James Fergusson

CONTENTS.

Mahavellipore—Kylas, Ellora.

Although it may not be possible to point out the origin of the Dravidian style, and trace its early history with the same precision as we can that of Buddhist architecture, there is nothing so mysterious about it, as there is regarding the styles of northern India, nor does it burst on us full blown at once as is the case with the architecture of the Chalukyas. Hitherto, the great difficulty in the case has been, that all the temples of southern India have been found to be of so modern a date. The great building age there was the 16th and 17th centuries of our era. Some structural buildings, it is true, could be traced back to the 12th or 13th with certainty, but beyond that all was to a great extent conjecture; and if it were not for rock-cut examples, we could hardly go back much further with anything like certainty. Recent investigations, however, combined with improved knowledge and greater familiarity with the subject, have now altered this state of affairs to a great extent. It seems hardly doubtful now that the Kylas at Ellora, and the great temples at Purudkul (Pattadkul), are anterior to the 10th century.[359] It may, in fact, be that they date from the 8th or 9th, and if I am not very much mistaken the “raths,” as they are called, at Mahavellipore are as early, if not indeed earlier, than the 5th or 6th, and are in reality the oldest examples of their class known, and the prototypes of the style.

One circumstance which has prevented the age of the Mahavellipore raths being before detected is, that being all cut in granite and in single blocks, they show no sign of wearing or decay, which is so frequently a test of age in structural buildings, and being all in the same material produces a family likeness among them, which makes it at first sight difficult to discriminate between what is old and what new. More than this, they all possess the curious peculiarity of being unfinished, whether standing free, as the raths, or cut in the rock, as caves, or on its face, as the great bas-relief; they are all left with one-third or one-fourth merely blocked out, and in some instances with the intention merely indicated. It looks as if the workmen had been suddenly called off while the whole was in progress, and native traditions, which always are framed to account for what is otherwise most unintelligible, have seized on this peculiarity, and make it the prominent feature in their myths. Add to this that it is only now we are acquiring that knowledge of the subject and familiarity with its details, which will enable us to check the vagaries of Indian speculation. From all these causes it is not difficult to understand how easily mistakes might be made in treating of such mysterious objects.

If we do not know all we would wish about the antiquities of Mahavellipore, it is not because attempts have not been made to supply the information. Situated on an open sea-beach, within one night’s easy dÂk from Madras, it has been more visited and oftener described than any other place in India. The first volume of the ‘Asiatic Researches’ (1788) contained an exhaustive paper on them by W. Chambers. This was followed in the fifth (1798) by another by Mr. Goldingham. In the second volume of the ‘Transactions of the Royal Asiatic Society’ there appeared what was then considered a most successful attempt to decipher the inscriptions there, by Dr. Guy Babington, accompanied by views of most of the sculptures. The ‘Madras Journal,’ in 1844, contained a guide to the place by Lieutenant Braddock, with notes by the Rev. W. Taylor and Sir Walter Elliot; and almost every journal of every traveller in these parts contains some hint regarding them, or some attempt to describe and explain their peculiarities or beauties. Most of these were collected in a volume in 1869 by a Lieutenant Carr, and published at the expense of the Madras Government, but unfortunately the editor selected had no general knowledge of the subject, nor had he apparently any local familiarity with the place. His work in consequence adds little to our previous stores.

In addition to all this, Colonel Mackenzie undertook to illustrate the place, and employed his staff to make detailed drawings of all the sculptures and architectural details, and a volume containing thirty-seven drawings of the place is in his collection in the India Office, and Daniell has also published some faithful representations of the place. Quite recently it has been surveyed by the revenue surveyors, and photographed by Dr. Hunter, Captain Lyon, and others, so that the materials seem ample; but the fact is, they have been collected at such distant times, and by individuals differing so essentially in capability or instruction, that it is almost impossible, except on the spot, to co-ordinate the whole. Any accomplished architect or archÆologist could do it easily in a month, and tell us the whole story. Meanwhile, however, the main features seem tolerably distinct, and ascertained within limits sufficient for our present purposes.

