Indian Architecture (Buddhist and Hindu Periods) c12: Chapter V: Buddhist Rock-Cut Architecture: The Early or Hinayana Phase, 2nd Century B.C. to 2nd Century A.D.
Indian Architecture (Buddhist and Hindu Periods) c12
- Language
- English
Chapter V: Buddhist Rock-Cut Architecture: The Early or Hinayana Phase, 2nd Century B.C. to 2nd Century A.D.
c12.1
BUDDHIST ROCK-CUT ARCHITECTURE: THE EARLY OR HINAYANA PHASE
c12.3
ABOUT the same time that the Buddhist communities at such places as Sanchi, Barhut, Budh Gaya and similar sites made sacred by Ashoka's stupas were elaborating these with structural additions, another and entirely different architectural development was becoming manifest in another part of the country. This took the form of Buddhist monastic establishments consisting of large halls and chambers, having notable architectural pretensions, not however, built of wood or stone in the customary manner but hewn out of the living rock by means of the pick, and finished off by the dexterous application of the chisel. It has been already shown that a beginning had been made in this direction by the rock-cut chambers in the Barabar hills, but after an interval of over half a century this method was revived on a much grander and more ambitious scale. Eventually as a form of architectural expression it assumed remarkable proportions, as there are as many as twelve hundred excavations of this nature, both large and small, in various selected localities. Rock-sculpture and rock-architecture have been practised in many countries in the past, particularly in Egypt and Assyria, by the Greeks in Lycia, and the Romans at Petra, while in Persia under the Achaemenids, and later by the Sasanids as seen at Naksh-i-Rustam, both found considerable favour. But in none of these instances did the art of the rock-cutter show so wide a range or such audacity and imaginative power as in India, where some of the most original examples of architecture produced in this manner may be seen. Specially does this apply to the great prayer-halls of the Buddhists, as within their pillared aisles there is something not only majestic but magical, as if they were the abode of spirits or supernatural beings, the carved and painted images on their walls giving substance to this impression.
c12.4
The fashioning of architectural forms out of the living rock, or rock-architecture, occupies a very prominent place in the development of creative art in India. Yet in spite of the admittedly high aesthetic character of these productions they have never been allowed that position in the field of art to which they are fully entitled. Ever since these examples of rock-architecture became a subject of study, it has been the custom to refer to them as "caves", implying that they were natural grottoes in the mountain side, the haunt of wild people, and still wilder animals. No word would be more misleading to designate these wonderful records of man's handwork, as many of them are large and well planned temples, skilfully wrought and chiselled out of the solid cliff, and to define which the term rock-architecture is the only one which can adequately describe their workmanship. If, however, the usually accepted definition of architecture as "good construction truthfully expressed" is applied, then on account of their technique alone they cannot be classed as architecture in the strict sense of the work. These rock-hewn forms are expert achievements but they involved no constructional principles, nor do they display any functional properties, their columns signify no adjustment of support to load, the arches carry no weight, nor do they counteract any thrust, in the whole operation no structural intelligibility is required, as no problems of this nature arise. In a word rock-architecture to all intents and purposes is not architecture it is sculpture, but sculpture on a grand and magnificent scale.
c12.5
It has already been shown that more than one form of early art in India was part of a widely spread cultural development in western Asia transmitted through the influence of Persia when that country was ruled by the Achaemenids towards the fifth century B.C. And it seems probable that the rock-architecture of the Buddhists originated from the same source For although a period of two hundred years separates the two movements, as this form of expression did not begin to appear in India until the third century B.C., the rock workmanship of each country bears a singular family resemblance. A comparison between the facade of the tomb of Darius at Naksh-i-Rustam near Persepolis, and that of the Lomas Rishi "chapel" in Bihar, although the latter is on a much smaller scale, reveals that the textural properties of the two productions have not a little in common. Rock-architecture appealed to the Indian mind for several reasons. In the first place its stability, as it was as immovable as the mountain of which it formed a part, was undoubtedly an attraction to a people living very largely in impermanent structures of wood and wattle. Secondly, it was acceptable to the Buddhists because from the earliest times natural caves and grottoes were the favourite abode of hermits and anchorites, a custom which even now survives in Nepal and parts of Tibet, where a Lamaistic form of Buddhism still prevails. Such habitations were therefore not only associated with religion but had also the sanction of tradition. But the principal reason was the great increase in the conventual life of the country at this time. From the ancient practice of asceticism, common in the Vedic period, it was but a step to that of monasticism, a system which all the world over has induced its followers to retire into rocky fastnesses, forest recesses, or lonely deserts, there to dedicate their lives to the rare worship of the self-absorbed. Some such convictions, perhaps accentuated by the pressure of religious intolerance, for the ruling powers belonged to the Brahmanical faith, were largely responsible for extensive monastic establishments flourishing within these secluded mountain retreats.
