Mediaeval Temples of the Dakhan c6: Architectural Note
Mediaeval Temples of the Dakhan c6
- Language
- English
Architectural Note
c6.1
WITH but few exceptions the ancient Hindu temples of the Dakhan belong, as a class, to the style denominated by Fergusson the Northern or Indo-Aryan, which was practised under the auspices of the various dynasties that ruled over that part of the country which extends from the borders of the Chālukyan and Drāviḍian districts through North Gujarāt, Central India and Rājputāna to the United Provinces and Orissa, during the period from the sixth to the thirteenth or fourteenth centuries A.D.¹ But it is, more particularly, with that branch of it which was fostered by the Yādava rulers of the Dakhan and their feudatories, to which, in many parts, the term "Hemāḍpanti" has been loosely and indiscriminately applied, that we are now interested. For the purposes of this account the term "Dakhan" is taken to embrace all that country between the Kṛishṇā river, on the south, and the Taptī on the north; and from the Koṅkaṇ, on the west, almost up to Haidarābād on the east.²
c6.2
¹ *History of Indian and Eastern Architecture*, Revised Edition, 1910, Vol. II, p. 84.
c6.3
² For a more precise definition of the term, and its extent at different periods in the past, see the *Bombay Gazetteer*, Vol. I, Part II, p. 133.
c6.4
In most parts of the country the building of these old temples is ascribed to a certain Hemāḍpant, just as those of North Gujarāt and the Kanarese districts are placed to the credit of Siddhārājā and Jakanācharya respectively, while those of Khāndesh are frequently assigned to the Gavali Rājās. Who the latter were, if not the early patriarchal rulers of the Yādava family, it is impossible to say. The earlier caves are ascribed to the Pāṇḍavas or Viśvakarma, and, like the temples, they are said to have been constructed during a single night. Hemāḍpant is sometimes described as a famous physician who, in return for some great cure which he effected, demanded the construction of three hundred temples;³ but, more correctly, as the famous minister of Rāmchandradēva, the Yādava king of Devagiri, who lived in the second half of the thirteenth century. This Hemāḍpant, or Hemādri, as he is also called, was a man of many parts who, in addition to his administrative duties, found time to compose several literary works; and, having been much attached to religion, he is credited with having erected some three hundred temples. Having been in a position to command the requisite funds for the purpose, it is quite probable that, like the brothers Tejaḥpāla and Vastupāla of North Gujarāt, in the thirteenth century, he caused a good number of temples to be built, and started a revival in temple building which was followed by others. But the earlier and more decorated examples, found in the Dakhan, were, as we shall see, erected long before his time. The true Hemāḍpanti temples are characterized by heaviness inclining to clumsiness, with severely plain exteriors. This absence of outside figure sculpture may be due to the want of sufficient funds, but that could easily have been met by building fewer and building them better. It is, however, more likely that it was due to the advance of the Muḥammadans, who were merciless iconoclasts and who mutilated and destroyed all they found in their path that savoured at all of idolatry. Images displayed upon the outside of temples would only have exasperated them and invited destruction. It was, no doubt, the uncompromising attitude of Islām which, about this time, caused the building of many temples to be temporarily suspended, and which was taken up again in later times, being finished off, or at least made usable, in a very rough and ready manner.
c6.5
³ Similar stories are told in other parts of India. Vikramāditya of Ayodhyā is said to have built 360 temples, and Hemachandra, the Jaina minister to Kumārapāla of Gujarāt, to have erected no less than 1,400.
c6.6
Compared with the great number of temple remains in North Gujarāt, the Central Provinces and the Kanarese country, those in the Dakhan are very few indeed, and this may be accounted for, in some measure, by the fancy there appears to have been for cave cutting, on which money and energy were exhausted, down to the tenth century at least. The Rāṣṭrakūṭas were cutting cave temples in the living rock in the Dakhan while the early Chālukyas, their neighbours on the south, were building massive shrines in cut-stone, and in the Central Provinces elaborate temples were being raised in moulded brickwork.
