Royal Patronage and Religious Tolerance The Formative Period of Gupta–Vākāṭaka Culture

Abstract Patronage by the royal court of religious institutions and foundations is one of the hallmarks of the development of India under the rule of the Gupta and Vākāṭaka kings (4th–5th centuries). This patronage was extended also to religious movements other than the king's own persuasion. The evolving culture of religious tolerance and enthusiasm is apparent in the temple monuments of the time. In this article we focus on four archeological sites where these developments become best visible: Udayagiri, Māṇḍhaḷ, Rāmagiri (Ramtek), and Mansar. The close relationship of the Gupta and Vākāṭaka realms is investigated in its local settings. Renewed attention is given to the ‘Mandhal Inscription, Year 5’ of the Vākāṭaka king Rudrasena II and the deity on whose authority the charter was issued: Muṇḍasvāmin. It is argued that the name Muṇḍa refers to no one else than the Gupta queen of Rudrasena II, Prabhāvatī Guptā. During the last decade, excavations in Mansar (5 km west of Ramtek) have brought to light the state sanctuary of the youngest son of Prabhāvatī, Pravarasena II. The findings there are placed within the tradition that can be traced back, through Rāmagiri and Māṇḍhaḷ, to the religious foundations in Udayagiri.

However, this rather insignificant Cave 8, which probably enshrined a liṅga, can hardly have been the main purpose of Candragupta's visit. 2 For, next door, around the corner, a truly grandiose scheme had just been, or was about to be accomplished, a work of art that was designed to broadcast the king's larger ambition: 'the winning of the entire earth', expressed allegorically in the great Varāha panel of Cave 5.
As Michael Willis (2004Willis ( , 2009) has shown, the Varāha panel was intrinsically connected through a system of hydrological engineering with a natural cleft in the central ridge of the mountain. Water once cascaded down through this narrow passage, flowing underneath a large image of Vis . n . u-Nārāyan . a. This panel shows the god in his cosmic sleep while resting on the primordial snake (Anantaśayana), floating on the cosmic ocean. At the bottom of the cascade, the water was collected in a reservoir at the feet of the Varāha sculpture of Cave 5, making the cosmic boar rise out of the waters.
It will not be our aim here to analyse the iconography and the allegorical double-entendre of these two panels, the Varāha and the Nārāyan . a, which most recently has been done by Willis in a penetrating study referred to above. 3 What concerns us here is the ecumenicity of the situation: a Vais . n . ava king, "whose prowess is [Vis . n . u's] Cakra", 4 selects a prominent spot at the centre of the holy mountain of Udayagiri for celebrating his faith and political mission, and allows at the same time two leading figures of his kingdom, a minister (Vīrasena) and a vassal king (the Sanakānika mahārāja Sod . hala (?)), to construct next door cave-sanctuaries (Caves 6 and 8) dedicated toŚiva, the other high god of Hinduism. Inscriptions in both shrines duly recognise the sovereignty of Candragupta. 5 Admittedly, we cannot be certain as to the original dedication of the sanctuary of Cave 6 -the cave no longer contains its original sculpture -but the two Vis . n . u images at the outer wall, flanking the two Dvārapālas, as well as the image of Gan . eśa overviewing the entrance at the left and a shrine of the seven mothers at the right of the sanctum, strongly suggest that Cave 6 was also aŚiva sanctuary, 6 in which the two Vis . n . u images, in the words of Joanna Williams, were merely "attendant upon the greater glory ofŚiva". 7 And, to complete the picture, adjacent to the Varāha cave on the other, southern side, was one moreŚaiva cave (4), enshrining an ekamukha liṅga. 8 Thus it cannot have escaped the attention of the contemporaneous visitor that, with the establishment of Gupta power over the area, the Bhāgavata religion had taken pole position, not by ousting itsŚaiva rivals, but rather by including them in a subordinate position. By doing so the three caves together 2 This cave is referred to as Cave 8 in the archaeological literature (Willis 2009, p. 39). Williams 1982 refers to it as Cave 7. 3 Willis 2004, 41ff.; Willis 2009, pp. 30-46. 4 Willis 2004, 55 n. 52. 5 For the inscription in Cave 6 see 'Udayagiri Cave Inscription of Chandragupta II. The Year 82', CII III (1888), pp. 21-25. The name of the king, 'Sod . hala,' is a conjecture of D.C. Sircar (Sel. Ins. I,279 n 5.). No king of that name is known otherwise. It is uncertain whether 'Sod . hala's' inscription refers to the construction of the cave sanctuary 6 as a whole. The inscription is placed above the Mahis .ā suramardinī and one of the Vis . n . u images on the outer wall. 6 Cf. Willis 2009, 142f. 7 Williams 1982 According to Joanna Williams (1982, p. 86) this cave may be slightly later than the excavations of caves 5 to 8, but she concedes that "the liṅga inside seems more conservative [ . . . ]. The round face, sharply defined features, and wide foliate necklace are still close to the Vis . n . u images of Cave 6 or of Besnagar". mirrored the situation in Cave 6 on a larger scale:Śiva (Caves 4 & 6) "attendant on the greater glory of" the Bhagavat (Cave 5).