The oldest and most interesting group of monuments at Mahavellipore, are the so-called five raths or monolithic temples standing on the sea-shore—one of these, that with the apsidal termination in the centre of the annexed woodcut (No. 181), stands a little detached from the rest. The other four stand in a line north and south, and look as if they had been carved out of a single stone or rock, which originally, if that were so, must have been between 85 ft. and 40 ft. high at its southern end, sinking to half that height at its northern extremity, and its width diminishing in a like proportion.

The first on the north is a mere Pansala or cell 11 ft. square externally, and 16 ft. high. It is the only one too that seems finished or nearly so, but it has no throne or image internally from which we might guess its destination.

181. Raths, Mahavellipore. (From a Sketch by the Author.)

The next is a small copy of the last to the southward, and measures 11 ft. by 16 ft. in plan, and 20 ft. in height. The third, seen partially in the above woodcut, is very remarkable: it is an oblong building with a curvilinear shaped roof with a straight ridge. Its dimensions are 42 ft. long, 25 ft. wide, and 25 ft. high. Externally, it seems to have been completely carved, but internally only partially excavated, the works being apparently stopped by an accident. It is cracked completely through, so that daylight can be seen through it, and several masses of the rock have fallen to the ground—this has been ascribed to an earthquake and other causes. My impression is, the explanation is not far to seek, but arose from unskilfulness on the part of workmen employed in a first attempt. Having completed the exterior, they set to work to excavate the interior so as to make it resemble a structural building of the same class, leaving only such pillars and supports as were sufficient to support a wooden roof of the ordinary construction. In this instance it was a mass of solid granite which, had the excavation been completed, would certainly have crushed the lower storey to powder. As it was, the builders seem to have taken the hint of the crack and stopped the further progress of the works.

The last, however, is the most interesting of the series. A view of it has already been given (Woodcut No. 66), and it is shown on the right hand of the last woodcut. Its dimensions are 27 ft. by 28 ft. in plan, 34 ft. in height. Its upper part is entirely finished with its sculptures, the lower merely blocked out. It may be, that frightened by the crack in the last-named rath, or from some other cause, they desisted, and it still remains in an unfinished state.

The materials for fixing the age of this rath are, first, the palÆographical form of the characters used in the numerous inscriptions with which it is covered.[360] Comparing these with Prinsep’s alphabets, allowing for difference of locality, they seem certainly to be anterior to the 7th century.[361] The language, too, is Sanscrit, while all the Chola inscriptions of the 10th and subsequent centuries are in Tamil, and in very much more modern characters.[362] Another proof of antiquity is the character of the sculpture. We have on this rath most of the Hindu Pantheon, such as Brahma and Vishnu; Siva too appears in most of his characters, but all in forms more subdued than are to be found elsewhere. The one extravagance is that the gods generally have four arms—never more—to distinguish them from mortals; but none of these combinations or extravagances we find in the caves here, or at Ellora or Elephanta. It is the soberest and most reasonable version of the Hindu Pantheon yet discovered, and consequently one of the most interesting, as well, probably, as the earliest.

None of the inscriptions on the raths have dates, but from the mention of the Pallavas in connexion with this place, I see no reason for doubting the inference drawn by Sir Walter Elliot from their inscriptions—“that the excavations could not well have been made later than the 6th century.”[363] Add to all this, that these raths are certainly very like Buddhist buildings, as we learn to know them from the early caves, and it seems hardly to admit of doubt that we have here petrifactions of the last forms of Buddhist architecture,[364] and of the first forms of that of the Dravidians.