c6.7
The earlier buildings are now, without exception, very shaky and in a more or less dilapidated state. This is chiefly due to the material used in their construction—the amygdaloidal trap of the country, quarried, as a rule, on the spot. Though a hard tough stone, it is full of flaws and minute cracks which render it very unsuitable for such parts as beams. The style of these buildings being trabeate, necessitating long beams in most cases, which were loaded with enormous masses of material above, the natural result has been that they have generally cracked through the middle. In the temple of Ambarnātha, for example, there is hardly a sound beam left. But, so long as the supports at either end do not give way, and remain vertical, they usually continue to stand, though cracked, until they eventually collapse through the constant grinding action, at the fracture, pulverizing and chipping the surfaces, unless supported by intermediate struts.
c6.8
A great deal of mischief has been done by people pilfering the material after the buildings had fallen into disuse, but not so much in these stone temples as in the older brick ones in other parts of the country. Brick was more easily handled than stone by the unskilled villager in the building of his hut. “Buildings are always best preserved in places little frequented, and difficult of access: for, when once a country declines from its primitive splendour, the more inhabitants are left, the quicker ruin will be made. Walls supply stones more easily than quarries, and palaces and temples will be demolished to make stables of granite and cottages of porphyry.”¹ This is true all over India.
c6.9
¹ Rassclas, Chapter XXXVIII.
c6.10
The more decorated style of temple building in the Dakhan began to decline soon after the Muḥammadan invasion and eventually died out; the hand of the salāt had forgotten its cunning, or his removal to more congenial and remunerative spheres left no one to carry on his work. Under the sultāns of Gujarāt the Hindu salāts, or sutradāras, were encouraged to practise their craft, with certain necessary restrictions, upon the magnificent mosques and tombs of their conquerors in that province,¹ and there, as may be seen in such buildings as the temple of Hatisingh at Aḥmedābād, the descendants of those builders still follow out, to some extent, the traditions and canons of their forefathers. Other states demanded their skill at this time, such as Māndu, Bidar and Bijāpur. and, in their practice upon Muḥammadan work they, probably. gradually lost their peculiar knowledge as temple builders. Modern temple work in the Dakhan is in a hybrid style partaking, in great measure, of local Muḥammadan feeling, a fair example of which is seen is the accompanying illustration. Fig. No. 2, and the temple of Sundara Nārāyaṇa at Nāsik on plate CXII.
c6.11
¹ See the Muhammadan Architecture of Ahmadābād, by Dr. J. Burgess, in two parts.
c6.12
Fig. 2.—A modern temple.
c6.13
The temples of the Dakhan are of no great size, that of Goṇdeśvara at Sinnar being about the largest now standing. One factor, in limiting the size of a temple. was the length of the shafts of its loftiest pillars. The shaft, or that portion of the pillar between the top of the base and the neck under the capital, was invariably a single block. In their best work the Hindu builders never constructed a shaft of two or more stones, hence the maximum length depended upon the greatest length of stone it was possible, with the means at their disposal, to get out of the quarry. As all parts of a temple are in strict proportion to one another, it follows that the maximum size of a temple was thus limited by the length of the pillar shaft.
c6.14
As already noticed, the material used in these buildings is, almost exclusively, the Dakhan trap, and the masonry has been put together without mortar or any cementing material whatever. The blocks have been dressed to fit one another upon level beds, their weight, and that of the superincumbent masses keeping them in position. Much, therefore, depended upon a firm foundation, and this was, without doubt, a weak point. Very little was done in the way of excavating for firm foundations, the hard black earth or the solid rock, so often found at the surface, being deemed sufficient. Upon this was laid a bed of great rough boulders from which the walls directly sprang, the lowest course of dressed masonry being roughly hammer-dressed to fit upon the uneven boulders (See Plate LXXV). Consequently, any settlement or yielding of the foundation, as has frequently been the case, has caused the collapse of the walls above. For reasons of economy or lightness, brick has been occasionally used for the towers and superstructure above the cornice, as in the temple at Kokamthān. There is reason to believe brick was in much greater use in early times, as we find it in the old Buddhist *chaitya* at Ter¹ and the other old temples at the same place, where they are provided with wooden door frames and beams which, in brick buildings, were afterwards in stone. The stone walls, which were of variable thickness, and far heavier than would be built at present, were run up in two shells—an outer and inner—the space between being filled in with loose boulders or dry rubble. There is a conspicuous absence of through or bond stones or clamps, and there are many instances where the outer shell has fallen away, leaving the inner intact and supporting the roof, so that a person standing inside the building would not notice anything wrong with it.
