Jaina Rock-Cut Caves in Maharashtra, India

Posted: 04.07.2017
Updated on: 21.12.2017

Centre of Jaina Studies Newsletter: SOAS - University of London

The hilly region of Maharashtra on the western coast of India is famous for a long and rich tradition of rockcut architecture with World Heritage Buddhist and Brahmanical cave-sites such as Ajanta, Ellora and Elephanta. Compared to these, the Jaina caves of the region, except for those at Ellora, are little known. The present study attempts to bring to light these lesser-known Jaina caves.1

The Jaina caves of Maharashtra have been briefly mentioned in early reports of the Archaeological Survey of Western India and in some articles published in the Indian Antiquary (Burgess 1876: 76-80, Burgess 1878: 4- 11, Sinclair 1872 & 1877). Some of the sites like Ankāi- Ṭankāi and Chāndor were reported in the exhaustive account of Wilson on the cave temples of Western India (Wilson 1847-48 & 1853). In these accounts, the caves have been mentioned very often as Buddhist caves due to general confusion of Jina icons with those of Buddha, or in some cases, the faith to which the caves belonged, has not been mentioned at all. Eventually, these caves were incorporated in the monumental work of Fergusson and Burgess on the caves of India, where some of the caves were described in detail (Fergusson and Burgess 1880). However, the caves at Chāndor, Añjanerī and Tringalvādī, though reported in earlier accounts, were not included here. These were dealt with in later reports (Burgess and Cousens 1897, Cousens 1931). The caves at Māṅgī-Tuṅgī were reported by Banerji with detailed description but without any drawings (Spooner 1921-22). The caves at Mohidā, Pāle and Daulatābād were noticed subsequently (IAR 1958-59, Sankalia and Gokhale 1969, Pathy and Dhavalikar 1987). While the Jaina caves at Ellora have been studied by various scholars (Pereira 1977, Pathy 1980, Soundara Rajan 1981 & 19882), the work on other caves has been almost negligible. A number of these caves are now almost in ruins. A holistic understanding of the Jaina caves of the region, placing them in the historical context was a desideratum.

This study was aimed at documenting each Jaina cave of Maharashtra in minute detail, proposing a chronological framework, highlighting general architectural and iconographic features of these caves and placing them in an historical context. The architectural, iconographic and stylistic features of the caves have been recorded along with detailed drawings of ground plans, doorways, pillars, ceilings and costume-ornaments of icons. The placement of icons has been plotted on the ground plan of each cave, while the icons have been recorded in minute detail with measurements. The chronological framework has been worked out by comparative analysis of caves with contemporary rock-cut and structural architecture of the region, iconographic development and epigraphic evidences, wherever available. The contemporaneous cultural milieu and issues of patronage have been discussed to situate caves in their historical setting. Historical development of Jainism in Maharashtra, prevalent sub-sects and important centres of the sect are discussed for understanding the role of these caves in religious development of the region.

Cave II at Ankāi-Ṭankāi, Nasik district, 12th century CE (Photo: Viraj Shah, 17.03.1999)

About seventy Jaina caves at nineteen sites scattered over western, hilly areas of Maharashtra had been reported. The study has revealed that these caves were excavated over a span of 1,500 years, from the 1st century BCE to 14th-15th century CE, with the largest number excavated during the 9th to 14th-15th centuries CE. All the caves belonged to the Digambara sect, most of which are concentrated in the Nasik district. Interestingly, two early Buddhist caves at Nasik and Junnar were converted to Jainism during the early medieval period. It has been revealed that these caves were excavated against the backdrop of reviving trade activities and varied socioreligious contexts of a changing society.

Architectural Features

These Jaina caves follow the contemporaneous regional architectural and stylistic trends. They form an integral part of the general pattern of the development of rockcut architecture of the region. Thus, the Dhārāśiva caves of the early 6th century are like the Buddhist caves at Ajanta in terms of ground plan, pillar and doorway ornamentation, type of pillars as well as pilasters and the treatment of the main icon in the shrine, while the caves at Kharosā and Ambejogāi of the late 8th century CE reflect similarities with eastern Chālukyan architectural style. The 9th-century Jaina caves at Ellora share similarities of ground plan, pillar type, pillar ornamentation and façade treatment with other Brahmanical caves of the site, especially Kailāśa and Lankeśvara. The presence of an elephant and a pillar or mānastambha in the court of Indra Sabhā points to architectural relation with Kailāśa, while Chhoṭā Kailāśa is a direct copy of Kailāśa on much smaller scale. The 12th to 13th-century caves at Ankāi-Ṭankāi, Tringalvādī, Daulatābād and Vāse share a number of features like types of pillars and doorways, treatment of fronts of benches in the caves with mouldings, use of more loose than carved icons, contemporaneous with Brahmanical caves at Panhāle in Sindhudurg district. Thus, there is nothing 'Jaina' about these caves except the icons.

