenth century, when it came to be used, though not sweepingly, in its physical sense meaning ëmeditating on the feet ofí implying a hierarchy of relationship between polities of different levels (Ferrier and Törzsök 2008). 4. D.C. Sircar was aware of only the last three rulers and allotted 65 years from AD 535 to 600 for the group (Sircar 1985b: 112). If Dvadasaditya is included as the first member of the group succeeded by Dharmaditya, Gopachandra and Samacharadeva, there is a distinct possibility of this chronological bracket being dragged roughly a decade back. Further, the discovery of the plate of Dvadasaditya has now proved that he was a separate ruler of the Kotalipara group and not identifiable with Maharajadhiraja Vainyagupta Dvadasaditya of coins (Islam 2011). 5. Sircarís suggestion that the use of ruling year instead of the more easily recognizable Gupta era in Gopachandraís records suggests that he was a rebel samanta of the royal Gupta lineage in Bengal is equally curious and unconvincing (Sircar 1985b: 109). The Sailodbhava king Madhavavarman II was a feudatory of Sasanka quite sometime after the end of Gupta rule in eastern India and yet he is found to have used the Gupta reckoning in his well known Ganjam copperplate (cf. Hultzsch 1981a). 6. The decipherer of the inscriptions reports that out of the three copperplates, two were engaved and the third was a blank with a seal bearing the name of the king, evidently, kept ready for a future donation; of the two engraved charters, one is badly corroded and beyond decipherment (Islam 2012: 61). 7. Sircar read in the photo of one these inscriptions the name of Rajaputra Balabhata having the epithet paramamahesvara mata-pitrpadanudhyata rajaputra, implying that Balabhata was a Saiva (Sircar op. cit. 148). One may note in this connection that the seal of the first Ashrafpur plate depicts the figure of a bull instead of the usual dharmachakra emblem of Buddhist rulers, although Rajaraja was a Buddhist. 8. The recent publication of a number of gold and silver issues of two kings named Prithubhata and Sarvanada has led Shariful Islam to incorporate these kings in the genealogy of the Khadga lineage, but the chronological status of these kings remains unknown (Islam 2008). 9. B.N. Mukherjee has identified the early medieval Harikela/Harikela in the light of exhaustive epigraphic and numismatic evidences with the present Chittagong tracts of Bangladesh (Mukherjee 1975). 10. Bhattacharya records that (Bhattacharya 2000: 473) ìThe Chittagong area over which he ruled is described as Khasha-maka [maka perhaps meaning ëdominioní] and his officers are described as asesha-Khasha-mak-adhikarana. This is perhaps the earliest epigraphist reference to the Khashas from eastern India...The Bhagavatapurana, 2.4.18 refers to the Khasas [sic] along with the Kiratas, Hunas, Andhras, Pulindas, Pukkasas, Abhiras, Suhmas and Yavanas who are called papas, sinners, and who were purified by embracing Vaishnavism. Our ruler, Devatideva was a Khasha king of the Chittagong area (ancient Harikela) who was a Buiddhist and not a Vaishnava.î The Pala-Sena and Others 205 D.C. Sircar recovered from epigraphic and literary references that the original territory of the Khasa tribe was apparently the Kumaun-Garhwal region (Sircar 1972-73: 50-1). 11. D.C. Sircarís estimate of Sasankaís eastern dominion extending even into the Vanga and Samatata sectors of eastern and southeastern Bengal, however, seems overemphasized (Sircar 1972-73: 116). 12. The name and location of the provenance of this copperplate and many others in Bengal have been matters of debate and confusion. It has recently been recovered that the village ëMalliaí reported by Lionel D. Barnett was actually Maliadanga in the Sagardighi Police Station of the Murshidabad district and the grant took place in areas along the border of this district with Birbhum to its west. The exercise of recovery of provenances of copperplate inscriptions from West Bengal has shown that proper location of find-spots and ancient settlements may provide a sound plank for the archaeological study of early medieval settlements (Sanyal 2010a). 