archaeology

Locating Domains of Devotion, Patronage and Power in the Eleventh Century

by MICHAEL WILLIS

Bijamandal and Bhojpur

The relative importance of Bhilsa vis à vis Udayagiri in post-Gupta times is indicated by the massive temple ruin known as the Bijamandal (Fig. 9). This is not dated but an inscription on one of the pillars belongs to the time of king Naravarman (c. 1093-1134). The inscription gives a short hymn in praise of the goddess Caccika.32 The eulogy, together with sculptures recovered from the debris piled against the lower mouldings, indicate that the Bijamandal was originally dedicated to this goddess and perhaps also to Siva. The inscription does not mention the foundation of the temple and the way it has been engraved on a pillar seems to indicate that it was added to a pre-existing structure. But was the temple relatively new when the inscription was incised on the pillar? Or was the building erected earlier than the reign of Naravarman? If it is earlier, when is it to be placed? These are not simple questions to answer given the present state of the evidence and the degree to which it has been analysed.

One site which potentially provides some answers is Bhojpur.33 The Siva temple there, much celebrated because it houses the largest linga in India, is not far from Bhilsa, standing as it does on the banks of the Betwa about forty kilometres upstream. Although a detailed comparative study would be needed to determine the matter scientifically, the huge base-mouldings at Bhojpur suggest that it is closely related to the Bijamandal, the two monuments perhaps even sharing the same architects and masons. Surviving sculptures from the two sites also share a number of decorative and iconographic conventions (Fig. 10). Because it was never completed, the Bhojpur temple has no dedicatory inscription. The name of the locality, however, points to an association with king Bhoja (c. 1000-55). This link is corroborated by a colossal Jaina image with an inscription mentioning [rajadhi] rajaparamesvara Bhojadeva.34 The Jaina image is in a shrine not far from the Siva temple and the location, if not the shrine’s current fabric, belongs to Paramara times. The implication is that there was considerable sculptural activity at Bhojpur when Bhoja was on the throne in the first half of the eleventh century.

9. Bhilsa, Bijamandal Detail of mouldings, probably eleventh century. Photograph courtesy of Adam Hardy.

10. Bhojpur, Siva temple. Niche on the plinth with a figure of Isana, probably mid-eleventh century. Courtesy of the Department of Oriental Antiquities, British Museum.

This evidence is malleable and proves nothing with certainty. However, the scale of Bhojpur is telling within the wider context of north Indian temple architecture. Parallels can be found at Khajuraho, Gwalior and other locations where buildings of analogous size were constructed in the eleventh century. The Sas Bahu temple at Gwalior is particularly well documented by a long inscription recording that it was founded by Mahipaladeva, the Kacchapaghata king.35 Not far from the Sas Bahu there are cave-shrines belonging to the Jaina faith. The analogy between Gwalior and Bhojpur (comparisons to other sites could easily be made) allows us to suggest that Bhojpur owes it origin to a royal patron, the most-likely candidate being Bhoja himself. If this is accepted, then architectural and sculptural parallels with the Bijamandal indicate that it too is a royal building belonging to the first half of the eleventh century. Whether we accept this dating or not, the fact remains that the Bijamandal was an urban temple of great size and importance. Its dedication to the goddess is perhaps a reflection of the long-standing importance of goddess worship in the region.36

The Sun Temple

11. Udayagiri, Schematic plan of the hill showing principal caves. Drawing by A. Searight. Courtesy of the Department of Oriental Antiquities, British Museum.

Aside from the goddess at Bijamandal, the other local deity that enjoyed a wide following was the Sun god called That Bhaillasvami. That Bhaillasvami was a name of Surya is directly stated in an inscription which, while praising the achievements of the Candellas, gives the following solicitation: ‘May the Sun named Bhaillasvami protect Krsnaraja, lord of the earth’.37 To understand something of this god, a useful starting point is the word bhailla.38 In Prakrit dictionaries we find two different meanings. The first is ‘ploughman’, ‘harvester’ or ‘farmer’.39 If this is accepted, then the name of the god, taken as a tatpurusa compound, could be glossed ‘Lord of Farmers’. Such a name might have arisen from the fact that the Sun temple was first established by farmers, who, in the ordinary course of their agrarian activities, would have been concerned with the seasonal movement of the sun.