c12.6
The production of this particular form of architecture was maintained for a very long period, as the first examples date from the third century B.C., and it continued to be practised throughout the whole of the first millennium only falling into, disuse in early mediaeval times. As associated with the Buddhist religion, rock-architecture in India resolves itself into two distinct movements, separated by a fairly long interval. The earliest or Hinayana phase lasted until the second century A.D., a period of some four hundred years, when for the time being it seems to have naturally declined. Then ensued three centuries of inactivity, but it revived again about the fifth century A.D , after which it was carried on for several centuries with great vigour and attained its richest and most varied form, not only in the hands of the Buddhists but also through similar efforts on the part of the Brahmans and Jains The present outline is concerned only with the first and earliest phase, as developed by the followers of the Hinayana sect, who observed what is known as the "Lesser Vehicle," or primitive aspect of the creed. In its art this primitive Buddhism is distinctive on account of one important convention, for it ordained that in all its productions the bodily form of the Buddha should not be portrayed, in the works of the *Brahmajala Sutta*, "on the dissolution of the body neither gods nor men shall see him "¹ It may be noted that similar restrictions occur in other creeds, as shown by the Hebraic commandment in consequence of which no graven image appeared in the Temple, and the Islamic prohibition regarding representations of the Prophet. In the Hinayana art the spiritual presence of the Master is attested by the introduction in the iconography of certain intimate personal attributes, such as an empty throne, a footstool, cushions, or foot impressions, these symbolized forms being readily recognised and understood by the initiated, but the figure of the Buddha, being sacrosanct, was always omitted. With this convention in mind the classification of Buddhist art and consequently its architecture into its two main divisions is not difficult. It should be observed however that in certain instances attempts were made by the later theistic hierarchy to convert Hinayana monasteries into the Mahayana order by superimposing figures of the Buddha on the earlier work. Such palimpsests may, however, be readily identified, as may be seen on the rock-cut chaitya hall at Kanheri and also at Nasik.
c12.7
¹ *Dialogues of the Buddha* trans by Rhys Davids, p 54.
c12.8
These Hinayana rock-cut monasteries represent a definite regional development, as they are limited to the western side of India, many of them lying within the Bombay State Taking Nasik as the centre they are all situated within a circle having a radius of less than two hundred miles from that town. Here in a configuration of rugged hills known as the Western Ghats, the terrain is naturally adapted to admit this form of architectural treatment, as much of it consists of horizontal strata of amygdaloid and cognate trap formations of considerable thickness and marked uniformity of texture. The edges of these strata terminate in nearly perpendicular cliffs, which provided an ideal surface for the type of rock-architecture contemplated. This architecture resolves itself into some ten separate groups of conventual establishments, each group being separate and self-contained, and each originally consisting of a prayer-hall and its accompanying monastery. The prayer-hall or Buddhist temple, and usually referred to as the *chaitya hall*, as it accommodated a chaitya or stupa, took the form of a large vaulted hall having an apsidal end and divided longitudinally by two colonnades into a broad nave and two aisles. In the apse stood the stupa, also carved out of the natural rock, consisting of an elaborated representation of the structural tumulus previously described. But the most striking fact in connection with the plan and general design of the Buddhist chaitya hall is its undeniable resemblance to the Graeco-Roman basilica, a type of structure which was being evolved in Europe about the same time. In spite of this similarity and their almost contemporaneous emergence, it is extremely unlikely that the two forms of hall were in any way related. The plan and arrangements of the basilica were such as would occur to anyone devising a commodious place of assembly. On the other hand the gradual process of formation to the chaitya hall from its very primitive beginning has been already explained, while its more advanced shape as seen in these rock-cut examples, was brought about by the growing requirements of the Buddhist ritual The side aisles and apsidal end were for the purpose of processions and circumambulation of the stupa, and the nave was provided for the congregational service (Plates XIX, XX and XXI.)
c12.9
The other architectural formation in these rock-cut retreats was the monastery proper, an arrangement of apartments for the accommodation of the monks, and known as a *vihara*. A typical vihara consisted of a square central hall entered by a doorway, in front of which was a vestibule, verandah, or portico. Out of the central hall doorways opened into square cells carried still further into the rock, each of which was the abode of one of the brotherhood. The original plan was for one of these viharas to be situated close to the chaitya hall, but as the priestly community increased, more cells were required to be excavated along the cliff side. These provided quarters for the ordinary members of the community, but there were others of a superior rank who felt it more seemly to live in chambers separated from the main group, of which an example is vihara number 4 at Karli. In addition there were those rare individuals on a still higher plane who claimed to be Arhats or Bodhisattvas, each of whom on account of his status lived alone in a single cell by itself, examples of which may also be seen at Karli. Communication between all these various forms of abode was maintained, where necessary by flights of steps, but in some of the larger monastic establishments, the cells are so numerous that the cliff side is honeycombed with them, recalling the nesting burrows of birds such as swifts in a disused quarry.
c12.10
The style of architecture employed in the rock-cut monasteries was of a very significant and evocative character. It took the form of reproductions, as far as the unusual conditions admitted, of existing structural originals, the general shape of such wooden buildings as well as every detail, even to the joints and fastenings of the carpentry construction, being exactly imitated in the natural rock. And as a proof that, except for the excursions into the rock medium, those who produced this form of architecture were still living in a wood-building age, the rock work itself was supplemented by a substantial amount of wooden construction attached to its surfaces. These additions are shown in both the exteriors and interiors of the rock-cut work, for on the facades of the large halls there are numerous mortice holes for fixing elaborate wooden frontages, while inside in several instances an intricate falsework of wooden groins even now adheres to the rock-cut vaulted roof. There is every evidence, therefore, that the earlier examples of this process were practically half-timbered Although one cannot fail to be impressed by the technical ability displayed in the production of the rock architecture of the Hinayana period, such a servile imitation of one method in another and entirely different material postulates that those concerned were in an early stage of their architectural experience and in the field of building construction still had much to learn.
