The main ancient elements are:

1) on the northern side, the temple of Mukte¿vara (No. 1), the largest and most beautiful monument on the site; to the east, in its east-west axis a later ma¸·apa;

2) on the eastern elevation of the ground three smaller shrines (2-4);

3) on the southern elevation four shrines in one row (6-9);

4) to the south a lofty monolithic stone mast (5);

5) six inscribed stelae recently installed in a modern shed (10; their original place visible on old photographs is indicated in No. 11).

The diagram of the site refers to the present state. When compared to Cousens plan, it shows a few modifications: the disappearance of the ground structure facing shrine 2 and the displacement of the inscribed stela.

1. Mukte¿vara a. garbhag¤ha

2. Kallid®va b. ¿ukan¡s¢

3. Li´ga c. ¤a´gama¸·apa

4. Li´ga d. mukhabhad¤a

5. K¢rtistambha e. later ma¸·apa

6. Li´ga

7. V¢rabhadra

8. Li´ga

9. Goddess

10. Inscribed stelae (new location in a modern shed; roman numbers refer to Epigraphy section)

11. Inscribed stelae (old location from Cousens plan)

12. V¢rabhadra

The undated photograph reproduced in áivadevavijayam, published in 1949 shows well the original location of the stelae bearing the inscriptions and the elements which have disappeared in front of shrine 2. The stela bearing the image of V¢rabhadra seen on top has been installed in the recent shed with the inscribed stelae.

To the south of the later ma¸·apa of the main monument there is a temple with one cella and an antechamber open to the west. On its axis Cousens plan shows a square platform which looks like the base of a ra´gama¸·apa, and traces of the foundations of a structure which we are not able to identify. This is a possible location for a porch which could have been erected in front of the ra´gama¸·apa. An old photograph which may be contemporaneous with Cousens plan shows in this very place a large stela bearing an image of V¢rabhadra in high relief. It can be identified as the stela which bears the ins-cription No. V on its other side. From the contents of the inscription we infer that this monument is the temple where KaÆn¡d®vi had a Li´ga installed, called Kallin¡tha after the name of her defunct husband, Kallid®varasa, in 1262. Nowadays the stela has been displaced; its original place has been levelled and the traces of the old foundations have disappeared; the traces of the ra´gama¸·apa platform seen on Cousens plan have also disappeared in a modern platform. On the west side of the surviving building we see clear indications of the existence of an attached structure, i. e. pilasters and corbels the presence of which imply the existence of the architraves of the lost structure. The original name, Kallin¡tha, of the Li´ga, is also forgotten. It was al-ready so in the time of Walter Elliot who recorded the name "temple of Gomuni" still in use among the villagers.

This shrine has kept its doorjambs and lintels. The outer walls of the cella have a double axial projection, but no other decoration. There is a tower above the cella and a projection, called ¿ukan¡s¢, coming above the antechamber. The tower is made of seven low strata of gradually diminishing size, with a small dome. The ¿ukan¡s¢ has six strata. Each strata has indentations in a regular spacing. We interpret this feature as a schematic representation of an eave with its attic windows and with turned up points at the angles. The small dome of square plan rests upon a flat moulding with a padma moulding between two recesses. It has a double axial projection on each face. Its graceful curve turns up at the base with upward points at the angles. The crowning motif is a padma moulding of square plan supporting a double circular base for a bulb shaped moulding and a pointed flower bud. The ¿ukan¡s¢ is covered with a flat roof. Its West side is closed with a gable-end motif.

To the south of this structure, a shrine consisting of only one cella open to the west, the outward wall of which has no projection, nor any decoration, is what remains of a small temple. There is no visible trace of any other structure attached to it. There is no superstructure. On an old photograph a post is seen in front of that shrine. It is a short post in stone, approximately 1.5 M high, with a lotus bud at the top, and on one side two mortices, at the bottom and towards the top.

To the south-east, there is a third structure consisting of only one cella of slightly larger size, open to the east. Only the inward facing of the wall remains. The outward facing, as well as the door frame, had already disappeared when Cousens drew his plan.

The southern group comprises four shrines. Cousens plan shows two parallel temples, with cella and antechamber, open to the north-west. At the level of their antechamber they are connected by one single rectangular cella, open to the same direction. To the west extremity of the row there is one more single cella of equal size. Only the inward facings of the walls remain. The three first shrines have preserved their doorframe. The whole has been modified by recent renovations.

The main monument is a structure comprising one cella (garbhag¤ha), one antechamber (¿ukan¡s¢) and a closed hypostyle hall (ra´gama¸·apa), all placed on the same east-west axis. All of them are open to the east. The ra´gama¸·apa has a second opening to the south, and a porch for each opening.

To the east one more ma¸·apa with four central pillars, of slightly larger size, has been added in later times. It has, in fact, no structural connection with the main monument. It is slightly out of its axis, even though its central low platform with four pillars is in the east-west axis. It is closed on its south and north sides. Only the internal facings of these walls remain.