The second dictionary definition of bhailla is that it represents Sanskrit bhagin and thus refers to one who enjoys a portion or share.40 In epigraphic usage bhagin, bhagika and bhagahara are royal officers who seem to have been responsible for collecting the king’s grain share.41 In this case, the name of the temple would be something like ‘Lord of the Royal Collector’ and possibly indicate that the shrine was founded and maintained by certain officers of state.

Setting dictionaries aside, we can approach the problem grammatically and analyse the constituents bha and illa. Prakrit grammarians teach that the suffix illa stands in the sense of mat and vat.42 If we follow this, then bhailla simply means bhasvat, ‘the luminous’, i.e. the sun. In this case Bhaillasvami would be a karmadharaya compound and be translatable as the ‘Lord who is the Luminous One’ or simply ‘Sun Lord’. This explanation gains credence from a Khajuraho inscription which calls the temple bhasvat and locates it on the banks of the Malavanadi (i.e. Betwa).43 While such testimony might seem to settle the matter in a definitive way, doubts linger because the Khajuraho inscription gives the impression of a Pundit’s learned word-play rather than a sympathetic knowledge of the Bhaillasvami temple and its ancient lore.

Adopting the model developed by Hermann Kulke, we could see all the explanations given above as partially valid and integral to the historical process by which Bhaillasvami moved from being a sub-regional autochthonous deity to one of supra-regional importance.44

12. Udayagiri, Serrated crown on the site of the Bhaillasvami temple, probably eleventh century. Courtesy of the Department of Oriental Antiquities, British Museum.

Whatever and whenever its ultimate beginning, the importance of Bhaillasvami by the eleventh century is undeniable. A series of records show that this god was not only the leading deity in the region but enjoyed a high reputation throughout India. The temple’s fame eventually drew the attention of the sultans of Delhi who attacked Bhilsa and pulled down the building in two campaigns, the first in 1234, the second in 1292.45 The destruction was so thorough that the building’s whereabouts was completely lost to memory. Inscriptions relating to the Sun god were first noted by D. C. Sircar who edited the records and discussed them on a several occasions.46 Despite these publications, the cultural and archaeological implications of the information they contain have not been fully assessed or assimilated. For the present purpose, two issues draw immediate attention: the original location of the Bhaillasvami and its status as a focal point of religious patronage.

13. Surya. Provenance unrecorded, probably twelfth century. Department of Oriental Antiquities, British Museum (Bridge Collection 1872. 7-1. 56).

The oldest surviving record of the Bhaillasvami is dated Vikrama year 935 (CE 878-79). This specifically describes the temple as Sri Bhallasvamyayatana.47 The most important point about this inscription is that it documents the existence of the building in the time of the Pratihara rulers and indicates that it was attracting significant donations in the reign of Mihira Bhoja (c. 836-85). That considerable building work was done on the Sun temple in Mihira Bhoja’s time is shown by a large sculpture of Surya preserved in the Archaeological Museum at Vidisha. A superb work of considerable iconographic sophistication, this image was probably a key part of the Bhaillasvami and may have come from one of the building’s cardinal niches.48 The exact provenance of the sculpture is not recorded, a problem that also applies to the relevant inscriptions (see appendix). Those from Paramara times attest to the growing importance of the Sun temple in religious terms, but their recorded find-spots yield no helpful information about the original site of the building. If anything, the inscriptions and sculptures only show that the fabric of the Bhaillasvami was thoroughly scattered after the building was destroyed in 1292. This is supported by the Muntakhab al-Tawarikh which tells us that the main image was taken away and thrown down before the gate of Badaun.49

With little direct help to be had from the archaeological literature, we need to examine the Sanskrit inscriptions for topographical clues about the location of Bhaillasvami. Acrucial record for this purpose is one which seems to belong to the eleventh century. Unfortunately it yields no historical information and little connected sense in its present condition.50 However D. C. Sircar was able to make out ambaracudamani, ‘the crest-jewel of the sky,’ a poetic description of the sun, and also the word vihaya, open space or sky. Along with the opening invocation omnamah suryaya, these words leave little doubt that this damaged record relates to the Sun temple. Critically for our concerns, the same inscription contains the word udayagiri, literally ‘sun-rise mountain’. This appears to be a direct reference to Udayagiri and suggests that the temple of Bhaillasvami once stood on the hill. The precise spot appears to be marked by a substantial mound on the central ridge (Fig. 11). The mound is covered with decayed brick and assorted stone fragments including an amalasaraka, the serrated disc used to crown temples in north India (Fig. 12). Nearby, in a collection of images at an open-air shrine, there is a much broken torso of Surya dating to the eleventh century. On the banks of the Betwa, a few hundred metres distant, there is another mound with further pieces, including a battered relief with rearing horses. This formed the base of a Surya image similar in some respects to one that is now in the collection of the British Museum (Fig. 13).51