In spite of forming an integral part of the general process of development of rock-cut architecture, these caves display some unique features. The enclosed courtyard in Cave II at Dhārāśiva is unparalleled. The rock-cut structure, possibly a base for a chaumukha or mānastambha, in front of the caves at Ambejogāi and Kharosā and perhaps a stūpa in the court of Cave II at Dhārāśiva are also unusual features. Similarly, the presence of lion heads on mattavārana and the addition of śikhara in the treatment of the façade as vimāna in the Jaina caves at Ellora, though not completely unknown, is a new feature at the site. The perforated screen and large lion figures at Ankāi-Ṭankāi II are also very unique as far as cave architecture is concerned.

The caves post-dating the 10th century CE reveal a tendency to imitate temple architecture of the period. The trend begins at Ellora where the frontage of the cave is given an appearance of vimāna from adhiṣṭhāna to śikhara, a feature already noticed in a number of Pallava caves. The type of pillars, doorways with chandraśilā and kakṣāsanas with dwarf pillars indicate architectural relations with Chālukya and Rāṣṭrakūṭa structural architecture of the same period. There must have been a mutual exchange between both modes of architecture. Still later, the 12th-century caves at Ankāi-Ṭankāi and Tringalvādī follow contemporaneous temple architecture of the Yādava period in ditto. The ground plan, type of pillars and doorways, ceilings and the decorative motifs are very similar to structural temples of the same period. More than anything, the treatment of the lower portion of the cave frontage as a plinth or adhiṣṭhāna with various mouldings indicates the effort to make a cave look like a structural temple, as functionally such plinths are useless in cave architecture. Apart from imitating features of structural architecture, some of these caves also make use of actual structural components. Thus, the verandah at Tringalvādī including plinth, back wall and the ornamental doorway as well as the hall pillars, the shrine doorway of Cave VI at Ankāi-Ṭankāi and the pillars at Vāse are structural.

A noticeable feature is the absence of many carved icons in the caves, which imitate the structural architecture and are decorative, such as Ankāi-Ṭankāi, Tringalvādī, Daulatābād and Vāse. A few loose icons found in the vicinity of these caves and the presence of benches in the caves suggests that many loose icons were installed. The reasons behind such an arrangement are difficult to determine. It could have been done due to the fear of iconoclasts since loose icons can be hidden and saved. It could also have been an attempt to imitate the structural temples in ditto. Compared to these, the architecturally plain and rough caves such as those at Chāndor, Bhāmer, Mohidā and Māṅgī-Tuṅgī have a large number of fine rock-cut sculptures. At some of the caves, both carved and loose icons were used.

Cave II at Mohidā, Nandurbar district, 13th century CE (Photo: Viraj Shah, 30.12.1997)

Most of the Jaina caves with a few exceptions are architecturally crude and rough excavations, though many of these display a large number of fine sculptures. Thus, in most of the Jaina caves the emphasis has been only on icons rather than the architecture or decoration of the caves, indicating the very focused interests of the patrons.

Iconographic Features

The icons of tīrthaṅkaras and yakṣa-yakṣī figures dominate the iconographic programme of most of the caves. Occasionally the figures of dikpālas, Ganeśa, Hanumān, Kṣetrapāla and yakṣa-yakṣī couples, or what is known as 'parents of Jina', also occur. The monk or ācārya figure with broom and kamaṇḍalu occurs at Māṅgī-Tuṅgī and is not found at any other site. The scenes of Kamaṭha's attack on Pārśvanātha and meditating Bāhubali are very popular themes at Ellora and continued to be represented until the12th to 13th centuries in caves at Ankāi-Ṭankāi and Mohidā, though not as twin scenes as at Ellora.