Although the majority of the temple structures on the Udayagiri can be assigned to Hindu gods, this by no means excluded other persuasions to establish their own sanctuary on the hill. At the top of the north side of the hill is a cave of the heterodox Jains, which an inscription dates to the reign of Candragupta's son Kumāragupta (GE 106 = AD 436). "The inscription records the dedication of an image of Pārśvanātha". 9 Jainism had been favoured in this region by Candragupta's ill-fated elder brother, Rāmagupta, who commissioned the three images that were reportedly found inside old Besnagar (Durjanpura), 1.5 km from Udayagiri. 10 Three inscriptions on the pedestals record basically the same dedication: This image of the Lord, the venerable Pus . padanta/Candraprabha, was commissioned by the MahārājādhirājaŚrī Rāmagupta, at the instigation of Celūks . aman . a, son of Golakyāntī, who is the pupil of the preceptor Sarpasenaks . aman . a and the grand-pupil of the pān . ipātrika 11 Candraks . aman . a, preceptor (ācārya) and forbearing monk (ks . aman . aśraman . a). 12 No direct evidence of a Buddhist presence in Udayagiri has come to light, the reason for which is not difficult to see: the great, ancient centre of Sanchi, 5 km to the south, naturally attracted all Buddhist devotion. This was, for instance, the case when another favourite of Candragupta,Āmrakārdava, donated in GE 93 (= AD 413), [A]n allotment of land, calledÍśvaravāsaka, and a sum of money to theĀrya-saṅgha, or community of the faithful, at the great vihāra or Buddhist convent of Kākanādabot . a, for the purpose of feeding mendicants and maintaining lamps. 13 Candragupta's favours are explicitly said to have contributed toĀmrakārdava's way and means of life. 14 Thus we possess within a circle of five kilometers from the Udayagiri and within a time frame of a few decades around AD 400, plenty of epigraphical and sculptural evidence that proves that the reign of Samudragupta's successors ushered in a period of royal patronage that was not limited to one religious denomination only, but was beneficial to all the four major religions of their times: Bhāgavata (Visnuism), Māheśvara (Sivaism), Jaina (Jainism) and Bauddha (Buddhism). This is not to say, of course, that the Guptas invented religious patronage, but their rule marked the emergence of kings and courtiers as a major class of patrons, whereas earlier "groups of lay people" were the prominent sponsors. 15 And in 9 Williams 1982, p. 87. 'Udayagiri Cave Inscription. The Year 106', CII III (1888), pp. 258-260. 10 The images are presently stored in the State Museum, Bhopal. See Gai 1969;Williams 1982, 25f, 28f. Bakker 2006 In an email dated 17 September 2009, Paul Dundas explained this term to me as follows: 'pān . ipātrika I would not take as a title but rather as a eulogistic epithet signifying an ascetic who uses his cupped hands as an alms bowl and thus follows the jinakalpa, the rigorous mode of life of the Jinas. The term usually designates, or came to designate, a Digambara monk'. 12 CII III (1981), pp. 231-234. The emended reading below is based on a collation of the three inscriptions (see Bakker 2006, p. 182 n. 9 (Miller 1992, 5f.).
contrast to most of the earlier patrons, their patronage extended to religions other than their personal persuasions, thus spreading an atmosphere of religious tolerance throughout the realm. 16 As we will see, this patronage could take the form of financing prestigious building projects as well as lending financial support to religious groups and individuals through gifts of money or land. We do not possess much direct evidence of courtly patronage of individual sculptors, decorators, architects or other artisans who worked in the visual arts, although this must have been substantial. They were paid, of course, and some artists more than others, no doubt. Since the king himself was the centre of the patronage system, it is remarkable that we do not possess epigraphical testimonies of Candragupta's own building projects. It looks as if he was satisfied with being credited by his subjects for his goodness, whereas the ones that appear to have been his own monumental plans, such as the Varāha and Nārāyan . a panels, were apparently so obvious that they needed no special commemoration in the form of dedicatory inscriptions.