c12.11
The remarkable manipulative skill displayed in fashioning these architectural forms in the rock—for the actual workmanship is incredibly precise—implies that the workmen employed were by no means mere beginners, but on the [Figure: CHAITYA HALLS, EARLY OR HINAYANA SERIES 2ᴺᴰ CENT B C TO 2ᴺᴰ CENT A D BHAJA KONDANE PITALKHORA NASIK CHAITYA NO IX AJANTA BEDSA
c12.12
SECTION IN PERSPECTIVE OF THE ROCK-CUT BUDDHIST CHAITYA HALL AT KARLI, BOMBAY PRESIDENCY. CIR. 1st CENT. B.C.
c12.13
[Figure: CHAITYA HALLS, EARLY OR HINAYANA SERIES 2ᴺᴰ CENT. B C TO 2ᴺᴰ CENT A D AJANTA SCALE OF 10 0 10 20 30 40 50 60 FEET JUNNAR KARLE KANHERI]
c12.14
Fig 1 Bhaja
c12.15
Fig 2 Kondane
c12.16
Facades of rock-cut Chaitya-halls . 2nd cent B C contrary had served a long apprenticeship to this highly specialised form of quarrying. But strangely enough there are no evidences of experimental undertakings or trial cuttings, no gradual growth, no progressive stages are recorded, the art emerges as does more than one architectural development in India, in a fully matured state. In fact the earliest examples of all these excavated halls, are the most perfectly aligned, planned, and wrought with every line mathematically straight and every angle true. What actually happened when the rock-cut monasteries were first contemplated is, however, tolerably clear, as it can be read in the manner of their formation. The priests who conceived the idea of these retreats called on the master rock-cutters to prepare halls and habitations suitable to their needs. As there was apparently no precedent for such a method of production, the only alternative which presented itself was to repeat in the rock those structures in wood which had already been found serviceable in the past. As from various causes several of the examples have been left unfinished, it is possible to follow each stage of their execution. A selected portion of the naturally steep scarp of the hill side having been roughly cleared until it was perpendicular, a level ledge or platform was thus provided from which the workmen could begin their operations. On this vertical surface the facade of the hall or monastery was marked out, the particular shape that it assumed being a material help in excavating the interior, as in the centre was a large window-opening through which access to the upper part could be obtained. By means of this wide aperture also, the debris from the interior could be removed without difficulty and on this account it was left entirely open until the completion of the work, when it was filled in with a framework of wood. With the spoil excavated from the interior the forecourt in front of the facade was built up.
c12.17
At an early stage in the proceedings, however it is evident that a small chapel with cells attached was first cut near the site of the main hall in which two or three monks entrusted with the direction of the work could live and conduct their usual ritual. Such chapels may be seen in the monasteries at Bhaja and Nasik, each a perfect specimen in miniature, and undoubtedly the earliest of their group. At Nasik also one chamber (No. 14) appears to have been cut first of all in order to provide accommodation for a priestly "resident architect" or "clerk of works", so that he could always be present to supervise the excavation of the main hall above. Having outlined the frontage, the work of hewing this into shape was begun from the top, as it was the practice to commence with the upper portions so as to have the advantage of the rough rock below for footholds, thus dispensing with scaffolding. In the interior the same procedure was followed, the ceiling or roof being first cut and completely finished, the work being gradually carried downwards. To be sure that a correct alignment was obtained the artificers gauged their direction by means of a rough driftway, and then progressed by cutting deep furrows with a tool resembling a heavy pickaxe, after which the intervening ridges were knocked out, the surfaces being dressed with a broad-edged chisel.
c12.18
From the markings left on the walls of rock it is possible to realise the sensitive understanding of the master craftsman, and to discern his hand making that "feathering" movement of the chisel, as he feels his way trying to avoid flaws in the strata or twists in the grain that might ruin the effect. There are evidences of that strong delicacy required so as not to crack or splinter the rock, and also to finish its surface pleasingly. This fine finish was obtained by means of a chisel a little over a quarter of an inch in width. Except for the preliminary work with the pick the cutting was done entirely by the aid of hammers and chisels, as there are no signs of such expedients as the drill or wedge. It seems that the same workmen who performed the rough quarrying also chiselled the carving, for had these two operations been the work of separate hands, portions of the rock would have been left boasted in readiness for the finer manipulation of the sculptor. If such a division of labour was the practice, it has certainly left no trace, although several of the halls remain in an unfinished state, from which it is inferred that the man who worked the pick also handled the chisel—he was quarryman and sculptor combined. As to the kind of tool used in the work there are no records except a few discarded chisels which were unearthed when investigating the foundations of the Heliodorus pillar at Besnagar. These are made of iron, the working of which was well known in India at an early date and there is little doubt that the picks and chisels were of the same metal.