13. One has to note here that the comparative chronologies of the Maukhari, Later Gupta and Gauda kings is an issue of considerable confusion. D.C. Sircarís suggestion that Grahavarman, Sasanka and Devagupta were contemporaries, when he himself places Devagupta at the end of the seventh century as a successor of Adityasena, suffers from obvious self-contradiction. There is no doubt from the mention of this king in Banaís composition and Harsaís inscriptions as the Malava king defeated by Rajyavardhana, elder brother of Sasanka, this Devagupta has to have been a near contemporary (or more precisely a predecessor) of Kumaragupta II of Malwa (c. AD 600-30), son and successor of Mahasenagupta (Thapliyal 1985: 42, Agarwal 2003; cf. Sircar 1985b: 192-3). 14. D.C. Sircar referred to this inscription rightly summarizing its contents, though it is not known if his reading was published in the Bangladesh Lalitkala as Sircarís list of forthcoming publication bearing a citation under ëA Matha of the Vaishnava Parivrajakasí indicates; there is no reference to this piece in Bhattacharyaís publication. Sircarís summary reads (Sircar 1982a: 45): ìWe come to know from an inscription discovered from Pabna in north Bengal that Pa¤cahuti was the lord of Bhattala mandala. Karkkaraja born in his lineage was the hero of Bhattala desa and was an associate/ friend of Yasovarman His grandson Pahila who served the kingdom by Devapala.î Gouriswar Bhattacharya demands that the inscription proves that Yasovarman was in occupation of northern Bengal for at least some time where Pahila donated a cloister (matha) to the Vaisnava saints more than a century later (Bhattacharya 2000:182-3). 15. Susan L. Huntington has tried to assign an image of seated Buddha from Bihar to the reign of Vigrahapala I based on her analysis of the stylistic features of the piece. 16. Sircar later revised his chronology after the discovery of the inscribed image of year 12 of Surapala I (Sircar 1982a: 174-5) and another image of year 37 of Rajyapala II (Sircar 1985b: 177-9). Thus the final version of his chronology is yet to be published. 17. The name ëTrilocanapalaí (in the place of Tribhuvanapala) appearing in Huntingtonís table is certainly a misprint. 18. This is an inscription where the statements of the imprecatory verses are illustrated as a sculptured composition at the bottom of the slab containing the inscription. Only a few examples of this type of epigraphic documents are known from central India as the first decipherer of the epigraph demands and in Bengal this is the sole specimen of its kind. 19. The date of accession of Dharmapala given in Bhattacharyaís publication as c. AD 765 is obviously a misprint as otherwise this suggests that he ruled conjointly with his father for the first one decade of his ruleñña custom never known in any period of Bengalís early history. Political History and Administration 206 20. A physical examination of the piece clearly reveals that the digit in the unit place is 2 and not 3. 21. However, D.C. Sircar reports that a Tibetan king named Mu-tig Btsan-po claims to have defeated Dharmapala between c. AD 804 and 815 (Sircar 1982a: 62), though in the absence of any reliable corroborative evidence this is difficult to believe in. 22. Dilip K. Chakrabarti differs from Sircar regarding the identity of the Yadu territory which he thinks ëneed not suggest Gujarat; some parts of Punjab or Mathura are a better probabilityí (Chakrabarti 2010: 98). 23. A Jain text called Harivamsa composed in Gujarat in the late eighth century suggested that this country was bounded by Avanti (western Malwa) to the east and the kingdom of Indrayudha to the north. A second reference is available in a passage of Rajasekharaís Karpuramanjari that reads ëSagaradatta went ot Kanyakubja, the capital of the illustrious Vajrayudha, king of Pa¤cala.í These statements have been taken by Sircar to argue that if the Ayudha kingdom had not extended considerably to the west of Kanauj to incorporate the contiguous territories of Rajasthan and Panjab, the reference to this kingdom with other ëmighty monarchsí of the time cannot be justified. 