Further references to Udayagiri seem to appear in other inscriptions and support our suggestion that the Bhaillasvami temple originally stood on the hill. The Pratihara inscription of 878-79 is particularly interesting in this respect for its purpose is to record donations in favour of Narayana and the divine Mothers at Bhaillasvami. Why such donations should be made in connection with a Sun temple is not immediately apparent until we think of the rock-cut images of recumbent Visnu and the mother goddesses at Udayagiri. These are located in Caves 6 and 13 immediately adjacent to the site we have proposed for the Bhaillasvami temple (Fig. 11). References to these Gupta-period images seems to recur in Chittapa’s eulogy to the Sun god.52 The poet Chittapa enjoyed the title  ahakavicakravarti and was quoted by king Bhojadeva in his Sarasvatikanthabharana.53 This seems to indicate that Chittapa was among the important poets who flourished at the Paramara court in the eleventh century. Chittapa’s poetic description of how the sun’s splendour falls on the hood of Sesanaga seems to allude to some of the relief panels at Udayagiri. As an example we quote one verse in which the order of Chittapa’s description follows features of the Varaha panel from bottom to top:54

phanamanisu s´esasya muktamanisutoyadhehtaramanisu ca vyomnas tava rocir virocate [8]

His rays illumine the crest-jewels on the hood of S´ es.a, the sparkling pearls in the sea and the adamantine stars in heaven.

Several points could be raised against the identification of Udayagiri as the site of Bhaillasvami. There is, firstly, no Mahatmya associated with the hill and we would certainly expect something for a place that was once so famous. A possible answer to this problem is that Mahatmya texts often carry misleading names and geographical affiliations are only revealed coincidentally. The discovery of some kind of Udayagiri Mahatmya thus remains a possibility. A second problem is the absence of Udayagiri as a place name in Gupta-period inscriptions, an especially notable gap given the rich corpus in the region. While arguments from silence are inherently weak, I would propose that the name Udayagiri is relatively late and probably dates only to the time of the Paramara ruler Udayaditya (c. 1070-93). This king was particularly active as an architectural patron at Udayapura, a town about 30 km north-east of Bhilsa, where a Varaha temple was built in addition to the Siva temple called Udayesvaradeva.55 The eponymous renaming of Udayagiri would have been in keeping with Udayaditya’s ambitious building programme and the cultural politics of his day. The renaming (if indeed this happened) suggests that Udayaditya was responsible for a campaign of refurbishment at the Bhaillasvami. Some of the undated inscriptions mentioned above could be assigned to such a refurbishment, as could many of the broken fragments still lying at or near Udayagiri. The name of Udayagiri before Udayaditya’s time is an open question, but the age of the Sun temple is not in doubt. Not only does the inscription of 878-79 demonstrate that it existed in Pratihara times, but a recent study of early photographs shows that the early lion capital, now in Archaeological Museum, Gwalior, was found lying just a few metres from the Sun temple mound.56

Udayagiri Cave 19

The foregoing information provides, albeit in a preliminary and somewhat contentious way, an epigraphical and archaeological context for the inscriptions inside Cave 19. The individuals mentioned in our records were operating on the periphery of a sacred landscape populated by gods of great prestige and antiquity. The marginal nature of Cave 19 is shown, in the first place, by its location at the northern edge of Udayagiri hill, considerably removed from the mound we have identified as the site of Bhaillasvami (Fig. 11). Secondly, Kanha, the man who ‘restored’ the cave, carries no title, a fact which suggests he had little standing in political terms. He did, however, bring an abandoned shrine back into religious use. Could he have been a wandering holy man who took up residence in the cave, reviving Vais. n. ava worship there?57 As to Somapala, Sodha and Vahilavahada, the donors Kanha appears to have mobilised, each gave one nivartana of land, a small gift compared to contributions recorded in other inscriptions.58 The monumental residue of these gifts is modest: a fragment with a standing figure in a niche is all that can be dated to the eleventh century (Fig. 14). Exactly how this was incorporated into Cave 19 is uncertain and underscores the changes that have taken place at the site since the thirteenth century.

The humble religious giving documented by the Udayagiri inscriptions has received little attention but it forms a crucial part of the historical record if we want to assess the full range of power, property and patronage under Paramara rule. The Cave 19 inscriptions, as already noted, do not define the relationship of our donors to royal authority. In this they are not unusual, a situation which has prompted much discussion about the social, economic and political constitution of India after the fifth century. The scholarly literature on this question is voluminous and filled with energetic debate. A synopsis of this literature would be out of place here but I would venture to make one observation which will probably be accepted by all who have contributed to the subject, namely that the various accounts we have of ‘medieval society’ have been construed using a rather limited corpus of epigraphical, archaeological and numismatic evidence. This evidence, such as it is, has been marshalled to produce radically different accounts of the post-Gupta world.59 Because no account is entirely convincing, the central question that emerges is whether a viable assessment of this period is possible or, in the final analysis, ever will be.

14. Udayagiri, Cave 19. Carved architectural fragment, late eleventh century. Present whereabouts uncertain. After Central India, years 1908-14, old negative number 50.

Aside from a traditional call for more attention to unedited inscriptions and the collection of related data, I would suggest that a two-part investigation is needed to resolve this issue. The first step should involve a review of the scholarly literature. This would have the specific aim of clarifying the empirical predictions which each historical model makes or implies. Let me give an oversimplified example to illustrate the point. If we accept the Marxist proposition that some sort of ‘feudal system’ developed after the fifth century, we are more or less accepting that a non-monetary, village economy emerged with the decline of urban centres and interregional trade. These developments, if they did take place, should have left specific physical traces in the archaeological, historical, numismatic and artistic record. The ‘segmental’ and ‘integrative’ explanations of post-Gupta India rest on assumptions that differ from Marxist models but which nonetheless posit certain types of social, economic and agrarian organisation. These in turn should have left quantifiable and discoverable residues.

Armed with predictions generated by this sort of historiographic analysis, an interdisciplinary team would be in a position to begin a theoretically-informed campaign of fieldwork. While it is impossible to anticipate either investigative methods or results, it is probable that future research would include a village-by village survey of temple fragments, inscriptions and associated archaeological material.60 This would show the age and distribution of settlements and the ways in which temple architecture defined the cultural landscape of post-Gupta India. I would imagine that this type of work could be coupled with archaeo-botanical studies aimed at understanding changes in forest cover and cropping patterns. Finally, archaeological excavation at appropriate city-sites (perhaps Besnagar) might focus on the degree to which cities declined or moved after the fifth century. These are just hints of the possibilities that lie ahead. Whatever the course of future research, these points reinforce the point with which I opened this essay, namely that much remains to be done, even at the most well-known sites.

NOTES

1. J. F. Fleet, Inscriptions of the Early Gupta Kings and their Successors, Corpus Inscriptionum Indicarum, Volume 3 (Calcutta, 1888), numbers 3, 6 and 61; D. R. Bhandarkar et al, Inscriptions of the Early Gupta Kings, Corpus Inscriptionum Indicarum, Volume 3, revised edition (New Delhi, 1981), numbers 7 and 11. The research presented in this article was supported by the British Museum and the Leverhulme Trust to whom many thanks are due. A research grant from the Leverhulme Trust in particular provided time away from my regular duties and allowed this article to be completed in a timely manner.
2. D. R. Patil, The Monuments of the Udaygiri Hill (Gwalior, 1948); Debala Mitra, ‘Varaha Cave at Udayagiri – An Iconographic Study,’ Journal of the Asiatic Society 5 (1963), pp. 99-103; J. C. Harle, Gupta Sculpture (Oxford, 1974), Figs. 8-17; Phyllis Granoff: ‘Mahisasuramardini An Analysis of the Myths,’ East and West 29 (1979), pp. 51; J. G. Williams: The Art of Gupta India: Empire and Province (Princeton, 1982), Figs. 35-39. M. A. Dhaky, Krishna Deva, Michael Meister, eds. Encyclopaedia of Indian Temple Architecture: North India, Foundations of North Indian Style (New Delhi, 1988), Figs. 15- 16 (hereinafter cited as EITA).
3. Illustrated in my Buddhist Reliquaries form Ancient India (London, 2000), Fig. 29.
4. See R. Salomon, ‘New S´ ankhalipi (Shell Character) Inscriptions,’ Studien zur Indologie und Iranistik 11-12 (1986), pp. 109-152, more recently but substantially less helpful, R. K. Sharma, ed. Studies in Shell Script (Delhi, 1990). There seem to be no published illustrations of the images and niches cutting through the shell inscriptions.
5. Harle, Gupta Sculpture, Fig. 3 for a general view including the tank. The eroded stone at the base of the Varaha sculpture (Ibid., Fig. 12) shows where the water washed across the base of the carving. These important observations were made by Meera Dass who is preparing a careful study of the site and to whom I am grateful for sharing many insights.
6. These problems are found in the survey books of Harle and Williams already cited. Patil, Monuments, follows the tradition (established by Cunningham) of numbering the caves and thus considering them one by one. EITA sits firmly in this framework, with a clear line drawn between monument and non-monument. The approach has had predictable effects, via conservation policies, on the environment of the remains.
7. Harle, Gupta Sculpture, Fig. 7, EITA, Figs. 19-22; see however discussion below which shows hitherto unnoticed monumental additions in the Parama-ra period.
8. For a general view of the pillar, see EITA, Fig. 21.
9. Cunningham, Archaeological Survey of India Report 10 (1880), p. 52 (hereinafter cited as ASIR).
10. J. F. Fleet, ‘Sanskrit and Old-Canarese Inscriptions,’ Indian Antiquary 13 (1884), p. 185, Annual Report on Indian Epigraphy (1952-53) B, number 131 (hereinafter cited as ARE); ARE (1956-57) C, number 93.
11. From photographs and in situ examination, as are all the readings given in this article. I am grateful to Richard Salomon who offered suggestions, particularly with regard to inscription number 1.
12. The reading is sam. vatu, but the u vowel on the last syllable is meant to be halanta (vira-ma).
13. The syntax is peculiar; the phrase beginning pas´ca -t would seem to be qualifying that the year in question belongs to the Vikrama era.
14. On the later tradition there is much literature and controversy; see D. C. Sircar, Ancient India and the Vikramaditya Tradition (Delhi, 1969). The inscription published here is not discussed by Sircar in his volume.
15. John Allan, Catalogue of the Coins of the Gupta Dynasties and of Sasanka, King of Gauda (London, 1914), pp. 34-5. ksitim avajitya sucaritair divam. jayati vikramadityah..
16. Frederick M. Asher, ‘Historical and Political Allegory in Gupta Art,’ in Essays on Gupta Culture, edited by Bardwell L. Smith (Delhi, 1983), pp. 53-66; also V. S. Agrawala, Matsya Purana – A Study (Varanasi, 1963), ibidem., ‘Guptayuga memmadhyadesa kakalatmaka citrana Nagaripracarinipatrika 48 (VS 2000), pp. 43-48. Also see the essential article by Hans Bakker, ‘The Footprints of the Lord’ in Devotion Divine: Bhakti Traditions from the Regions of India, Studies in Honour of Charlotte Vaudeville (Groningen and Paris, 1991), pp. 20-37.
17. Sir car, Select Inscriptions, 1: 89, for an improved edition, Richard Salomon, Indian Epigraphy (Oxford, 1998), p. 266.
18. Cunningham, ASIR 10 (1880), p. 36 briefly mentions the tirtha ‘where two small hollows in the rock are believed to be the charan, or foot prints of Vishnu’.