The want of interiors in these raths makes it sometimes difficult to make this so clear as it might be. We cannot, for instance, tell whether the apsidal rath in the centre of woodcut No. 181 was meant to reproduce a chaitya hall, or a vihara like that of woodcut No. 48. From its being in several storeys I would infer the latter, but the whole is so conventionalised by transplantation to the south, and by the different uses to which they are applied for the purposes of a different religion, that we must not stretch analogies too far.[365]

There is one other rath, at some distance from the others, called Arjuna’s rath, represented in the above woodcut (No. 182), which, strange to say, is finished, or nearly so, and gives a fair idea of the form these oblong temples took before we have any structural buildings

of the class. This temple, though entering in the side, was never intended to be pierced through, but always to contain a cell. The large oblong rath, on the contrary, was intended to be open all round, and whether, consequently, we should consider it as a choultrie or a gopura is not quite clear. One thing, at all events, seems certain—and it is what interests us most here—that the square raths are copies of Buddhist viharas, and are the originals from which all the vimanas in southern India were copied, and continued to be copied nearly unchanged to a very late period. Woodcut No. 183, for instance, represents one from MÁdura, erected in the 18th century. It is changed, it is true, and the cells and some of the earlier features are hardly recognisable; but the wonder rather is that twelve centuries should not have more completely obliterated all traces of the original. There is nothing, however, in it which cannot be easily recognised in intermediate examples, and their gradual transformation detected by any one familiar with the subject. On the other hand, the oblong raths were halls or porticoes with the Buddhists, and became the gopuras or gateways which are frequently—indeed generally—more important parts of Dravidian temples than the vimanas themselves. They, too, like the vimanas, retain their original features very little changed to the present day, as may be seen from the annexed example from a modern Tamil temple on the opposite shore of the Gulf of Manaar (Woodcut No. 184). To all this, however, we shall have frequent opportunities of referring in the sequel, and it will become much plainer as we proceed.

The other antiquities at Mahavellipore, though very interesting in themselves, are not nearly so important for our history as the raths just described. The caves are generally small, and fail architecturally, from the feebleness and tenuity of their supports. The southern cave diggers had evidently not been grounded in the art, like their northern compeers, by the Buddhists. The long experience of the latter in the art taught them that ponderous masses were not only necessary to support their roofs, but for architectural effect; and neither they nor the Hindus who succeeded them in the north ever hesitated to use pillars of two or three diameters in height, or to crowd them together to any required extent. In the south, on the contrary, the cave diggers tried to copy literally the structural pillars used to support wooden roofs. Hence, I believe, the accident to the long rath, and hence certainly the poor and modern look of all the southern caves, which has hitherto proved such a stumbling-block to all who have tried to guess their age. Their sculpture is better, and some of their best designs rank with those of Ellora and Elephanta, with which they were, in all probability, contemporary. Now, however, that we know that the sculptures in cave No. 3 at Badami were executed in the 6th century[366] (A.D. 579), we are enabled to approximate the date of those in the Mahavellipore caves with very tolerable certainty. The Badami sculptures are so similar in style with the best examples there that they cannot be far distant in date, and if placed in the following century it will not probably be far from the truth.

The great bas-relief on the rock, 90 ft. by 40 ft., is perhaps the most remarkable thing of its class in India. Now that it is known to be wholly devoted to Serpent worship,[367] it acquires an interest it had not before, and opens a new chapter in Indian mythology.[368] There seems nothing to enable us to fix its age with absolute certainty; it can hardly, however, be doubted that it is anterior to the 10th century, and may be a couple of centuries earlier.

There is one other antiquity in a place called Saluvan Kuppan, two miles north of Mahavellipore, which has not yet been drawn or described, but deserves notice as a lineal descendant of the tiger cave at Cuttack (Woodcut No. 73). Here not one but a dozen of tiger heads welcome the anchorite to his abode. Here, too, they are conventionalised as we always find them in Chalukyan art; and this example serves, like every other, to show how the Hindu imagination in art runs wild when once freed from the trammels of sober imitation of natural things, which we find to be its characteristic in the early stages of Buddhist art.

Kylas, Ellora.

From the raths at Mahavellipore to the Kylas at Ellora the transition is easy, but the step considerable. At the first-named place we have manifest copies of structures intended originally for other purposes, and used at Mahavellipore in a fragmentary and disjointed manner. At Ellora, on the contrary, the whole is welded together, and we have a perfect Dravidian temple, as complete in all its parts as at any future period, and so far advanced that we might have some difficulty in tracing the parts back to their originals without the fortunate possession of the examples on the Madras shore.