24. It has already been argued elsewhere that (Sanyal 2009a: 313): ìthe Siyan stone inscription of Nayapala refers to the construction of a vadabhi for the goddess ëCarccikaí by a king named Mahendrapala (Sircar 1982a: 48). It will be pointless now to criticize Sircar for identifying this Mahendrapala with ëprobably the Gurjara-Pratihara monarch of that nameí, quite in keeping with the historiography based on the then known sources related to the Pala-GP struggle. One question that never received attention by either Sircar or other scholars, however, is why a Prasasti type of inscription of a Pala king would eulogize the royal patronage of a rival Pratihara monarch towards the construction of temples simultaneously with many other Pala rulers? ì 25. The Bargaon plate of Ratnapala claims that this king defeated the contemporary Gaudendra without any specific reference to his name (Sharma 1978: 157, 163). Nayanjot Lahiri rightly identifies this Gauda king with either Mahipala I or his son Nayapala (Lahiri 1991: 83). The Gachtal plate of Gopalavarman makes a more clear statement by referring to the victory of Ratnapala precisely against Gaudaraja Rajayapala (Sharma 1978: 214, 218). But the identification of this Rajyapala has been a subject of controversy. Ratnapala is dated between c. AD 1010 and 1050 by Hoernle (Hoernle 1898). D.C. Sircar seems to have accepted the statements made in the Gachtal inscription (Sircar 1982a: 77). But since Rajyapala II, son of Narayanapala, is dated in the first half of the ninth century, P.C. Choudhury (who places Ratnapala between c. AD 1010 and 1040) found it convenient to identify this Rajyapala with the Kamboja king of Priyangu, predecessor of Nayapala, ruling in the southwestern part of the delta. More recently Dimbeswar Sarma has reconsidered the chronology of the Pala lineage of Assam and rightly aregued that Ratnapala should be dated between c. AD 930 and 980 since his son Indrapala was a contemporary of Trailokyachandra ruling in eastern Bengal from c. AD 975 to 1000. But he has unfortunaly equated Rajyapala II of the Pala family with his namesake of the Priyangu seat (Sarma 2003: 103-9). While his review of the chronology of Ratnapala is quite juastifiable, the absurdity of his contention in equating Rajyapala of two different polities demands no explication. Suchandra Ghosh has categoricaly and rightly located the defeat of Rajyapala II of the Pala dynasty by Ratnapala in the broader atlas of BengalAssam power struggle in the early medieval period (Ghosh 2011b). 26. Mahipala I could not have annexed the southwestern delta taking advantage of the Chola invasion because firstly there is no evidence for such an event and secondly, the assumption will imply that the Pala king had enough time to subdue the territory in the interval provided by Rajendraís strikes in Vangala and northern radha, the latter The Pala-Sena and Others 207 shaking Mahipala I himself along with his subordinate Sangu. Thus Sircarís argument seems unreasonable. by Soma, depended on the support of the strong arms of his obedient cousins, viz. the sons of his maternal uncle (Mahana).î 27. Shastri further equates this Pala king with the Mahipala of the drama Chandakausikam composed by Ksemisvara in the twelfth century (Shastri 1935-36). However, this assumption has been rightly rejected by Sircar on the ground that the Karnnata referred to in this composition must have been the RK and in that case the Mahipala of Chandakausikam must have been the GP king of that name. 32. Extensive architectural activities in brick found during my exploration of both the areas around Garh Mandaran and Dumni clearly show that there was steady growth of settlements in the eleventhtwelfth century horizon in these regions. Detailed archaeological studies of these areas are expected to throw welcome light on the pattern of early medieval settlements in the region. 28. These are the 1. Gaya Krishnadwarika temple inscription of year fifteen (Chakravarti 1900, Maitra 2004: 110-20, Sircar 1967b); 2. Gaya Narasimha temple inscription of the same year (Banerji 1915: 78); 3. Gaya Gadadhar temple inscription (Banerji op. cit., Sircar op. cit.); 4. Siyan inscription (Sircar 1982b: 102-22) and 5. Bangarh inscription of Murtisiava (Sircar 1982b: 85-101). 29. Both the Bhera-Ghat and the Rewah stone inscriptions offer reliable demands on Karnaís victory over Bengal (Keilhorn 1894b, Mirashi 1942). 30. D.C. Sircar did not, however, accept S.K Saraswatiís arguments and suggested that ëconsidering the rule of Mahipala I in AD 1026 and the coronation of Madanapala in 1143, it is a lesser probability to think of Mahipalaís reign for five years (Sircar 11982b: 87). 