The hilltop monastic settlement at Giriyaka is situated on an extension of the Rājagriha hills. Rājagriha was destined to be the earliest capital of Magadha as well as epicentre of multiple religions, especially śramanic. The political as well as religious glory of the city is visible in the physical remains of historical structures, archaeological artefacts, solitary caves, and lofty hills. The geographical propensities of the region like hills, forests, caves, hot springs, water resources, and vast agrarian lands were the catalyst to the kings and mendicants alike to cherish their dreams. The ancient city was surrounded by five major hills-Ratnagirī, Vipulagirī, Sonagirī, Udaigirī, and Vaibhārgirī and the city was protected by a fortification wall (Bimbisāra or Cyclopean wall) which is 40 to 45 km in length and approximately 17.5 feet thick. (Figure 1) The wall had sixteen rectangular bastions meant for defence. At one side the river Bānagaṅga is situated along with Jethian forest and Veṅuvana and at other Giriyaka hill with river Paṅchāna at its foot (Marshall 2002: 87–89).1 The name Paṅchāna is told because of her five streams meet together in the south of Rājagṛiha. The river has been identified as Sappinī of Pāli literature (Law 2016: 38).2 The name indicates the serpentine movement of the courses of the river. It meanders through the Rājagriha and Nālandā and finally merges in the Gaṅgā. The Rājagriha hills are composed of quartzite rock and extended in two parallel ranges originating from Oriar Khurd-Bhindas villages in the neighbourhood of Gayā which further moves to the north-east up to 58 kilometers to Rājagriha. The eastern termination of the southern range is comparatively flattened, but the northern hills keep their height, and meandered through two lofty peaks drooping towards the river Paṅchāna is mainly identified as Giriyaka. The lower hill of this range in the east has a brick stūpa locally known as Jarāsaṅdha-kī-Baithak (abode of King Jarāsaṅdha) and the higher peak has an quadrilateral walkway enclosed with remains of stūpa/vihāras facilitated with a steep flight of steps leading through a pillared room (Cunningham 2000, I: 16). In the local tradition, the eastern hill is acknowledged as Mamubhagṅā or Phulwariā Pāhar and the western ridge is known as Hawelia Pahār (Broadley 1872: 272). These names seem to be mythologically related with Jarāsaṅdha as Phulwari in local dialects is known as garden and the site is identified as Jarāsaṅdha’s garden.
Haweli means mansion or palatial house showing that the remains of monastery and stūpa were later accepted as the palace of Jarāsaṅdha. Mamu-bhagnā is also associated with maternal uncle-nephew dyad and had some mythological link of Jarāsaṅdha to King Kaṁsa of Mathurā. Buchanan linked this place to Girivraja, the old name of Rājagriha, and says that in passing time it got vulgarised to be known as Giriak/Giriyaka. On basis of Faxian’s account of an isolated hillock, Cunningham accepts its name as Giriyaka. Beglar seeks the mythological connection of this hill and says that the hill was named on queen Girikā, wife of King Vasu who once ruled this region (Patil 1963: 148–149). The archaeological remains of NBPW phase puňca-marked coins, moulded terracotta figurines, and the existence of ring wells suggest the beginning of state formation in Rājgriha since the 6th century BCE (Barba 2004). Ghosh has explored the archaeological artefacts scattered in the valley and found evidence of Northern Black Polished Ware (N.B.P. W.), fragments of coarse undefined pottery, and two post-cremation burials. The pottery in Rājgir is likely to go to the earlier period. The horizon with the coarse ware and the post-cremation burials may be linked to the local prehistoric culture (Chakrabarti 1976: 265; Agrawal, and Kusumgar 1974: 136). Giriyaka was well associated with this socio-economic formation.
From the village side, a pathway was created connecting the habitation with the monastic structures lying on the two hills (Figure 2).
Buchanan informs that a connecting road was built with stone blocks taken from the hill. It is approximately twelve feet wide and meandering through the hill to reach the top (Jackson 1925: 112). Technically, the construction of the path way is very crude showing its antiquity and hoary past. The closer examination of the pathway suggests that at one nodal point, it break opens trespassing the Bimbisāra wall to create the passage for the road to reach the monastic settlement on the top of the hill. The same stones were used as in the fortification wall. It suggests that this path up to the monastic site was built in the same period when the fortification wall was built. The presence of small rectangular bastion-like structures in this area gives the possibility that monks who resided at the hills were permitted to move for their alms round from this path with the permission of the existing authority. The remains of the Bimbisāra wall and bastion like small structures in this area suggests that it was protected by state.