Independently, however, of its historical or ethnographical value, the Kylas is in itself one of the most singular and interesting monuments of architectural art in India. Its beauty and singularity always excited the astonishment of travellers, and in consequence it is better known than almost any other structure in that country, from the numerous views and sketches of it that have been published. Unlike the Buddhist excavations we have hitherto been describing, it is not a mere interior chamber cut in the rock, but is a model of a complete temple, such as might have been erected on the plain. In other words, the rock has been cut away, externally as well as internally. The older caves are of a much more natural and rational design than this temple, because, in cutting away the rock around it to provide an exterior, the whole has necessarily been

placed in a pit. In the cognate temples at Mahavellipore (Woodcut No. 181) this difficulty has been escaped by the fact that the boulders of granite out of which they are hewn were found lying free on the shore; but at Ellora, no insulated rock being available, a pit was dug around the temple in the sloping side of the hill, about 100 ft. deep at its inmost side, and half that height at the entrance or gopura, the floor of the pit being 150 ft. wide and 270 ft. in length. In the centre of this rectangular court stands the temple, as shown in the preceding plan (Woodcut No. 186), consisting of a vimana, between 80 ft. and 90 ft. in height, preceded by a large square porch, supported by sixteen columns (owing probably to the immense weight to be borne); before this stands a detached porch, reached by a bridge; and in front of all stands the gateway, which is in like manner connected with the last porch by a bridge, the whole being cut out of the native rock. Besides these there are two pillars or deepdans (literally lamp-posts) left standing on each side of the detached porch, and two elephants about the size of life. All round the court there is a peristylar cloister with cells, and some halls not shown in the plan, which give to the whole a complexity, and at the same time a completeness, which never fail to strike the beholder with astonishment and awe.

As will be seen from the view (Woodcut No. 187) the outline of the vimana is at first sight very similar to that of the raths at Mahavellipore, but on closer inspection we find everything so modified at Ellora as to make up a perfect and well understood design. The vimana with its cells, and the porch in front of it with its side cells, make a complete Hindu temple such as are found in hundreds in southern India, and instead of the simulated cells that surround the hall in the Madras example, they again become realities, but used for widely different purposes. Instead of being the simulated residences of priests, the five or rather seven cells that surround the central object here are each devoted to a separate divinity of the Hindu Pantheon, and group most pleasingly with the central vimana. It is, however, so far as is now known, the last reminiscence of this Buddhist arrangement in Hindu architecture; after the year 1000 even these cells disappear or become independent erections, wholly separated from the temple itself.

Though considerably damaged by Moslem violence, the lower part of the gopura shows a considerable advance on anything found at Mahavellipore, and a close approach to what these objects afterwards became, in so far, at least, as the perpendicular parts are concerned; instead, however, of the tall pyramids which were so universal afterwards, the gopura in the Kylas exhibits only what may be called the germ of such an arrangement. It is only the upper member of a gopura placed in the flat roof of the gateway, and so small as not to be visible except from above. In more modern times from five to ten storeys would have been interposed to connect these two parts. Nothing of the kind however exists here.[369]

On either side of the porch are the two square pillars called deepdans, or lamp-posts, before alluded to, the ornament at the top of which possibly represents a flame, though it is difficult to ascertain what it really is, while the temptation to consider them as representatives of the lion pillars of the Buddhists (Woodcut No. 6) is very great (Woodcut No. 188).

In the south of India, however, among the Jains, as mentioned above (p. 276), such pillars are very common, standing either singly or in pairs in front of the gopuras, and always apparently intended to carry lamps for festivals. They generally consist of a single block of granite, square at base, changing to an octagon, and again to a figure of sixteen sides, with a capital of very elegant shape. Some, however, are circular, and, indeed, their variety is infinite. They range from 30 ft. to 40 ft. and even 50 ft. in height, and, whatever their dimensions, are among the most elegant specimens of art in southern India.