c12.19
Of the two kinds of structure, the chaitya hall and the vihara that were copied in the rock-cut manner, the more important from every point of view was the chaitya hall. There are eight of these belonging to the Hinayana period as follows: Bhaja, Kondane, Pitalkhora, Ajanta (No. 10), Bedsa, Ajanta (No. 9), Nasik and Karli, probably executed in the order named. The two at Ajanta are part of that long series of both Hinayana and Mahayana monasteries all on one site comprising altogether as many as four chaitya halls and over twenty viharas, but this instance of grouping is exceptional and will be explained in due course. All the eight examples were excavated just previous to the Christian era, the first four in the second century, and the remainder in the first century B C. To these may be added two chaitya halls from a numerous and miscellaneous group at Junnar, one of which is small but complete while the other is unfinished, but they are both of the same type and date as that at Nasik. Finally, executed towards the middle of the second century A.D., is the chaitya hall at Kanheri on the island of Salsette adjacent to Bombay, which marks the end of the Hinayana movement as far as its rock architecture is concerned.
c12.20
The chronological sequence in which the chaitya halls were executed is based mainly on the principle that the earlier the example the closer it copies wooden construction. Another guide is the slope of the interior pillars, as it is assumed that their inward inclination was derived from wooden posts so slanted in order to counteract the outward thrust of a heavy timber roof, thus the greater the angle of the rake, the nearer the relation to its wooden prototype Moreover there is the particular shape of the arch forming the main feature of the facade, the curve of which is the subject of change as the style progresses, so that the rule also may be applied that the more rudimentary the curves the earlier the example This may be proved by reverting for the moment to the previous group in the Barabar hills containing the initial effort in rock architecture, that of the chapel of Lomas Rishi, the facade of which shows a gable-end consisting merely of an arrangement of laminated boards, bent rather than curved to a suitable angle (Plate XXV, Fig. 1) About a century later in the earliest example of the Hinayana series now being described, that at Bhaja, the characteristic horse-shoe arch first appears in the facade, but it is of the stilted variety and its indecisive curves suggest that this distinctive feature was undergoing a process of formation (Fig. 2). The facade of Kondane followed and denotes an advance, as there is a slight inward return to the spring of the arch, and the whole is more vigorous and firm in outline At Ajanta (No. 9) and at Karli (Fig. 3) the chaitya arch has arrived at its early maturity, being composed of a subtle and refined combination of curves, a shape which it retained for the remainder of the Hinayana period. Afterwards, in the succeeding Mahayana style as in chaitya hall No. 19 at Ajanta (Fig. 4), the arched aperture tends towards exaggeration, it becomes constricted at the base, and more florid in its curvatures until, as in the seventh century facade of Visvakarma at Ellora, it is no longer a horse-shoe arch but almost a complete circle (Fig. 5). Finally, when the chaitya arch motif appears as a decorative feature in the Brahmanical temple, it assumes a variety of forms of which Fig. 6 is an illustration. As an ornamental accessory known as the *kudu*, a kind of miniature acroterium, it is frequently used to break the line of the cornice in the Brahmanical architecture of the south.
c12.21
A reference to the diagrams on Plate XX, will indicate that the chaitya halls varied considerably in size, and although perhaps at a first glance, notably at Karli, the interior suggests a Norman nave, none of them approach the average dimensions of an ordinary abbey-church in the Gothic style. For purposes of comparison it may be noted that the size of that at Ajanta (No. 10), one of the largest, is appreciably less than that of the somewhat similarly shaped hall in Henry VII’s chapel, Westminster Abbey, London. Although the apsidal end of the chaitya hall is a characteristic feature of its composition, one of the series, namely Ajanta (No. 9), is square-ended but even in this instance the interior colonnade carries out the apsidal plan. The pillars forming the colonnades in all the earlier examples appear as copies of plain wooden posts, chamfered into an octagonal section, and without either capitals or bases. It was only towards the end of the style that it became the custom to elaborate each different member of the pillar and its entablature, thus forming it into some semblance of an “order.” From the first, however, the exterior and frontal aspect of these chaitya halls was regarded as of great architectural significance, and accordingly considerable ingenuity and artistic skill was expended on the design of the facade. In most instances the actual front was composed of a massive pillared portico or vestibule which in the earlier examples has disappeared. Behind this was the facade which consisted of a great horse-shoe archway above, with a wall or screen below, having one or more doorways giving access to the nave and aisles. But undoubtedly the most conspicuous feature of the front-age was the opening or window occupying the centre of the archway, and through which the light was admitted into the body of the hall. This chaitya or sun-window a form of rose-window and fan-light combined, was clearly an enlargement and elaboration of the dormer window of the Vedic hut, with its projecting cowl to keep off the sun and the rain. In shape it is a stilted semi-circular aperture divided into lunettes by means of curved wooden transoms held in place by braces, also of wood, and radiating like the spokes of a wheel. As an architectural conception, and at the same time as an artistic and effective method of introducing light, this very early form of a traceried window is a remarkable achievement.
