Indian Architecture (Buddhist and Hindu Periods) c13: Chapter VI: Rock-Cut Architecture

Indian Architecture (Buddhist and Hindu Periods) c13

Language: English

Chapter VI: Rock-Cut Architecture

(A) The Hinayana Buddhist Viharas or Monasteries

(B) The Rock Architecture of Orissa (2nd Century B.C.)

c13.s2.1

THE rock-cut viharas, or monastic "houses" of the Hinayana Buddhist order, although not of such architectural importance as the chaitya halls or temples described in the previous chapter, contain many significant features. Among other interesting facts, their planning and treatment generally throw considerable light on the system which prevailed in these retreats, and reveal the manner in which the practical requirements of the community were provided for. In their broad aspect they also demonstrate that the Buddhist monachism of India had much in common with the monastic establishments of Europe, a condition due to the similarity in their aims. For instance the Buddhist monks, as did their Cistercian brethren, planted the houses of their order in wild and desolate places, for apparently the same reasons that they might conduct their observances undisturbed by the distractions of any human environment. In a like manner their habitations had a similar beginning, for just as the cloister with its simple lean-to roof on stone pillars was the first step in the construction of the Benedictine monastery, so the early Buddhist vihara consisted of an open court, corresponding to the cloister-garth, enclosed also by a lean-to roof propped up by wooden posts. Then, in connection with the former were developed the "carralls" or small study-rooms, whereas in the latter these took the form of cells leading out of the central court. In the course of time the full complement of each type of monastery, whether European or Buddhist, was composed of a dormitory, a common room, a refectory or frater, a kitchen with other service amenities, and, in the case of the former a fish pond, and the latter a tank for the water supply. As a proof of the fact that both communities occasionally broke away from the rigid discipline of their order to indulge in a little relaxation, certain sinkings in the shape of small cups in the stone benches in the one instance, and in the rock flooring in the other, indicate that the Benedictines in their spare moments played some such game as Nine Men's Morris, and the Buddhists the national Indian game of *pachisi*. Finally, and in the same way that, after the Dissolution, the monasteries in England lay in ruins, so the rock-cut viharas on the decline of the Buddhist faith were completely deserted and left to crumble into decay; where at one time these numerous halls pulsated with spiritual life, they have now for centuries been the home of the jackal and the bat.

c13.s2.2

room was a large square compartment, its space uninterrupted by any formation of pillars or colonnade. (Plate XLIX, Fig. 1.) Further, the cells opening out from the central hall always contain couches or beds also formed in the rock, and there is often a small recess for use as a locker cut in some convenient position. Owing to the situation of the couch in such small cells, which average only nine feet square, the doorway is not in the centre but to one side of the outer wall. In these rock-cut copies of the structural original it is quite easy to see that the central hall corresponded to the open courtyard, while the facade, vestibule, and cells were all translations, in the rock medium, of the conventional wooden type of viharas, of which owing to their impermanency, no examples have survived.

c13.s2.3

An illustration of the Hinayana system of monastic retreat may be seen at the famous Buddhist settlement of Ajanta, where, for a long period, a relatively small group of the early type were the only examples. This first series consisted of five excavations in all, two of which, Nos 9 and 10, are chaitya halls, while there are three viharas, Nos. 8, 12, and 13 (Plate XXV). It will be noted that No. 11 is omitted from the series as it is a Mahayana vihara rather awkwardly interposed at a much later date. Chaitya hall No. 10, with its attached vihara No. 12, were the first to be cut, vihara No 13 being added shortly afterwards to accommodate the increasing body of monks. Then, most probably because the priestly community had been still further augmented to justify the production of another house of prayer, chaitya hall No. 9 was excavated, together with vihara No. 8 as its accompanying monastery. Of these early viharas, No 12 provides a simple but typical example of the single-storied variety, although its facade has almost entirely disappeared. Around its square central hall is carved that horse-shoe arcading characteristic of the work of this phase, its upper portion resolving itself into a pleasing frieze. Every feature is planned and cut with remarkable precision, the facility with which the artificers chiselled out the surfaces and finished off the ornamentation being noteworthy.

c13.s2.4

Another vihara in one story is that attached to the left side of the chaitya hall at Kondane, a representative example, save that it is one of the rare exceptions of the Hinayana type in which the central hall is not plain, but pillared. The exterior, although much of its lower part has broken away, is a very interesting production, as originally it consisted of a pillared portico, the end walls of which still remain. Projecting over this portico is a massive cornice, together with a feature corresponding to an entablature, every detail of which is an exact copy of intricate wooden construction. Within the portico is a screen wall with three square-headed openings, forming the doorway and a window on each side Inside is a large hall measuring 23 feet by 29 feet, surrounded by a colonnade, and with cells opening out from the three interior sides. The pillars of the colonnade support roof

c13.s2.5

As already indicated the architectural significance of the excavated monasteries of the Hinayana sect lies in the fact that they are facsimiles, in the rock, of structural buildings devised to meet a similar demand, and which undoubtedly existed in considerable numbers. All these rock-cut viharas were by no means alike in their design, they took a variety of forms, but a representative example of the earlier or Hinayana type may be distinguished from that of the later or Mahayana, by several well-defined characteristics Chief among their typical features was the open simplicity of the central hall, for, with one or two exceptions, this assembly beams with bridging joists and other structural details, all faithful imitations of a timber original An instructive motif carved on the end wall of the portico is a conventional representation of a chaitya hall facade, from which it is possible to gain no little information as to the final appearance of these frontages. The other Hinayana monastery with pillars in its central hall is that at Pitalkhora, but of this example all that remains are a few of the cells These cells, however, are not plain square rooms as in all the other viharas, but small vaulted chambers with ribbed roofs and lattice windows, reproductions in miniature of the work of the carpenter (Plate VIII, Fig. 8) The exteriors are as ornate as the interiors, as between the overarch of each doorway is a Persepolitan pilaster surmounted by addorsed animals or gryphons.

c13.s2.6

The single storied type of Hinayana vihara in its most decorative form may be seen in a series of three examples at Nasik, particularly in the treatment of their exteriors (Plate XXIX). Judging by the style of their design and workmanship, supported by inscriptional evidence, they appear to have been executed in the first century A. D. According to their inscriptions they have also been named, Gautamiputra (No 3), Nahapana (No. 8), and Sri Yajna (No. 15), Nahapana having been the first to be excavated, followed shortly afterwards by the two others. All have columned porticos and large central halls without pillars, out of which open the usual range of cells containing in most instances stone beds. In their general appearance the porticos of these three viharas are much alike, but the variations in the details are notable, particularly in the design of the pillars. The series of four, with a half pillar at each end, which form the Nahapana facade, are almost exact copies of those in the interior of the Ganesh Lena chaitya hall at Junnar, both of which in their turn having been derived from the portico pillars at Bedsa, every detail, from the lotus-base on the stepped pedestal below to the animal groups on the abacus above, being the same

c13.s2.7

In the facade of the Gautamiputra example at Nasik the designer has worked out a somewhat similar scheme, but with elaborations, as in addition to the range of pillars forming the portico, he has placed their bases behind a richly carved dwarf wall, while below he has sculptured a row of giant figures appearing to carry the entire structure, by means of projecting beams, on their shoulders like an immense tabernacle Under its enormous weight these stalwart atlantes, rising out of the earth, with bulging muscles stagger along, seeming to represent elemental beings from an underworld, forced into the service of the creed Above the portico is a broad architrave supported on the superstructure of the pillars, each of which consists of either a pair of elephants, bulls, gryphons or other beasts, while there is a a fine border of animals, alternating with a scroll of foliage, the whole rendered with exceptional spirit. The entrance doorway in the inner wall of the portico is a square-headed opening surrounded by an elaborate ornamental composition in which are included figures of an unusual character. In some respects the treatment of this doorway recalls that of the Sanchi toranas, as it has lintels or crossbars with voluted ends, but like most of the rock-carving at Nasik, although it retains the principles of the style, shows a marked originality and independence in certain details

c13.s2.8

The remaining example of the three large viharas at Nasik, the Sri Yajna, may have been the last to be executed, but it was produced in much the same form as the others of the group. Then, several centuries afterwards, when the Mahayana priests took over these early monasteries, the interior of this particular vihara appears to have been considerably altered in order to make it suitable for the performance of the later theistic ritual. Such a procedure presented no special difficulties, as the floor was merely sunk so as to provide a square dais towards the middle of the central half, and a cella excavated at the far end with a pillared antechamber for the accommodation of a large image of the Buddha; the style of this alteration indicates that the seventh century was the date of its conversion.

c13.s2.9

Contemporary with the monasteries excavated in the Western Ghats, another group of rock-cut halls and cells was being produced on the eastern side of the country, near Cuttack in Orissa This resolves itself into a collection of chambers, not Buddhistic, but attributed to the opposing belief of the Jains, as in their treatment there are certain features implying a connexion with the latter creed. Nor are there any chaitya halls, for they consist in most instances of a formation of cellular retreats recalling in some respects the viharas of the western Hinayana type. The close grouping on two low hills, not far removed from the famous Brahmanical fanes of Bhubaneshwar, supplies additional proof, if such were required, that this area was of special sanctity, the whole country around having sacred and historical associations To Orissa, as its share in the precious relics of the Buddha, fell the left canine tooth, and the holy city of Dantapura, the "Town of the Tooth", where this priceless possession was at one time deposited, lay in the vicinity of one of the neighbouring towns, either of Bhubaneshwar or Puri, although all traces of it are now lost. As a token of the antiquity of these parts, near at hand is Dhauli hill, where is inscribed one of the rock edicts of Asoka, guarded by a fine sculpturesque representation of an elephant. The two tree-clad hills in which the rockcut chambers are situated, are, together, locally known as Khandragiri, but the northern elevation is called Udaigiri. Separated by a defile, between them wound the "Pilgrims Sacred Way", or *Via Sacra*, to Bhubaneshwar, where it is likely there stood, in early days, a stupa—the pilgrims' goal.

c13.s2.10

According to an inscription carved on the rock of a shallow natural cave on Udaigiri, called the Hathi Gumpha, the mean date of these excavated chambers corresponds to about B C 160. This inscription records, among other information, the chief episodes in the reign of king Kharavela, ruler of Kalinga, the ancient Orissa, including his conquests of a great portion of the north and east of India. Kharavela was by religion a Jain, and appears to have been personally interested in the priestly community who had selected these hills as a place of retreat. It is just possible that the small group of Ajivika hermits responsible for the excavated chapels in the Barabar hills, having lost the protection of Asoka on the death of that monarch, migrated to Orissa, not only to be under a Jaina ruler, but in order to continue their system of living in cells cut in the rock.

c13.s2.11

All the excavations of the Orissan group appear to have been made within the 150 years previous to the Christian Era, after which the production ceased, although on the Khandragiri hill, a short revival took place as late as the medieval period, when a few Jaina cells were added In its general character this group seems to imply an independent development, having but little in common with any other rock architecture of the country. Yet there are certain fundamental forms, both architectural and decorative, found in all this excavated work, which indicate that whether Buddhist or Jain, or whatever the locality, it was all part of the same movement, originating from one source, the beginnings of which had been introduced into the country by Asoka a century before.