31. The names of these rulers in the text of Sandhyakara appears in the following manner (Sastri 1969: 37): ìHe (Ramapala) was competent to conquer the earth, having been joined by those (warriors) having large armies, viz. Vandya (Bhimayasas), Guna (Viraguna), Simha (Jaasimha), Vikrama (Vikramaraja), Sura (Laksmisura, as also Surapala), Sikhara (Rudrasikhara), Bhaskara (Mayagalasimha) and Pratapa (Pratapasimha)... Ramapala, who enhanced thew prosperity of Arjuna (Narasimharjuna, as also Chandarjuna) and Vijaya (Vijayaraja) coming to him (as allies), who solicited the help of Vardhana (Dvorapavardhana), and who had (other) associates (samantas) headed 33. It is certain that the statements made in the text called Kalingattu Parani was a flight of panegyry, as Sircar also categorically suggests (Sircar 1982a: 91). However, repeated raid of Kulottunga Chola in Kalinga is s historical reality. 34. The inclusion of these rulers will also imply that the king Bhimapala referred to in the text Sabdapradipa was a member of the pala dynasty, inasmuch as his chronological closeness to Ramapala is quite apparent from the comtext in which he is mentioned in the text (Sircar 1982a: 91). Therefore, such statements demand sound corroborative evidence. 35. Richard Salomon cautions that ëa single instance of a portable instance found outside the known limits of a rulerís realm should not be automatically accepted as evidence of his vast conquests, but a consistent pattern of this may be historicallyí (Salomon 1998: 229). 36. The observation of D.C. Ganguly is also echoed by Niharranjan Ray who remarks on the administrative system during the Sena rule that (Ray 1994: 285) ññ[i]n at least one case the division between mandala was the khandala; elsewhere the vithi followed the mandala. In one case we see that after the mandala there was the chaturaka, which elsewhere was a division of an avritti, but we cannot ascertain of what an avritti was a division, although it is not entirely unlikely that it was a subdivision of a mandala. 37. It may be noted here that the only Pala inscription that records the existence of a avritti is the Rajibpur Political History and Administration 208 plate of year 2 of Madanapala that locates the village Budhavada in the Srigatika avritti of the Kotivarsa visaya (Mukherjee, 1990-91). 38. Ranabir Chakravarti has shown that from the mid tenth century the Navya and the Yola territories of early mediaeval Vanga played a major role in the context of Indian Ocean trading networks along the eastern littorals (Chakravarti 1996). Now, if the port of Vangasagara-sambhandariyaka was located within the orbit of the Yola mandala and the Vasusri and Navasamgraha chaturaka formed part of Navya (that also formed earlier a mandala under the Chandra dominion) in the Vanga territory, there is every reason to believe that these chaturakas were located within the orbit of these territories. 39. D.C. Sircar (Sircar 1966: 68-9) subscribes to the view of the editors of Lekhapaddhati (Dalal and Srigondekar 1925: 9) that a chaturaka was a ëpolice stationí or ëtax for the maintenance of a police station.í 40. N.R. Ray took chaturaka as a combination of four villages, deriving it evidently from the root of four (Ray 1994: 285). Brajadulal Chattopadhyaya seems to hold a somewhat similar view and believes further that the name Kantallapura chaturaka mentioned in the Bakultala plate was a ëvillageí that headed a group of four villages Chattopadhyaya 1990: 36). Ranabir Chakravarti rightly hunches at its derivation from the root ëchatuhí and further suggests that this place-name had an association ëwith a place (neither a village nor a town) where converged roads from different directionsí (Chakravarti 2002b). 41. The earliest reference to pataka as a land measurement unit is found in the Gunaighar copperplate of the time of Vainyagupta. For an impressive study on the changing denominations of pataka in different parts of early Bengal, refer Rita Ghosh-Ray (Ghosh-Ray 1993) and Chitrarekha Gupta (Gupta 1996). Evidences clearly suggest that the existence of pataka as a measure of land does not seem to have been at stake at any point of time, the last known evidence of the kind being attested in the Rajabadi copperplate of Laksmanasena. 42. It is curious to note that the same inscription also refers to the use of pataka as a land measuring unit. Names of villages with padraka/padrika/ padrika name-ending also occur in inscriptions of seventh century in southwestern west Bengal, for example Kumbharapadraka of the Antla plate and Kaparddipadrika of the Egra plate of the time of Sasanka (for a discussion, Sircar 1965a: 376-77). 