The natural vegetation, deep forest, isolated hills, and rocky caves around Rājagriha have always been a fascination for the monks and nuns. Some of these places were the abodes of prominent monks and some important Vinayas were composed and sermons were delivered here (Cullavagga IV.4.4). Even the Buddha had high praise for the city of Rājagriha and Gijjhakuta hills (Dīgha Nikāya II: 116). Mahākassapa resided in Pippali-guhā. Once the Buddha also visited this place to meet him when he fell sick. This cave was also used by the later monks for meditation (Udāna I: 4; Saṁyutta Nikāya V: 79). On basis of its proximity to the Asura palace, B C Law identified it at Vaibhāra hill to the west of the hot springs (Law 2016: 252). The caves at Gijjakūta, Asura, Giddhdwāra, etc were few of them which became dwellings of numerous ascetics. Description of the two caves, one at Giriyaka hill which I call ‘Devadatta-guhā’ and other popularly known as Indrasāla-guhā will be a valuable addition to examine antiquity of Buddhism in Rājagriha.
Identification of Devadatta-guhā
When one proceeds to Giriyaka hill from Rajgir by Nawada-Hiswa road, reaching the nodal point that turns towards the hill just before the river Paṅchāna, after moving about 200 feet on the right side on the hill, a cave exists still hidden from the keen eyes of the archaeologists/explorers (Figure 3).
This cave, therefore, is not described earlier by any colonial archaeologist. The cave is one of the best caves of the region and can be rated only after the Sattapanni, Barabara, and Nāgarjunī clusters of caves. The physical verification of the cave suggests that it is not a natural cave but man-made. The chiselled marks are still showing on the wall of the cave. It has a wide hall with one small cavern at the right side, meant for the meditation by the monk who occupied it. Outside the cave, a small cavern is also excavated. The cave is undoubtedly Buddhist as the Dharmachakra with six lotus petals is engraved at the entrance of the cave (Figure 4). The most unique feature is the stairs (steps) i.e. ten in number to get access to the main hall of the cave which is like a parallelogram measuring sixteen feet long and ten feet wide (Figure 5 & 6). Some imprints of vermilion, candaṇa, and akṣata have been visible on the outer wall of the cave (Figure 7).
Later on, this cave might have been be used by some tantrika ascetics. The cave had chiselled marks but does not exhibit polish which shows its pre-Maurya origin. This whole monastic establishment was on periphery of Rājagriha and in all respect, was excavated by Ajātśatru for his spiritual mentor Devadatta. I would not hesitate to call it ‘Devadatta-guhā’. This extraordinary structure required extensive labour and huge financial resources which only the state could bear. The structural arrangement suggests Ajātśatru replicate the model of Gijjhakuta cave where the Buddha resided in one cave, Ᾱnanda always stayed as an attendant in an adjoining cave. The same structure is developed in Devadatta-guhā. Just outside the cave, a small cavern was intentionally developed for the attendant monk. The Pāli literature mentions that Devadatta replicate this idea from the Buddha (Dhammapada Aṭṭhakathā, I: 122). Faxian mentions the stone cell of Devadatta and flat rock near it outside Rājagriha (Beal 2005: 118). Xuanzang also saw a cave known as the Devadatta samādhi that was placed outside Rājagaha and besides it, there was a stūpa. He also informs about the flat rock of black colour where an ascetic performed suicidal practices and oozed blood, the red stain of the blood is visible (Watters 1973, II: 155). Adjacent to the cave, moving upward the rocks are black in colour and a large natural platform is still present with the comparatively smooth surface may be the apparent rock seen by the Chinese travellers but they were confused with red vermillion marks as blood stains. The red stains may be a mark of vermilion spotted on the surface of the rocks that were used by tantric practitioners of later period. Some of the marks are still visible on the walls of the cave. The monks seen by them could be the followers of Devadatta who used to practice dhutāṅga (extreme self-mortification) meditation.
Beglar on Cunningham’s description of Xuanzang tries to identify a cave on Vipulagirī as Devadatta-guhā (Beglar 1878: 185). Though, Cunningham does not give any real identification or description of the cave. Ghosh identifies a cave on the northside of Vipulagirī near Makhdum kund and hot springs as place for Devadatta. He says that a piece of stone with crimson marks is believed to be blood stain seen by Xuanzang and it could be the platform where the the monks practiced meditation. This cave with flight of steps was later on used by Makhdum Shah Sharfuddin for meditation (Ghosh 1987: 40). But it can be refuted on many grounds. Xuanzang mentions that Devadatta-guhā was situated outside Rājagriha (Watters 1973, II: 155–56). Vipulagirī was always the seat of the Jains. Tirthānkara Mahāvīra spent many vassavāsa here and Muni Suvrat spent his entire life on this hill. They both were contemporary of the Buddha. It is a little possibility that Buddhists would encroach the caves and forest zones wandered by the Jains. The cave mentioned by Ghosh is near to cluster of caves occupied by Mahākassapa and his followers. Devadatta would not live near the sacred space dominated by his opponents. The cave that I have described is still in outskirt of modern Rājgir, lying on the border of district Nawada of Bihar can be acknowledged as Devadatta-guhā.