This large-scale and broad-minded religious patronage practised by the court, which involved the king, his courtiers and officials (amātya), favourites and so-called feudatories (sāmanta), 17 enhanced the development of the classical forms of the Indian religions, as well as their sacred architecture and art. It formed part of a wider system of patronage, which operated as a powerful political strategy, through which the king could advance and visualise a religiously sanctified social order (dharma), from which his authority as a ruler by God's grace was deduced. 18 Partly due to this policy, Northern India became covered with religious monuments of various sorts, a small number of which survived the ages. The Gupta court and its satellites promoted a lifestyle and devotional ethos that obviously was appealing to a broad section of the population; it proved not only to be politically effective, but it lent stability and grandeur to Gupta reign, thanks to which some still speak of 'the golden age', or 'the classical age'. 19 16 Cf. A. K. Narain in Religious Policy and Toleration in Ancient India: 'Toleration' in this context means a state's recognition and protection of the right of private judgement in matters of faith and worship, and a lack of persecution or obstruction of the beliefs and practices of sects and religions other than the king's own. Often the king of ancient India does not get credit for his act of toleration by modern historians beause it is wrongly assumed that toleration is an essential part of his religious practice or of rājadharma. (Narain 1983, p. 50). 17 Ali 2004, p. 8, "It was the activities of the king's court, composed of dependents and retainers, and attended by underlords and vassals, which constituted 'government' rather than a putative 'administration'".
18 See Bakker 1992. This should not be taken as support of a legitimation theory which, in the words of Daud Ali, "suggests the rather unlikely and even anachronistic scenario of the court acting collectively on the basis of certain principles, and then representing them back to itself in order to legitimate them". On the contary, we subscribe to Ali's insight that "the ideas enshrined in art and literature are in fact identical to the key concepts found in the texts which urge the king and his men to constitute their political actions, which is to say that the supposedly non-ideational realm of power and politics is in fact already ideational". (Ali 2004, p. 14). 19 Although Daud Ali admits that he has "bracketed out the question of religion", he makes an important observation that is relevant to our subject: "In fact, religious and political notions of lordship differed more in degree than kind. They formed part of a continuous and homologously structured 'chain of being' which linked the entire cosmos. This, on the one hand, meant that the king's authority and mystique resembled and participated in that of the temple god, giving a theological dimension to relationships at court. On the other hand, however, it meant that the life of gods, housed in their sumptuous palaces, shared striking resemblances to those of princes". (Ali 2004, p. 104). I have earlier argued in a similar line, pointing out, however, that this notion of 'lordship' was itself part of a historical process, or, that the 'degree' of homology between gods and kings can be demonstrated to have evolved from an analogy in The Gupta/Vākāt . aka age to a full homology in the centuries before the establishment of the Delhi Sultanate (Bakker 1992, p. 99).

Mandhal
This successful model set an attractive example for other dynasties to follow. Influence emanating from the Gupta realm becomes best visible in a dynasty to the south of the Vindhyas with which close relations were established, the Eastern Vākāt . akas of Vidarbha.
Candragupta had married a princess of the Nāga dynasty (an old enemy of his father Samudra) named Kuberanāgā, by whom he had a daughter. This princess, who came to be known from her own inscriptions by her dynastic name Prabhāvatī Guptā, was married to the Vākāt . aka prince Rudrasena II in about AD 388. 20 The Nāgas had been a major force in Eastern Malwa and the region around Udayagiri-Vidiśā till the Gupta conquest, and the marriage of a Gupta-Nāga princess with the crown-prince of the Vākāt . akas established a triangular political relationship between the centre of Gupta power in the Gaṅgā-Yamunā doab in the North, Vidiśā in the West, and the Eastern Vākāt . aka kingdom in the South. 21 Rudrasena, once king, inspired it would seem by the example of his in-laws, initiated a tradition of large-scale religious patronage within the Vākāt . aka kingdom. 22 Two inscriptions of Rudrasena himself bear testimony to this: the Deotek Inscription and the Mandhal Plates, Year 5. The Deotek palimpsest stone inscription (Rudrasena overwriting an inscription that apparently pertains to the reign of Aśoka) records the erection of a Dharmasthāna, evidently in the vicinity of present-day village Deotek on the Ben . n .ā or Wainganga River. No further details of this building are known. Forty kilometers to the north of it is the site of the second Rudrasena inscription discovery, which was found while ploughing a field near the village of Mandhal. 23 The locations of both inscriptions indicate that the earliest political heart of the Vākāt . aka kingdom was along the Wainganga, to the south of Nagpur. This may have been the Padmapura district that is known from two inscriptions, one of which is Rudrasena's Mandhal Plates. Although the exact size and location of this district is unknown, the area around the present village of Mandhal no doubt belonged to it. 24 Padmapura itself must have been the capital of this district. This capital features in the so-called 'unfinished Durg' or Mohallā Plate. 25 Closer consideration of the Mohallā Plate leads to the hypothesis that Padmapura was the political centre of Rudrasena II's reign.

