An Introcution to Ajantā Walter M. Spink   The arcing ravine at Ajantā, which "dead-ends" dramatically in a beautiful sequence of seven connected waterfalls, is one of India's most awe-inspiring sights. (fig.1) Long the abode of yakṣas and nāgas--the divinities of earth and water--it is hardly surprising that it was taken over as an ideal monastic retreat by the devotees of the Buddha once that new religion had spread from the Gangetic plain to Western India. Between 100 BC and 100 AD two great caitya halls (halls of worship) along with three associated vihāras (monastic residence halls) had been excavated at a conveniently low level of the scarp, (fig.2) close to the bank of the Waghora ("tiger") river, and just a few kilometers from the busy trade route crossing the nearby plain. (fig.3) Fig. 2
	Facade. The big Caitya hall is visible in the centre.

Indeed, it may have been the tigers in this once-wild region that were responsible for the site's discovery by the distant world. According to tradition, a local boy led a group of English soldiers, who were tiger hunting in the hills above, to a point overlooking Ajantā's gorge, saying that the deep holes therein were tiger's lairs. Among these "holes", the vast vault of Cave 10 must have stood out boldly, and it was in this old "Hinayāna" (early Budddhist) caitya hall (fig.4) that one of our amazed adventurers, having descended into the deep gorge, scratched his record--John Smith 28th Cavalry, 28 April, 1819. (fig.5) One can still find it today, written across the chest of a fifth century Buddha image painted on one of the pillars (R13) in the old cave.

Fig. 4
	Hall > nave. Stūpa in the centre. The colonnade, the triforium and the ambulatory is apsidal whereas the vault-shaped ceiling is hemispherical at the rear.

The fact that John Smith's inscription, instead of being normally positioned, is found seven and a half feet up on the pillar is easily explained, and tells us something about the interim history of the site. For years, over the course of the centuries following Ajantā's eventual abandonment in the fifth century AD, debris fell down from the cliff above, particularly during the rainy season. It flooded into the caves, building up around the lower level of the walls and pillars, and literally dissolving the fragile plaster and paint against which it settled. It was this debris, mingled with the accumulation of centuries of the droppings of bats and birds until it reach a depth of about a meter, upon which John Smith stood.¹

Obviously debris would not have built up in caves that were in use; and the fact such accumulations are found in nearly every cave at Ajantā confirms the fact that after its heyday in the fifth century AD this great site was abandoned--indeed, almost forgotten--for over thirteen hundred years. Indeed, it is to this surprising circumstance, as well as to the protection that its excavated character provided, that we owe Ajantā's remarkable preservation. (fig.6) In many of the caves--particularly those which were still being decorated when the site was precipitously abandoned in about 480 AD, the paintings are in an amazingly pristine condition. By contrast, at nearly every other early site in India, the continuity of worship--with old lamps and incense, to say nothing of devotional activity--has left the interiors damaged and heavily begrimed.

Fig. 6
	Hall > rear wall > right of shrine-antechamber. BODHISATTVA KING IN A MOUNTAINEOUS LANDSCAPE. Detail of Bodhisattva King's bust. Figure on the left is possibly a vassal king.

Throughout western India, the next few centuries, from about 150 AD to 450 AD, were as hard on Buddhism as those before had been supportive. The Chinese pilgrim Fa Hien did not even bother to come to the site during his journey through the subcontinent in the early fifth century. Although he reported that monks were still living in the caves, and (the roads being dangerous) were sometimes even seen flying into the monastery through the skies, he complained that the local people were hostile, and alien to the Buddhist faith. (Fa Hien, 1923, 63)

Fa Hien's account being from hearsay, it is not surprising that he merges reports about the slowly spiraling descent of the water at the end of the ravine with someone's recollection of the caves themselves. The monastery, he says, was made "by hollowing out a great rock. It has five storeys in all; the lowest being in the form of an elephant, with five hundred stone chambers....At the very top there is a spring of water which runs in front of each chamber, encircling each storey, round and round, in and out, until it reaches the bottom storey...." (Fa Hien, trans. H. Giles 1923, 62) Although it is clear from Fa Hien's account that monks were still residing at the site in his day, his stress is obviously on the site's problems and its virtual abandonment, for he tells us nothing about the activities or rituals or beliefs of its sparse community. Nor did he go there himself. Ajantā clearly was living in the shadows during those difficult centuries between its early ("Hinayāna") phase and its later ("Mahāyāna"--or we might better say "Vākāṭaka") phase.²

The site's renaissance, which turned out to be as brief as it was awesome, started in the reign of the Vākāṭaka emperor Harisena, and was primarily due to his approval and his involvement. One of those lost heroes of antiquity still little known to the history books, Harisena was arguably the greatest ruler in the world during the 460's and the 470's--the period of his brief but glorious ascendancy.

The emperor Harisena had inherited vast domains (including the Ajantā region) from his father, and it appears that at his accession in about 460 his realm was at peace. Thus his reign had an auspicious start; and his prospects could only have been helped by the tribulations of the rival Gupta dynasty. Attacked both by the aggressive Hunas from the north, and by the Vākāṭakas from the south, by the mid-fifth century the renowned Gupta house was already beginning its long decline.

By contrast, it was the Vākāṭaka emperor Harisena, insistently extending his empire toward both the western and the eastern sea, who was now holding aloft the burgeoning vessel of India's Golden Age. (fig.7) And by a generous gift of fate, the treasures within that golden jar were now to be poured out in abundance, within the rocky fastnesses of Ajantā's deep ravine. Thus it is Ajantā itself, and Ajantā alone, that offers to the world an endlessly illustrated history of these happy, and then in the end tragic, times.

As if to celebrate the potentials of both Harisena and of the Vākāṭaka empire when this forceful new ruler came to power, a weighty consortium of high-placed patrons early made the decision to initiate a renaissance at the old Buddhist site of Ajantā. Indeed, Ajantā was to be the first great rock-cut monument of the Golden Age; for such excavations had not been done anywhere in India during the last three hundred years.³

Needless to say, the new caves would be very different from those of the past. They would, understandably, mirror the lavish esthetic of the day, and emulate (as their inscriptions suggest) the fabled palaces of heaven as well as the luxurious contemporary palaces and temples described in detail in Ajantā's murals. (fig.8) As if to assure exacting standards as well as control over its growth, the patronage of the site was now highly selective. A half-millennium before, developments at Ajantā itself, as at so many other sites in western India, or at early monuments such as Sanchi, were community efforts, with the costs typically supplemented by some mode of subscription. The early Caitya Cave 10 has separate inscriptions recording the gift of the facade, the left wall, and (apparently) a portion of the ceiling. (Ghosh,A. EI #43, 241-244)⁴

Fig. 8
	Hall > rear wall > left of shrine. SUMĀGADHĀ  on left and MAHOṢADHA (commonly known as MAHĀJANAKA JĀTAKA) on right and below.

By contrast, Ajantā's renaissance in the 460s and 470s was the exclusive province of an elite, many of whom were connected either with the imperial court, or of courts feudatory to the imperial power. No one else was allowed to donate a single image, at least while the site was flourishing, right up to the moment of Harisena's death in c.477. Even the craftsmen, who probably worked variously as architects, sculptors, and painters, as developments required, must in many cases have been sent down from the cities, where they had been working on palaces and temples for this same elistist group of patrons for years; so they were well aware of the high expectations of the site's patrons. (fig.9)

Fig. 9
	Facade > upper portion. The Caitya window is flanked by figures of Yakṣa on either sides. This part of the facade, architecturally as well as sculpturally, was emulated at Cave 26 Caityagṛha started subsequently.

Essentially, everything at Ajantā began in one grand burst of enthusiastic activity. At least three quarters of the more than two dozen Vākāṭaka excavations were begun in the first half-decade of activity. (See Time Chart (fig.10)) As we might expect, the most prestigious patrons in this inaugural group must have been able to select the most desirable locations and probably the most sufficient crews of workmen too.

It will be interesting to meet some of these rich and eager donors and to get a sense of the enthusiasms that so dramatically energized the site during the years of its brief flowering. Fortunately, some of their long and laudatory inscriptions still remain, intentionally located in highly visible spots outside the caves. Through them, as well as through the excavations themselves, we can learn much of the actions and the aspirations of the patrons, which help us to better understand these remarkable times.

Since the emperor Harisena himself did not start his own great Cave 1 until the site had been under development for nearly five years, the most prestigious member of the high-placed consortium that inaugurated the site would have been his imperial Prime Minister, Varāhadeva. It was he, if we can believe the inscriptional record, who handled most of the empire's affairs on behalf of the emperor. Not surprisingly, he located his own cave (Cave 16) at the exact central point of the curving scarp, a position from which it dominates the ravine. (fig.11) The great minister's power and prestige is also reflected in the fact that the still-functioning Elephant Gate through which devotees climbed on the way up from the river below was the "entrance to the site", according to reports given to the Chinese pilgrim Zuan Xang.(Zuan Xang, 1881, 451) (fig.12) Inside, as Varāhadeva says in his inscription, the cave "with its beautiful picture galleries....(resembling) the palaces of the lord of the gods" afforded the "enjoyment of well-known comforts in all seasons" to the monks residing within it (Ajantā Cave 16 inscription verses 24, 26). In this cave, there were probably thirty-two monks residing for each of the cave's sixteen cells had space for two residents. (fig.13)

Fig. 11
	Facade (on far left). The facade has six pillars. It is approached through a Gajadvār (Elephant Gate) seen in the picture. Initially the Gajadvār's steps were going down to the ravine below.
Fig. 11
	Facade. The cave is situated on far left. It was among the first few to have begun during the Vākāṭaka Phase. Perhaps, the local Aṣmaka King, Upendragupta, was its donor besides Caves 17, 18 and 19.
Fig. 12
	Facade (six-pillared) with the Elephant Gate. Originally, there were steps coming up from the ravine below through the Gajadvār (Elephant Gate), and turning up to the left to reach the outer court of the Cave. These elephants are mentioned in Huen Tsang's account. There is the Nāgarāja panel.

Significantly, the prime minister's architects not only cut "a temple (i.e. a shrine) of the Buddha inside", (fig.14) but (at the very point where one enters the cave) excavated a "shrine of the lord of the nāgas" as well; thus they wisely propitiated the ancient snake divinities who, according to tradition, long had ruled over the deep ravine, and hold their place in the imagination of the local people even today. (fig.15) We will find them given places of honor in nearly every cave. This was surely because of their innate powers; but the fact that the Vākāṭaka rulers traced their origins back to the ancient Nāga dynasty could also explain the honor that they gave to these tutelary divinities.

Fig. 14
	Shrine > central Buddha image. Front-view. The colossal image is enshrined in Dharmacakra Mudrā. The Pralamba-padasana posture and gigantic scale are typically late features at Ajanta meaning that the shrine had not been carved with the hall, but afterwards during the last years of Vākāṭaka Phase.
Fig. 15
	Facade > Gaja-dvāra > Nāga shrine. The Nāga King has a central position ahead of the Elephant Gate, so that the worshipper must pray to him first, then climb up to the cave for the Buddha's shrine.

If the Prime Minister Varāhadeva put the power of the Vākāṭaka imperium behind this sudden renaissance at Ajantā, it was the local feudatory king, Upendragupta, who supplied the driving enthusiasm and the most lavish funding for the site. Shocked by the sudden death of his beloved younger brother, which filled him with "the consciousness of transience", he suddenly desired "to make the great tree of religious merit grow, (by) adorning the earth with stūpas and vihāras" (Cave 17 inscription 13, 22). To testify to his obsessive piety, however, nothing now remains within his ancient kingdom except those monuments that were cut into Ajantā's sheltering scarp. Of these hopefully indestructible monuments, all intended to "cause the attainment of well-being by good people as long as the sun dispels the darkness by its rays," he gave no less than five impressive excavations: Caves 17-20; and Cave 29, (fig.16) a second unfinished caitya hall on the slope just above.⁵
Upendragupta's beautiful vihāra, Cave 17, (fig.17 & 18) neighbors the impressive excavation of his friend the Prime Minister, with whom he even shares a vast cistern, whose reservoir, filled by water coming from the hills above, extends far beneath the two contiguous caves. Visitors today still find refreshment from this old and auspicious cistern. However, Upendragupta's Cave 17 has two cisterns, one at either side of the court. It is probably the other (the left one), fronted by a fine (now badly reconstructed) colonnade, which the king's inscription mentions with pride as always being filled with "sweet, light, clear, cold and copious water", even though this is not the case today.⁶

Fig. 16
	Facade > front-view from across the Waghora River. This Caityagṛha was probably the last undertaking on site, that's why it was forced up above when most favorite space in the row of caves below had been occupied. The cave is inaccessible.
Fig. 17
	Facade > Gaja-dvāra > Nāga shrine. The Nāga King has a central position ahead of the Elephant Gate, so that the worshipper must pray to him first, then climb up to the cave for the Buddha's shrine.
Fig. 18
	Hall > rear aisle and shrine. View from the hall. Advanced than Cave 16, the pillars of this cave are ornate.

The most sumptuous of all Upendragupta's offerings, his gem-like caitya hall (Cave 19), was planned as the ceremonial center of the site, even though, as we shall see, it was fated for abandonment. (fig.19) Its lavish decor, surely the work of the best artists in the empire, provided a model upon which the site's various artists would willingly draw as the site developed. (fig.20) Originally, Cave 19, like most of the caves, could be approached by a long rock-cut stairway coming up from the river below. The fact that nothing remains of this approach other than one of a splendid pair of over life-size snake divinities which guarded the entrance into the court area suggests how much of the cliff face, throughout the site, has fallen away.

Fig. 19
	Facade. middle. Front-view.
Fig. 20
	Hall > nave > Stūpa. This "Gandhakuti" (perfumed hall), donated possibly by Upendragupta, the local Aṣmaka King under the Vākāṭakas, is highly innovative. The standing Buddha image on the Stūpa reflects the smooth transition from Hinayāna to Mahāyāna periods.

We know far less about the patron of Cave 4, one Mathura, but he must have been very rich and well connected, for he donated by far the largest vihāra at the site; its great hall is over eighty feet square. (fig.21) It was in part these extreme dimensions that caused a section of Cave 4's ceiling, weakened by a pernicious geological flaw in the rock above, to collapse unexpectedly. (fig.22) Since the whole cave had been excavated at the ceiling level before the collapse occurred, the collapse must have occurred after 468, during Ajantā's "Recession", when the cave had been temporarily abandoned. Although it is likely that no one was killed by the fall, after 475, when work started up again, most of the still-intact portions of the cave's ceilings were raised to a safe level above the flaw line out of concern for any further collapse. As if in compensation for these great problems, the heightening of the interior was finally responsible for the creation of a shrine image of awesome dimensions. Cave 4's seated Buddha, rushed to completion in 478, just after the emperor Harisena's death, is nearly 18 feet high, its size reminding us how an interest in the colossal was now pervading the trade routes from Rome to China. But except for this huge Buddha and two of the six huge attendant Buddhas standing in the shrine antechamber, the cave is very unfinished. It was never painted, while the impressive carved panels in the porch were put there by new and later donors in 479, after the original donor had to give up his connection with the cave during the first months of 478.

Fig. 21
	Veranda. View from proper left-end.
Fig. 21
	Hall > ceiling. View from rear of the hall. This view shows the extant remains of the original ceiling level on left, which collapsed during excavation thus exposing a layer of the flow of volcanic lava. There was subsequently an attempt to raise the ceiling level as seen between the central two pillars.
Fig. 22
	Hall > ceiling. View from rear of the hall. This view shows the extant remains of the original ceiling level on left, which collapsed during excavation thus exposing a layer of the flow of volcanic lava. There was subsequently an attempt to raise the ceiling level as seen between the central two pillars.

It is evident that pride as well as piety ruled over Ajantā's renaissance; and this potent combination surely explains the ambitious undertakings of the prestigious monk Buddhabhadra, responsible for the huge Cave 26, with its upper and lower wings. (fig.23) "Born of a noble family", he had been attached to the powerful minister of Aṣmaka "through many successive births", and in fact dedicated this impressive caitya hall in the latter's honor. (Ajantā Cave 26 inscription, verses 9 - 16 passim) Buddhabhadra probably made good use of these courtly connections from the start, for the competition at the burgeoning site for the best locations and for the best workmen must have been acute. Particularly during its first few years, when so much was happening, and when word was just going out about the employment opportunities at the site, it seems clear from the evidence of unfinished work in lesser caves, that the supply of excavators and artisans was far from sufficient to satisfy the demand. Indeed, even Buddhabhadra was affected by these realities, for by the time the excavators had started to work in the central hall, work on the slightly earlier and less important upper wings had to be temporarily abandoned⁷

Fig. 23
	Caityagṛha > facade. View from across the Waghora River. Photo shows the Caityagṛha, the main edifice, and the subsequent expansion of its wings, courtyard, entrance Gate and steps.

Buddhabhadra's caitya hall, with its two upper and two lower wings, is in fact a caitya-complex; it is far more than just a simple single cave. Indeed, there is good reason to believe that this whole western extremity of the site, which includes the important caves 21, 23, and 24, was controlled by Buddhabhadra, with his Aṣmaka connections, for they were developed in close concert with the associated caitya hall. (fig.24) This is proved by the transformations that these three large vihāras underwent. All were planned at first, in 466, with simple single cells at their porch ends, and all three would have fitted into the space allotted to the right of the great Caitya complex. However, in 467, taste demanded complex pillared (doubled) cells at all porch ends, and to satisfy this new "requirement" (and allow for Cave 21's expansion), the enlarged Cave 23 had to be displaced to the left, as the abandoned blank wall at the right of its facade reveals. Then, as if this were dominoes falling, Cave 24 (also enlarged by the new porch complexes) also had to be shifted leftward, which meant that it penetrated at least twenty feet into the Cave 26 caitya complex. What is clear is that Buddhabhadra made the conscious decision to sacrifice the embarrassingly "primitive" right wings of his complex, in order to allow these new and later vihāras to be truly up-to-date.

Fig. 24
	Caityagṛha > facade. View from across the Waghora River.
Fig. 24
	Facade. View from across the Waghora River. The two pillared hall second from lower right.

Not only the vigor of their patronage, but the very size of the Aṣmaka undertakings was ominous, particularly when we come to know the developing history of the site. In terms of its grandiosity, Buddhabhadra's caitya hall (Cave 26) alone--even without its wings--far outdoes the beautiful little caitya hall (Cave 19) made by Upendragupta, the feudatory king of the region. (fig.25) And as if this disparity were fateful, within a half-decade of the time when these caves had been begun, the threatening imbalance had become not merely architectural but political. Although relations between the two feudatory rulers of Riṣika (the Ajantā region) and Aṣmaka (which lay just to the south) had been cordial at the start, by 468 war between them was imminent. And it was at this moment that Buddhabhadra, because of his connections with the threatening Aṣmakas, was forced to halt all work on his great cave complex, which was still merely roughed out, with no decoration whatsoever. All of its fine and complex decoration, including its plethora of Buddha images, is very late, having been added in 475 and after. (fig.26) It is for this reason that the hall, traditionally, has been seen as belonging to the site's latest phase. But the truth is: this whole body of decoration is a late overlay, upon a notably early core; the Cave 26 architect clearly opted to get all of the general cutting done before adding any decoration whatsoever.⁸

Fig. 25
	Caityagṛha > hall > nave > mid left portion.
Fig. 26
	Caityagṛha > hall > ambulatory > apse. One of the BUDDHA panels carved intrusively after the original patron, Buddhabhadra, had abandoned the cave, following certain catastrophy. Individual donors who were members from the laity, grabbed these spaces for earning their own religious merit. Some have also left dedicatory inscriptions.

It was just at the end of 468, by which time Cave 26 and its wings, along with many other caves at the site, was just being revealed from the rock, that a "Recession" began at the site, in response to the increasingly troubled political situation in the area. (See Time Chart) Although Buddhabhadra's excavations, with their Aṣmaka connections were affected the most suddenly and the most drastically of all, work soon had to stop on most of the other caves at the site as well; it is reasonable to assume that the funds which were being used upon them were now diverted, by the decree of the local king, to preparations for war. Only the excavations of the "super-elite"--the local king himself (Upendragupta); his overlord the emperor Harisena; and the imperial prime minister Varāhadeva-were able to continue underway during this time of mounting concern about the building-up of Aṣmaka military power.⁹

The "Recession" which started in 469 was ominous but not yet shattering. However, by the beginning of 472, a total Hiatus in the site's patronage occurred, obviously reflecting the flaring up of war in the region. Many workers had already left Ajantā, because of the Recession, but now work at Ajantā went into total abeyance, and the remainder must have been forced to leave as well. The majority of them found related employment at the Bagh caves near Indore, while some migrated to lesser sites in other peaceful areas. (fig.27) Bagh was also a Vākāṭaka site, but under direct imperial control, and thus unaffected by the local Riṣika-Aṣmaka conflict.¹⁰

By 475, it is evident that control of the site had undergone a drastic change, due to a total takeover by the aggressive (even if pious) Aṣmakas. Not surprisingly, work on the Aṣmakas' long-abandoned Cave 26 complex now was taken up again with great vigor, (fig.28) whereas the handsome group of caves that had been started by the local king was never touched again. Although Upendragupta's fine vihāra, Cave 17, continued in use (for accommodations for the monks were badly needed at this point) worship was totally disallowed in his caitya hall (Cave 19), despite its central location and despite the fact that (unlike the Aṣmaka's Cave 26) it was finished enough by 471 to be utilized for worship immediately. However, there was not enough time to get Upendragupta's prasasti inscribed in the recess reserved for it just inside and above the caitya hall's doorway. (fig.29) Thus, as if in compensation, the paintings in Cave 19 are not at all begrimed today!) Indeed, the victorious Aṣmakas went so far as to break a passage through two of the cells in Cave 19's court, in order to be able to walk to their own cave complex more conveniently. (fig.30) But as if in compensation for these humiliations, the paintings in Cave 19 are not at all begrimed today, because they were not subjected to years of use; for the same reason, there are a number of places over the capitals were hooks for garlands were either not inserted, or were never used.¹¹

Fig. 28
	Caityagṛha > hall > ambulatory > left wall. MĀRAVIJAYA panel.
Fig. 30
	Facade > left pillared subsediary-shrine. The two-pillared subsediary-shrine has a votive Buddha image on rear wall, and cells on side-walls. It has a counterpart on symmetrically opposite side. The front cells of both subsediary-shrines are broken through for passage. The Nāgarāja panel on right of the shrine's facade possibly had a counterpart on left.

However--again ominously--the Aṣmaka feudatories were not satisfied merely with taking over the Ajantā region itself. All too soon, when the power situation in the empire suddenly changed, they were able to dream of taking over the whole Vākāṭaka empire itself! What occasioned this decision--ultimately disastrous to both the Aṣmakas and the empire was the sudden death (late in 477) of the powerful Vākāṭaka overlord, the emperor Harisena, who had so dramatically expanded his empire during his all-too-brief reign (460-477). (fig.31) Danḍin's historical novel, the Daśakumāracarita, written nearly one hundred fifty years later, recalls the events of this tragic time with a startling precision. Such "total recall" is indeed surprising in such an account, for typically and unapologetically mix fact with fable. However, for Dan�in, whose forebears had come from Vidarbha and had apparently served at the Vākāṭaka court at this critical moment, the story was literally unforgettable. After all, it describes nothing less than the sudden ending of India's Golden Age--a traumatic event in which Danḍin's own forebears may have played their own tragic part.

Remarkably, it is Ajantā itself that confirms and explains this very precise reading of the final flowering and the sudden fall of the Vākāṭakas in the late 470s. We could never understand the situation, never interpret the evidence, without the authoritative aid of the caves. And the crux of the evidence is to be found in Cave 1, the most splendid of all of the vihāras at the site, upon which work was able to proceed without a break both when Upendragupta ruled the site, and then again when the rival Aṣmakas took it over. (fig.32) That is, Cave 1's patron must have been the overlord of both of these contentious feudatory powers who could--except when war flared between them during the Hiatus (472-474)--rise about the altercations of such local subsidiaries, and continue his own proud and pious work. (fig.33)

Fig. 32
	Facade. Front-view. Extant remains of the portico visible in the centre of entablature.
Fig. 33
	Hall > left half.

On many other grounds too, Cave 1, even though it was never either fully finished nor inscribed, convincingly qualifies as the intended dedication of the emperor Harisena himself. Not only is it the only cave at the site with an uninterrupted history--at least in those times when the site was flourishing--but it is by far the most spectacular of the vihāras in its ornamentation containing much of the highest quality work at the site. (fig.34) Its vast courtyard, and unique facade wings, also may have been conceived with elaborate imperial ceremonies in mind. Moreover, Cave 1's compelling iconographic program would further support the conclusion that the cave was the grand donation of the emperor Harisena himself. All of its painted jātakas are centered around stories of kingship; (fig.35 & 36) even the two involving snakes have as their protagonists snake kings, as the incarnations taken by the Buddha in previous lives. Similarly, the uniquely decorated facade depicts the various duties and prerogatives of kingship-war, the hunt, and dalliance. Even the important carved sequences showing the Buddha's decision to leave the "material" for the "spiritual" world, the life of the palace for the life of the forest, from prince to monk, represents the vocation to which the emperor, like the princely Siddhārth, would ideally aspire. (fig.37)

Fig. 34
	Shrine with antechamber.
Fig. 35
	Hall > left wall > centre. JANAKA; episode 1: The king of Videha called Janaka is getting his wife used to the idea that he intends to renounce the worldly life.
Fig. 35
	Hall > left wall > centre. JANAKA; episode 2: Consequent upon Janaka, the king of Videha, deciding to follow the path of renunciation, a dance performance in the palace pavilion is organised to make him aware of the pleasures of worldly life.
Fig. 36
	Hall > rear wall > left of shrine. SUMĀGADHĀ  on left and MAHOṢADHA (commonly known as MAHĀJANAKA JĀTAKA) on right and below.
Fig. 37
	Facade > left pillared-vestibule > entablature. Relief showing three out of four ominous sights which Gautama encountered during an excursion leading to his renunciation: a sick man, an old man and a corpse.

However, Cave 1 presents a surprising situation. It was abandoned when the last of its paintings--on the front and right walls--were so near to completion that they surely could have been finished in a week or two. Such a circumstance, together with the lack of any inscription and of any trace of grime in its shrine (which is the invariable product of ritual use), suggests that the cave's patron died very suddenly indeed; then, either through neglect or trauma its further development was summarily abandoned. (fig.38) In fact, from a ritual point of view, Cave 1-the finest vihāra in India-is "dead". And this is why, when the Period of Disruption began, not a single intrusive image was placed in it.¹²

Fig. 38
	Hall > left half.

Perhaps one could not conclude that Cave 1 was the emperor Harisena's own cave by reference to the above evidence alone; but when we analyze the development of the other caves at Ajantā, we find that it is at exactly and immediately after the Emperor Harisena's death (that is, in c. 478) that a shock wave surges through the site. Except in Cave 1 itself, in cave after cave the patrons, as if they realize that time is desperately short, rush to complete their shrine Buddhas alone, totally abandoning the developing programs of their often still very unfinished caves. (fig.39) (See Spink 1991, 86-91) Only the monk Buddhabhadra, with his Aṣmaka connections, is briefly immune to such concerns; his Cave 26 complex, as well as his fine Caves 21 and 23, all go on relatively unperturbed throughout the course of this sobering year (478).

Fig. 39
	Central shrine > right wall. Intrusive Buddha figures.
Fig. 39
	Shrine-antechamber > left wall, ceiling and rear wall. MAHĀPRĀTIHĀRYA on left and rear wall. Painted decorations on ceiling.

What seems clear is that when the great Harisena died the ambitious Aṣmaka feudatories decided to take advantage of this wrenching political change for their own purposes, for they knew the character of the new king and perhaps could benefit too from the problems surrounding his succession. In contrast to Harisena, who is described in Danḍin's Daśakumāracarita as a "partial incarnation of the God of Justice", Sarvasena III, his son and successor, "unluckily held the science of politics in little esteem". (Kale 1966, Chapter 8, passim) In fact, he allowed himself to be dominated by false friends, among who was numbered no less a person than the scheming son of the Aṣmaka minister. This wolf in sheep's clothing, "under the pretext of being expelled by his father" had come to Vidarbha "with a numerous train of musicians and dancing girls..., numerous retainers and spies" specifically to infiltrate the Vākāṭaka court and to corrupt Harisena's young successor, working his perfidy from within while military preparations were being made from without.

During the first months of 478, the site's "Vākāṭaka" patrons, expecting the worst, were anxiously hurrying to complete their Buddha images while there was still time.¹³ By contrast, the Aṣmaka patrons, hurried but not worried, continued to work throughout the year in a somewhat normal manner, even after the Aṣmakas made their storm-laden assertion of independence from the new imperial overlord, the ineffectual Sarvasena III. (fig.40)

Fig. 40
	Caityagṛha > hall > ambulatory > left wall. MAHĀPARINIRVAṚA. It is the only depiction of this theme at this site.

Neither Sarvasena himself, nor the great Vākāṭaka house, are given as much as a mention in Buddhabhadra's Cave 26 inscription--an inscription which, because of this "unthinkable" omission, could itself be seen as a declaration of war.¹⁴ In any case, the political realities that Buddhabhadra's record reflects brought the vigorous decade and a half of established patronage at the site to an abrupt end. (fig.41)

Fig. 41
	Caityagṛha > hall. View from rear. Grossly incomplete excavation.

The Aṣmakas now must have turned their attention, and all available funds, purely to the needs of the widespread insurrection that, according to the Daśakumāracarita, they themselves were busily instigating against the central Vākāṭaka house, hoping in the end "to swallow the whole plunder for themselves". After 478, the old programs of work throughout the site were totally abandoned while, with a very few exceptions, the old "Vākāṭaka" patrons (now of course on the wrong side politically) had no say whatsoever in the fate or the future of the excavations upon which they had "expended (such) abundant wealth". (Ajantā Cave 17 inscription, verse 25) (fig.42)

Fig. 42
	Hall > shrine with antechamber. During the very last phase of excavation, work had stopped in this cave, leaving the shrine-doorway incomplete, while painting was already done at many places in the cave.

It is significant that five of the known inaugurators of the site were still alive and active when its course was run, while the sixth-the emperor Harisena-had died barely months before.¹⁵ At the same time, since most of the other caves parallel the inscribed ones in their development, it is reasonable to assume that they too were, in most cases, the work of single donors, even though they apparently never got inscribed. The very eagerness with which most of these donors rushed their shrine Buddhas to completion just after Harisena's death, in order to get the merit, would confirm their intimate connection with such caves, while it is something of a truism in any case, that patrons who start a cave intend to finish it, unless death or taxes intervene. Thus for nearly all of the Vākāṭaka donors, the total development of their caves, from start to finish, takes place within a span of patronage which would typically have started when they were mature and well-established in life.¹⁶

That is to say, the span of Ajantā's development has to be surprisingly brief, since at least ten, and probably more, patrons were present at both the Vākāṭaka phase's inauguration and at its demise. Particularly in a period when life expectancies were far shorter than now-even though these particular donors probably lived a life of privilege-it would be most surprising, from an actuarial point of view, that they would all be alive more than twenty years after the time when they inaugurated the site.¹⁷

Of course, the old view-that Ajantā's development took ten or even twenty times that long-is not unreasonable, even though it is not correct. The speed of stylistic, iconographic, and technological developments depends entirely upon the temper of the times, with political and economic and personal factors all playing their part. Remarkably, in Harisena's reign, all such factors cohered to render the reign of the emperor one of a truly startling productivity and to sponsor a site with perhaps no parallel.

However, before 478 was out, the powers and privileges of the past were gone. Suddenly, with no strong central authority controlling developments at the site, the hold of its old elitist patronage was broken. During the Period of Disruption (479 - 480), when all of the site's previous well-laid plans were in disarray, in principle anyone could put any image anywhere they wished. Furthermore, since most of the caves were very unfinished-in fact, not a single Vākāṭaka cave was ever truly completed-it would seem that there was considerable space available for such donations.¹⁸

However, even if it seems, at first impression, that this disruptive period allowed a kind of free-for-all, the fact is that this was not the case. An absolute rule determined where such intrusive images could go. They could be painted or carved only in caves in which the shrine Buddha had been completed and dedicated-no matter how expediently.¹⁹ For only such caves were "alive".²⁰ Thus although every cave in which the shrine had been finished has either a few or many such donations, these new and uninvited donors carefully avoided caves without a dedicated image, even if there were large areas of seemingly ideal wall space which could have been utilized for their offerings.

It is true, of course, that the elitist barriers had now hhbeen shattered. However, in these days of mounting troubles, who would now come to the collapsing site? The roads, upon which the caravans had traveled so confidently in Harisena's time, were now in jeopardy. One can well understand why numerous representations made in praise of Avalokiteśvara as the Lord of Travelers prominently figure among the many intrusive donations in these difficult days, for the whole countryside must now have been obsessively focused upon the demands and the expectations of war, rather than upon those of the spirit. (fig.43)
 

Fig. 43
	Veranda > rear wall proper right of doorway. LIYANY SCENE of Avalokiteśvara saving the devotees from eight perils (Aṣṭabhaya). It is an intrusive panel.

The truth is, this was really a time to depart from, rather than to come to, this troubled site. This helps to explain why, judging from the many hastily painted inscriptions on their intrusive and helter-skelter votive images, most of the "new" donors in this so-called Period of Disruption (479-480) were the monks still resident at the site. Now, finally at the eleventh hour, they had the chance to make their own donations, appropriating to their own purposes whatever spaces they found desirable. (fig.44) The abundance of such "intrusions" shows that there must also have been a considerable number of sculptors and painters (or sculptor/painters) at the site even after the major work programs were so peremptorily halted at the time of Harisena's death. But, after all, where else had they to go, when the whole empire was falling apart? It seems reasonable to assume that the monks "hired" these artists to make their images in exchange for whatever offerings of food they might be given, or perhaps even in exchange for prayers, in these difficult days.²¹

Fig. 44
	Facade > lower right portion. The side pillared-shrine on right was added later as an afterthought. All the Buddha sculptures on the facade are intrusive. The Buddha figures carved on rear walls of these outer pillared-vestibule are intrusive.

Finally, after about 480, even this helter-skelter donative activity came to an end, most probably because the artists ultimately had either to leave or to starve. (fig.45) The monks themselves, who could beg food in the villages, held out for a few more years, as is evident from the significant wear in the pivot holes in the doorways of many of the site's latest cells-cells which could not even have been put into use until at least 477. However, these cells too had been abandoned, along with the site itself, long before this once so radiant, now so darkening, century was done.

Ironically, it was only this abandonment-this ceding of the site to "the chirping of the birds and the chattering of the monkeys"(Ajantā Cave 26 inscription, verse 18)-that was responsible for its remarkable preservation.²² (fig.46) A few Sādhus, known from trident symbols that they left, lived at the site in later centuries, but otherwise it was waiting, with some concern, for that fateful day in 1819 when its rediscovery would begin.

 

ADDENDUM TO CHAPTER I

A NOTE ON THE DATING OF AJANTĀ'S MAIN, OR VĀKĀ�AKA, PHASE

Including Time Chart and Brief Reconstruction of Events

The Emperor Harisena's regnal dates (c.460-c.477), like the dates for Ajantā's main or later phase (c.462-c.480), cannot be determined absolutely, for there are no dated inscriptions within the two decade span in question. But fortunately there are external termini that frame these twenty years quite closely.

We know that Vākāṭaka patronage did not start at Ajantā until after Harisena's accession, which we place at c.460, since his father was still reigning in 458 (our terminus post quem); it is evident, if one is going allow Ajantā's Vākāṭaka development a span of nineteen years, and if this development had to have ended by about 480 (see below), that Harisena's father's reign could hardly have continued much after 458.

The terminus ante quem is 486, by which time the great Vākāṭaka empire had already fallen, since Anūpa, the last stronghold of that great house, was already in the hands of Mahārāja Subandhu of Mahiṣmati by that date, according to his Barwani inscription.

As the text explains, the end of Ajantā's patronage coincided with the beginning of the great insurrection instigated by the Aṣmakas. This led to the Vākāṭaka's fall sometime prior to 486. Since this tragic course of events under Harisena's son and successor took a few years to run its course-as the evidence of the Daśakumāracarita confirms-it is unlikely that Ajantā's development could have continued past 480.²³

Yet it could hardly have ended much before that date, for although the nineteen year development suggested here (462-480) is certainly reasonable, and is supported by many historical synchronisms, it is hard to compress it much more, or to push it any closer to our terminus post quem.

Thus, even though the many inscriptions at Ajantā itself are not dated, we can determine the beginning and the end of the sequence of developments at Ajantā--both stylistic and historical--rather closely. When we see them as all taking place within a period of twenty years or less--following the so-called "Short Chronology"--it emphasizes the truly remarkable quality of the Vākāṭaka achievement at Ajantā, which preserves so much of the glory of India's Golden Age. The old view, that Ajantā's later development went on for two hundred, or three hundred, or even four hundred years, under a succession of different dynasties, can no longer be sustained, for it totally confuses the history of the site. Ajantā's later phase developed in an incredibly concentrated burst of creative fervor almost totally within the great emperor Harisena's reign. When Harisena died in c.477, and the established control of the site collapsed, there was a brief spate of activity by new and anxious donors; but this too soon fell victim to the times. After the year 480, not a single image, not a single sculpture, not a single painting was ever added to the slumbering site again.

When we see Ajantā's burgeoning Vākāṭaka development as concentrated within the short span of less than twenty years; when we consider how progressive and how consistent this development was; and when we analyze the many distinct breaks in the course of its patronage (see Time Chart) which act as revealing "markers" in its evolution, then we can assign quite specific dates to the various events and the various images which fall within this crowded span. Indeed, in the end we should we should be able to put almost everything which is to be found at the site into a very precise, almost year by year sequence, because the overall span is so short and because the evidence with which we have to work is so abundant. But, unless specifically dated inscriptions from the period of Ajantā's development turn up, we will never be able to be absolute in our dating, but will have to allow a small margin of error--even if only a year or two--for every date that we give. The reader should be aware of this since, for convenience, I date both images and events to specific years herein.

References:

For an explication of the evidence bearing on Ajantā's "Short Chronology", the reader can refer to: 1. W. Spink, "The Vākāṭakas' Flowering and Fall" in R. Parimoo, et als, The Art of Ajantā: New Perspectives (New Delhi, 1991), Vol. 1, pp 71-99. 2. W. Spink, "The Archaeology of Ajantā", Ars Orientalis, Vol. 21, 1991, pp. 67-9. 3. W. Spink, "The Achievement of Ajantā", in The Age of the Vākāṭakas, (ed. A. M. Shastri), New Delhi, 1992, pp. 177-202.

Reference :

1.	The accumulation of debris is also responsible for damage to carvings at the lower levels in the caves, and particularly to the pillars. Nearly all pillars at the site have horizontal flaws reflecting the successive lava deposits. When damp debris builds up around the pillars it further weakens these flawed areas, and when it alternately gets damp and then dry, season by season, it can gradually push out the lower sections, leaving the capitals above attached. This explains why old photographs of Ajantā show pillar sections lying in the porches; though often very beautifully carved, they were all thrown into the ravine during a cleanup campaign many years ago. Old photographs of the great cave at Elephanta also show many "hanging" capitals, with fallen shafts beneath. Also at Elephanta the action of the debris seriously damaged the lower levels of most of the panels, leaving the upper levels quite unaffected. The popular notion that the Portuguese, with their cannons, did this damage, is retired to the realm of myth when we realize that only the feet, never the faces, are damaged.
2.	The later phase is generally called Mahāyāna, but this is hardly precise. As Schlingloff and others have pointed out, it would more properly be identified as "Mulasarvastavadin"-a developed Hinayāna sect.
3.	Among the few precedents, which are hardly comparable in terms of scale or elaboration, are the strong and simple cave at (Eastern) Vākāṭaka excavation at Patur, probably dateable to the early fifth century, as well as the small Gupta excavations at Udayagiri, also ascribable to the early fifth century.
4.	The great stūpa at Sanchi has hundreds of donative inscriptions, especially on its railings.
5.	Cave 29 was buried by debris in the nineteenth century, when the other caves were numbered, so it is out of sequence.
6.	The more elaborate cistern, which may also have served the adjacent Cave 19, was probably the one referred to in the inscription. Like the other (and all at the site) it is filled by water channeled down from the mountain above.
7.	Also, at about this same time, Buddhabhadra had made the decision to allow the huge Cave 24, the courtyard of which was being exposed by about 466, to penetrate into the left side of his Cave 26 complex, essentially putting a stop to most further work on the upper and lower right wings. An expedient shrine was started in the lower wing in 468, but was not completed until a decade later, when Buddhabhadra tried (with only partial success) to get shrines finished in all four wings of the complex.
8.	Evidence for this can be seen in the early portions of Cave 25, which were being simply roughed out. By contrast, note how the porch of Cave 24 was being elaborately decorated (after 475) while the interior was still being exposed.
9.	Actually, the Prime Minister Varāhadeva also ran into difficulties. Although he had clearly expected to rush his cave to completion during the first years of the Recession, he had to abandon it, and finish it after 475.
10.	See following chapter for discussion of Bagh caves.
11.	Actually, Upendragupta had managed to dedicate the cave, which is proved by the presence of numerous later intrusions; it is possible that these new donors did in fact worship in the cave briefly after 479, but this was either very brief, or without much use of ritual lamps. The garland hook (now missing) in the front aisle may actually have been put in place by the "intruders"; the plaster around it did show signs of wear, although it was later filled in (and then partially removed at my suggestion) by the authorities at the site.
12.	See below for discussion
13.	Since the "Vākāṭaka" patrons of the caves along Ajantā's main scarp must have left by mid-478, it is conceivable that some of the intrusions in that area date from late 478; however, until 479, the Aṣmakas were clearly ruling and working at the site, and may have disallowed such unstructured activity. The presence of carved pralambapadasana Buddhas, as intrusions in various attractive and convenient (i.e. early) locations (Cave 4 porch, Cave Upper 6 court, Cave 11 porch, Cave 19 court, Cave 20 hall) would suggest that the "Vākāṭaka" caves were not thus "invaded" until 479 (not late 478), since pralambapadasana Buddhas appear as intrusions only after the Period of Disruption was well underway.
14.	This is not to say that the Cave 26 inscription was the only declaration; it must reflect more "official" announcements made at very much the same time.
15.	Harisena, the minister Varāhadeva, the monk Buddhabhadra and his associates Dharmadatta and Bhadrabandhu, and Mathura.
16.	The very unfinished Caves 3 and 14, as well as the unfinished Aṣmaka caves 22, 23A, 24A,28, must be left out of the calculation, since they were all started after 475, and abandoned before 478
17.	There must have been about a dozen donors who were involved in the fifth century work at the site from start to finish, if as I suggest, the monk Buddhabhadra himself was responsible (at least as overseer) of all of the caves at the western extremity. Of the sixteen caves inaugurated in the early 460s along the main scarp, the doomed local king sponsored five, so for actuarial purposes we can consider only eleven. If each was the work of a separate donor, then counting Since the "Vākāṭaka" patrons of the caves along Ajantā's main scarp must have left by mid-478, it is conceivable that some of the intrusions in that area date from late 478; however, until 479, the Aṣmakas were clearly ruling and working at the site, and may have disallowed such unstructured activity. The presence of carved pralambapadasana Buddhas, as intrusions in various attractive and convenient (i.e. early) locations (Cave 4 porch, Cave Upper 6 court, Cave 11 porch, Cave 19 court, Cave 20 hall) would suggest that the "Vākāṭaka" caves were not thus "invaded" until 479 (not late 478), since pralambapadasana Buddhas appear as intrusions only after the Period of Disruption was well underway. Buddhabhadra, we have twelve donors who inaugurated the caves and who were still alive when their patronage collapsed-or in the case of the emperor Harisena, only the year before. Of course, a few of these donors could have sponsored more than one cave, but there is no proof of this.
18.	The only Vākāṭaka cave at Ajantā that, perhaps in the eyes of the patron, was considered finished is Cave Lower 6. However, we might note that its shrine ceiling and left shrine wall were left partly unpainted, and the rear shrine wall totally unpainted, in the rush to get the Buddha image dedicated in 469. Since there was no attempt to complete these areas in the later phase of work on the cave, when the cells were fitted out, one could (arguably) call it finished.
19.	Cave 7 might seem to be an exception, because it seems clear that the Buddha image was painted along with the many intrusions in the shrine and the shrine ceiling, during the Period of Disruption. But the presence of the intrusions can be explained by the fact the Buddha image, although being remodeled after 475, had already been completed and dedicated in 469; that is, it was already "alive".
20.	The Aṣmaka Cave 22 presents a special problem, for it is filled with intrusions, but had been abandoned in 478 before its shrine was even started. It is possible that it had been started as the donation of a monk (like Buddhabhadra and his two associates) who was remaining at the site in the Period of Disruption, and therefore continued to develop it, perhaps making its fine central Buddha which was clearly painted at the very same time as the adjacent intrusions. Or, since there were many separate shrinelets made in the Period of Disruption as separate offerings, sometimes even connected with "dead" caves (court shrine of Cave 24), perhaps Cave 22, hardly larger than they, was conceived as the same.
21.	Surprisingly, most of the plastering in the cells was done during the Period of Disruption (479-480); perhaps there was merit to be made from such gifts to the sangha. The very idea was very new; prior to 478, not a single cell at the site had been so treated.
22.	Cave 26 inscription, verse 18.
23.	For a detailed discussion of the Daśakumāracarita, see Chapter 2.