Parvati-guhā as Indrasāla-guhā: some arguments
Cunningham identifies Giriyaka hill as Indrasāla-guhā where the Buddha preached Sakka. He proposes that both Faxian and Xuanzang inform that the cave is situated on the southern side of the hill, which could be identified as Gidhadvar. It is 3.5 km to the southwest of the Giriyaka and one and a half km from Jarasandha’s stūpa. Cunningham says that ‘Hill of Isolated Peak’ corresponds to Giri+eka, single hill (Cunningham 1871: 472–73). Faxian mentions that from Pataliputra to south-eastern direction after travelling nine yojanas a small rock hill standing by itself on the top of which is a stone cell facing the south. On one occasion when the Buddha was in the middle of the cell, Sakka visited him with his attendant musicians and asked 42 questions to the Buddha, writing each of them simply with his finger on stone. The traces of writing still exist (Beal 2005: 110–111). Xuanzang says that Indasāla-guhā has two peaks with flat gorges covered with vegetation. The south side of the West had a cave in which the Buddha used to reside. Here Indra asked 42 questions to remove his doubts (Watters 1973, II: 172–73). A. M. Broadly has contradicted Cunningham’s discovery and proposed hill of Bihar Sharif as Indrasāla-guhā. He on the description of Faxian, says that Cunningham’s identification of Indrasāla-guhā is hypothetical and based on inaccurate data. Both hills are entirely different (Broadley 1872: 262–65). Broadley says that Giriyaka was not a solitary hill and here the rocky ranges of the Rajgir hill get terminated. Secondly, from the solitary hill, Faxian proceeded south-west one yojna to Nālā but it is only exactly ten kilometres north-west of Giryaka. For this region, the hill of the solitary rock could not be Giryaka but the hill of Bihar Sharif (Broadley 1872: 284–286). Based on Viradeva’s inscription of 9th century CE, he identifies Indrasāla-guhā near Ghosarwan, thirteen kilometres to the great Nālandā Mahāvihāra, eleven kilometres southeast of the isolated rock of Bihar, and ten miles north-east to the ruins of Rājagriha. The inscription found from Ghosarwan informs that Viradeva in devotion of the saṁgha, erected for the good of the world two crest-jewels in the shape of the caityas on the crown of Indrasāla (Broadley 1872: 266–73). However, Cunningham did not accept Broadley’s argument and countered that Faxian description of direction and distance is not accurate and on basis of Xuanzang accounts of Pataliputra, Nalanda and Indrasāla-guhā the exact location is Giriyaka.
The Pāli literature mentions that Indasāla-guhā is in Vediyaka mountain which is a natural cavern in the rock. It is narrow, dark, and uneven. The cave is surrounded by a wall fitted with doors and windows and decorated with scrolls and floral designs. The blessed once resided here (Dīgha Nikāya II: 268–69).3 Pārvati can be the most suitable place for this cave as it has two distinct hills, one is flatter and presently occupied with a Muslim grave and the second is comparatively high with monastic remains. At the cliff of this second hill lies a natural cave but some modification as chiselled marks are apparent. The cave is more approachable to live and meditate. It is not steep like the Gidhadwar and the Buddha visited this region frequently. Buddhaghoṣa informs that it was a natural cave situated between two hills but the hill where the cave waw existing known as Vediyaka or Vediya hill because it was surrounded from all sides with blue rocks (Sumaṅgalavilāsinī III: 697). Buddhist literature inform that the Buddha preached the Sakkapaṅṅa Sutta near a brāhmin village Aṁasaṅda. This brāhmin village can be identified with village Apsadh which is still dominated by the bhumihars and brāhmins. A M. Broadley has raised two important points in his archaeological discoveries in the Bihar region. On basis of Viradeva’s inscription found from Ghosrawan, which mentions that the king erected two crest jewels in the shape of caityas on the summit of Indrasāla, he identifies Ghosrawan as the place where the Buddha preached the Sakkapaňňa Sutta. Secondly, he identifies Ᾱṁasaṅda as Ghosrawan which was inhabited by people of ‘Babhan’ (brāhmin) caste till 1857. They raised a banner of revolt against the British but were subdued brutally by them. Their whole village was destroyed, many people were killed and their properties were burnt down (Broadley 1872: 263–266, 273). Ghosarwan is not a hill and it may be a possibility that Viradeva raised crystal jewels at Parvatī. Hirananda Sastri on basis of the stone inscription of the reign of Yaśovarmadeva identifies Ghosarwan with Yaśovarma Vihāra (Sastri 1999: 78–79). It reaffirms that Indasāla-guhā was at Parvatī.