c6.15
¹ *Archæological Survey of India Annual Report* for 1902-03, p. 195.
c6.16
With respect to the orientation of temples it was the usual practice to place those dedicated to Śaiva worship to face either east or west; for instance, Ambarnātha and Aundha face the west while Gondeśvara at Sinnar and Maheśvara at Pāṭṇa face the east. Vaiṣṇava and Jaina shrines may face these points and also the north, temples to Kṛiṣṇa being generally found facing the latter direction. Temples to goddesses are also found facing the north, while those to Gaṇapati generally face the south.
c6.17
Although many features are common to both, there is a very marked difference between the style of the early Dakhan temples and the Chālukyan² of the same age, and this reveals itself mostly in the *śikhara*, or towers, and the pillars. The Chālukyan style was confined to a much smaller extent of country, and, the land of its origin and development being wedged in between that of the older Drāviḍian of Southern India and the Indo-Aryan of Northern India, it is, as might have been expected, a composite or mixture of both, there is, as partaking of the characteristics of both. Between the temples of the Dakhan and those of the north there is but little difference, and this is mainly apparent in the towers and pillars. The towers in North Gujarāt,³ for instance, are clustered *śikharas*, which are made up of a central main tower with smaller ones clustered around and applied to its sides and corners (See Fig. 3). This pattern also pervades Central and Northern India. The Dakhan *śikhara* rises as a single tower, with fretted vertical bands running up each of its four faces. The corners, between the bands, are filled up with horizontal tiers of miniature *śikharas*, diminishing in size to the top, which, in the treatment they have received, have lost their individuality and have become mere decorative detail. There is a third variety of the northern *śikhara* which is made up entirely of vertical
c6.18
² See *The Chālukyan Architecture of the Kanarese Districts*, by Henry Cousens.
c6.19
³ See *The Architectural Antiquities of Northern Gujarāt* by Dr. J. Burgess.
c6.20
THREE VARIETIES OF THE NORTHERN TYPE OF ŚIKHARA
c6.21
Fig. 3—Types of śikharas.
c6.22
bands, the corners being a repetition of a modified form of *śikhara* but so disguised as to be hardly noticeable at first sight. This is, perhaps, one of the oldest varieties of the northern type and is found upon seventh and eighth century temples in Orissa, the Bombay Presidency, Central India and as far north as the Kulu valley.
c6.23
The Dakhan temples occur singly, in groups of individual temples and in *pañchāyatana* groups, that is groups of five shrines—a central main building surrounded by four subsidiary ones arranged symmetrically around it. A complete temple consists of a shrine containing the object of worship, an antechamber before it, a large hall in advance of this again, with either closed in or open sides, and one or more entrance doorways, each with its porch before it. Śaiva temples generally have a small detached pavilion standing out before the main doorway in which was placed a Nandi or Śiva's bull, Vaiṣṇava temples sometimes have a similar shelter for a *varāha* or boar.
c6.24
The plans of these buildings are either rectangular in their general outline or star-shaped, and revel in a great multiplicity of angles which are carried up through the walls and the towers; and these, with the numerous horizontal mouldings crossing them, break up the wall surfaces into a bewildering mass of projections and recesses with their sparking lights and deep shadows. A reference to the plans illustrated in the following plates will shew, at a glance, how these are designed. The star-shaped plan is arrived at by revolving a square about its centre—in this case the centre of the shrine or hall—the angles being thus all right angles. There is, therefore, nothing very different between the plans of the Dakhan temples and those of Gujarāt and the north generally. Chālukyan temples have, as a rule, larger pillared halls, and often two—an inner closed one and an outer open hall. There are other variations of plans, such as those with three shrines arranged around the common hall, and others where a number of shrines line each side of the hall, both of which are found at Balsāne.
c6.25
The walls and wall mouldings differ little from those of North Gujarāt, but more so from those of the Chālukyan temples. There are four marked varieties, namely, (1) the much decorated with sculptured panels filled with an imposing array of deities and other images, such as the Ambarnātha and Balsāne shrines; (2) those equally decorated but with few images, most of the panels being filled with lozenge-shaped ornament, little pilasters or arabesque, as in the temple of Gondeśvara at Sinnar; (3) those decorated mainly with running bands of arabesque as in the temples of Sangameśvara, Baleśvara at Pedgāon, Ratanvādi, Devalāna and others; and (4) the latest with plain heavy exteriors and little or no attempt at decoration—the Hemadpanti class—like those at Limpangāon and Karjat.