c13.s2.12

The principal object of these excavations was obviously to provide a monastic retreat, the chambers and cells being cut out of the coarse sandstone of which the hills are composed As the country around consists mainly of a laterite formation, the selection of this sandstone outcrop, on account of its more workable character, is understood. There are in all some thirty-five excavations, large and small, but only half of them are of any significance, some sixteen of which are in the Udaigiri hill, while there is only one of any importance on Khandagiri. Apparently laid out on no regular plan, they were evidently cut in convenient places and connected by paths, still traceable through the glades of trees. Two of these retreats consist of single cells only, three or four combine several cells having a portico with the remains of a courtyard in front, while the four largest and most elaborate are composed of galleries of chambers in double stories and overlooking a quadrangle None of the courtyards is covered in, as are most of those in other parts of India, but the fact that they are open to the sky may be due to the shallow configuration of the rock in which they are cut. Compared with the Asokan examples, or those in the Western Ghats, the workmanship on the Orissan monasteries is clumsy and crude, but such conditions may be partly accounted for by the rough texture of the sandstone. Moreover it seems to have little or no connexion with the remarkably fine phase of structural architecture and plastic art which developed in its vicinity at Bhubaneshwar and elsewhere under Brahmin patronage some centuries later. The local generic name for these Orissan monasteries is "gumpha", meaning a cave, and each is distinguished by a picturesque or appropriate prefix, as for instance the Ganesh (or Elephant) Gumpha, so called on account of two sculptured elephants forming a prominent part of the conception

c13.s2.13

The architectural features of the more developed Orissan monasteries consist of their facades of pillared verandahs and the cells leading out of them. In the treatment of the former most of the pillars have simple square shafts with bracket capitals, some of the bracket forms being of a very special character. For instance in one example, that known as the Rani Gumpha, or Queen's Cave, the largest and most important of the entire series, there is a bracket of a very primitive order, not unlike the curved branch of a tree On the other hand in the Manchpuri Gumpha the portico pillars support intricately carved struts made up of figures riding hippogryphs and other compositions of a similarly fanciful nature (Plate XXXII, Figs 5 and 6) It may be noted that this form of bracket is the prototype of those which are a prominent feature of the Brahmanical rock-cut temples at Badami in Dharwar, produced at least six centuries later. As already observed a distinctive element in all the early rock-cut viharas is the arcading which decorates the walls, and which in the Orissan examples is of an exclusive kind. Instead of being of the horse-shoe variety, the arches of the arcades are almost invariably semi-circular, and their lower ends, corresponding to the "springer" of a true arch, are expanded to enable them to be supported on pilasters These pilasters have capitals formed of pairs of recumbent animals, a decadent derivative of the Persepolitan type, and a number of them have vase bases. Another feature in some of the Orissan viharas is a ledge or podium carved like a continuous bench around certain of the compartments. Here is seen an early appearance of the *asana*, a stone seat with a sloping backrest, which in a more developed and highly decorated form became prominent in the temples of Central and Western India of the early medieval period. The cells comprising the interiors are not square as in most of the other viharas, but oblong in plan, and some are long chambers entered by several doors, in shape more like dormitories than single rooms; in place of the stone bed, differentiating the early type of cell, the floor in each compartment is sloped so as to form a couch, and as in many instances the height of the room is only four feet, they can only have been intended for sleeping.

c13.s2.14

A few of the single cells of the Orissan group are of a very primitive character, and one known as the Bagh Gumpha, or Tiger Cave, is a fanciful production indicative of a somewhat morbid imagination (Plate XXXII, Fig. 4). Carved out of a shoulder of rock projecting from the hill side, the exterior is shaped like the mask of a tiger, the antechamber simulating the gaping mouth, and the cell door within this, the gullet. On the door-jambs, which slope inwards, are pilasters with winged creatures as capitals, and pots for bases The interior consists of a room only 3½ feet high, but some six feet deep and nearly eight feet wide. Over the doorway is an inscription in characters anterior to the Christian era, stating that it was the abode of an anchorite named Sabhuti, who, reclining in this narrow cell resembling a tiger's maw, seems to have passed his life literally in the jaws of death