Stūpas and Vihāras at Giriyaka: Antiquity and Structure
Buddhist settlements discovered at Giriyaka hill in the colonial period have brought to the limelight many early stūpas along with monasteries. On the western peak which is very steep, a whole monastic settlement has been identified. The majority of parts of this structure are under the heap of broken bricks and other earthen materials Broadley measures it twenty feet high and one hundred and seven feet wide with visible remains of two or three stairs (Figure 8). This place is acquainted with a ledge and was well protected by a strong parapet of stone blocks (Figure 9).
The site also has remains of a stūpa on a 45 square feet quadrangular base and adjoining the stūpa, the remains of monastic complex and cells raised on stone pillars are quite noticeable (Broadley 1872: 209–215, 263). Buchanan also informs the remains of two or three steps and the top of this steep structure is a hollow space with a copious ledge around it. It could be an open courtyard well protected by a stone wall. The terrace surrounds the whole structure from all sides. He also informs the existence of a tapering mass of bricks i.e., stūpa with the observable square platform (Jackson 1925: 112–113). The site gives the impression that two stūpas were built one small with a square base and the other still under the heap that may be stūpa in the middle surrounded by a terrace from all sides. It could be an elongated structure formed of a succession of steps with niches meant for decoration, alternating with pillars, and in the middle lies a stūpa. Such types of stūpas were prevalent right from the Magadha region to Odisha, Andhra Pradesh, and Gandhāra. The material remains of stūpas and vihāras, masonry work in stone, size of bricks of this area suggest that the first structure was raised in the pre-Maurya period and continued up to the early medieval age.
A huge protective wall of chiselled stone blocks was raised from the bottom of the hill from the western side to balance and protect the stūpas and adjoining vihāra on the hill. This monastic establishment was connected by a causeway to a water reservoir in which the same stone masonry was used. The same category and size of chiselled stone blocks were used in the Bimbisāra fortification wall which leaves no doubt that the earliest structure on the hill and the parapet to protect it was built by King Ajātśatru as he was a builder of new Rājagriha and this site falls in this area. About a little distance from the foundation of the parapet, a natural pond is existing showing that water of it could be used by the monks residing upside the hill. This monastic complex was well connected with other Buddhist settlements of Ghora-Katora and other centers of Vaibhār hill. Broadley reported the remains of the monastery on the steep hill covered by broken bricks and stones near the causeway. He says that the foundation made of bricks (18 inches x 9 inches x 2 inches) was intact. The brick was joined by clay mortar. The ruins of pillars supported the roof are scattered in the area (Broadley 1872: 263; Jackson 1025: 111–113). The remains belong to the Gupta period however the first monastery was built along with the stūpas and parapet wall in the pre-Maurya period (Figure 10). The pillars are square at the base and octagonal in the upper part showing the Gupta characteristics (Figure 11). The chiselling marks on the surface suggest some kind of ornamentation on the pillars. One of the pillars found in the existing temple is well ornamented with numerous stūpas.
Such a glorious structure could be compared to the stūpa built on the bank of river Bāṅgangā, and excavation of the Sattapanni caves by Ajātśatru. It keeps us wondering that to whom this stūpa complex was dedicated. My observation zeroing in for Devadatta. It is also possible that after the death of Devadatta this place was occupied by Ᾱnanda and his followers. Chinese travellers mention that Ajātśatru made relic stūpas in Rājagriha for Devadatta and Ᾱnanda. It can be a suitable place for Devadatta as he emphasised to live in the forest (āraññaka); not to the neighbourhood of a settlement; to get alms by begging (piṇḍapātika); to dwell at the foot of a tree (rukkhamūlika) etc., (Vinaya Pitaka, III: 171). The Dadhi-Vāhana Jātaka and Pañca-Paṇḍita Jātaka narrate an instance that Ajātasattu built a monastery for Devadatta. The king used to send five hundred pots of sumptuous dishes for him and his wandering community. Even this luxurious food was also shared by the followers of the Buddha (Jātaka 186; 508). Thus, it appears that Devadatta continued to have the support of Ajātśatru. Though Devadatta was considered as a man of demonic habits in early Pāli literature but his dissidence was indeed a kind of his adherence to certain austerities, which the conventional community of the bhikkhus was not keen to follow (Sarao 2014: 31–48). The Sarvāstivāda-Vinaya informs that for twelve years after his admission into the saṁgha, Devadatta conducted himself with faultless deeds and judgements. He got perfection in the Vinaya and the Dhamma and strictly followed truthful and disciplined life (Mukherjee 1966: 120). In the Saddharmapuṇḍrīka Sūtra, Devadatta is depicted in his previous life as a forest renunciant who helped the Buddha Sākyamuni to attain Buddhahood (Saddharma-Puṇḍrīka XI: 46). Giriyaka was an ideal place which had all requisites demanded by him. In a similar fashion story of Ᾱnanda has been mentioned in the Atthakathās. When Ᾱnanda desired to take mahāparinibbāna and proceeded towards Vaiśalī. On hearing this, Ajātśatru and Vajji saṁgha both requested Ᾱnanda to take mahāparinibbāna in their territory. To avoid any kind of conflict between the two Buddhist kingdoms, Ᾱnanda performed a miracle on the bank of river Rohinī and his body was amalgamated in the flame. His relics were divided into two parts and Ajātśatru made a huge stūpa at Rājagriha over his share (Dhammapada Aṭṭhakathā, II: 99ff). Giriyaka could be ideal place for it.