Padmapura
After mentioning Padmapura as the place of issue, padmapurāt, the Mohallā Plate begins the dynastic genealogy, i.e., the praśasti, of the Eastern Vākāt . aka kings in words that are identical to Rudrasena's Mandhal Plates up to bhavanāgadauhitrasya (ll.  Shastri), 29 or to "a successor of Pravarasena II" (proposed by Shrimali). 30 The discovery of Rudrasena II's Mandhal Plates has revealed that the praśasti of the dynasty of the Eastern Vākāt . akas was already fully developed by Rudrasena II's time, thus taking the edge off Mirashi's argument that the praśasti of the Mohallā Plate resembles that of the grants of Pravarasena II. 31 The wording of the praśasti resembles indeed the one found in Pravarasena II's inscriptions, but it equally resembles the one of Rudrasena II's charter. The box-headed palaeography of the Mohallā Plate does also not significantly differ from the Mandhal Plates of Rudrasena, nor from Pravarasena's plates for that matter. 32 However, Pravarasena's charters were issued from either Nandivardhana or Pravarapura. 33 Pr . thivīs . en . a I, the most likely author of the Mohallā Plate according to Shastri,34 is an unlikely candidate in my view, since it is questionable whether this king already used the medium of copper plates to record his charters 35 -at least nothing has been found -although, admittedly, he might have been the founder of Padmapura. 36 To finish our review of possible authors of the Mohallā Plate, there is Pr . thivīs . en . a II, the only successor of Pravarasena II of whom we have inscriptions. His charters were issued from, a) the Rāmagiri (Mandhal Plates, Year 2), or b) Pr . thivīsamudra (Mahurjhari Plates), which is unidentified, but may have been a new residence of Pr . thivīs . en . a II, or c) the 26 Cf. CII V, 78 and Shastri & Gupta 1997, p. 155. 27 CII V, 77; Shastri 1997, p. 6. 28 CII V, 77. 29 Shastri 1997, p. 8. 30 Shrimali 1987, p. 113 n. 109. 31 CII V, 77. 32 Shastri 1997, p. 7. 33 Two charters of Pravarasena were issued from military camps (vāsaka), Tripurīvāsaka (Indore Plates, Pravarasena Year 23), and Hiran . yānadīvāsaka (Wadgaon Plates, Pravarasena Year 25). One other was issued from what most likely was a holy place: Narattaṅgavāristhāna (CII V, 49f.). Pravarasena may have made an offering for the salvation of his mother Prabhāvatī at this tīrtha. I take the fact that only the "increase of merit of his mother" (mātr . pun . yopacayārtham, CII V, 50 l. 16) is mentioned, not that of his father, not as an indication that she was still alive, as does Mirashi in CII V, 49, but, on the contrary, as an indication that she might have recently passed away (c. AD 444). This interpretation is suggested not only by the name, which connects the (holy) place with water, but especially by the day on which the grant was made: the twelfth of the dark fortnight of the month Māgha. i.e. S . at . tilādvādaśī, a day particularly suitable for a sesame and pin . d . a offering to the ancestors (Kane V,434). In the same twenty third year of his reign, Pravarasena II had invaded Gupta territory, as we know from the charter that was issued from the military camp at Tripurī (Bakker 1997, p. 25), and it is very unlikely that he had done so when his mother was still alive. 34 Shastri 1997, pp. 6-9. 35 For the earliest Sanskrit copper plates, which belong to the second half of the fourth century, see Salomon On account of the above considerations it is our best guess to ascribe the unfinished Mohallā Plate to the same Rudrasena II who issued the Mandhal Plates, Year 5. However, instead of mentioning Padmapura as the place of issue, as one might have expected on the basis of this hypothesis, these Mandhal Plates specify the god 'Mondasvāmin' as the issuing agency. Yet, there is archaeological and historical evidence that makes it plausible that this authority relates to a Vis . n . u temple in the Mandhal area, that is, arguably, in the vicinity of Padmapura, as we will see below.

Mun . d . asvāmin
In his Mandhal Plates Rudrasena II styled himself as someone "whose rule is based on the cakra, the emblem of the Bhagavat", 39 i.e. Vis . n . u; in other words, just like his father-inlaw Candragupta, Rudrasena professed to be a bhāgavata. The Vākāt . aka kings before and after him, with the exception of Pr . thivīs . en . a II, were all devotees of Maheśvara (Śiva), i.e. Māheśvaras, so it seems that his conversion to Bhagavatism had been part of an antenuptial contract.