The main structure is a remarkably well-built monument. It is clearly the product of one well-thought architectural project. Though an inscription mentions a renovation made in the 13th century, in the construction, as we see it now, we do not observe any recast of an initial project, nor any later modification or important repair. The renovation may have been such an important reconstruction, that no earlier parts, nor any trace of an older stage have been left visible.

The main monument shelters a Li´ga called Mukte¿vara "Lord of the Released (Souls)". The inscriptions which contain a number of references to the Li´ga called Mukte¿vara and express the feelings and thoughts of the worshippers, in the form of refined literary eulogies of the god, reveal a noteworthy evolution in their attitude and conceptions. The two first inscriptions describe mostly the mythological aspect of áiva. The following ones keep the same theme and introduce new concepts. In inscription III the foundation of a temple is told to be the bringing of Kail¡sa on the earth (l. 33):

"áivadeva having brought the Kail¡sa, said to be beyond imagination, with the army of all gods, to the earth, had a pleasant abode built to áiva, to produce áiva's pleasure."

Similarly the temple of áiva is declared to be a M®ru, MuktikÀetra being a Kail¡sa (III l. 23-24).

The inscription III introduces a new religious figure, áivadeva. With him comes a new concept of the deity, which includes the mythological representation, but does not give to it the same emphasis. The emphasis is put on devotion. A larger place is given to the view of the worshipper áivadeva is told to be an incarnation of Nandin or Amara-Ga¸a i. e. a courtier of áiva in Kail¡sa. He performed worship mostly in the form of heroic vows described in Inscription III. In fact he belongs to the class of saints and yogins practicing meditation and pursuing mystic experiences. He is turned more towards the unmanifested aspects of God. His attachment to the manifested aspect appears in his vow not to move out of MuktikÀetra. May be that was due to his belief in the sanctity of this place. His belief in the sacredness of ár¢¿ailam in Ëndhra, where he resided before his coming to MuktikÀetra in 1225, is also recorded (III l. 15b-16a). The attachment to the Li´gas of these two particular temples seems to be the main link maintained by him with the manifested.

The attitude of the worshipper who directs his mental representation towards a concrete object, who places the unmanifest on the manifest is directly expressed (III l. 16b-17a):

"áivamuddud®va who says: "I shall show to this world that the invisible áiva is manifested here in Dhava½ali´ga" believes in it firmly."

Later the emphasis is placed more on the practice of meditation, than on ritual practices. It goes upto the point to declare that the mental representation creates the concrete form. In inscription No. V stanza 7 which describes the sage áivadeva meditating on the áivali´gatattva and causing it to take form:

"This áivadeva who caused the very rich essence áivali´ga to take form by meditating intensely over it."

The essence called áivali´ga is in V¢ra¿aiva doctrine the supreme entity, not accessible to the senses. The qualification ghanatara is often given to the supreme principle to indicate the plenitude of its essence (compare with the epithet cidghana "full of consciousness". áivadeva practices meditation over the supreme. The intensity of his meditation gives a concrete form to the abstract entity. By the force of meditation the yogin gets a direct perception. Thus the non-manifest and the manifest are linked by the religious practice.

The identification of the concrete Li´ga with the abstract concept of God is also remarkably illustrated in inscription VIII l. 9-10:

"The god Mukte¿vara who is the Li´ga-soul of the siddha áivadeva."

From the date of 1265 onwards áivadeva is called a siddha, i. e. one who has acquired supernatural powers by y°ga, who has attained liberation in the form of union with the supreme áiva. In V¢ra¿aiva vocabulary he has become the principle of áiva Li´ga. His soul or livening power called pr¡´a is a Li´ga, one with the supreme. The demise of the saint took place in MuktikÀetra. This is his sam¡dhi or fusion with Li´ga. From that time the soul of the saint equated with God is placed on the Li´ga in the temple. The soul of the saint fused with God is the unmanifest placed in the manifest by the event of the demise and liberation of the saint. In inscription VII l. 10-11 the name of the saint is transferred to the Li´ga:

"In the presence of the god Mukte¿vara who has the name ár¢ Siddha áivadeva."

One more fact about the Li´ga Mukte¿vara deserves to be recorded. Today the villagers declare that the deity of their temple is Mukte¿vara in the form of an udbhava li´ga. This is the common designation in Kanna·a of a Li´ga which is believed to have come out spontaneously, by itself, in a particular place. The installation is not ascribed to men, but to the god himself, who is considered as having come to the earth in a particular place by his own will. In Sanskrit this type of temple image is called svayaÆbh£ - li´ga. N¢laka¸tha áiv¡c¡rya defines it in his Kriy¡s¡ra:

"Satisfied with the penance of gods and sages, in order to make himself present, áambhu who is inside the earth as a seed in the form of n¡da, like an immobile germ, breaks the ground and becomes manifest. Because he is bh£ta i. e. born from himself, he is known as svayambh£ "self-born". By the worship of this li´ga, knowledge grows by itself.