Buddhist literature suggests that Ᾱnanda had great respect for Devadatta and Mahākassapa. When the Buddha visited Kapilvastu, Devadatta along with Ᾱnanda, Bhagu, Kimbila, Bhaddiya, Anuruddha, and Upali were converted into saṁgha. Devadatta was the first to achieve arhathood (Vinaya Pitaka II: 182). The association and respect among them never faded and despite Devadatta’s ire against the Buddha, he was never abandoned by their cousins. The proceeding of the First Buddhist Council suggests that despite Ᾱnanda’s prominence, he was not taken kindly by Mahākassapa and their followers. Several trivial charges were levelled against him to demean his image. Though he pleaded his innocence and achieved arhathood to be a compiler of the Sutta Pitaka. Even, he was not treated enlightened by the other senior arhats. Malalasekera points out that despite his great acumen and erudition, Mahākassapa felt jealous of other prominent monks like Ᾱnanda and others. It may be the reason that he did not welcome him in the first Council. Mahākassapa did not have positive vibes for the bhikkhunis as evident from his spat with bhikkhunis Thullatissa and Thullananada (Malalasekera 2007, II: 480–81).4 Both the nuns had a great veneration for Devadatta and Ᾱnanda but hatred against Mahākassapa. Once Mahākassapa with Ᾱnanda visited to preach bhikkhuni saṁgha, Thullatissa became annoyed by his attitude and mocked Mahākassapa ‘as the needle pedlar try to sell a needle to the needle maker (Saṁyutta Nikāya, II: 215ff.). Thullananada was an erudite nun with vast expertise in Dhamma. She was an ardent follower of Ᾱnanda and also had respect for Devadatta. Once Mahākassapa called Ᾱnanda ‘a boy’, then Thullananda scolded Kassapa. She became so infuriated that she left the saṁgha (Vinaya Pitaka IV: 66, 182, 211–15, 233, 239–40, 248–50, 335). Though Ᾱnanda became the successor of Mahākassapa frequent interactions and sharing of common abode were probably avoided by the followers of both groups. The southern vicinity of Rājagriha i.e. Ekanāla and Dakkhinagirī was dominated by Purāṅa and his followers who even challenged the proceedings of the First council. The Dharmaguptaka and Mahīsāsaka Vinayas have referred this incident and Purāṅa of Dakkhinagiri has been mentioned as the person who introduced seven rules in the Vinaya of Mahīsāsaka (Datta 1998: 122). It seems Purāṅa ushered in great tradition of the Porāṅṅikas and his followers later founded Mahaiśāsaka sect. The followers of Sarvāstivadins seek their lineage from Ᾱnanda and his disciple Sambhuta Sanvasi founded its center to Mathurā and Majjhantika led it Kashmir and Gandhara (Datta 1998: 128–129). At Vaibhār hill Mahākassapa and his followers had widespread influence. Gijjhakūta, Pippali, Asura, and Sattapanni groups of caves belonged to Mahākassapa and his followers. It was a traditional bastion of the bhikkhus who later developed the Kaśyapiya sect. The vicinity of Rājagriha from Giriyaka, to Ᾱṁasaṅda (Apsadha) village, was much frequented by the Buddha and Ᾱnanda. It might be a possibility that Ᾱnanda and his followers had a main stay on this hilltop monastery. The textual tradition indicates that just after the mahāparinibbāna of the Buddha, the doctrinal difference among the prominent Theras developed and their followers subsequently accentuated to their Master’s view that gave rise to three important sects from Rājagriha-Mahaiśāśaka by Purāṅa and his followers from Dakhinagirī, Kaśyapiya-Mahākassapa and his followers from Gijjhakūta-Pippali-Asura-Sattapanni Caves, Sarvāstivadin by Ᾱnanda and his followers from Giriyaka-Aṁasaṅda region. Despite differences, personality clashes, and diverse interpretations of some rules, the saṁgha remained united and it took hundred years to come out openly to oppose some of doctrinal interpretations and practices. But soon after it, the eminent elders put their best effort to unify the saṁgha as visible from the Kathāvaṭṭhu and the schism edicts of Aśoka.