c6.26
As the Dakhan temples are found spread over that part of the country in which the bulk of the older cave-temples occurs it would be natural to expect to see some analogy between the architectural details of both, but it is surprising to find how little there is. A greater likeness is found between the cave work and that of the seventh and eighth century temples of the early Chālukyas. In the first place, the cave-temples, beyond a small amount of ornamental façade, have no exteriors, and the principal features of the interiors, that we can compare with the later work, are the pillars, doors and decorative detail.
c6.27
In the pillars we find little in common so far as their general outlines are concerned; but, in their details, the cave pillars present some earlier forms. The characteristic pot and the cushion capitals, and the pot bases of the earlier caves, such as at Kārli and Elephanta, are not found in those early forms in the structural temple work, but those of the later can, in some cases, be traced back to the earlier forms. The pot and foliage feature which, in the cave-temples and in its more stereotyped form in Gujarāt and the Dakhan, is so universally used, does not occur upon the pillars of Chālukyan temples, but it is found as a very conspicuous ornament alone upon the walls of porches and in similar positions. Figure 118 in Fergusson's *History of Indian and Eastern Architecture* (Volume I, Revised Ed.) shows this in its simplest and most natural form as it occurs upon a pillar in cave XIX at Ajantā, and it may be compared with the same feature upon the pillars in the Tringalvādi cave (Fig. 10) and the temple of Lakshmī-Nārāyana at Pedgāon (Plate LXXXIII).¹ Above the pot and foliage in the first is seen the *kīrtimukha* face which is so much used in the ornamental detail of both the caves and the later temples. This is noticed more fully further on.
c6.28
¹ See the remarks upon the pot and foliage member in *The History of Indian and Eastern Architecture*, Vol. I, p. 300.
c6.29
The prevailing type of pillar used in the Dakhan is that which is found in the Pāṭṇa temple (Plate XXX), but there are numerous varieties as will be seen in the plates which follow. They differ considerably from those of North Gujarāt, and still more from those of the Chālukyan temples to the south. On the brackets, under the beams above the capitals, are found the little fat kichaka figures, a kind of squat Atlantes (see Fig. 11), and sometimes rolled brackets with a cobra head upon the rolls. The former do not occur upon the cave or Chālukyan pillars, but they do on those in Gujarāt. A handsome type of pillar is that in the temple of Lakshmī-Nārāyaṇa at Pedgāon where the pot and foliage occupies a very definite position upon the shaft. In late temples this is replaced by a plain or slightly decorated square block. The style of pillar seen in the temple of Āeśvara at Sinnar, which is scored into a bundle of recessed angles all the way up the shaft, is rare. It may be seen again in the porch before a small shrine near the top of the defile at Lonār lake in Berār. Somewhat like them, though not nearly so much corrugated, are those in the porches of the Ambarnātha temple.
c6.30
There is less difference in the style of the doorways of the temples, in the three provinces compared, than in any other feature, except, perhaps, the decorated ceilings. Yet no two are exactly alike in their detailed reatment. Over the doors of the Dakhan Śaiva temples Gaṇapati is usually the presiding deity, just as he is in Gujarāt, whereas in the Chālukyan temples it is Gaja-Lakshmī. The Jaina temples have, of course, as elsewhere, a Jina. Gaṇapati has always been a favourite deity in the Dakhan, and he is generally invoked at the beginning of a new year or any new undertaking. New ledgers are opened in his name, and his birthday is kept up with considerable parade, when his images are well to the fore. Being the god of wisdom and Śiva's favourite son, it is appropriate that he should be found in this position. As in Gujarāt the kirtimukha mask adorns the threshold of the doorway. The proportion of height to width of the door openings is, roughly, two to one, which is a very pleasing proportion; and the width of the vertical mouldings on either side is often about the same as the opening.