43. The term parisara may also be taken to mean ëcontiguousí or ëattached toí; in that case it would imply that the actual place of issuance of the charter was close to the village mentioned. From the context of its occurrence of the term, however, ëan enclosed or fenced areaí seems more probable (Monier-Williams 2002: 604). 44. Ramacaricam describes Ramavati of Ramapalaís time as the city ëcarrying an immense mass of gemsí, ëthe city of gods and wealthy residentsí, the city having ëa series of lofty temples of godsí and a city ëfit to be enjoyed by Kubera, the lord of Yaksas and was excessively rich on account of its Sevadhií (Sastri 1969: 72, 78). 45. A recent archaeological study on the early mediaeval settlements in the Sabhar region is a further indication of such local linkages between eastern and southeastern territories of Bengal (Khan, Majid and Rahaman 2004). An intensive study of archaeological material continuously hailing from the present Khari regions of southeastern West Bengal may throw considerable light on the commercial linkages of this territory with areas farther east-southeast in the concerned period of time. 46. In deltaic formations, soil in coastal alluvial tracts around navigable water bodies and estuaries is often highly fertile for cultivation (Singh 1993: 28687). The Pala-Sena and Others 209 REFERENCES Agrawal, A. 2003. A New Copperplate Grant of Harsavardhana from the Punjab, Year 8. Bulletin of the School of Oriental and African Studies 66(2): 220-8. óñ. 2000. Essays on Buddhist Hindu Jain Iconography & Epigraphy. Studies in Bengal Art Series: No. 1, pp. 471-80, Dhaka: The International Centre for Study of Bengal Art. Allinger, E. 2006 (December 27). Astasahasrika Prajnaparamita Manuscript from the Yarlung Museum, Tsethang, TAR. www.asianart.com/ articles/ allinger/index.html. Bhattacharya, S.C. 2007. The Jagjibanpur Plate of Mahendrapala. Journal of Ancient Indian History 23 (2005-06): 61-125. Altekar, A.S. 1934. The Rastrakutas and their Times. Pune: Oriental Book Agency. Bhattasali, N.K. 1918. Some Image Inscriptions from East Bengal. Dacca Review O Sammilan 7(11-12): 28391. Banerji, R.D. 1915. The Palas of Bengal. Memoir of the Asiatic Society of Bengal. Vol. 3 Calcutta: The Asiatic Society. óñ. 1983a (reprint).The Ghugrahati Copperplate Inscription of Samachara Deva. Epigraphia Indica 18 (1925-26): 74-86. óñ. 1918. The Pratihara Occupation of Magadha. Indian Antiquary 47: 109-11. óñ. 1983b (reprint). The Kedarpur Plate of SriChandradeva. Epigraphia Indica 17 (1923-24): 188-92. óñ. 1987 (reprint). Bangalar Itihas. Vol. I (i.e. History of Bengal , in Bengali). Calcutta: Deyís Publishing. Bautze-Picron, Claudine. 1999. Buddhist Painting during the Reign of Harivarmadeva (end of 11th century) in Southeast Bangladesh. Journal of Bengal art 4: 159-97. Beal, S. 1981 (reprint). Si-Yu-Ki: Buddhist Record Records of the Western World. Delhi: Motilal Banarsidas. Bhandarkar, D.R. 1982 (reprint). Chatsu Inscription of Baladitya. Epigraphia Indica 12 (1913-14): 10-7. Bhattacharya, G. 1991. The Munificence of Lady Catuhsama. In Gouriswar Bhattacharya (ed.), Aksayanivi: Essays presented to Dr. Debala Mitra in Admiration of the Scholarly Contributions. Pp. 313-22 (=2000: 281-91). Delhi: Sri Satguru Publications. óñ. 1997. Bangladesh National Museum Prasasti of Pahila (9th century AD). Journal of Bengal Art 2: 111-20 (= 2000: 481-7). óñ. 1998. Genealogy of the Pala Dynasty, 8th-12th c. In Claudine Bautze-Picron. Art of Eastern India in the Collection of the Museum f¸r Indische Kunst (Inscriptions by Gouriswar Bhattacharya). Berlin: Dietrich Reimer Verlag 123. Bühler, G. 1892. The Madhuban Copperplate of Harsha, Dated Samvat 25. Epigraphia Indica 1: 67-75. Chakrabarti, Amita. 1991. History of Bengal (c. 550750 AD). Burdwan: University of Burdwan. Chakrabarti, D.K. 2010. The Geopolitical Orbits of Ancient India: The Geographical Frames of Ancient Indian Dynasties. New Delhi: Oxford University Press. Chakravarti, R.K. 1999 (reprint.). Gauder Itihas (i.e. History of Gauda, in Bengali). Calcutta: Deyís Publishing. Chakravarti, M.M. 1900. An Important Inscription of the Time of Nayapala from Krishnadwarika Temple at Gaya. Journal of the Asiatic Society of Bengal 69(1): 190-5. Chakravarti, R. 1996. Vangasagara-sambhandariyaka: A Riverine Trade-centre of Early Medieval Bengal. In Debala Mitra (ed.). Explorations in Art and Archaeology of South Asia. Calcutta; Directorate of Archaeology and Museums, Government