Some of the early stūpas constructed were mainly of stone, earthen, and then brick stūpas. The remains of stone/earthen stūpas are visible on both sides of the pathway in significant places. Marshall noted existence of many brick, clay, and earthen stūpas in Rājagriha and its surroundings. Many of them had antiquity since the Maurya or earlier period (Marshall 2002: 96–97). Cunningham reports stone foundations of small brick stūpas on nodal points of Giriyaka pathway measuring from 5 to 6 feet in height with circumference of 12 feet (Cunningham 2000, I: 16–17). I have noticed many stone and earthen stūpas with foundation of stone on both side of pathway (Figure 12). Many have them have same circumference i.e. 12 feet as measured by Cunningham but one stone stūpa is very unique having diameter of 18 feet, larger than other such kind of stūpas (Figure 13). It is situated on the right side of the pathway and just opposite it on the other side of the road a square platform was raised. This stūpa has special significance as the platform may be meant for some ritual purpose. It was a kind of memorial or śārīrika stūpa. The material remains indicate its pre-Maurya origin. Some of the other mud/stone stūpas may belong to the Maurya period and some of them were getting renovated in the age of the Kuṣānas. The monastic structure can be traceable when one starts climbing from the village side and after covering around 500 meters when reaching on a first flat surface on the right side of the hill.
The monastery though very small is lying near the earthen stūpas. The foundation wall and remains of the wall are still visible. It may have a late origin between the Gupta to later Gupta periods.
On the hill at least two structural remains of the stūpas are visible. The site is known for one main stūpa that has been recently renovated by the Archaeological Survey of India. Broadley informs the existence of stūpas near the tank:
‘it brings us to a semi-circular platform about thirty feet in radius, on which is another conical building quite ruined. East from thence and adjacent is an area forty-five feet square, the center of which is occupied by a low square pedestal twenty-five feet square divided into the sides by compartments like the paneling on wainscot, and terminating in a neat cornice. On this pedestal, rises a solid column of brick sixty-eight feet in circumference. About thirty feet up, this column has been surrounded by various moldings, not ungrateful, which have occupied about fifteen feet, beyond what remains of the column, perhaps ten feet, is quite plain. A deep cavity has been made into a column probably in search of treasure, and this shows that the building is solid. It has been constructed by bricks cemented by clay, and the outside has been smoothed with a chisel and non-plastered. Part of the original smooth surface remains entire, especially on the east side. The weather on the west side has produced many injuries.’ (Broadley 1872: 261)
Adjacent to the water reservoir, two stūpas one still intact and the other only has a foundation is existing. Both of the stūpas are seemed to be intentionally built together with narrow passages. The mound of the second stūpa is lower than the cylindrical stūpa and the foundation of the structure suggests that it was square with a projection of center in each face. It has been identified as a shrine but a stūpa with a square base and hemispherical dome (Figure 14). The upper part is damaged as it was excavated by Cunningham but the shape of the stūpa indicates that its dome was 6–7 feet high. Such kinds of stūpas with enlarged Medhi and less projected dome have been found in different parts of Magadha, Bengal, Odisha, and South India. This stūpa does not have exposed niches like the adjacent one but projections in all directions with subdued niches and floral designs in form of lotus petals have been found (Figure 15). The dome is broken but heap of bricks and square platform indicate that the stūpa was less in size.
The larger stūpa discussed by Broadley i.e., Jarāsaṅdha-ki Baithak is a brick structure and cylindrical (Figure 16). It was placed on a square brick platform 14 feet high. The cylindrical shape stūpa is 23 feet in circumference and 21 feet high. Cunningham reports that the upper part of the stūpa was decorated with a hemispherical dome which is now broken except one segment 6 feet. This part of harmikā may also be decorated with chattra. So, the total height of the whole stūpa architecture could be around 55 feet. He sank a shaft from top up to 41 feet probably to discover relics but got nothing out of it (Cunningham 2000, I: 17). The structure of the stūpa may be of the Gupta age and it might be constructed between the 5th–6th century CE. The first three brackets from the base of the stūpa represent adhishthāna. The square platform rising from adhisthāna symbolizes jagatī upon which stūpa has been raised. The jagatī has many brackets, the lower one has eleven niches each on all sides (Figure 17).
Eleven niches at each wall of the platform indicate some spiritual significance. Because the whole complex is dominated by the stūpa cult, the eleven must represent the triple gem (triśarana-Buddha, Dhamma, Samgha) and atthāṅgika-magga i.e. sammā-ditthi, sammā-saṁkappa, sammā-vāchā, sammā.-kammanta, sammā-ājīva, sammā-vyāyāma, sammā-sati, sammā-samādhi. Though eight aniconic symbols (Lotus, mandala, golden fish, dhvaja, Dhammachakra, treasure vase, chatra, and śankha) related to the Mahāyāna along with trikāya doctrine may also be associated but has less possibility because remains of decorative motifs represent only floral designs especially lotus petals showing insignificant evidence of Mahāyāna. The structure was also adorned with beautiful moldings and the remains of lime plaster show that it was well treated with plaster. The cylindrical part is like śikhara or deul representing the Buddha and his prominence. The dome may be capped with āmalaka or chattra but its remains are not available. Adjacent to these remains, a small monastic settlement is visible. The brick structure of the monastery suggests that it had very late origin probably in the early medieval period.