Unfortunately, there is no inscription or other date from which the age of the Kylas can be ascertained with precision. It is safe, however, to assert that it was erected by the southern Dravidians, either the Cheras or the Cholas who held sway here during the eclipse of the Chalukyas, or between A.D. 750 and 950; and Mr. Burgess’s recent researches in Dharwar enable us to assert with tolerable confidence that its age must be nearer the first than the second of these dates. The great temple at Purudkul—his Pattadkal—is covered with inscriptions, none of which unfortunately are dated, but from their import and the form of their characters, both Bhau Daji[370] and himself ascribe to the 8th or 9th century,[371] and I see no reason for doubting the correctness of the date assigned by Mr. Burgess to this temple, which, according to him was erected during the 8th century. In plan it is almost exactly a duplicate of the Kylas, as may be gathered from the annexed woodcut (No. 189), but there is some little difficulty in instituting such a comparison of their architecture as would enable us to feel sure of their relative dates[372]—in the first place, because the one is structural the other rock-cut, but also because we hardly know what allowance to make for distance of locality. On the whole, however, I am inclined to believe the southern temple is the elder of the two, but certainly not distant in date. If, consequently, it were necessary to fix on a date which should correctly represent our present knowledge of the age of the Kylas, I would put down A.D. 800, with considerable confidence that it was not many years from the truth either way, allowing, of course, some thirty to fifty years for the execution of so important a monument.

Considerable misconception exists on the subject of cutting temples in the rock. Almost every one who sees these temples is struck with the apparently prodigious amount of labour bestowed on their excavation, and there is no doubt that their monolithic character is the principal source of the awe and wonder with which they have been regarded, and that, had the Kylas been an edifice of masonry situated on the plain, it would scarcely have attracted the attention of European travellers. In reality, however, it is considerably easier and less expensive to excavate a temple than to build one. Take, for instance, the Kylas, the most wonderful of all this class. To excavate the area on which it stands would require the removal of about 100,000 cubic yards of rock, but, as the base of the temple is solid and the superstructure massive, it occupies in round numbers about one-half of the excavated area, so that the question is simply this—whether it is easier to chip away 50,000 yards of rock, and shoot it to spoil (to borrow a railway term) down a hillside, or to quarry 50,000 cubic yards of stone, remove it, probably a mile at least to the place where the temple is to be built, and then to raise and set it. The excavating process would probably cost about one-tenth of the other. The sculpture and ornament would be the same in both instances, more especially in India, where buildings are always set up in block, and the carving executed in situ. Nevertheless the impression produced on all spectators by these monolithic masses, their unalterable character, and appearance of eternal durability, point to the process as one meriting more attention than it has hitherto received in modern times; and if any rock were found as uniform and as easily worked as the Indian amygdaloidal traps, we might hand down to posterity some more durable monument than many we are now erecting at far greater cost.

Before leaving this branch of the subject there is one other rock-cut example which deserves to be quoted, not either for its size or antiquity, but from the elegance of its details. It is situated at a place called KÛmÛlÛlÛ,[373] thirty-five miles south-west from Shivelliputtun, and consequently twice that distance north from Cape Comorin. Like the examples at Mahavellipore, this one never was finished, probably because the person who commenced it did not live to complete it, and it was nobody’s business to finish what was of no use, and intended only to glorify him who made it. It is not cut out of a separate boulder, but out of a ridge, as I fancy those at Mahavellipore to have been, and if successful, any number of others of any dimensions might have followed. The other side of the hill had been occupied by the Jains, and numerous images of their Tirthankars are carved upon it, with inscriptions that could easily be read if any one cared to do so. It was evidently to mark the triumph of Siva over Mahavira that this little shrine was undertaken, probably in the 10th or 11th century, and if it had been completed it would have been one of the most perfect gems of the style. For some reason unexplained it was only blocked out, and the upper part only carved, when it was abandoned, and is now entirely forsaken. From its details, it certainly is more modern than the Kylas—how much we cannot yet say with certainty.