c13.s2.15

It has been already remarked that there are no chaitya halls in this monastic retreat, but the example which more nearly corresponds to a temple, is the Rani Gumpha, probably executed about B C. 150 (Plate XXXI, Fig 1). This "abbey church" is a double storied production with its cells ranged around three sides of an open courtyard, the fourth side comprising the frontal approach. A broad terrace, projected from the upper story, was originally supported on structural pillars, either of wood or stone, and these formed the verandah of the ground floor. A stairway cut in the rock at the side gave access to this upper terrace, on one face of which there is a spacious throne, with arms and a foot rest, evidently the seat of honour of the abbot or other high dignitary. While the general arrangements of the cells indicate that most of them were for the accommodation of the monks, there are several supplementary chambers and recesses evidently devised for special purposes connected with the ritual, either as robing rooms or for the preparation of offerings. In addition there are rooms for storing the sacred vessels and vestments, with a place for the custodian or sacristan. (Plate XXXI, Fig. 1).

c13.s2.16

That the open courtyard and its overlooking terraces were specially designed for some spectacular kind of ceremonial seems fairly clear, and a clue to the form that it took is also provided. For around the walls of the upper story there is a long frieze consisting of figures engaged in a series of connected episodes of a distinctly dramatic character (Plate XXXI, Fig. 2). As these same scenes are repeated in part, in one or more of the other viharas on this site, they evidently represent some vivid epic in the heroic age of the people It may be inferred therefore that this arrangement of courtyard and terraces forming the Rani Gumpha constituted an open air theatre in which the scenes depicted in the sculptured frieze around it were brought to life by being performed on festival occasions as a kind of Passion Play, in the same way that the so-called Devil Dances are celebrated in the monastery (*gumpha*) quadrangles of Tibet. If so, the peculiar formation of the gumpha is at once explained, and its various parts fall into their proper place. Moreover it is not difficult to picture the courtyard occupied by the actors in this drama, while seated on the terraces, like an amphitheatre, with the high priest enthroned in the central position, would be a closely grouped background of spectators, the whole forming a brilliant and moving pageant amidst the dark encircling groves.

c13.s2.17

The architectural treatment of Rani Gumpha is characteristic of the Orissan style of rock-architecture as a whole, and provides the key to the entire movement. In the design of the pillars supporting the verandahs and the pilasters of the mural arcading two traditions are represented, the former on the one hand shown by their simple bracket capitals that they are of indigenous origin derived from a wooden prototype; on the other hand the pilasters on the walls with their capitals formed of addorsed animals are debased descendants of the Persepolitan order, still identifiable although rough and coarse in their technique. Yet in spite of the appearance of this classical motif, there is much in the decorative nature of the arches to connect it with the early structural art of the country. It is fairly clear that this ornamental arcading was suggested by the wooden barge-boards at the gable-end of the ordinary house, adapted and repeated so as to become mere embellishment. It is possible to see by the manner in which each member of the arcading fulfils its purpose, the arch "riding" firmly on the pilaster capital, the "railing" string-course fitting securely into the expanded "springer", with the central positions of this rail-moulding supported on corbels, all of which are reminiscences of a structural experience of no common order. Of the plastic decoration of a less conventional nature on the Rani Gumpha, there are certain floral forms boldly modelled and singularly Gothic in their execution, but the chief interest centres in the figure scenes, the frieze already referred to being full of spirit and dramatic power. There are also single figures of life-size proportions sculptured in prominent positions, some of them armed with spears acting as guardians or doorkeepers Of striking appearance are two burly individuals mounted on ponderous animals, the one seated on a bull being strangely Assyrian in its modelling and conception as a whole Within the cells of this gumpha there are the remains of a system of water-supply, apparently conveyed from a rock-cut cistern above, and distributed by means of channels throughout the entire composition, but whether for ablutions, or connected with the ritual is not clear.