19. ASIAR WC (1913-14), number 2635; ARE (1957-58) C: 238B.
20. The small mark at the end is either a visarga or diminutive dan. d. a.
21. The term nivartana is a land measure the size of which varied according to the region and period; see Sircar, Indian Epigraphy (Delhi, 1965), pp. 409-10. The term is used in Paramara plates, see H. V. Trivedi, Inscriptions of the Paramaras, Chandellas, Kachchapaghatas and Two Minor Dynasties, Corpus Inscriptionum Indicarum, Volume 7, (Delhi, 1978-91), number 16 and 11 (where a village is described as having at least 100 nivarttana). Trivedi’s Inscriptions is hereinafter cited as Trivedi, CII 7.
22. ASIAR WC (1913-14), number 2635; ARE (1957-58) C: 238B.
23. This term appears, for example, in the twelfth century plates of Mahabhavagupta, see Epigraphia Indica 4 (1896-97), p. 258, line 8; further discussion in Sircar, Indian Epigraphy, p. 376 and Indian Epigraphical Glossary (Delhi, 1966) s.v. palli, pallika.
24. Archaeological Department, Gwalior State, Archaeological Report for VS 1993/AD 1936-37, number 1. The slab was recovered from a wall of a house in Bhilsa.
25. Trivedi, CII 7, number 16 (line 9).
26. Because Udayagiri inscription 5 might be later, it is excluded from the present discussion. I am grateful to Daud Ali at the School of Oriental and African Studies for offering comments on the concluding section of this essay.
27. D. C. Sircar, Studies in the Geography of Ancient and Medieval India (Delhi, 1971), pp. 264-65; B. C. Law, Historical Geography of Ancient India (Paris, 1954) p. 336.
28. Dpvs 6: 16 set.t. hidhitadevinamati. Abbreviations of Buddhist texts follow my Buddhist Reliquaries from Ancient India, p. 10. The Dpvs and those texts cited in the following note are, of course, much later than the events they record; their composition was embedded in political and historical debates which complicate their representation of historical ‘facts’, see Jonathan Walters, ‘Buddhist History’, Querying the Medieval (Oxford, 2000), pp. 99-164, reviewed below in the present number of South Asian Studies.
29. For Devi’s residence in Vedisa, see Mhvs 13, pp. 7-12; Dpvs 12:15; Thvs, p. 192.
30. To my knowledge the lintel and stepwell have not been published.
31. Trivedi, CII 7, p. 149 and 207.
32. For the Caccika inscription, see Trivedi, CII 7, number 36.
33. K. K. Chakravarty, Bhojpur Temple: A Vision of Harmony (Bhopal, 1991) for plans and some illustrations of sculpture.
34. Trivedi, CII 7, number 17. Jaina images are illustrated in Chakravarty, Bhojpur Temple, passim.
35. See Willis, Inscriptions of Gopaks.etra (London, 1996), p. 6 and ibidem., ‘Architecture in Central India under the Kacchapaghata Rulers’, South Asian Studies 12 (1996), pp. 13-32.
36. See, for example, Harle, Gupta Sculpture, Figs. 30-3.
37. bhaillasvaminamaravir avatu bhuvah svaminam krsnarajam, see Fitz-Edward Hall, ‘Three Sanskrit Inscriptions,’ Journal of the Asiatic Society 2 (1862), p. 111, note. The inscription was found by Hall ‘within the fort… in the outer wall of a modern house’. This would seem to refer to the old walled town of Bhilsa. Krsnaraja in this record was a Candella prince who ruled the western territories of the dynasty in the circa late tenth century. See Trivedi, CII 7, p. 93 and 354. The location of this inscription has not been traced as far as I am aware.
38. Trivedi, CII 7, p. 122, note 6, following the Prakrit grammarian Vararuci, suggests bha- and Prakritilla means ‘one who possesses or is a storehouse of lustre’. Hall, ‘Three Sanskrit Inscriptions’, p. 112 suggests bha and root il, to cast, thus the ‘thrower of light’ (but a note by Hall’s editors cite Prakrit grammarians and the fact that Prakritilla stands for Sanskritmat). In these accounts no attempt is made to explain how this is to be interpreted in conjunction with the termination svami (or isvara) which appears as the ending of the names of gods who are eponymously named. Sircar, ‘Two Inscriptions’, p. 214 feels that it is simply a personal name.
39. Paia sadda mahannavo (Varanasi, 1963), s.v.
40. Ibid., s.v.
41. Sircar, Indian Epigraphical Glossary, s.v.
42. R. Pischel, Grammatik der Prakrit-Sprachen (Strassburg, 1900), paragraph 595.
43. Trivedi, CII 7, (verse 45), malavanaditirasthitat bhasvatah
44. Hermann Kulke, Kings and Cults, State Formation and Legitimation in India and Southeast Asia (Delhi, 1993), pp. 4-7.