The development and the trend of iconographic norms in Maharashtra broadly followed the general pattern as noticed at the pan-Indian level. Some of the features such as the introduction of śāsanadevatās as well as lāñchana appear late in this region. In the early caves of the 6th to the late 8th century CE, only Jina figures were depicted. Though śāsanadevatās or yakṣa-yakṣī figures attending upon Jina were introduced in the 5th century CE in other parts of the country, they appeared as late as the 9th century at Ellora in the caves of Maharashtra. Sarvānubhūti and Ambikā were the most popular yakṣa-yakṣī, depicted as attending upon all Jina figures in a cave. Eventually each Jina was given a separate pair of yakṣa-yakṣī. However in spite of the introduction of a different pair of yakṣa-yakṣī for each Jina, only Gomukha-Chakreśvarī and Dharaṇendra-Padmāvatī occur and no other pair was represented. This is because in most of the caves no distinction was made between Jina figures except Pārśvanātha, who has the snake-hood above the head and Ṛṣabhanātha, who is distinct with hair on the shoulders. Also wherever Jinas were distinguished by lāñchanas, yakṣa-yakṣī pairs were not provided. The yakṣīs were prominent and popular, as some of the goddesses such as Chakreśvarī, Padmāvatī, Sarasvatī and Ambikā also appeared as independent figures. The representations of goddesses as independent deities, Ganeśa, Hanumān and Kṣetrapāla were obviously in response to the growing threat of other sects and attempts to assimilate local deities into the Jaina pantheon for more popular mass appeal.

Ṛṣabhanātha in the shrine of Cave I at Mohidā, Nandurbar district, 13th century CE (Photo: Viraj Shah, 04.06.1999)

In early caves Jina figures were represented with chaurī-bearers, garland bearers, halos and simhāsana, while from 9th century onwards, triple chhatra, drum players, musicians, elephants, devotees and lāñchana were added. Lāñchana, though present in some cases, was not a very regular feature. While early Jina figures were depicted in ardhapadmāsana and did not have śrīvatsa, later Jina figures were invariably depicted in padmāsana and had śrīvatsa. This clearly indicates that the early caves were influenced by the southern art idiom, while later caves were more influenced by western and central Indian style. There are a few regional variations and innovations, both in general iconographic development and the execution of particular icons. But overall, Jaina iconography in Maharashtra as displayed in the caves is simple compared to the elaborate and complex iconography as noticed in some regions like Gujarat and Rajasthan.


There are traces of paintings in some of the caves, most of which are very crude geometric designs and belong to a much later date than the cave. The small portion of a painted Jina figure in Cave II at Dhārāśiva is a fine piece of art, in the style of Ajanta paintings. The Jaina caves at Ellora were profusely painted, large sections of which are still extant on ceilings and sidewalls. The themes of these paintings are mostly flying celestial beings in pairs engaged in various activities such as dancing, playing musical instruments, carrying pātra with offerings, garlands or paying adoration to Jina with hands in añjali. These celestial beings were meant to participate as attendants or devotees in the iconic representation of Jinas in the cave, carved both in the shrine and the hall. There are a few narrative and iconic panels too. These paintings form part of the sculptural composition and supplement the plastic art, continuing the tradition of Indian art as seen at Ajanta and Badami. In fact, here the dominance of plastic art over graphic art is more prominent than at Ajanta.

Location of Cave-site

Most of these caves thrived near large, prosperous and apparently 'urban' towns. While some of these were seats of political power, some were important religious centres and some were commercial emporiums. In the case of many of the cave-sites, there is a fort on the same hill as the caves or on a nearby hill, which, though built in the later Muslim period, reveals the strategic importance of the area. Rivers and high hills with peculiar shapes and seemingly inaccessible nature were specifically preferred. Many of these sites developed as tīrthas and continued to be worshipped until the 17th to 18th centuries CE.

Reviving Cave-sites as Tīrthas

While at present most of these sites are deserted, some like Māṅgī-Tuṅgī, Chāmbhāra Lena and Ellora are living tīrthas, while some like Junnar, Chāndor and Ankāi- Ṭankāi are worshipped as devī or goddess shrines by the local population, and the Jaina cave at Kharosā has been converted into a Buddhist cave by the neo-Buddhist population of the area. Interestingly, today these ancient caves are 'used' by the Jaina population to assert their position and 'show off' their wealth by lavishly 'doing up' the caves and providing modern facilities for the pilgrims against the stark contrast of relatively poor, agricultural surrounding villages. With the publications of 'māhātmya' booklets that glorify the antiquity and sanctity of some of the sites relating them to legendary Jaina personalities as well as tīrthas and empowering them with magical forces, all attempts are made to keep these sites alive as tīrthas.