Both the stūpas are well protected by an enclosure wall of brick. It was approximately 75 cm thick and bricks were joined by mud mortar. The pradkshinapatha was made by rammed bricks and was developed in such a way that pilgrims can circumundulate either one stūpa or both stūpas together (Sharma, Manjul, and Tewary 2015: 449–456, 454). Two stūpas of the same period with narrow passage suggest that these two stūpas are dedicated to two same persons for whom earlier stūpas were dedicated. I can infer that main cylindrical stūpa was for Ᾱnanda and the smaller one for Devadatta. The difference between the two is that the former group may be relic stūpas but these are votive.
Cunningham says that the Jarāsaṅdha stūpa is that one made in the devotion of the goose. It can be further corroborated by a broken figurine carved on a pedestal and an impression of goose on lac seals found from the stūpa remains (Cunningham 2000, I: 18–19). On basis of Xuanzang’s account, Cunningham identifies it as Haṁsasaṁghārama and says that earlier it was associated with Theravāda which allowed meat-eating on condition of tikotopariśuddha. He says that one of the seals bears figure of a goose on its top and a broken figure with a large goose carved on a pedestal have also been found (Cunningham 2000, I: 17–18). Once the resident monks could not find alms for many days, When saw a hoard of geese flying over the monastery, the monks desired to get some of the swans to satisfy their hunger. The leader of the Haṁsas listened and realised their problem. He intentionally fell from the sky and died in front of them to satisfy their hunger. The monks realised their fault and buried the Haṁsa with due homage. After that these monks left the Theravāda and became a follower of Mahāyāna to follow the complete nonviolence. In memory of the Haṁsa, they built a monastery known as Haṁsasaṁghārama (Beal 2000: 119–120). The physiology of the site suggests that it cannot be Haṁsasaāghārama. Giriyaka hill since its inception to decline seems to a dominated by the stūpa cult. There is a rare presentation of images, mandalas, or structures showing the dominance of Mahāyāna at this place. The water reservoir cannot be accepted as a tank or sarvovara used for the domestication of Haṁsa. All Buddhist monastic establishments on the hilltop across the Indian sub-continent had the provision of a water reservoir for everyday purposes. The monastic site of Apsadh (Aṁasaṅda) can be the most identifiable place for Haṁsasaṁghārama. The archaeological artefacts at Apsadh show strong presence of Mahayana-Vajrayana along with Vaishnavism in the Gupta and post-Gupta periods. The village is existing about 5 km to the southeast of Pārvatī and had an abundance of Buddhist and Brahmanical sculptures and monastic remains (Chaudhuri 1936: 307). The Haṁsasaṁghārama can be well identified with it. Apsadh tālāb is still the most suitable place for swans. The Apshad inscription of Adityasena informs that this pond was excavated by queen Koṅadevi for the welfare of people of this area along with living creatures like alligator, fish, and birds (Fleet 1888, III: 205–208). The talab which is still existing seems to be a natural pond re-excavated by the queen who was a worshipper of Lord Viṣnu. Earlier it was a Buddhist site as obvious from material remains but later on Vaiṣnavism also prospered here with the patronage of the Gupta kings.
At the receding end of the higher slope within 50 feet of the main stūpa, an oblong water reservoir has been found. This tank was excavated from two sides from the natural rock of the hill and on the other two sides plucked up with stone blocks (Figure 18). It is about 100 square feet (Patil 1963: 152).
Buchanan observed that this tank was surrounded all along its length by an 18 feet wide paved causeway connecting it to the stūpa complex at the hill. It was five hundred feet long and forty feet wide. Opposite the tank, a narrow ridge is found extending east and west and rises high on both sides from the tank but more towards the west. At the west end of this causeway is a steep slope of brick, twenty feet high and 157 feet wide (Jackson 1925: 111–112). There is second tank at short distance to the north formed by the excavation of the rock for construction purposes. Both tanks were dried when Cunningham saw them (Cunningham 2000, I: 16–17). One natural pond was at the bottom of the hill where the parapet had the foundation. The reservoir had no ritual purpose as suggested by some scholars. The basic purpose of this tank was to harvest rainwater for the routine work of the monastery. Making reservoirs for rain harvesting was important feature of the Buddhist monastic settlement. This structure was one of the earliest experiments in the land of Magadha. This technique was experimented in a broader way in the monasteries situated on main land of the Magadha and Ahara-Paine system of irrigation (based on rain harvesting) was developed. For monasteries on the hills, it was difficult to get water, so all such monastic establishments had provision of making cisterns and reservoirs. The masonry and use of stone blocks suggest that its excavation was pre-Mauryan and definite beneficence by state patronage. The reservoir gives an empirical basis for re-assessing the development of monastic settlements and the role of the state. The reservoir offers interesting indications that how kings played an active role to set up monasteries, and building stūpas. Water management played a major role in this monastic settlement to survive for a longer period.
Cunningham discovered eighty-four oval seals of lac of different sizes. He describes that the seals are different in size but usually 3 inches long and two inches wide with the impression of the large stūpa with four smaller stūpas on each side with the inscription of Buddhist creed Ye Dhamma hetu prabhava in medieval Nagarī character. The seals were all oval, but of different sizes, generally about 3 inches long and 2 inches broad, All, however, bore the same impression of a large stūpa (Lawson 1982: 514). The plaque/sealing (Anon 2020) looks good and an impression is visible (Figure 19). This is a ritual object for offering as a model of piety by lay or monastics. These sealings may be put inside small mud stūpa as a mode of commemoration. The seal suggests a stūpa cult associated with this site. The eight great stūpas on the surrounding may have representatives of eight great places i.e. either atthamahāṭṭhani (Luṁbinī, Bodhgayā, Sārnāth, Kuśinagara, Vaiśālī, Rājagriha, Srāvastī, and Saṅkasya) or eight places where relic stūpas (Ajātśatru of Magadha; Licchavis of Vaiśalī; the S̄akyas of Kapilavastu; Bulis of Allakappa; the Koliyas of Rāmagāma; the brāhmins of Vethadvīpa; to the Mallas of Pāvā; and the Mallas of Kuśinagar) were made just after the mahāparinibbāna of the Buddha.
This seal belonged to a larger number of similar sealing which was used to be put inside small mud stūpas as a mode of commemoration. A beaded border gives stūpa a special status. The central stūpa would be with these eight great places, but specific reference to the find-spot. The stūpa shown in the center can be the Giriyak stūpa’s visual representation, which is held high and with more emphasis by its enlarged portrayal. It appears floral decorations were hanging from the upper part of this stūpa. The central stūpa is shown with elaborate architectural details in its Medhi, anda, and harmikā. A small projected feature is shown in the front part of the central stūpa. The multiple layers of the chatra and yasti are intricately depicted. These features indicate that the central feature was being regarded as a caitya incorporating a ritual veneration space and not a mere stūpa. These seals could be dated to the Gupta-post-Gupta period and not earlier. The pinnacles of the central stūpa resemble like śikhara which had seven stūpa like visible figure that may represent seven relic stūpas made of the Kṣatriya princes of the seven early stūpas except for stūpa of Rāmagāma dug out by the emperor Aśoka to make 84000 stūpas.
Giriyaka Buddhist complex can be said as the first fully developed hilltop monastery of Magadha. Since the age of the Buddha, hill regions in Magadha were occupied by the ascetics of different religious traditions. At Rājagriha hills, Mahāvīra and his followers spent many vassavāsas. Some of the prominent monks of Buddhism either frequently visited hills for meditation or made their abode in natural caves that existed there. Even the Buddha and other bhikkhus visited frequently at Gijjhakūta for meditation and discourses. John Marshall has reported that at Gijjhakūta, remain of brick structures of monasteries, and stūpas of the Maurya age are found (Marshall 2002: 92–93). Similar brick structures showing remains of the monastery have been noticed at Nāgārjunī hills which were earlier carved out and donated by king Daśratha to the Ᾱjīvika monks. These caves were later appropriated by the Buddhists who raised brick monasteries in front of caves. At Giriyaka monastic activities began during the age of the Buddha and continued up to the early medieval age. The existing artefacts, brick remains, stūpas, pillars, and monastic remains leave no doubt that it was the first fully developed hilltop monastery of Magadha.
Giriyaka hill monastic complex is dominantly embedded with the stūpa cult. The remains show an absence of images and dominance of stūpas. Numerous stūpas, symbol of stūpas on pillars and seals show that the site belongs to the Sthahviras.
The cave on the side of a hill is meant for a prominent Buddhist monk. The topography of the region, contexts given in travelogues and literature indicate that the cave belongs to Devadatta. The cave at Pārvatī hill can be the real Indasāla-guhā as it is very near to Apsadha (Ᾱṁasaṅda) village. The other cave i.e. Gidhadwara cannot be Indrasāla-guhā because it is too steep and not easily accessible. This cave may be for the dhutāṅga practices, not for the Buddha.
The inferences drawn in the paper are engrossing in nature and promise a lasting influences in Buddhist archaeology in times to come. The observations would open up new avenues of explorations which would enthral the future scholars luring them to venture further in the sacred space of the Giriyaka hilltop monastery and its surroundings.