45. See the important discussion in Richard M. Eaton, ‘Temple Desecration and Indo-Muslim States’, in Essays on Islam and Indian History (New Delhi, 2000), pp. 94-132. My colleague Dr D. Maclean informs me that that the translations of the relevant Indo-Islamic texts are very unreliable; this has not stopped writers from repeatedly citing old translations without reference to the original texts.
46. Sircar, ‘Two Inscriptions’, pp. 210-19; ibidem., ‘Bhayillasvamin and Bhillama ladeva’ in Studies in the Religious Life of Ancient and Medieval India (Delhi, 1971), pp. 115-32; ibidem., ‘Chittapa’s Prasasti of the Sun-god Bhilsa’, in Studies in Yugapurana and other Texts (Delhi, 1974), pp. 107-13. The last two of these articles are near verbatim repetitions of what was first published in Epigraphia Indica; the 1971 publication introduces ‘Bhayillasvami’ but I can find no authority for this spelling. Some of the inscriptions were subsequently re-edited in Trivedi, CII 7, see appendix to this article.
47. Sircar, ‘Two Inscriptions’, p. 214; also listed here as number 1 in the appendix.
48. Kirit Mankodi, ‘An Image of Surya ‘Bhayillasvami’ from Bhilsa’, Journal of the Indian Society of Oriental Art 10 (1978- 79), pp. 41-47. For other monuments in Malwa dating to the last quarter of the ninth century see my Temples of Gopaks.etra, pp. 76-8; the Surya in Vidisha seems to be no later than circa 875. Before studying this question I had assumed that the Bijamandal marked the site of Bhaillasvami. I am particularly grateful to Dr Mankodi who first pointed out to me that the surviving images show this cannot have been the case.
49. Cited in Sircar, ‘Two Inscriptions’, p. 211.
50. Sircar, ‘Two Inscriptions’, p. 215 and number 4 in the appendix given here. Sircar had no doubt that this inscription and the others relating to the Sun god were embedded in the walls of the Bha-illasva-mi. We share this view but are obliged to note that there is no direct evidence.
51. This sculpture was collected by Charles ‘Hindoo’ Stewart (d. 1828) who is known to have travelled in the Bhilsa area; for a brief account of Stuart see my ‘Sculpture from India’, in A. W. Franks: Nineteenth Century Collecting and the British Museum, edited by Marjorie Caygill and John Cherry (London, 1997), pp. 250-61.
52. Trivedi, CII 7, number 37. Also listed here in the appendix, number 5.
53. Ibid., number 37 (introductory matter).
54. Ibid, number 37. The Varaha panel faces east and is illuminated by the rising sun; the description could also apply to Cave 13. Illustrated in Williams, Art of Gupta India, Figs. 37, 39. Chittapa’s text is full of double entendre which I pretend to translate only in a loose way. The importance of this text (and others) as raw material for understanding the visual culture of Indian images has been missed by art historians who are more concerned with their own descriptive vocabulary and categories of thought.
55. Trivedi, CII 7, number 19 (Udayesvara), p. 24 (Varaha); for the architecture of the Siva temple, Krishna Deva, ‘Bhumija Temples,’ in Studies in Indian Temple Architecture (Poona, 1975), pp. 90-113.
56. Illustrated, as already noted above, in Buddhist Reliquaries form Ancient India, Fig. 29. Prints of the old photographs were procured by Meera Dass and during the course of fieldwork we were able to determine where the capital lay in the nineteenth century from a study of the background and rock formations. Aco-authored article with illustrations is planned on these findings.
57. There is not a shred of evidence for this but it was suggested to me by a long inscription on the side of the cave door which indicates that the same thing happened again in later times. I have not had time to work this later record in detail.
58. For example Trivedi, CII 7, number 16.
59. For historiographic perspective on these questions, Brajadulal Chattopadhyaya, The Making of Early Medieval India (Delhi, 1994) and Chattopadhyaya, Aspects of Rural Settlements and Rural Society in Early Medieval India (Calcutta, 1990). Chattopadhyaya advocates an ‘integrative’ model in opposition to the Marxist ‘feudal’ position; for the ‘segmentary’ model, Burton Stein, Peasant, State and Society in Medieval South India (Delhi, 1980).
60. EIT A marks an important beginning in this but the art historical preoccupation with complete monuments and with style (which has influenced most studies of temple architecture to date) means that questions of distribution and geographical context in the wider landscape have not been properly addressed; see further my review of Packert Atherton, Sculpture of Early Medieval Rajasthan (Leiden, 1997) in South Asian Studies 15 (1999).