c12.22
Taking the chaitya hall of Bhaja first, as undoubtedly representing the initial effort, this example is singularly instructive (Plate XXII, Fig. 1). The entrance, and facade generally, owing to the action of time, and the climate, have now become a great open archway, bringing the entire interior of the hall into view. Such, however, was no more the designer’s intention than would be a Gothic church with its west front removed. Originally the whole of the open space was filled in with a highly finished and appropriate wooden construction, which completely screened the lower portion, and affected to no little extent the appearance of the upper parts of the facade From the shape and position of the mortice holes, aided by decorative representations in the bas-reliefs, it is possible to reconstruct the scheme of this wooden frontage. The simple but substantial framework on which it was built took the form of a letter H, the two uprights fitting into each side of the rock-cut archway, the cross-piece being a horizontal beam connecting them and holding them into position. The lower half was then filled in by a screen containing one central and two side doorways, while above the cross-beam was projected a hanging balcony supported on four pillars, the whole constituting a spacious and elegant portico. The facade of wood was ornamented in the same manner as the rock-cut oriels still existing on each side, with lattices and brackets in keeping with the remainder of the scheme. There is little doubt that this timber and rock conception was an architectural composition of considerable merit, ingenious in its construction, and not lacking in artistic effect.
c12.23
In the interior of the hall at Bhaja, woodwork was also freely applied, the closely ranked roof ribs as well as the finial (*harmika*) of the stupa with its umbrella being all originally of this material. But even with these additions it must have had a definitely austere appearance, although its proportions are good. It measures 55 feet long and 26 feet across, the side aisles being 3½ feet wide. The pronounced slope of the pillars which are five inches out of the perpendicular in a total height of 11 feet produces an effect which is not altogether pleasing, but the high stilted vault above, 29 feet from the floor level, is a fine piece of work and to some extent compensates for the unusual slanting formation of the colonnade. As to the stupa this central feature in its present condition is rather a plain conception in two simple parts consisting of a cylindrical base supporting a tall domical body with a “railing” finial. It is more than probable however that the stupa, together with most of the surfaces of the hall itself, were freely decorated with wooden additions, plaster reliefs, and figure subjects in coloured fresco.
c12.24
The next example, the chaitya hall at Kondane, displayed a facade of much the same type as the preceding, except that the upright beams on either side of the archway instead of being entirely of wood were in this instance partly carved in the rock, thus implying a small but progressive step in the evolution of the style (Plate XXII, Fig. 2). Much of the interior has been destroyed but enough remains to show that its dimensions were slightly larger than those of Bhaja, as it is 66 feet long 26½ feet across and its height is rather more than 28 feet. The more finished line of curvature of the facade archway has been already described. Out of the entire series of Hinayana chaitya halls, those of Bhaja and Kondane, being the oldest, are most informative, as they record in a striking manner the methods and expedients, not only of the rock-carver but also of the wood-worker at this early date.
c12.25
The two remaining examples of this class, that of Pitalkhora and Ajanta (No. 10), although considerably ruined, appear to have been of much the same order as the two previously described. In the case of the former, another advance in the development may be seen by the treatment of the roof-ribs in the side-aisles which instead of being of wood are carved out of the rock, evidently a further attempt to reduce the wooden attachment. Owing to a fault in the strata some of the pillars had to be completed structurally, and the stupa having disappeared the probability is that it was built up in the same way, either of stone or wood. The length of this chaitya hall when complete must have been over 50 feet, and it is 34½ feet wide with a height of 31 feet. The last of the series, Ajanta (No. 10), by its large size shows that the artificers were becoming bolder in their conceptions, as it measures as much as 100 feet in length, is 40 feet wide and 33 feet high; in this instance also the ribs of the side-aisles are not of wood but cut out of the rock and they radiate from the end pillars singularly like groins from a Gothic boss. As to the stupa this is somewhat elaborated, its circular base being in two tiers, and the dome is rather more than a hemisphere, indicating a step towards that elongation of its elevation which characterised the later forms.
c12.26
Of the next class of chaitya halls, judging mainly by the design and treatment of the facade, Ajanta (No. 9) and that known as Pandulena at Nasik are the two principal examples, as both have no wooden additions to their frontage, the whole having been carved out of the rock. The Ajanta facade is a singularly well-balanced design recalling in some respects of Gothic rood-screen, with a doorway in the centre and a window on either side, each opening being protected above by an elegant cornice thrown out on brackets like a shallow portico. Over this is what corresponds in the Gothic parallel to the rood-loft, a broad sill or ledge used as a minstrel's gallery, and rising above the whole is the sun-window within a chaitya arch of graceful curves. On the flat surface around the archway are carved as objects of decoration several small lattice windows, conventional renderings of the projecting casements copied so realistically from wooden originals as seen on the previous type at Bhaja and Kondane. As already mentioned the plan of this hall is a rectangle and moreover the ceiling of the side-aisles instead of being curved is flat, while its pillars are perpendicular. Originally the vault of the nave was braced with wooden ribs which, however, appear to have been regarded as superfluous and accordingly were removed at an early date. This left a broad surface of the wall above the colonnade similar to a triforium, which was seized upon by the sixth century painters of the famous Ajanta school as an appropriate space for their mural frescoes.