And there is more in this inscription, which may be dated to c. AD 400, that links up with Candragupta, who, at about this time, was engaged in his grand Udayagiri project. The opening of the Mandhal inscription runs as follows: Success! Approved. By order of Lord 'Mondasvāmin', God of gods, who holds the conch, discus and sword; after He had made the Nāga king stretch out on the waters of the ocean, He has entered a state of yogic sleep, while lying on the serpent coils of that (king), the Infinite One (Ananta) -a bundle of expanded hoods bending (over Him by way of canopy). 40 The reading of the first lines of the first plate are difficult due to severe damage. The editors Gupta & Shastri read in the middle of line 2: mondasvāminas • ; in the published photograph 37 Shastri 1997, pp. 97-106. 38 See below n. 62 on p. 11. The exact meaning of the terms dharmasthāna and vaijayikadharmasthāna stands in need of further investigation. According to Sircar's Indian Epigraphical Glossary s.v. it means 'temple', 'a holy place'. Mirashi CII V, 4 translates 'place of religious worship'. Shastri 1997, 4f. argues that the term 'was never used in ancient times in the sense of a temple or place of worship'; it refers to a 'court of justice', whereas vaijayika is to be taken as 'an honorific'. 39 Shastri & Gupta 1985, 227 l. 11: bhagavataś cakralaks . mapratis . t . hitaśāsanasya. This expression is changed in the inscriptions of Rudrasena's son Pravarasena into: "whose royal fortune was obtained by the grace of Cakrapān . i, the Lord": bhagavataś cakrapān . eḩ prasādopārjitaśrīsamudayasya (CII V, 12 ll. 13-14) -not merely because, as Shastri observes, this is "grammatically more acceptable", but also because Pravarasena, who had again embraced theŚaiva faith of his ancestors, thought it politically more acceptable; for as a Māheśvara he could not officially declare that the rule of his father, or of his House for that matter, was based ( of Plate 1, I can read • svāminas • , but the reading 'monda' is impossible to confirm. The word monda gives apparently little sense. Dental and cerebral nd are identical but for their top part, in which the square head-mark is replaced by two short curved strokes. Assuming that the reading mon . d . a is correct, we could think of Prakrit influence by whichǒ replaced u before a double consonant without much change of pronouncation. 41 If mon . d . a is not the correct reading -the upper and lower part of the aks . ara are illegible -one could think of reading mau instead of mo, the difference between mo and mau being only that the right top stroke is making an upward curve rather than being a horizontal curve. Or, more likely perhaps, mu, in which the straight stroke down forming the short syllable u has become erased. I tentatively conjecture the reading mun . d . asvāminas • , but mo(au)n . d . asvāminas • cannot be excluded either.
The image described is that of the great Nārāyan . a relief in Udayagiri. But who is this Monda-or Mun . d . asvāmin who gave the instruction (sam . deśa)? Certainly this refers to Vis . n . u-Nārāyan . a, but it remains unclear whether it signifies a local deity (i.e., image) installed in a temple in Mandhal/Padmapura. The editors of the inscription do not come forward with a satisfactory explanation of this name. 42 I would like to suggest that the first part of the name mun . d . a refers to an eponymous person who envisaged and installed this manifestation of the Bhagavat.
As Isaacson and I have shown elsewhere, the 'Kevala Narasim . ha Temple Inscription' on the Rāmagiri mentions the princess that was wedded to Rudrasena by the name of 'Mun . d .ā ' -"a younger daughter who resembled the lustre of the moon", 43 the same lady who, from her own inscriptions, came to be known as the 'Lustrous One', Prabhāvatī. This queen may have used the name Mun . d .ā as long as her husband was still alive and may have changed it into the dynastic 'Prabhāvatī Guptā' after she had assumed power in Vidarbha as the regentess of her minor sons in about AD 405. 44 Mun . d . a-or Mon . d . asvāmin, and even Maun . d . asvāmin interpreted as a Karmadhāraya compound, may thus be taken to mean 'Lord of Mun . d .ā ', 41 Von Hinüber 2001, p. 121. 42 Shastri & Gupta 1997 propose to connect this with a toponym by pointing to two present-day villages in the neighbourhood called Maudā and D . oṅgar Maudā. It seems more likely, however, that the toponyms preserve the ancient name of the temple than that the temple-name derived from these toponyms. 43 Kevala Narasim . ha Temple Inscription (KNS) vs. 7: y avīya sī candra(ma)saḩ (p)ra(bh) eva sā mun . d . anāmnī tana(yā) (Bakker & Isaacson 1993, p. 53). 44 Bakker & Isaacson 1993, p. 53;Bakker 1997, p. 16. The KNS inscription may hint at the dynastic name, when it compares the girl Mun . d .ā with the light (prabhā) of the moon (candra): Prabhāvatī is 'the one who possesses the lustre', scil., of Candra, her father. Tedesco 1945 proposes to derive MI mun . d . a from Skt. vr . ddha. This is not taken over by Mayrhofer, EWA s.v., who gives as first meaning: 'kahl' ('-Unklar'). As an appellation of a princess, the name 'mun . d .ā ' ('the bald one') is odd. One may conjecture that the name refers to the tribal background of the princess, the Mun . d . as, in which case these Mun . d . as must have been part of the Nāga confederation mentioned in Prabhāvatī's inscriptions; the Mun . d . as are known, for instance, from MBh 3.48.21 (present at Yudhis . t . hira's consecration), MBh 6.52.9b (App.) and MBh 7.95.20; VāP 45.123. Cf. Murun . d . a-devī/svāminī, mother of the Uccakalpa kingŚarvanātha (CII III (1888), 127, 131), a queen obviously named after the Murun . d . a people from which she came (this people is mentioned among the 'foreign countries' in Samudragupta's Allahabad Pillar Inscription (CII III (1888), 8)). We possess no epigraphical corroboration of the existence of the Mun . d . as, however. Another explanation for the name Mun . d .ā could be: the girl was bald (mun . d . a), i.e., suffered from alopecia. Her bald head may have resembled the lustre of the full moon. Lüders classified the name Mun . d . a (Prakrit: Mud . a), which is found as the name of a donor on a railing at Bharhut (A 102 (827)), in the category of 'names derived from the appearance of the body' (CII II.2 (1998), pp. 53, 6). That the name Mun . d . a could be a brahmin name (Prabhāvatī belonged to the Dhāran . a gotra) is proven by the Grant of Svāmidāsa, Year 67 in the copper-plate hoard from Bagh, which has preserved the name of a donee, Mun . d . a, a brahmin belonging to theŚān . d . ilya gotra (Ramesh & Tewari 1990, p. 65).