Thus, it is evident that along with numerous Buddhist caves and elaborate Brahmanical cave-temples, large numbers of Jaina caves were also carved over a long period of time in this region with its prominent tradition of rock-cut architecture. Interestingly, with Jainas, this tradition continued to as late as the 15th century and consequently, the Jaina caves in this region form important evidence of the later phase of rock-cut architecture, of which these are the only examples.

Viraj Shah received her PhD from Deccan College PostGraduate and Research Institute, Pune, India. She has been working on different projects on temple institution in medieval Maharashtra, India and archaeological remains in the coastal areas of Konkan in Maharashtra. Her areas of interest are socio-economic-religious history, religious architecture, iconography, and pilgrimage and landscape studies. She is specifically interested in Jaina art, iconography and the history of Jainism.


Banerji, R. D. Archaeological Survey of India Annual Report. Delhi: Archaeological Survey of India, 1921- 22, pp. 66-69.

Burgess, J. "The Dharasinva Rock Temples". Indian Antiquary, Vol. V, 1876, pp. 76-80.

Burgess, J. Antiquities of the Talukas of Parner, Sangamner, Ankole and Koparguam- with Revised Lists of Remains in the Ahemadnagar, Nasik, Puna, Thana and Kala'dgi Zilla. Archaeological Survey of Western India, No. VI, Bombay: Government Central Press, 1877.

Burgess, James and Cousens Henry. Revised Lists of Antiquarian Remains in the Bombay Presidency, Vol. VIII, Archaeological Survey of India, New Imperial Series, Vol. XVI, 1897.

Cousens, H. Medieval Temples of the Dakhan. Archaeological Survey of India, Vol. XLVIII, Imperial Series, Calcutta: Central Publication Branch, 1931.

Fergusson James and Burgess, James. The Cave Temples of India. Delhi: Munshiram Manoharlal Publishers Pvt. Ltd, Second edition 1988, First published in 1880.

IAR Indian Archaeology- A Review, Delhi: Archaeological Survey of India.

Pathy, T. V. Elura: Art and Culture. Delhi: Sterling Publishers Pvt. Ltd, 1980.

Pathy, T. V. and Dhavalikar M. K. "Daulatabad Caves" in Kusumanjali Vol. I (M. S. Nagaraja Rao Ed.), Delhi: Agam Kala Prakshan, 1987, pp. 191-195.

Pereira, José. Monolithic Jinas. Delhi: Motilal Banarasidass, 1977.

Sankalia, H. D. and Gokhale Shobhana. "A Brahmi Inscription from Pale". Epigraphia Indica, Vol. XXXVIII, 1969, pp. 167-168.

Shah, Viraj. Jaina Rock-Cut Caves in Western India with Special Reference to Maharashtra. New Delhi: Agam Kala Prakashan, 2008.

Sinclair, W. F. "The Caves of the Bronzen Glen and other Remains about Mauje Patna, Taluka Chalisgaon". Indian Antiquary. Vol. I, 1872, pp. 295- 297.

Sinclair, W. F. Architectural and Archaeological Remains in Khandesh. Archaeological Survey of Western India, Bombay: Government Central Press, 1877, p. 22.

Soundara Rajan, K. V. Cave Temples of the Deccan. Delhi: Archaeological Survey of India, 1981.

Soundara Rajan K. V. Ellora Monoliths. Delhi: Gian Publishing House, 1988.

Spooner D. Brainerd (Ed.). "Notes on Places visited by Mr. R. D. Banerji". Annual Report of the Archaeological Survey of India. New Delhi: Archaeological Survey of India, 1921-22, pp. 66-69.

Wilson, John. "Memoir on the Cave Temples and Monasteries and Other Ancient Buddhist, Brahmani cal and Jaina Remains of Western India". Journal of Bombay Branch of Royal Asiatic Society. Vol. III, Pt. II, 1847-48, pp. 32-108.

Wilson, John. "Second Memoir on the Cave Temples and Monasteries and Other Ancient Buddhist, Brahmanical and Jaina Remains of Western India". Journal of Bombay Branch of Royal Asiatic Society. Vol. IV, 1853, pp. 340-379.

Share this page on: