The Ajanta Caves (Ajiṇṭhā leni; Marathi: अजिंठा लेणी) in Aurangabad district of Maharashtra, India are about 30 rock-cut Buddhist cave monuments which date from the 2nd century BCE to about 480 or 650 CE. The caves include paintings and sculptures described by the government Archaeological Survey of India as “the finest surviving examples of Indian art, particularly painting”, which are masterpieces of Buddhist religious art, with figures of the Buddha and depictions of the Jataka tales. The caves were built in two phases starting around the 2nd century BCE, with the second group of caves built around 400–650 CE according to older accounts, or all in a brief period between 460 to 480 according to the recent proposals of Walter M. Spink. The site is a protected monument in the care of the Archaeological Survey of India, and since 1983, the Ajanta Caves have been a UNESCO World Heritage Site.
The caves are located in the Indian state of Maharashtra, near Jalgaon and just outside the village of Ajinṭhā 20°31′56″N 75°44′44″E), about 59 kilometres (36 miles) from Jalgaon railway station on the Delhi – Mumbai line and Howrah-Nagpur-Mumbai line of the Central Railway zone, and 104 kilometres (64 miles) from the city of Aurangabad. They are 100 kilometres (62 miles) from the Ellora Caves, which contain Hindu and Jain temples as well as Buddhist caves, the last dating from a period similar to Ajanta. The Ajanta caves are cut into the side of a cliff that is on the south side of a U-shaped gorge on the small river Waghora (or Wagura), and although they are now along and above a modern pathway running across the cliff they were originally reached by individual stairs or ladders from the side of the river 35 to 110 feet below.
The area was previously heavily forested, and after the site ceased to be used the caves were covered by jungle until accidentally rediscovered in 1819 by a British officer on a hunting party. They are Buddhist monastic buildings, apparently representing a number of distinct “monasteries” or colleges. The caves are numbered 1 to 28 according to their place along the path, beginning at the entrance. Several are unfinished and some barely begun and others are small shrines, included in the traditional numbering as e.g. “9A”; “Cave 15A” was still hidden under rubble when the numbering was done. Further round the gorge are a number of waterfalls, which when the river is high are audible from outside the caves.
The caves form the largest corpus of early Indian wall-painting; indeed other survivals from the area of modern India are very few indeed, though they are related to 5th-century paintings at Sigiriya in Sri Lanka. The elaborate architectural carving in many caves is also very rare, and the style of the many figure sculptures is a highly local one, found only at a couple of nearby contemporary sites, although the Ajanta tradition can be related to the later Hindu Ellora Caves and other sites.
Like other ancient Buddhist monasteries, Ajanta was a kind of college monastery, with a large emphasis on teaching, and divided into several different colleges for living and for some of the education, under a central direction. The layout of the site reflects this organizational structure, with most of the caves only connected via the exterior. The 7th-century travelling scholar Xuanzang informs us that Dinnaga, the celebrated Buddhist philosopher and controversialist, author of well-known books on logic, lived there in the 5th century. In its prime the settlement must have accommodated several hundred teachers and pupils. Many monks who had finished their first training may have used Ajanta as a base to return to during the monsoon season from an itinerant lifestyle.
The caves are generally agreed to have been made in two distinct periods, separated by several centuries.
The earliest group of caves consists of caves 9, 10, 12, 13 and 15A. According to Walter Spink, they were made during the period 100 BCE to 100 CE, probably under the patronage of the Satavahana dynasty (230 BCE – c. 220 CE) who ruled the region. Other datings prefer the period 300 BCE to 100 BCE, though the grouping of the earlier caves is generally agreed. More early caves may have vanished through later excavations. Of these, caves 9 and 10 are stupa halls of chaitya-griha form, and caves 12, 13, and 15A are vihāras (see the architecture section below for descriptions of these types). The first phase is still often called the Hinayāna phase, as it originated when, using traditional terminology, the Hinayāna or Lesser Vehicle tradition of Buddhism was dominant, when the Buddha was revered symbolically. However the use of the term Hinayana for this period of Buddhism is now deprecated by historians; equally the caves of the second period are now mostly dated too early to be properly called Mahayana, and do not yet show the full expanded cast of supernatural beings characteristic of that phase of Buddhist art. The first Satavahana period caves lacked figurative sculpture, emphasizing the stupa instead, and in the caves of the second period the overwhelming majority of images represent the Buddha alone, or narrative scenes of his lives.
Spink believes that some time after the Satavahana period caves were made the site was abandoned for a considerable period until the mid-5th century, probably because the region had turned mainly Hindu.
The second phase began in the 5th century. For a long time it was thought that the later caves were made over a long period from the 4th to the 7th centuries CE, but in recent decades a series of studies by the leading expert on the caves, Walter M. Spink, have argued that most of the work took place over the very brief period from 460 to 480 CE, during the reign of Emperor Harishena of the Vakataka dynasty. This view has been criticized by some scholars, but is now broadly accepted by most authors of general books on Indian art, for example Huntington and Harle.
The second phase is still often called the Mahāyāna or Greater Vehicle phase, but scholars now tend to avoid this nomenclature because of the problems that have surfaced regarding our understanding of Mahahayana.
Some 20 cave temples were simultaneously created, for the most part viharas with a sanctuary at the back. The most elaborate caves were produced in this period, which included some “modernization” of earlier caves. Spink claims that it is possible to establish dating for this period with a very high level of precision; a fuller account of his chronology is given below. Although debate continues, Spink’s ideas are increasingly widely accepted, at least in their broad conclusions. The Archaeological Survey of India website still presents the traditional dating: “The second phase of paintings started around 5th – 6th centuries A.D. and continued for the next two centuries”. Caves of the second period are 1–8, 11, 14–29, some possibly extensions of earlier caves. Caves 19, 26, and 29 are chaitya-grihas, the rest viharas.
According to Spink, the Ajanta Caves appear to have been abandoned by wealthy patrons shortly after the fall of Harishena, in about 480 CE. They were then gradually abandoned and forgotten. During the intervening centuries, the jungle grew back and the caves were hidden, unvisited and undisturbed, although the local population were aware of at least some of them.
During the intervening centuries, the forest grew back and the caves were hidden, unvisited and undisturbed, although the local population was aware of at least some of them.
On 28th April 1819, a British officer for the Madras Presidency, John Smith, of the 28th Cavalry, while hunting tiger, accidentally discovered the entrance to Cave No. 10 deep within the tangled undergrowth. There were local people already using the caves for prayers with a small fire, when he arrived. Exploring that first cave, long since a home to nothing more than birds and bats and a lair for other larger animals, Captain Smith vandalized the wall by scratching his name and the date, April 1819. Since he stood on a five-foot high pile of rubble collected over the years, the inscription is well above the eye-level gaze of an adult today. A paper on the caves by William Erskine was read to the Bombay Literary Society In 1822. Within a few decades, the caves became famous for their exotic setting, impressive architecture, and above all their exceptional, all but unique paintings. A number of large projects to copy the paintings were made in the century after rediscovery, covered below. In 1848 the Royal Asiatic Society established the “Bombay Cave Temple Commission” to clear, tidy and record the most important rock-cut sites in the Bombay Presidency, with John Wilson, as president. In 1861 this became the nucleus of the new Archaeological Survey of India. Until the Nizam of Hyderabad built the modern path between the caves, among other efforts to make the site easy to visit, a trip to Ajanta was a considerable adventure, and contemporary accounts dwell with relish on the dangers from falls off narrow ledges, animals and the Bhil people, who were armed with bows and arrows and had a fearsome reputation.
Today, fairly easily combined with Ellora in a single trip, the caves are the most popular tourist destination in Mahrashtra, and are often crowded at holiday times, increasing the threat to the caves, especially the paintings. In 2012 the Maharashtra Tourism Development Corporation announced plans to add to the ASI visitor centre at the entrance complete replicas of caves 1, 2, 16 & 17 to reduce crowding in the originals, and enable visitors to receive a better visual idea of the paintings, which are dimly-lit and hard to read in the caves. Figures for the year to March 2010 showed a total of 390,000 visitors to the site, divided into 362,000 domestic and 27,000 foreign. The trends over the previous few years show a considerable growth in domestic visitors, but a decline in foreign ones; the year to 2010 was the first in which foreign visitors to Ellora exceeded those to Ajanta.
Mural paintings survive from both the earlier and later groups of caves. Several fragments of murals preserved from the earlier caves (Caves 9 and 11) are effectively unique survivals of court-led painting in India from this period, and “show that by Sātavāhana times, if not earlier, the Indian painter had mastered an easy and fluent naturalistic style, dealing with large groups of people in a manner comparable to the reliefs of the Sāñcī toraņa crossbars”.
Four of the later caves have large and relatively well-preserved mural paintings which “have come to represent Indian mural painting to the non-specialist”, and fall into two stylistic groups, with the most famous in Caves 16 and 17, and apparently later paintings in Caves 1 and 2. The latter group were thought to be a century or more later than the others, but the revised chronology proposed by Spink would place them much closer to the earlier group, perhaps contemporary with it in a more progressive style, or one reflecting a team from a different region. The paintings are in “dry fresco”, painted on top of a dry plaster surface rather than into wet plaster.
All the paintings appear to be the work of painters at least as used to decorating palaces as temples, and show a familiarity with and interest in details of the life of a wealthy court. We know from literary sources that painting was widely practised and appreciated in the courts of the Gupta period. Unlike much Indian painting, compositions are not laid out in horizontal compartments like a frieze, but show large scenes spreading in all directions from a single figure or group at the centre. The ceilings are also painted with sophisticated and elaborate decorative motifs, many derived from sculpture. The paintings in cave 1, which according to Spink was commissioned by Harisena himself, concentrate on those Jataka tales which show previous lives of the Buddha as a king, rather than as an animal or human commoner, and so show settings from contemporary palace life.
In general the later caves seem to have been painted on finished areas as excavating work continued elsewhere in the cave, as shown in caves 2 and 16 in particular. According to Spink’s account of the chronology of the caves, the abandonment of work in 478 after a brief busy period accounts for the absence of painting in caves such as 4 and 17, the later being plastered in preparation for paintings that were never done.
The paintings have deteriorated significantly since they were rediscovered, and a number of 19th-century copies and drawings are important for a complete understanding of the works. However the earliest projects to copy the paintings were plagued by bad fortune. In 1846 Major Robert Gill, an Army officer from Madras presidency and a painter, was appointed by the Royal Asiatic Society to replicate the frescoes on the cave walls to exhibit these paintings in England. Gill worked on his painting at the site from 1844 to 1863 (though he continued to be based there until his death in 1875, writing books and photographing) and made 27 copies of large sections of murals, but all but four were destroyed in a fire at the Crystal Palace in London in 1866, where they were on display.
Another attempt was made in 1872 when the Bombay Presidency commissioned John Griffiths, then principal of the Bombay School of Art to work with his students to make new copies, again for shipping to England. They worked on this for thirteen years and some 300 canvases were produced, many of which were displayed at the Imperial Institute on Exhibition Road in London, one of the forerunners of the Victoria and Albert Museum. But in 1885 another fire destroyed over a hundred paintings that were in store. The V&A still has 166 paintings surviving from both sets, though none have been on permanent display since 1955. The largest are some 3 × 6 metres. A conservation project was undertaken on about half of them in 2006, also involving the University of Northumbria. Griffith and his students had unfortunately painted many of the paintings with “cheap varnish” in order to make them easier to see, which has added to the deterioration of the originals, as has, according to Spink and others, recent cleaning by the ASI.
A further set of copies were made between 1909 and 1911 by Christiana Herringham (Lady Herringham) and a group of students from the Calcutta School of Art that included the future Indian Modernist painter Nandalal Bose. The copies were published in full colour as the first publication of London’s fledgling India Society. More than the earlier copies, these aimed to fill in holes and damage to recreate the original condition rather than record the state of the paintings as she was seeing them. According to one writer, unlike the paintings created by her predecessors Griffiths and Gill, whose copies were influenced by British Victorian styles of painting, those of the Herringham expedition preferred an ‘Indian Renascence’ aesthetic of the type pioneered by Abanindranath Tagore.
Early photographic surveys were made by Robert Gill, who learnt to use a camera from about 1856, and whose photos, including some using stereoscopy, were used in books by him and Fergusson (many are available online from the British Library), then Victor Goloubew in 1911 and E.L. Vassey, who took the photos in the four volume study of the caves by Ghulam Yazdani (published 1930–1955).
The monasteries mostly consist of vihara halls for prayer and living, which are typically rectangular with small square dormitory cells cut into the walls, and by the second period a shrine or sanctuary at the rear centred on a large statue of the Buddha, also carved from the living rock. This change reflects the movement from Hinayana to Mahāyāna Buddhism. The other type of main hall is the narrower and higher chaitya hall with a stupa as the focus at the far end, and a narrow aisle around the walls, behind a range of pillars placed close together. Other plainer rooms were for sleeping and other activities. Some of the caves have elaborate carved entrances, some with large windows over the door to admit light. There is often a colonnaded porch or verandah, with another space inside the doors running the width of the cave.
The central square space of the interior of the viharas is defined by square columns forming a more or less square open area. Outside this are long rectangular aisles on each side, forming a kind of cloister. Along the side and rear walls are a number of small cells entered by a narrow doorway; these are roughly square, and have small niches on their back walls. Originally they had wooden doors. The centre of the rear wall has a larger shrine-room behind, containing a large Buddha statue. The viharas of the earlier period are much simpler, and lack shrines. Spink in fact places the change to a design with a shrine to the middle of the second period, with many caves being adapted to add a shrine in mid-excavation, or after the original phase.
The plan of Cave 1 (right) shows one of the largest viharas, but is fairly typical of the later group. Many others, such as Cave 16, lack the vestibule to the shrine, which leads straight off the main hall. Cave 6 is two viharas, one above the other, connected by internal stairs, with sanctuaries on both levels.
The four completed chaitya halls are caves 9 and 10 from the early period, and caves 19 and 26 from the later period of construction. All follow the typical form found elsewhere, with high ceilings and a central “nave” leading to the stupa, which is near the back, but allows walking behind it, as walking around stupas was (and remains) a common element of Buddhist worship (pradakshina). The later two have high ribbed roofs, which reflect timber forms, and the earlier two are thought to have used actual timber ribs, which have now perished. The two later halls have a rather unusual arrangement (also found in Cave 10 at Ellora) where the stupa is fronted by a large relief sculpture of the Buddha, standing in Cave 19 and seated in Cave 26. Cave 29 is a late and very incomplete chaitya hall.
The form of columns in the work of the first period is very plain and un-embellished, with both chaitya halls using simple octagonal columns, which were painted with figures. In the second period columns were far more varied and inventive, often changing profile over their height, and with elaborate carved capitals, often spreading wide. Many columns are carved over all their surface, some fluted and others carved with decoration all over, as in cave 1 (illustrated left).
The flood basalt rock of the cliff, part of the Deccan Traps formed by successive volcanic eruptions at the end of the Cretaceous, is layered horizontally, and somewhat variable in quality, so the excavators had to amend their plans in places, and in places there have been collapses in the intervening centuries, as with the lost portico to cave 1. Excavation began by cutting a narrow tunnel at roof level, which was expanded downwards and outwards; the half-built vihara cave 24 shows the method. Spink believes that for the first caves of the second period the excavators had to relearn skills and techniques that had been lost in the centuries since the first period, which were then transmitted to be used at later rock-cut sites in the region, such as Ellora, and the Elephanta, Bagh, Badami and Aurangabad Caves.
The caves from the first period seem to have been paid for by a number of different patrons, with several inscriptions recording the donation of particular portions of a single cave, but according to Spink the later caves were each commissioned as a complete unit by a single patron from the local rulers or their court elites. After the death of Harisena smaller donors got their chance to add small “shrinelets” between the caves or add statues to existing caves, and some two hundred of these “intrusive” additions were made in sculpture, with a further number of intrusive paintings, up to three hundred in cave 10 alone.
A grand gateway to the site, at the apex of the gorge’s horsehoe between caves 15 and 16, was approached from the river, and is decorated with elephants on either side and a nāga, or protective snake deity.
In the pre-Christian era, the Buddha was represented symbolically, in the form of the stupa. Thus, halls were made with stupas to venerate the Buddha. In later periods the images of the Buddha started to be made in coins, relic caskets, relief or loose sculptural forms, etc. However, it took a while for the human representation of the Buddha to appear in Buddhist art. One of the earliest evidences of the Buddha’s human representations are found at Buddhist archaeological sites, such as Goli, Nagarjunakonda, and Amaravati. The monasteries of those sites were built in less durable media, such as wood, brick, and stone. As far as the genre of rock-cut architecture is concerned it took many centuries for the Buddha image to be depicted. Nobody knows for sure at which rock-cut cave site the first image of the Buddha was depicted. Current research indicates that Buddha images in a portable form, made of wood or stone, were introduced, for the first time, at Kanheri, to be followed soon at Ajanta Cave 8 (Dhavalikar, Jadhav, Spink, Singh). While the Kanheri example dates to 4th or 5th century CE, the Ajanta example has been dated to c. 462–478 CE (Spink). None of the rock-cut monasteries prior to these dates, and other than these examples, show any Buddha image although hundreds of rock-cut caves were made throughout India during the first few centuries CE. And, in those caves, it is the stupa that is the object of veneration, not the image. Images of the Buddha are not found in Buddhist sailagrhas (rock-cut complexes) until the times of the Kanheri (4th–5th century CE) and Ajanta examples (c. 462–478 CE).
The caves of the second period, now all dated to the 5th century, were typically described as “Mahayana”, but do not show the features associated with later Mahayana Buddhism. Although the seeds of Mahāyāna teachings were sown right from the 1st century there is little art and archaeological evidence to suggest that it became a mainstream cult. In Mahayana it is not Gautama Buddha but the Bodhisatva who is important, including “deity” Bodhisatva like Manjushri and Tara, as well as aspects of the Buddha such as Aksobhya, and Amitabha. None of them, except for some Bodhisatva, are depicted at Ajanta. Even the Bodhisatva images of Ajanta are never central objects of worship, but are always shown as attendants of the Buddha in the shrine. If Bodhisatva is shown in isolation, such as the Astabhaya scenes, these were done in the very last years of activities at Ajanta, and are mostly ‘intrusive’ in nature, meaning that they were not planned by the original patrons, and were added by new donors after the original patrons had suddenly abandoned the region in the wake of Emperor Harisena’s death (Spink).
The contrast between iconic and aniconic representations, that is, the stupa on one hand and the image of the Buddha on the other, is now being seen as a construct of the modern scholar rather than a reality of the past. The second phase of Ajanta shows that the stupa and image coincided together. If the entire corpus of the art of Ajanta including sculpture, iconography, architecture, epigraphy, and painting are analysed afresh it will become clear that there was no duality between the symbolic and human forms of the Buddha, as far as the 5th-century phase of Ajanta is concerned. That is why most current scholars tend to avoid the terms ‘Hinayana’ and ‘Mahayana’ in the context of Ajanta. They now prefer to call the second phase by the ruling dynasty, as the Vākāţaka phase.
Cave 1 was built on the eastern end of the horse-shoe shaped scarp, and is now the first cave the visitor encounters. This would when first made have been a less prominent position, right at the end of the row. According to Spink, it is one of the latest caves to have been excavated, when the best sites had been taken, and was never fully inaugurated for worship by the dedication of the Buddha image in the central shrine. This is shown by the absence of sooty deposits from butter lamps on the base of the shrine image, and the lack of damage to the paintings that would have been happened if the garland-hooks around the shrine had been in use for any period of time. Although there is no epigraphic evidence, Spink believes that the Vākāţaka Emperor Harishena was the benefactor of the work, and this is reflected in the emphasis on imagery of royalty in the cave, with those Jakata tales being selected that tell of those previous lives of the Buddha in which he was royal.
The cliff has a more gentle slope here than at other caves, so to achieve a tall grand facade it was necessary to cut far back into the slope, giving a large courtyard in front of the facade. There was originally a columned portico in front of the present facade, which can be seen “half-intact in the 1880s” in pictures of the site, but this fell down completely and the remains, despite containing fine carving, were carelessly thrown down the slope into the river, from where they have been lost, presumably carried away in monsoon torrents.
This cave has one of the most elaborate carved façades, with relief sculptures on entablature and ridges, and most surfaces embellished with decorative carving. There are scenes carved from the life of the Buddha as well as a number of decorative motifs. A two pillared portico, visible in the 19th-century photographs, has since perished. The cave has a front-court with cells fronted by pillared vestibules on either side. These have a high plinth level. The cave has a porch with simple cells on both ends. The absence of pillared vestibules on the ends suggest that the porch was not excavated in the latest phase of Ajanta when pillared vestibules had become a necessity and norm. Most areas of the porch were once covered with murals, of which many fragments remain, especially on the ceiling. There are three doorways: a central doorway and two side doorways. Two square windows were carved between the doorways to brighten the interiors.
Each wall of the hall inside is nearly 40 feet (12 m) long and 20 feet (6.1 m) high. Twelve pillars make a square colonnade inside supporting the ceiling, and creating spacious aisles along the walls. There is a shrine carved on the rear wall to house an impressive seated image of the Buddha, his hands being in the dharmachakrapravartana mudra. There are four cells on each of the left, rear, and the right walls, though due to rock fault there are none at the ends of the rear aisle. The walls are covered with paintings in a fair state of preservation, though the full scheme was never completed. The scenes depicted are mostly didactic, devotional, and ornamental, with scenes from the Jataka stories of the Buddha’s former existences as a bodhisattva), the life of the Gautama Buddha, and those of his veneration. The two most famous individual painted images at Ajanta are the two over-life size figures of the protective bodhisattvas Padmapani and Vajrapani on either side of the entrance to the Buddha shrine on the wall of the rear aisle (see illustrations above). According to Spink, the original dating of the paintings to about 625 arose largely or entirely because James Fegusson, a 19th-century architectural historian, had decided that a scene showing an ambassador being received, with figures in Persian dress, represented a recorded embassy to Persia (from a Hindu monarch at that) around that date.
Cave 2, adjacent to Cave 1, is known for the paintings that have been preserved on its walls, ceilings, and pillars. It looks similar to Cave 1 and is in a better state of preservation.
Cave 2 has a porch quite different from Cave one. Even the façade carvings seem to be different. The cave is supported by robust pillars, ornamented with designs. The front porch consists of cells supported by pillared vestibules on both ends. The cells on the previously “wasted areas” were needed to meet the greater housing requirements in later years. Porch-end cells became a trend in all later Vakataka excavations. The simple single cells on porch-ends were converted into CPVs or were planned to provide more room, symmetry, and beauty.
The paintings on the ceilings and walls of this porch have been widely published. They depict the Jataka tales that are stories of the Buddha’s life in former existences as Bodhisattva. Just as the stories illustrated in cave 1 emphasize kingship, those in cave 2 show many “noble and powerful” women in prominent roles, leading to suggestions that the patron was an unknown woman. The porch’s rear wall has a doorway in the center, which allows entrance to the hall. On either side of the door is a square-shaped window to brighten the interior.
The hall has four colonnades which are supporting the ceiling and surrounding a square in the center of the hall. Each arm or colonnade of the square is parallel to the respective walls of the hall, making an aisle in between. The colonnades have rock-beams above and below them. The capitals are carved and painted with various decorative themes that include ornamental, human, animal, vegetative, and semi-divine forms.
Paintings appear on almost every surface of the cave except for the floor. At various places the art work has become eroded due to decay and human interference. Therefore, many areas of the painted walls, ceilings, and pillars are fragmentary. The painted narratives of the Jataka tales are depicted only on the walls, which demanded the special attention of the devotee. They are didactic in nature, meant to inform the community about the Buddha’s teachings and life through successive rebirths. Their placement on the walls required the devotee to walk through the aisles and ‘read’ the narratives depicted in various episodes. The narrative episodes are depicted one after another although not in a linear order. Their identification has been a core area of research since the site’s rediscovery in 1819. Dieter Schlingloff’s identifications have updated our knowledge on the subject.
The Archeological Survey of India board outside the caves gives the following detail about cave 4:”This is the largest monastery planned on a grandiose scale but was never finished. An inscription on the pedestal of the buddha’s image mentions that it was a gift from a person named Mathura and paleographically belongs to 6th century A.D. It consists of a verandah, a hypostylar hall, sanctum with an antechamber and a series of unfinished cells. The rear wall of the verandah contains the panel of Litany of Avalokiteśvara”.
The sanctuary houses a colossal image of the Buddha in preaching pose flanked by bodhisattvas and celestial nymphs hovering above.
Caves 9 and 10 are the two chaitya halls from the first period of construction, though both were also undergoing an uncompleted reworking at the end of the second period. Cave 10 was perhaps originally of the 1st century BCE, and cave 9 about a hundred years later. The small “shrinelets” called caves 9A to 9D and 10A also date from the second period, and were commissioned by individuals.
The paintings in cave 10 include some surviving from the early period, many from an incomplete programme of modernization in the second period, and a very large number of smaller late intrusive images, nearly all Buddhas and many with donor inscriptions from individuals. These mostly avoided over-painting the “official” programme and after the best positions were used up are tucked away in less prominent positions not yet painted; the total of these (including those now lost) was probably over 300, and the hands of many different artists are visible.
Cave 3 is merely a start of an excavation; according to Spink it was begun right at the end of the final period of work and soon abandoned. Caves 5 and 6 are viharas, the latter on two floors, that were late works of which only the lower floor of cave 6 was ever finished. The upper floor of cave 6 has many private votive sculptures, and a shrine Buddha, but is otherwise unfinished. Cave 7 has a grand facade with two porticos but, perhaps because of faults in the rock, which posed problems in many caves, was never taken very deep into the cliff, and consists only of the two porticos and a shrine room with antechamber, with no central hall. Some cells were fitted in.
Cave 8 was long thought to date to the first period of construction, but Spink sees it as perhaps the earliest cave from the second period, its shrine an “afterthought”. The statue may have been loose rather than carved from the living rock, as it has now vanished. The cave was painted, but only traces remain.
Walter M. Spink has over recent decades developed a very precise and circumstantial chronology for the second period of work on the site, which unlike earlier scholars, he places entirely in the 5th century. This is based on evidence such as the inscriptions and artistic style, combined with the many uncompleted elements of the caves. He believes the earlier group of caves, which like other scholars he dates only approximately, to the period “between 100 BCE – 100 CE”, were at some later point completely abandoned and remained so “for over three centuries”, as the local population had turned mainly Hindu. This changed with the accession of the Emperor Harishena of the Vakataka Dynasty, who reigned from 460 to his death in 477. Harisena extended the Central Indian Vakataka Empire to include a stretch of the east coast of India; the Gupta Empire ruled northern India at the same period, and the Pallava dynasty much of the south.
According to Spink, Harisena encouraged a group of associates, including his prime minister Varahadeva and Upendragupta, the sub-king in whose territory Ajanta was, to dig out new caves, which were individually commissioned, some containing inscriptions recording the donation. This activity began in 462 but was mostly suspended in 468 because of threats from the neighbouring Asmaka kings. Work continued on only caves 1, Harisena’s own commission, and 17–20, commissioned by Upendragupta. In 472 the situation was such that work was suspended completely, in a period that Spink calls “the Hiatus”, which lasted until about 475, by which time the Asmakas had replaced Upendragupta as the local rulers.
Work was then resumed, but again disrupted by Harisena’s death in 477, soon after which major excavation ceased, except at cave 26, which the Asmakas were sponsoring themselves. The Asmakas launched a revolt against Harisena’s son, which brought about the end of the Vakataka Dynasty. In the years 478–480 major excavation by important patrons was replaced by a rash of “intrusions” – statues added to existing caves, and small shrines dotted about where there was space between them. These were commissioned by less powerful individuals, some monks, who had not previously been able to make additions to the large excavations of the rulers and courtiers. They were added to the facades, the return sides of the entrances, and to walls inside the caves. According to Spink, “After 480, not a single image was ever made again at the site”, and as Hinduism again dominated the region, the site was again abandoned, this time for over a millennium.
Spink does not use “circa” in his dates, but says that “one should allow a margin of error of one year or perhaps even two in all cases”.