'Mun . d .ā 's Lord' that is the 'God of Prabhāvatī'. As far as I am aware, this is the first attestation of the practice to name the installed deity after a living founder (svanāmnā). 45 The aim of Rudrasena's Mandhal inscription is not the installation of this deity, but rather the donation by the king of four villages in the eastern division of Padmapura to a pair of settlements (adhivāsa) of the Sātvata caran . a (branch/school), which came from Vatsagulma, the capital of the Western Vākāt . akas. 46 This is the earliest attestation of royal support of a Bhāgavata organisation named Sātvata, a religious sect known, for instance, from the  and the 'Tusam Rock Inscription'. 47 The Sātvatas worshipped Vāsudeva and the four other members of the Vr . s . n . i clan; they originated from western India. 48 The invocation of the god of his wife strongly suggests that Rudrasena was acting on her behalf: the queen hiding herself behind the god Mu(o/au)n . d . asvāmin at whose command the grant was made. The heuristic principle cherchez la femme has again proven its value.
The Mandhal inscription thus seems to prove that the Vākāt . aka queen (née Gupta) followed in her father's footsteps by favouring religious organisations and building temples. She may have been the major source of inspiration of her husband and have brought the Sātvata ritual experts from Vatsagulma to Padmapura in order to serve in her new temple.
Archaeological excavations in Mandhal have brought to light three temples, the most prominent of which stood on top of a hillock named Bom . gī Hud . kī (BHK II); this temple was built on a massive platform (11.70 × 14.70 m), with a garbhagr . ha and man . d . apa. The platform was partly built into the surface of the rock, which was levelled by 32 layers of brick at the southern side. 49 At the foot of this hillock excavations revealed another building (BHK I); underneath its brick floor a dozen of Hindu sculptures were found, among which two or three images of the Sātvata deities: Vāsudeva, Sam . kars . an . a, and (possibly) Sāmba (Āditya). 50 Whether or not this Vāsudeva image represented the Mun . d . asvāmin is impossible to say. The fact that, next to these three Vais . n . ava images, five stylistically very similar images of markedlyŚaiva nature were discovered, 51 testifies to the apparently harmonious way by which the Bhāgavatas and Māheśvaras existed side by side, tapping the same royal source of subsistence. 45 Cf. the Prabhāvatisvāmin mentioned in the 'Kevala Narasim . ha Temple Inscription,' installed after the same queen's death for the sake of merit transfer (Bakker & Isaacson 1993, pp. 54, 69;Bakker 1997, 30, pp. 140-142). The shortening of the feminine suffixesā andī to a and i is permitted by P. 6.3.36. This naming practice created a legal fiction, individualising the god so that he could be endowed with property and authority (cf. Sanderson 2003-04, n. 250). For naming deities after the donors see also Willis 2009, p. 139. The two liṅgas (?) Upamiteśvara and Kapileśvara mentioned on the Mathura Pilaster (AD 380) were named after the deceasedācāryas Upamita and Kapila (CII III (1981) Bakker 1997, pp. 59-62. 49 Shastri 1975Bakker 1997, 80f. 50 Bakker 1997, Plates XIX-XXIV. 51 Bakker 1997, Plates VII-XIV, XXVa.

Rāmagiri
Rudrasena died young. His sudden death may account for the fact that the Mohallā Plate remained unfinished. His Gupta queen assumed power, officially as regentess of her minor sons. Her Poona Plates, dated in the thirteenth year of the yuvarāja Divākarasena, i.e., around AD 418, show that momentous decisions had been taken since the Mandhal foundations. 52 The dowager queen had her residence moved northwards, to Nandivardhana at the foot of the Rāmagiri, which hill she was developing into a state sanctuary that could match her father's Udayagiri. Again a hilltop was chosen; two minor caves were excavated on its slope, but the main Bhāgavata sanctuaries were stone temples on the top, five of which survive till today. The Poona Plates convey the impression that Prabhāvatī was personally involved in promoting the cause of the Bhāgavata religion. The charter was issued on the twelfth of the bright half of Kārttika, after breaking the fast of Prabodhinī Ekādaśī, Vis . n . u-Nārāyan . a's awakening from his cosmic sleep (earlier referred to in Rudrasena's Mandhal Plates). This festival of cosmogonic re-awakening had an important political dimension, as has been shown by Michael Willis in the case of Candragupta and Udayagiri, 53 a case that must have been very well known to the queen. Styling herself as entirely devoted to the Bhagavat (atyantabhagavadbhaktā), at this occasion she bestowed land upon another bhagavadbhakta, thē acārya Canālasvāmin, who may have been one of the priests responsible for the waking rite and who may have been the queen's own teacher. 54 The grant was first ritually offered (nivedya) to the bhagavadpādamūla, the feet of the Lord. As we learn from Prabhāvatī's R . iddhapur Plates, which mention the pādamūla of the Lord of the Rāmagiri (rāmagirisvāmin), 55 this may be taken to refer to the main sanctuary on top of the Rāmagiri. 56 A parallelism with Udayagiri can even be extended further, since the oldest image that survives on the hill is that of the Boar, Varāha, lifting the world (Bhūdevī) on his tusk out of the cosmic ocean. That the idea was taken over but not slavishly copied, appears from the fact that the Rāmagiri Varāha is theriomorphic instead of half man half boar. 57 At Nagardhan and environs, the area identified with Prabhāvatī's residence Nandivardhana, (fragmentary) images have been found of Vis . n . u, but also of Gan . eśa and the goddess Mahis .ā suramardinī, whereas the discovery of a hoard of three splendid Buddhist bronzes, three prabhāvalis, a parasol, a bell and an incense-burner at Hamlapuri (a village near Nagardhan) testifies to the presence of the Buddhist Saṅgha close by the court of the Vākāt . aka queen. 58 52 Poona Plates of Prabhāvatī Guptā, CII V, pp. 5-9. 53 Willis 2004, 37-41;Willis 2009, p. 46. 54 CII V, 7f. Willis 2009 Sircar has pointed out (Sel. Ins. I, 512 n. 2) that pādamūla, literally meaning 'sole of the foot,' is frequently used in inscriptions as a metaphor of a deity in a temple to whom respect is paid. This deity may have been represented by an image (iconic), not necessarily as his aniconic footprints. cf. Bakker 1997, 136f. 57 Bakker 1997, 138f. 58 Jamkhedkar 1985. The loose pedestal is inscribed in nail-headed 4th/5th century script ( Within this court Prabhāvatī's three sons grew up, but of only the youngest of them Pravarasena, do we possess records. These show that the prince had returned to the faith of his ancestors, since he called himself entirely devoted to Maheśvara. 59

Mansar
For several years Bhāgavatas and Māheśvaras were living together in the Nandivardhana residence, but somewhere between the eleventh and sixteenth years in the reign of Pravarasena, the latter had decided that he should construct his own residence, Pravarapura, and his own state sanctuary, Pravareśvara, both named after the king himself. 60 For these two building projects wonderful sites were found, which mirrored, as it were, the old residence and the Rāmagiri: five kilometers to the west on a hill, again, the Pravareśvara complex was erected, facing the Rāmagiri, whereas at its foot the residence Pravarapura arose. Excavations at these two sites near the village of Mansar (MNS III and II) during the last ten years have brought to light a wealth of material, which has been the subject of a symposium in the British Museum in 2008. 61 The personal seals, both, of Pravarasena and Prabhāvatī found in MNS II, suggest that the Vākāt . aka king stayed close to his mother and that she, at an advanced age, remained a powerful figure behind the scenes, 62 also in the new palace of her youngest son, which was her third and last residence since she had left the Gupta court of her youth -Padmapura, Nandivardhana, Pravarapura. When the Gupta-Nāga princess, the queen mother of the Vākāt . akas, eventually died in her early seventies in c. AD 443-444, a funeral monument (ed .ū ka) was erected, as I have argued elsewhere, at a location in between the Rāmagiri and Pravarapura, a monument that was appropriately guarded by a Nāga. 63 In the present context it may suffice to evaluate the recent discoveries at Mansar in as far as they throw light on the theme of this essay, royal patronage, religious tolerance, and the formation of Gupta-Vākāt . aka culture.