of West Bengal, pp. 557-72. óñ. 2002a. Trade and Traders in Early Indian Society. New delhi: Manohar. Political History and Administration 210 óñ. 2002b. Between Villages and Cities: Linkages of Trade in Early India (c. AD 600-1300). In Georg Berkemar et al. (eds.), Explorations in the History of South Asia: Essays in Honour of Dietmar Rothermund. New Delhi: Manohar. Fleet, J.F. 1882. Sanskrit and Old-Canarese Inscriptions No. CXXV. Indian Antiquary 11: 156-63. Chakravarti, T.N. 1956. Narayanpur Image of Vinayaka. Indian Historical Quarterly 32(2-3): 324-8. óñ. 1863. Corpus Inscriptionum Indicarum. Vol. 3. Delhi: Indological Book House. Chanda, R.P. 1975 (reprint). Gaudarajamala (i.e. The Kings of Gauda, in Bengali). Calcutta: Navabharat Publishers. French, J.C. 1928. The Art of the Pala Empire. London: Humphrey Milford. Chattopadhyay, D.P. (ed.). 1980. Taranathaís History of Buddhism in India (Translated from the Tibetan by Lama Chimpa and Alaka Chattopadhyaya). Calcutta: K.P. Bagchi & Co. Chattopadhyaya, B.D. 1990. Aspects of Rural Settlements and Rural Society in Early Medieval India. Calcutta: K.P. Bagchi & Co. óñ. 1993-94. Urban Centres in Early Bengal: Archaeological Perspectives. Pratna Samiksha 23: 169-92 (=2003: 66-101). óñ. 2003 Studying Early India: Archaeology, Texts and Historical Issues. New Delhi: Permanent Black. Chowdhary A.K. 1971. Early Medieval Village in NorthEastern India (AD 600-1200): A Socio-economic Study. Calcutta: Punthi Pustak. Ferrier, Cédric and Judit Törzsök 2008. Meditating on the Kingís Feet? Some Remarks on the Expression padanudhyata. Indo-Iranian Journal 51: 93-113. Furui, Ryosuke. 2008. A New Copperplate Inscription of Gopala II. South Asian Studies 24: 67-75. óñ. 2011. Indian Museum Copperplate of Dharmapala, Year 26: Tentative Reading and Study. South Asian Study 27 (2): 145-56. Ganguly, D.C. 1966 (second edition). Northern Indian during the Eleventh and Twelfth Centuries. In R.C. Majumdar (ed.), Struggle for Empire. Second edition. Bombay: Bharatiya Vidya Bhavan. Ghosh, S. 2006. Character of the Pala Army as Reflected in Inscriptions and Literature. Journal of the Asiatic Society of Bangladesh (Humanities) 51(1): 41-7. óñ. 2011a. The Trans Meghna Region: Making of a SubRegional Identity. Journal of Ancient Indian History 27: 220-31. Chowdhury, A.M. 1967. Dynastic History of Bengal (c. 750-1200 AD). Dacca: The Asitic Society of Pakistan. óñ. 2011b. Kamarupa and Early Bengal: Understanding Their Political Relationship. Proceedings of the Indian History Congress. 71st Session, Malda 201011: 110-8. Cowell, E.B. and F.W. Thomas (translated) 1897. The Harsa-Carita of Bana. London: Royal Asiatic Society. Ghosh-Ray, R. 1993. Prachin Banglay Pataker Hisab (i.e. Denominations of Pataka in Ancient Bengal, in Bengali). Itihas Anusandhan 8 (1993): 199-207. Cunningham, A. 1968 (reprint). Archaeological Survey of India Report of Tours in the Gangetic Provinces from Badaon to Bihar in 1875-76 and 1877-78. Vol. 11. Delhi: Indological Book House. Dalal, C.D. and G.K. Srigondekar (ed.). 1925. Lekhapaddhati. Gaekwad Oriental Series, 20. Gupta, C. 1996. Land Measurement and Land Revenue System in Bengal under the Senas. In Debala Mitra (ed.), Explorations in Art and Archaeology of South Asia. Calcutta: Directorate of Archaeology and Museums, Government of West Bengal, pp. 573-93. Das, S. R. 1968. Rajbadidanga 1962. Calcutta: The Asiatic Society. Gupta, P.L. 1963. Nesarika Grant of Govinda III, Saka 727. Epigraphia Indica 34 (1961-62): 123-34. Dhar, D. 1904. Ballalcariter Banganubad (i.e. Bengali Translation of Ballalacarita, in Bengali). Calcutta: Hare Press. Haque, M.M. and A. Kuddus. 2005a. Lakshmansenner Bagbari Prashastio: Varendra Gabeshana Jadughare Ek Mulyaban Sanyojan (i.e. The Bagbari Eulogy of The Pala-Sena and Others 211 Laksmanasena: An Improtant Collection of the Vaarendra Research Museum, in Bengali). IBS Journal (Journal of the Varendra Researcvh Museum) 11:7-20. óñ. 1994. Exploring Indiaís Sacred Art (edited, with a biographical essay by Barbara Stoler Miller). New Delhi: Indira Gandhi National Centre for the Arts, pp. 204-40. Hultzsch, E. 1885. The Sarnath Inscrition of Mahipala. Indian Antiquary 14: 139-40. Lahiri. N. 1991. Pre-Ahom Assam: Studies in the Inscriptions of Assam between the Fifth and the Thirteenth Centuries AD. New Delhi: Munshiram Manoharlal Publishers Pvt. Ltd. óñ. 1981 (reprint). Plate of the Time of