c6.31
Like those of the Northern Gujarāt and Chālukyan temples, the ceilings are a very conspicuous feature in the general scheme of decoration of the interiors. Being almost invisible in the gloom which ever broods over the deity enthroned in the innermost shrine, the ceiling of the sanctum is, as a rule, designed in a very simple manner; but that of the great hall or sabhamaṇḍapa is often a veritable work of art. In the older temples these are usually domical in outline, though the domes are not constructed with radiating voussiors as western domes are. A ceiling, in its simplest form, is no more than one or more flat slabs laid across above four beams forming a square, which are supported upon four pillars or four walls. This is the case in ceilings of the smallest span, but, as the span increases, other devices are used to cover in the larger space. Triangular slabs are then laid across the corners of the square formed by the beams, so as to reduce the opening to a smaller square set diagonally to that of the beams, and, if necessary corner slabs are again placed above these, and so on until the opening is small enough to be closed up by a single square slab. These may then be carved and decorated, a favourite ornamental detail for these simpler ceilings being the rosette or conventional lotus. Such ceilings are generally found in the side bays and corners of the hall.
c6.32
In the largest class of domical ceilings the distance apart at which the four main pillars are set necessitates the insertion of two intermediate pillars on each side of the square, and these are so placed that their points of support form the corners of an octagon within the square. From this, triangular slabs filling in the corners, the octagon is worked up to a circle which forms the base of the dome. This is constructed of ascending concentric and diminishing rings of stones laid upon horizontal beds, each ring being corbelled forward until they meet at the crown, the keystone, together with one or more of the rings immediately around it, forming a pendant. These rings are kept in position by being locked within the mass of the filling-in of the haunches above. The under sides of these stones, which were, no doubt, roughly dressed before being put in position, were carved into rings of cusped and ribbed mouldings, each little cusp having its own small pendant.
c6.33
Much of all this fine work is hardly seen owing to the defective lighting of these temples; and, whether this was designedly so or not, it is hard to say. There is no doubt that an added amount of mystery and a feeling of the occult are gained by an object of worship with a mantle of gloom, the powers of darkness deeply impressing the simple-minded worshipper with the requisite fear and awe upon approaching the sanctuary. Windows are very seldom to be found, and where they have been used they are so filled up with stone grating or perforated arabesque that very little light filters through.
c6.34
The first thing that strikes one, when examining the image sculpture upon these temples, is the apparent absence of drapery. Its existence is only detected, in most cases, by an odd line here and there crossing the legs at the ankle or the breast. The sculptor was always anxious to display the form, and this would have been excusable, in the case of the female form, if they had been able to delineate the human figure with a tolerable amount of success; but their images are not remarkable for beauty of outline or truthfulness to nature. The limbs are without any indication of muscle, the legs being mere drumsticks. It is true that the Indian does not possess the developed muscle of the European, yet he has enough to be plainly seen in the natural limb of a well-grown healthy individual. The idea of muscular display did not occur to the Indian sculptor even in the portrayal of his heroes and warriors. Their strength and prowess were rather expressed by exaggerated action and size of body compared with surrounding figures; and, in their action, there is much more natural vigour and abandon in the earlier cave images. Female charms were made to centre in great hips and breasts; faces are particularly expressionless. There can be little doubt that women of high caste and good social standing wore less clothing above the waist than they do at present in most parts of India, and this is plainly seen in the wall paintings of the Ajanta Caves, where, beyond jewellery and beads, they are represented with, at most, a very thin diaphanous muslin, only noticed from the pattern of the small spots upon it, which more expressed than hid them. The women who were fully clothed from the neck to the ankles were dancing girls and those of that class, the idea being that, with the better class, there was no shame where there was no sin.
c6.35
Another marked feature with these mūrtis, or images, is the amount of jewellery they are represented as wearing. This, of course, denoted wealth. The ropes of pearls and strings of precious stones are carved with a minuteness and accuracy which demanded endless patience on the part of the sculptor. They are disposed about the figure as necklaces, armlets, waistbands and girdles, earrings, bracelets, and festoons hanging from the girdle. Many an image, indeed, appears to be clothed in nothing else but jewellery.