The magnificent state sanctuary (devakulasthāna) erected on the Hid . imbā Tekd .ī (Pravareśvara) near the village of Mansar (MNS III), served as the place where the ceremony of issuing royal charters was performed (dharmasthāna), as attested by the 'Pān . d . urn .ā Plates of Pravarasena II, Year 29', i.e., AD 441. 64 A generous land-donation was granted to a group of Vājasaneya brahmins, and the same inscription records that one of the donees (Somārya) received some more land later, after a ceremonial offering of sesamum (tilavācanaka), i.e. a srāddha ritual, had been performed in that dharmasthāna. 65 The king's largesse towards his subjects was immortalised in the benign smile of the mainŚiva image of this temple, the god who bestowed life on his devotees by extending his right hand to them, filled with flowers, in a gesture of benevolence (varadamudrā). 66

Conclusion
In the four decades or thereabouts following Candragupta's inauguration of his Udayagiri monument at the turn of the fifth century, his daughter (Mun . d .ā /Prabhāvatī) had consecrated her sanctuary in Mandhal, probably on the Bom . gī Hud . kī, dedicated to Mun . d . asvāmin, and after that the sanctuaries on the Rāmagiri (Rāmagirisvāmin), and his grandson (Pravarasena) had founded his own state sanctuary, Pravareśvara, on a hilltop near the village of Mansar. Yet, as is the case in the Gupta kingdom, we have no documentation of the founding of these royal sanctuaries by the Vākāt . aka monarchs themselves. 67 The fact that, 1) these sanctuaries bore the names (Mun . d . asvāmin, Pravareśvara), 2) were alternatively dedicated to either the Bhagavat or Maheśvara, and, 3) served as hallowed places for making large donations to various religious groupings (Sātvatas and Vājasaneyins), shows however, that the policy of religious patronage initiated by Candragupta and his court was successfully continued in the Eastern Vākāt . aka kingdom, where it took on more and more the form of land donations. 68 And later a similar policy was practised by the Dharmamahārājas of the Western Vākāt . aka branch in Vatsagulma, in whose kingdom we find the famous Buddhist cave monuments of Ajanta. Though here too we witness the same curious phenomenon: the lack of charters of foundation issued by the reigning king, Haris . en . a himself.
Although the kings showed clearly their religious orientation, we have no evidence whatsoever that could point to great enmity between the religious organisations, although rivalry for patronage as well as clients no doubt did play a role. The first four decades of the fifth century, dominated by the rule of Candra-and Kumāragupta in the north and Prabhāvatī Guptā and Pravarasena in the south, may be counted among the most stable and peaceful periods in ancient Indian history. It is this stability established by good policy and tolerance that created the conditions for an unprecedented development of the religious imagination, reflected in the literary and visual arts, achieving a perfection that still impresses today.
The iconographic programme developed on the Udayagiri mountain was only partly taken over in the Vākāt . aka kingdom: images of Vis . n . u, Varāha, Narasim . ha, Gan . eśa, and Mahis .ā suramardinī are found at both locations, all nearly in the classical form that would remain normative for the centuries to come. However, in Vidarbha this sophisticated artistic tradition blended with a local form of craftsmanship which lent to it its peculiar, somewhat rustic flavour and which gave rise to creative experiments that were not all equally successful. 65 CII V, 66. Kane 1930-62, IV, 418f. 66 Bakker 1997 We possess a founding document of Skandagupta, viz. his Bhitari Stone Pillar Inscription, in which the Gupta king records his installation of the deity Kumārasvāmin (name uncertain, see Bakker 2005, p. 249;cf. Willis 2009, 144ff.), and of Prabhāvatī's daughter Atibhāvatī (name uncertain), viz. the KNS Temple Inscription, which records the installation of Prabhāvatisvāmin beside an artificial lake Sudarśana (Isaacson & Bakker 1993, p. 54;Bakker 1997, pp. 30-33). 68 For a survey of the Vākāt . aka land donations see Shrimali 1987. The sculptures found in Mandhal are a good example of this. But in the course of a few years since the Mandhal inception, the classical images on the Rāmagiri of Varāha, Trivikrama, and Narasim . ha were created in the mature Vākāt . aka idiom, which surpassed their Udayagiri counterparts in artistic quality and refinement. The workshop set up by Prabhāvatī for the production of high quality, red-sandstone sculptures continued to work under her son and reached an unmatched level of sophistication and originality in the sculptures that once adorned the Pravareśvara Temple. Although the research of these sculptures is still in its infancy and their iconography is still little understood, their appeal lies in their tranquility that expresses mood and emotional state (bhāva) rather than action (vikrama), sentiment and emotion (rasa) rather than force (bala), and as such they are characteristic of the best that Indian art has produced.