APPENDIX

INSCRIPTIONS RELATING TO BHILSA AND THE BHAILLASVAMI TEMPLE

1. INSCRIPTION DATED V.S. 935
Records donations in favour of Narayana and the Mothers at Sri Bhaillasvamyayatana. Found at Mahalghat, Bhilsa.
Bibliography:
D. C. Sircar, ‘Two Inscriptions from Bhilsa’, Epigraphia Indica 30 (1953-54), p. 214.

2.INSCRIPTION OF YAS´ OVARMAN DATED V.S. 1011
Records the construction of a Visnu temple by Yasovarman and recounts his achievements and those of the Candella dynasty. Records also (in verse 45) that the Candella dominion reached up to Bhasvat (i.e. Bhaillasvami). Found at Khajuraho.
Bibliography:
H. V. Trivedi, Inscriptions of the Paramaras, Chandellas, Kachchapaghatas and Two Minor Dynasties, Corpus Inscriptionum Indicarum, Volume 7, (Delhi, 1978-91), number 98.

3.INSCRIPTION OF KR. S. N. ARA- JA
Fragmentary inscription equating Bhaillasvami with Ravi and recording that a minister of Krsnaraja, a Candella lord, dwelt on the banks of the Vetravati. Found in the wall of a house in Bhilsa.
Bibliogaphy:
Fitz-Edward Hall, ‘Three Sanskrit Inscriptions,’ Journal of the Asiatic Society 2 (1862), p. 111, note.

4.INSCRIPTION MENTIONING THE SUN GOD
Heavily damaged inscription mentioning the Sun god and Udayagiri. Datable to the eleventh century. From Bhilsa but precise find-spot not recorded.
Bibliography:
D. C. Sircar, ‘Two Inscriptions from Bhilsa’, Epigraphia Indica 30 (1953-54), pp. 215-16.

5.INSCRIPTION WITH EULOGY TO THE SUN GOD
Records the merits of a distinguished person (perhaps a king) who is compared to the sun. Contains a eulogy to the sun by the poet Chittapa; composed at the order of dandanayaka Sri Candra. Dateable to the eleventh century. From Bhilsa but precise find-spot not recorded.
Bibliography:
D. C. Sircar, ‘Two Inscriptions from Bhilsa’, Epigraphia Indica 30 (1953-54), p. 219; H. V. Trivedi, Inscriptions of the Paramaras, Chandellas, Kachchapaghatas and Two Minor Dynasties, Corpus Inscriptionum Indicarum, Volume 7, (Delhi, 1978-91), number 37.

6.COPPER-PLA TE OF MADANAVARMAN DATED V.S. 1190
Records the donation land to a brahman. a by the Candella king Madanavarman from his camp near Bhailasvami. Found in Banda distirict, Uttar Pradesh.
Bibliography:
H. V. Trivedi, Inscriptions of the Parama-ras, Chandellas, Kachchapaghatas and Two Minor Dynasties, Corpus Inscriptionum Indicarum, Volume 7, (Delhi, 1978-91), number 118.

7.COPPER-PLA TE OF HARIS´ CANDRA DATED V.S. 1214
Records the division of the village of Dadarapadra into sixteen shares and its donation to nineteen bra-hman.as by the Paramara Mahakumara Haris´candra. Mentions (in line 9) Bhailasvamidevapura. Found in Bhopal.
Bibliography:
H. V. Trivedi, Inscriptions of the Paramaras, Chandellas, Kachchapaghatas and Two Minor Dynasties, Corpus Inscriptionum Indicarum, Volume 7, (Delhi, 1978-91), number 44.

8.INSCRIPTION OF TRAILOKYAVARMAN DATED V.S. 1216
Records the establishment of a temple of Murari in the boar incarnation and other images together with a garden by Trailokyavarman, a Paramara prince. Further records the imposition of certain taxes in the temple’s favour and mentions the river Vetravati. Found near the Jain temple, Bhilsa, in the door of a house.
Bibliography:
H. V. Trivedi, Inscriptions of the Paramaras, Chandellas, Kachchapaghatas and Two Minor Dynasties, Corpus Inscriptionum Indicarum, Volume 7, (Delhi, 1978-91), number 42.

9.INSCRIPTION OF JAYASIM. HA DATED V.S. 1320.
Records the donation of a village named Dvormela (?) by Sanumati (Bhanumati?) in Bhailasvamidevapura during the reign of Jayasim hadeva. From Bhilsa but exact provenance not recorded.
Bibliography:
H. V. Trivedi, Inscriptions of the Parama-ras, Chandellas, Kachchapaghatas and Two Minor Dynasties, Corpus Inscriptionum Indicarum, Volume 7, (Delhi, 1978-91), number 58.