c6.36
Figure sculpture embraces the modelling of gods and goddesses, *rakshasas* or demons, kings, queens and their attendants, warriors, *jogis* or religious mendicants, dancing girls and a few animals. Among the last are most frequently found the elephant, lion, tiger, horse, bull and the animal vehicles of the gods. The best modelled of these is the elephant; and, as this came easiest to their hand, sculptors just revelled in its repetitions. The worst executed is the horse, and, consequently, it is not so often found upon the walls of their temples. The lion and tiger are mixed up in very heraldic attitudes, and it is not always easy to say which is which, excepting when the mane is distinctly shown. The camel is not found on temples in the Dakhan, nor elsewhere in early work, but it occurs on thirteenth, fourteenth and fifteenth century work in Gujarāt and Rājputāna. In the earlier temples, down to the eighth century, of which we have no examples in the Dakhan except the cave temples, the human figure was more natural and vigorous in its poses, and there was more diversity in its representations. But by the eleventh century these had become crystallized into stereotyped forms—a set form for each character—and these were monotonously repeated in every building in which they appeared at all. The sculptor's hand was tied, and he was thenceforth allowed little or no liberty of action or originality of design. All the deities, except Sūrya, the Sun-god, who has but two, are represented with four or more arms, and the reason for this exception is that sun-worship was introduced from Persia. This is further confirmed by the fact that he is almost always represented as wearing high Persian boots, being the only deity with his feet covered. It is thus easy to differentiate between gods and men. Some of the Hindu deities are blessed with more than one head, while one, at least, has been accommodated with three legs. To describe all these would require another chapter upon mythology, which would carry us beyond the scope of the present volume. Individual deities will be noticed as they occur in the following descriptions of the temples.
c6.37
In ornamental detail there is very considerable variety; but, in the earlier temples, as exemplified in the temple of Ambarnātha, there was, perhaps, too much of it crowded over the surfaces of the walls. There were not enough plain surfaces left to act as a foil to the ornament, as there was in the temples of the seventh and eighth centuries. Compare, for instance, the walls of the great temple of Virūpāksha at Pattadakal¹ with those of Ambarnātha, or with that at Halebid, one of the most profusely decorated temples in India.
c6.38
¹ See *The Chalukyan Architecture of the Kanarese Districts*.
c6.39
Two most notable objects in ornamental detail are the *makara* and the *kīrtimukha*, and these are of universal occurrence all over India. On the *makara* an article has already been written in the Annual Report of the Archæological Survey of India, for the year 1903-4. in which an attempt has been made to connect the animal with the tapir or rhinoceros. A good example of this conventionalised beast is that given on plate XLVI. It is certainly not intended for the alligator or crocodile to which the term "maggar" is now applied. The animal not having a tail to speak of in its natural state, the fertile imagination of some early artist has supplied the deficiency with the most wondrous creation in arabesque and floral whorls.
c6.40
The *kīrtimukha* occurs alone or in combination with the *makara*. This curious "Face of Fame," or mask, is found repeated as a regular band of moulding round the basement of temples. In Gujarāt it is called by the *salāts*, or native builders, a *garāsamukha*, and the moulding a *garāspatti*. It is found as a prominent detail upon many pillars and at the crowns of little ornamental arched niches upon the walls. It also occurs in the corners of ceiling panels, and on the front of the thresholds of doorways. In most cases it appears to be a grotesque caricature of the human face, but, often, as the face of a conventionalised or heraldic lion.¹ In the earlier examples, the face, though grotesque, is a well formed one; in later work the *motif* seems to have been forgotten and it is often difficult to discover any resemblance to a face at all. Some good examples of the earlier type are found at the caves of Ajantā. The origin of the word *kīrtimukha*, which means "Face of Fame," is given in one of the accounts of
c6.41
¹ The lion, and the separate mask, are seen side by side on the temple of Kedāreśvara at Balagāmve in Maisur. See *The Chalukyan Architecture of the Kanarese Districts*. Plate CXIV.
c6.42
Fig. 4—The Kīrtimukha mask.
c6.43
the doings of the demon Jalandhara. When he attempted to seduce the wife of Śiva, that god produced a terrible being from his third eye who straightway demanded something to devour. Śiva commanded him to devour himself, which he did. all but his head. Śiva being pleased with this performance, declared that henceforward he should be called Kīrtimukha, and that his head should always adorn the doorways of his temples, where everyone should worship it before entering. On account of the presence of the face on the doorstep, it is usual for worshippers to step over and not *upon* the threshold.
c6.44
Minor ornamental detail is crowded upon the mouldings around door frames, and on pillars and beams, amongst it being some very pretty bands of diaper, scroll and lozenge, and bead ornament. There is also much fretted surface tracery.