c13.s2.18

Other monasteries on this site, treated in much the same manner as the preceding, but simpler in their formation, are the Ganesha Gumpha, the Manchapuri Gumpha, and the Ananta Gumpha. The Ganesha, or Elephant cave, is probably one of the earliest examples and displays several interesting features It is excavated in a ledge, or terrace, of rock, the exterior consisting of a columned verandah, some thirty feet wide and six feet in depth, and approached by steps flanked with figures of elephants. This scheme of sculptured animal guardians to the entrance to a rock-cut hall appears here for the first time, but it was afterwards developed with considerable effect in the Brahmanical temples excavated much later at Ellora and Elephanta, with the elephants, however, replaced by lions. The pillars forming the facade of the Ganesha Gumpha were originally five in number, and are of a type frequently found in this group, the shafts being, in section, square above and below, but octagonal in the centre, with a figure bracket at the top to support the overhanging cornice or eave. At each end of the facade is projected a pilaster *in antis*, not however repeating the conventional design of the pillars, but boldly carved in the shape of a figure—a doorkeeper, or *dwarpala*, armed with a huge spear, and above him, forming the capital to the pilaster is a kneeling humped bull. Executed a little later, the Manchapuri Gumpha or "House of Earth", and the Ananta Gumpha, the latter being on the southern, or Khandragiri hill, are composed of elements and motifs not unlike those in the Ganesha Gumpha just described, and which are common to the group as a whole These sculptured forms, especially when used in the bracket capitals, as for instance the figures astride hippogryphs, have a very ancient appearance, and seem to have emerged from an early phase of art and building construction, of which all other records have disappeared. As already remarked a similar motif was revived several centuries later in the rock-architecture of Badami, but it is so rare and distinctive as to suggest a lost tradition of some importance. The productions at Khandragiri are coarsely rendered, and not of a high standard of design or workmanship Moreover they seem to have been an end in themselves, as except in the few details referred to, they led to no further development, their forms died early in the Christian era leaving no heritage. On the other hand this Orissan rock architecture has every appearance of being a final copy, or the last stage, of a cultural movement which at one time had no little significance a method of expression strictly regional, but of a profound and moving order. What is left merely represents it in its decay.

c13.s2.19

At the ancient city of Junagadh, in Kathiawar, the existence of a considerable group of "caves" indicates that here at one time was settled a thriving religious community manifestly Buddhist and Jain. One of the Asokan inscriptions on a rock at Girnar a mile east of the city is proof of the antiquity of this settlement and also of its Buddhistic origin. The remains of cells suggest that here was at one time a monastery for the accommodation of a large fraternity, while some of the more elaborate excavations, particularly those in the Uparkot, or citadel, by their design seem to signify some special form of ceremonial or ritual. These halls, apparently for communal purposes, are in two stories connected by a winding staircase, with a lower story having broad recesses all round its walls surmounted by a typical frieze of Buddhist chaitya arches. The upper chamber to which is attached a small refectory, also contains a tank and is surrounded by a corridor. But the most striking feature of this compartment are the six columns supporting the roof, which on account of their rich carving, from base to capital, stand out both in their design and technique as the production of one or more experienced craftsmen. No other workmanship quite of this character is known in these parts, and it seems as if here, early in the Christian era probably about 300 A. D., a small group of rock-carvers of exceptional ability flourished for a time and then to have disappeared

c13.s2.20

REFERENCES

c13.s2.21

Bengal District Gazetteer (*Puri*) Cal. 1903 Cunningham, A. *Archaeological Survey of India*, Vol XIII, Cal 1882 Fergusson, J. and Burgess, J., *Cave Temples of India*, London, 1880 Gangoly, M, *Orissa and Her Remains, Ancient and Medieval*, Cal. 1912. Mitra, Rajendralal, *Antiquities of Orissa*, Calcutta 1875, etc