c12.27
The other example of this small class, the Pandulena at Nasik, with which may also be included the unfinished Manmoda chaitya hall at Junnar, although both differ considerably in their details from the foregoing bear a general resemblance as they are also entirely carved out of the rock and have no porticos or wooden vestibules. A feature of the facade in both instances is a lunette carved with symbolic design, in the Pandulena this is above the doorway, but in the Manmoda it fills the upper space of the archway over the sun-window. In both frontages the decoration on the wall around the main archway contains fresh motifs particularly that on the Pandulena where sculptured in relief there is an arcade containing pilasters of the "Persepolitan" order. The interior of the Nasik example also shows that its designers were beginning to recognise the decorative value of the pillar as instead of these being treated merely as posts, they have introduced on some of them a rudimentary capital in the form of a square abacus. A singular fact however in this connection is that more attention was paid to the embellishment of the bases of the pillars than to the capitals, as here for the first time the pot-shaped base is met with, a motif presumed to have been derived from the primitive expedient of embedding a wooden post in a pot to protect its lower extremity from the ravages of insects or the effect of damp. Another noticeable feature which distinguishes the pillars of this chaitya hall are their proportions. Instead of the broad and massive piers of most of the rock-cut temples, these are tall and slender, with a diameter of one eighth of their height, a proportion very near to the finest Greek and Roman models of approximately the same date. Immediately within the interior of the Pandulena above the doorway there are grooves and sockets for the attachment of a musicians' gallery, which was fixed inside the hall and not as in other examples in the form of a balcony across the outside of the facade. In the design of the stupa there is expressed by the tall proportions of the cylindrical base a still further desire for height in this central object which became more marked as the style advanced.
c12.28
The maturity of the early series of chaitya halls is shown in the two very fine examples at Bedsa and Karli, in which the scheme of the facade differs very materially from any of the preceding. In both instances the outstanding feature of the exterior takes the form of a massive propylaeum carved out of the rock face and serving as a kind of vestibule to the arcaded screen in its rear. That at Bedsa which is probably the earlier, is composed of two columns between pilasters, planned on the same principle as the entrance to a *distyle in antis* type of Greek temple. Only a restricted view of this notable conception is possible as the approach to it is marked by masses of rock left in the rough on either side, for some purpose which is not clear. It is the design and execution of the pillars and pilasters of the portico which make the facade such a remarkable production, as by means of an elaborately carved superstructure they act as supports to the main beam of the roof, all however in one solid piece as the entire frontage is carved out of the natural rock. Derived from the Asokan free-standing monoliths the pillars show the extent of their departure from the classical model during the interval of two hundred years The companiform capital remains only slightly altered, and there is the sculptured group above corresponding to the carved Buddhist symbols on the original. But the vase-shaped base, the octagonal shaft, and various additions to the capital are innovations, several of which suggest a survival from a primitive and indigenous source. The eye is instinctively attracted by the group of figures and animals surmounting the capitals, each of which consists of a male and female figure lightly clad but with heavy ornaments seated astride kneeling animals, on the one side horses and on the other elephants. These groups are exceptionally vivid and spirited examples of rock sculpture, evidently the work of a master craftsman in this medium. Above the figures are the beams, binding joists, and other wooden features of the portico roof all copied most literally in the rock the whole being completed by a parapet of railing design carried across the entire front. Within the portico is the inner facade with its arrangement of archway, sun-window, and entrance doorway all placed according to the conventional disposition of these essential elements. The interior of the chaitya hall is severely plain, the pillars being merely copies of octagonal posts without bases or capitals, although on some of the shafts are carved Buddhist symbols. An appreciable amount of woodwork in the form of vault ribs and other additions was originally attached, but most of it has disappeared within recent years and at one time there were traces of fresco painting with which the pillars and also the stupa were decorated The hall is small in size, being only 45½ feet long and 21 feet wide.
c12.29
It was however at Karli that the Hinayana type of chaitya hall reached its culmination, as this example is a most impressive production The exterior, owing either to a flaw or fall in the rock face is now unsymmetrical, as much of the right side of the facade has become completely obscured, but no doubt as first conceived it was a well balanced composition of somewhat the same style and massive proportions as Bedsa. But at Karli the two large columns, which are the main feature of its form of frontage, were differently disposed, not comprising part of the "construction" so to speak, as they were wholly detached. They were therefore free-standing pillars, one on each side of the facade and slightly in advance of the entrance, which with their lion capitals they appear to guard. The plan of raising two pillars in front of a temple is a very ancient one—before that of the moon god at Ur, 3000 B.C., two wooden columns were raised, and it was probably the same thought that moved the Egyptians to erect a pair of obelisks at the entrances to their temples, while in the porch of Solomon's temple at Jerusalem two pillars of brass were set up as objects of veneration¹ The pillars at Karli appear to have been the result of the same belief, their great size striking the imagination of the worshipper and filling his soul with reverence and awe. Seen from the ample terrace facing them made up of the debris obtained from the excavation, these colossal columns and their symbolic superstructure formed an imposing introduction to the wonders beyond. Each pillar stood on a wide cylinder of rock and with its campaniform capital and group of addorsed lions supporting a large wheel possibly of metal, it was at least fifty feet in height, equalling that of the Asokan column at Sarnath by which the entire conception was certainly inspired. The addition of the base, the sixteen sides to the shaft, the fluted abacus above the capital with its *harmika* forming a square pedestal for the lions, are however all deviations from the original model and indicate a further Indianization of the Persepolitan prototype
c12.30
¹ Kings, Chap. 7.