Sasankaraj; Gupta-Samvat 300. Epigraphia Indica 6 (190001): 143-6. Huntington, S.L. 1984. The Pala-Sena Schools of Sculpture. Leiden: Brill. óñ. 1985. The Art of Ancient India: Buddhist, Hindu, Jain. New York: Weather Hill. Islam, S. 2008. Recently-Discovered Coins and the History of the Khadgas. Journal of the Asiatic Society of Bangladesh (Humanities) 53(1): 1-11. ñó. 2011. Kotalipada Copperplate of Dvadasaditya. Journal of the Asiatic Society 53 (2): 71-82. óñ. 2012. Udisvara Copperplate of Sridharanarata. Journal of the Asiatic Society of Bangladesh (Humanities) 57 (1): 61-72. Jayaswal, K.P. 1934. An Imperial History of India in A Sanskrit Text. Lahore: Motilal Banarsidass. Keilhorn, F. 1888. A Buddhist Stone Inscription from Ghosrawa. Indian Antiquary 17: 307-12. Maitra, A.K. 1987. The Fall of the Pala Empire. Darjeeling: The University of North Bengal. óñ. 2004 (reprint). Gaudalekhamala (i.e. Epigraphs of Gauda, in Bengali) Kolkata: Sanskrita Pustak Bhandar. Majumdar, R.C. 1971 (reprint). Rise of Gauda and Vanga (Ch. 4), pp. 47-76, and Administration (Ch. 10), pp. 263-89. In R.C. Majumdar (ed.) History of Bengal, Volume I: Hindu Period. Patna: N.V. Publications. óñ. 1998 (reprint). Bangle Desher Itihas (i.e. The History of Bengal). Calcutta: General Printers and Publishers Private Limited. Mirashi, V.V. 1942. Rewah Stone Inscription of the Time of Karna: The [Chedi] Year 800. Epigraphia Indica 24 (1937-38): 101-15. Mishra, V.B. 1966. The Gurjara-Pratiharas and their Times. Delhi: S. Chand & Co. óñ. 1894a. Badal Pillar Inscription of the Time of Narayanapala. Epigraphia Indica 2: 160-67. Monier-Williams, M. (edition) 2002. A Sanskrit-English Dictionary. New Delhi: Motilal Banarsidass Publishers, 2002. óñ. 1894b. Bhera-Ghat Stone Inscription of the Queen Alhanadevi. The [Chedi] Year 907. Epigraphia Indica 2: 7-17. Morrison, B.M. 1970. Political Centers and Cultural Regions in Early Bengal. Tucson: University of Arizona Press. óñ. 1981a (reprint). Radhanpur Plates of Govinda III: Saka Samvat 730. Epigraphia Indica 6 (1900-01): 239-51. Mukherjee, B.N. 1975. Original Territory of Harikela. Bangladesh Lalitkala 1(2): 115-9. Khan, A., Z. Majid and S.M. Rahaman. 2004. Prachin Sabhar Anchale Manav Basati: Ekti Puratattvik Paryalocana (i.e. Human Settlements in Ancient Sabhar: An Archaeological Discourse, in Bengali). Pratnatattva: Bulletin of the Department of Archaeology, Jahangirnagar University 10: 51-69. Kramrisch, Stella. 1929. Pala and Sena Sculpture. Rupam 40: 207-26 (=1994: 204-40). Mukherjee, S.C. 1990-91. Palarajasya Madanapaadevasya Sampratikakale Labdhayoh Tamrasasanayoh (Sam 2, Sam 32) Vivecanam (i.e., Two Recently Discovered Copperplates of the Pala Ruler Madanapaladeva, in Sanskrit). Sanskrita Sahitya Parishat Patrika 73: 27-37. óñ. 1999. The Royal Charters of King Madanapala and the Chronology of the Pala Kings of Bengal and Bihar. Journal of Bengal Art 4: 61-5. Political History and Administration 212 Pal, S. 2008. Matsyanyaya of Khalimpur Inscription: Revisiting Its Geo-Historical Significance. Journal of the Asiatic Society 50(2): 21-36. Paul, P.L. 1939. The Early History of Bengal from the Earliest Times to the Muslim Conquest. Indian History Series No. 2. Calcutta: The Indian Research Institute Publications. Puri, B.N. 1957. The History of the Gurjara-Pratiharas. Bombay: Hind Kitab. Ray, H.C. 1936. The Dynastic History of Northern India: Early Medieval Period. (Vols. I & II). Calcutta: University of Calcutta. Ray, N.R. 1994. The History of the Bengali People: Ancient Period (Translated with an Introduction by John W. Hood). Calcutta: Orient Longman. óñ. 2001 (reprint). Bangalir Itihas: Adi Parva (i.e. The History of the Bengali People: Ancient Period). Kolkata: Deyís Publishing. Reu, B.N. 1933. History of the Rashtrakutas (Rathodas) from the Beginning to the Migration of Rao Siha towards Marwar. Jodhpur: The Archaeological Department. Roy, K. 2009. Socio-Economic and Cultural Study of the Rajbhita (Greater Dinajpur District, Bangladesh) Stone Slab Inscription of the Time of Mahipala (I), Regnal Year ñ 33. Journal of Bengal Art 13-14: 271-7. Culture in Honour of Gouriswar Bhattacharya. Vol. II. New Delhi: Kaveri Books, pp. 302-18 óñ. 2009b. Geo-polity in Early Mediaeval Bengal under the Sena Rule: Rereading Epigraphic Sources. Journal of Ancient Indian History 25: 94-113. óñ. 2010. Copperplate Inscriptions of West Bengal: Finding Find-spots and Locating Localities. Pratna Samiksha: A Journal of Archaeology, New Series 1: 107-34. Saraswati, S.K. 1978. Paljuger Chitrakala (i.e. The Art of Painting in the Pala Period, in Bengali). Calcutta: Ananda Publishers Pvt. Ltd. Sarma, D. (ed.) 2003. Kamarupasasanavali. Guahati: Publication Board, Assam. Sastri, H.P. (ed.) 1969 (reprint). Ramacaricam of Sandhyakaranandin (Revised with English Translation and Notes by Radhagovinda Basak), Memoirs of the Asiatic Society 3(1). Calcutta: The Asiatic Society. Sen, B.C. 1942. Some Historical Aspects of Inscriptions of Bengal. Calcutta: University of Calcutta. Sen, B.C. and D.P. Ghosh. 