c12.31
Behind the two lion columns or *sinha-stambhas*, was the vestibule to the hall, the front face of which was formed by a rock-cut screen having a triple entrance below and a pillared clerestory above On the space between these upper and lower openings, corresponding in some respects to a triforium, numerous mortice holes may be seen for the attachment of a minstrel's gallery of wood which extended right across the front. The wooden addition has completely perished but it was an important feature in the design of the facade and would improve its effect; access to this hanging gallery was obtained by means of a stairway at the back of the lion column on the left side. Passing through one of the doorways in the screen the vestibule is entered and the whole inner frontage with its sun-window, decorated arcading, and sculptured figure compositions, comes into view (Plate XXIII). The greater part of the inner wall of the vestibule is occupied by the towering horse-shoe archway within which is recessed the sun-window, portions of the wooden tracery in the opening being still in situ, but the lower section with its semi-circular screen of wooden trellis-work is missing. In the spandrils on each side of the great archway, and also on the narrower ends of this vestibule are carved in the rock tiers of chaitya arcading separated by bands of decorative railings, a very dignified method of ornamenting these subsidiary spaces. Below, some of the panels are filled with figures in relief, but several of them are later Mahayana impositions, and as they were not part of the original scheme they do not harmonise with the rest. Perhaps the most spirited motif in the entire composition is the series of supporting elephants at the ends, each half life-size in high relief and originally provided with tusks of ivory. By such artistic treatment and skilful distribution of all the decorative and symbolic elements, some of them repeated and alternated like a theme in music, the minds of the votaries were attuned and thus prepared for admission into the mysterious and sacred hall of the stupa.
c12.32
Entrance to the chaitya hall was obtained by means of three doorways, the one in the centre being reserved for the priests, and others of high standing, and was approached by a raised pathway on each side of which the floor was sunk to form shallow cisterns filled with water. The ordinary devotees stepped through these shallow pools on their way to the side doors, a procedure which compulsorily cleansed their feet from the contamination of the outer world. Of these who have been privileged to enter this grand interior there must be few who have not been profoundly moved by the solemn permanency and appropriateness generally of its pillared aisles (Plate XXIV, Fig. 1). Although the largest of its kind, its dimensions are not great as it has only 124 feet long, 46½ feet wide and 45 feet high, so that it would go inside the choir of either Wells or Rouen cathedral, yet within its hoary precincts one ceases to think in terms of space and only in those of time. Many more than a thousand years have passed since the rustle of pilgrims' feet in procession last echoed through its shadowed colonnades, or priests filled the high vault of the nave with their chanted liturgy, but the atmosphere of religious emotion which was then created still lingers. There can be nothing but admiration for those who, urged by their passionate devotion for the Great Teacher, could conjure out of the bare hill-side such a majestically imposing and at the same time supremely artistic place of worship.
c12.33
The interior of this chaitya hall owes much of its architectural content to the manner in which its three principal elements have each in themselves been treated and then combined to produce a co-ordinated whole. These are the colonnade, the vaulting, and the great sun-window. Taking the colonnade first, the pillars forming it are thirty-seven in number, and are closely set, much more so than any previous example, as the space between each is but little more than the width of the column itself Those encircling the apse have plain octagonal shafts, as in all the other halls of the series, while the fifteen on each side are highly carved and decorated, repeating in their design, the theme of the larger free-standing lion-columns (*sinha-stambhas*) of the exterior. Each pillar consists of a vase base on a plinth, an octagonal shaft, a campaniform capital with spreading abacus, and each is finished off above by a fine group of sculptured statuary. To the conception of this colonnade the interior owes much of its strength in repose, but it is from the figure composition it supports that the principal vitality of effect is obtained (Plate XXIV, Fig 2). These carved groups are all somewhat similar, but there is sufficient variation in [Figure: Fig. 1 Ajanta, Facade of Chaitya-hall, No 26, 6th Century A D.]
c12.34
[Figure: Fig. 2 Karli Chaitya-hall, Vestibule, c 2nd cent B C]
c12.35
Fig 1
c12.36
[Figure: Fig 2 Karli Chaitya-hall; 2nd cent. B.C.]
c12.37
[Figure: DEVELOPMENT OF CHAITYA ARCH, FROM 250 B C TO 9TH CENT A D
c12.38
LOMAS RISHI CIR 250 B C 2 BHAJA CIR 150 B C 3 KARLI CIR. 50 B C 4 AJANTA CIR 500 A D 5 VISVAKARMA CIR 600 A D 6 BHUBANESWAR 8TH TO 9TH CENT. SITE PLAN OF ELLORA TEMPLES AND MONASTERIES FROM 5TH CENT TO 8TH CENT. SCALE OF FEET 100 0 100 300 500 700 SITE PLAN OF AJANTA 2ND CENT B C TO 7TH CENT A D BED OF STREAM
c12.39
[Figure: Fig 1 Junnar (Poona Dist) Rock-cut facades Manmoda group, Nos 1 and 3 c 1st cent BC]
c12.40
[Figure: Fig 2 Ajanta, general view *(By Courtesy of Anton Schroll & Co., Vienna)*]
c12.41
their design to prevent monotony, and they merge themselves into a kind of sculptured frieze of great beauty. Each group consists of two kneeling elephants, each animal bearing a male and female rider both wearing resplendent head dresses and loaded with a wealth of jewellery. On the reverse side of these groups, within the side aisles, they are also richly carved, but here horses take the place of the elephants, the former having originally been decked with metal trappings, and the latter provided with silver or ivory tusks. The figures were allegorical and intended to signify the opulence and power of the rulers of the earth on their lordly elephants who kneel in humility before the sacred shrine of the Buddha.
c12.42
Above the rich zone of chiaroscuro produced by this frieze of figures rises the high arched vault of the roof, the mystery of its shadowed space vaguely defined by the narrow projecting ribs or wind-braces, not carved out of the rock, but fashioned out of separate pieces of wood attached to the surface by means of plugs or socketed into grooves. These braces are wide flat planks with a curved outline, approximating the shape of an inverted ship's hull, the grain of the wood being usually continuous in the curve as in the mediaeval roofs of Europe, while at the apsidal end they converge to a centre as do the ribs of a Gothic vault. Remains of coloured decoration on the wooden accessories imply that at one time much of the interior was painted. Under the semi-dome of the apse is the stupa, a plain yet dignified composition, its cylindrical base in two diminishing tiers, the sole embellishment being two bands of railing in imitation of balustrades, clearly suggested by the custom of an upper and lower processional path around this symbol as at Sanchi. The body, a plain hemisphere, is surmounted by a massive *harmika*, or finial, decorated with a rail pattern in low relief, and high up over the whole spreads a large wooden umbrella carved in the pattern of a lotus.
c12.43
Such are the arrangements of the colonnade and the vaulting, with the stupa as the focal point of both these essential parts of the scheme. But although the architectural, decorative, and symbolic fundamentals have been so carefully worked out, there is little doubt that the interior of the hall owes much of its appearance to the position and treatment of that most characteristic feature, the sun-window of the facade. On this great aperture depends the whole system of lighting, as the interior is produced on the same principle as a tunnel with one end closed—it could only be illuminated from one direction. But the method by which the open end was manipulated to make the most of the conditions thus imposed shows no little thought and skill. The aim of the designer was to deflect the glaring sunshine outside through this opening to be distributed, modified, and subdued, so that it is projected not so much as light, but as an effulgence, into all parts. It was, therefore, first filtered between the openings of the clerestory in the fronton screen, then broken up by the wooden grille of the sun-window as through a mullioned oriel, finally to fall evenly on the stupa itself like a delicate mantle of light yet edging it with a black shadow. From this central feature it shaded off imperceptibly by modulated half-tones among the surrounding pillars, to be lost in the comparative gloom of the aisles, thus conveying the impression of a fathomless cavern extending indefinitely into the depths of the mountain. There are few lighting effects more solemnly beautiful than the soft luminous atmosphere diffused in this manner through the sun-window at Karli.
c12.44
With the great effort at Karli, the Hinayana phase of rock architecture seems to have spent itself as although viharas for the accommodation of the still increasing communities of monks continued to be added to these retreats, no more chaitya halls were excavated for some time. At the monastery of Junnar, however, a number of chambers, most of them of a relatively small order, were in the course of production over a considerable period, some perhaps as early as the second century B.C., and others in the first century A.D. In one of the groups on this site known as the Tulja Lena, there is a small chaitya hall, which is different from any other and may be a survival of a very ancient type. It is circular in plan and 20 feet in diameter with a stupa in the centre over which rises a dome about 18 feet high supported on twelve plain octagonal pillars, the whole composition being on the same principle as the monopteron style in Roman architecture, of which the temple of Vesta in the Forum Boarium of Rome is an example. Apparently a reproduction in the rock of a primitive structural shrine, it was a form probably not uncommon at one time and of which there are representations on some of the early reliefs, as for instance on the railing at Barhut.
c12.45
There are several other small chaitya halls at Junnar, not apsidal, but consisting of rectangular chambers without pillared aisles and with a stupa at their further end; their exteriors contain neat little vestibules with pillars *in antis* similar to those at Bedsa although on a much smaller scale. Light is admitted into the chamber through the doorway, as the chaitya arch although forming part of the facade is not perforated. In another group on this site, that of the Ganesh Lena (No. 6), there is one chaitya hall, small, yet so perfect in its design and production as to merit special reference. It is of the same dimensions as the hall at Nasik, and the treatment of its pillars is similar to that on some of the viharas there, while the groining also is not wood but rockcut, the style of the whole indicating the first century A.D. as the probable date of its execution. The interior measures only 45 feet in length, but the five pillars on each side of the aisles are carved in the same manner and almost as elaborately as the much larger examples at Karli. Instead however of the elephant and rider motif of the latter, the capitals support animals only, such as pairs of tigers, elephants, and sphinx. The admirable proportions of this hall and the richness of its decoration, although all compressed within a limited space, make it a typical example in miniature of Buddhist rock architecture.
First 44 of 51 passages. Read on