1934. A Dated Copperplate from Sundarban. Indian Historical Quarterly 10: 321-31. Sharma, M.M. 1978. Inscriptions of Ancient Assam. Gauhati: Department of Publication, Gauhati University. Salomon, R. 1998. Indian Epigraphy: A Guide to the Study of Inscriptions in Sanskrit, Prakrit and the Other Indo-Aryan Languages. New York: Oxford University Press. Shastri, H.N. 1982 (reprint). Haraha Inscription of the Reign of Isanavarman: [Vikrama samvat] 611. Epigraphia Indica 14 (1917-18): 110-20. Sanyal, N.B. 1950. Madhuk Inscribed Image of Ganesa of the Reign of Gopala II. Varendra Research Societyís Monograph 8: 4-6. óñ. 1983 (reprint). Barah Copperplate of Bhojadeva; Vikrama Samvat 893. Epigraphia Indica 19 (192728): 15-9. Sanyal, R. 2005. Sen Vamsiya Tamrasasaner Sakshye Caturaka (i.e. Caturaka in the Light of Sena Copperplate Inscriptions, in Bengali). Itihas Anusandhan 19: 169-75. Shastri, K.A. Nilakanta. 1935-36. Mahipala of Chandakausikam. Indian Culture 2: 797-99. óñ. 2009a. Dedicatory Inscriptions of the Time of Mahendrapala: A Fresh Appraisal. In J.R. Mevissen Gerd and Arundhati Banerjee (eds.), Prajnahara: Essays on Asian Art, History, Epigraphy and ñó. 1955 (second edition). The Cholas. Madras: Madras University. óñ. 1937. Rajendra Chola and Mahipala of Bengal. Indian Historical Quarterly 13(1): 149-52. Singh, S. 1993. Physical Geography. Allahabad: Prayag Pustak Bhawan. The Pala-Sena and Others 213 Sircar, D.C. 1942. Narayanpur Vinayaka Image Inscription of King Mahipala, Regnal Year 4. Indian Culture 9: 405-16. óñ. 1979. Some Epigraphical Records of the Medieval Period from Eastern India. New Delhi: Abhinav Publications. óñ. 1951. An Important Date in the Chronology of the Palas. Journal of the Asiatic Society, Letters 17(1): 29 óñ. 1982a. Pal-Sen Yuger Vamsanucharit (i.e. Genealogy and Chronology of the Pala-Sene Era, in Bengali), Calcutta: Sahityalok. óñ. 1952. Pala Rule in the Tippera District. Indian Historical Quarterly 28(1): 51-7. óñ. 1982b. Silalekha-Tamrasasanadir Prasanga (i.e. On Stone-and-Copperplate Inscriptions, in Bengali). Calcutta: Sahityalok. ñó. 1954. Madanapala and His Successors. Journal of the Asiatic Society, Letters 20(1): 43-8. óñ. 1958. Three Inscriptions from Balgudar. Epigraphia Indica 28 (1949-50): 137-45. óñ. 1963. Note on the Nesarika Grant of Govinda III, Saka 727. Epigraphia Indica 34 (1961-62): 13540. ñó. 1965. Three Pala Inscriptions. Epigraphia Indica 35(5) (January 1964): 225-38. óñ. 1983. Select Inscriptions Bearing on the History and Civilization of Ancient India form the Sixth to the Eighteenth Century AD. Vol. 2, New Delhi: Motilal Banarsidass. óñ. 1985a. The Kanyakubja-Gauda Struggle from the Sixth to the Twelfth Century AD. Calcutta: The Asiatic Society. óñ. 1966. Indian Epigraphical Glossary. Delhi: Motilal Banarsidass. óñ. 1985b. Pal-Purva Yuger Vamsanucharit (i.e. Genealogy and Chronology of the Pre-Pala Era, in Bengali). Calcutta: Sahityalok. óñ. 1967. Inscriptions of Two Brahmana Rulers of Gaya. Epigraphia Indica 36(2) (April 1965): 81-94. óñ. 1985c. Siyan Stone Slab Inscription of Nayapala. Epigraphia Indica 39(2) (July, 1971): 439-56. óñ. 1972-73. Three East Indian Inscriptions of the Early Mediaeval Period. Journal of Ancient Indian History 6: 39-59. Thapliyal, Kiran Kumar. 1985. Inscriptions of the Maukharis, Later Guptas, Puspabhutis and Yasovarman of Kanauj. Delhi: Agam Prakashan. óñ. 1973. Epigraphic Discoveries in East Pakistan. Calcutta: Sanskrit College. Venis, Arthur. 1894. Copperplate Grant of Vaidyadeva, King of Kamarupa. Epigraphia Indica 2: 348-58. óñ. 1975-76. Indological NotesññR.C. Majumdarís Chronology of the Pala Kings. Journal of Ancient Indian History 9: 200-10. Westmacott, E. Vesey 1872. Correspondence, Notes and Queries: Dinajpur, 16th February 1872. Indian Antiquary 1: 127-8. [RS]