"The denomination of svayambh£ or udbhava does not occur in the inscriptions, nor is it told in any of them that the Li´ga was installed by any historical figure. References are to the construction of temples, not to the creation of a Li´ga. However two passages give clear indications that the belief in the spontaneous presence of áiva in MuktikÀetra was current in the times of áivadeva. In inscription IV a myth of the descent of áiva in MuktikÀetra is narrated with the words (l. 16b-18a):

"Staying there I shall purify all in the three worlds" saying thus the beloved of mountain's daughter appeared as Mukt¢¿a and aroused love in the Lord of the Earth. "Li´ga has already appeared on the Earth" (See also Inscription V l. 9b-11a).

Therefore nothing goes against the local oral tradition which we record nowadays. And we may consider that Mukte¿vara has been regarded as being a svayambh£-li´ga in the past also.

It also has the low shape characteristic of this type of Li´gas.

East West section looking South, Vim¡na.

The vim¡na is the architectural volume dedicated to the Li´ga Mukte¿vara. The abstract entity, áiva, is represented by the Li´ga, which is here of small size and on a very short base, as in many other temples of the same region and time. The Li´ga is at the centre of a cella of square plan (2.20 x 2.20 M). This cella is almost cubic, the height is 2.45 M under the ceiling, 2.20 M under the architraves. It is called garbhag¤ha as it is the innermost part of the monument. It has only one opening, a door on the eastern side. The walls of considerable thickness (1.50 M at the plinth level in the east-west axis) are made of a filling, which we could not observe, between two facings of long stone slabs. Those of the inward facing have the full size of the wall on each side of the cella. In each angle there is a pilaster. On the west and north walls there is a cyma for the purpose of keeping cult instruments. The architraves have a chamfered edge. The ceiling is made of two series of triangular slabs placed in the angles and leaving a square space in the centre. That square space is covered by a slab decorated with three circles of lotus petals and a flower bud.

The cella is topped by a tower of pyramidal shape with two steps and a small dome. We had no possibility of observing the interior of it.

The outward facings of the walls bear the superimposed secondary structures, which will be described later. In the recesses between them parts of what may be considered as belonging to the main structure, are seen. The elements which it is made of are the same as those of the secondary shrines, so that there is agreement and continuity of both structures. For the cella they are a double base, a wall with pilasters, an eave; for each step of the tower there is a wall with pilasters and an eave. The dome of square plan rests upon a base and a recess. Its outward shape is that of lotus petals curved downwards and with turned up points at the base. It has three axial projections on each side. The last is the architectural motif of a square gable-end of a roof; it is now a flat stone; it may have been the support for a sculptured stela. There is one more crowning motif: three rows of lotus petals and a lotus bud on a double base.

The whole, cella and tower, bears the name of vim¡na or pr¡s¡da. These two terms are used for the dwellings of gods or kings, or more technically refer to buildings having a superstructure. The word vim¡na is sometimes used in reference to the superstructure alone (Mayamata) (VIII 195) and this use which seems to be rare in ancient times, seems to become more current in modern times.

Thus the facings of the walls are of a totally different nature. The inward facing is made of seemingly thin stone slabs placed on edges and devoid of decoration; it is a pure stone architecture. The outward facing is made of thicker stone slabs bearing a sculptured representation of wooden architectural models. The ground level is the same for the cella and the courtyard outside, so that the two facings have the same height.

We have not been able to observe the foundations. Their last layer, the top of which is apparent at the level of the paving of the courtyard, is made of long stone slabs. That may be understood as the component called up¡nah in áilpa¿¡stra. It is the basic layer, the strength and horizontality of which ensures the tightness of the construction. The plinth is a few centimetres in recess. Following the most common prescriptions of áilpa¿¡stra the base is to be considered as made of two parts, lower and upper, called upap¢¶ha and adhiÀ¶h¡na. The former comprises three major protruding mouldings: a flat plinth (jagat¢ or p¡duka) repeated in slight recess and topped by a lotiform moulding (padma), a chamfered band (kumuda), a terminal moulding (Kapota) in the shape of an eave, i. e. a sloping roof with jutting out attic windows cal-led n¡sik¡ or n¢·a. In between these three mouldings there are two deep recesses (ka¸¶ha). A finer square moulding, which may be called Kampa, together with small padmas provide graceful transitions between them. There are deep hollow joints at the base and between the plinths. This series of mouldings forms a unit, as shown by the up¡nah which is a lower element and by the kapota which is an upper and terminal element. That is why it bears the distinctive name: upap¢¶ha. It is the proper base of the building.

All the three major mouldings bear a refined decoration sculptured in very low relief: a creeper motif on the first plinth, a frieze of haÆsa on the second one, a geometrical jewellery motif on the outward edge of the kumuda, a series of n¡sik¡, symmetrically disposed, with the lion-head motif on the kapota.

The upper part of the base, or adhiÀ¶h¡na, is a base for the pillars of the wall. Therefore it may be considered more properly as a part of the wall itself. It consists of two protruding mouldings separated by low recesses. The first moulding above the kapota of the upap¢¶ha is an original feature, which is proper to this type of temple in Karn¡¶aka; it is the representation of a series of extremities of beams with connecting bands. Their number and disposition depend upon the representations of secondary structures and will be described later with them. The last moulding is a square band supported by a large inverted padma. It may be called prati which in áilpa¿¡stra is a common designation of the uppermost moulding of the adhiÀ¶h¡na. It supports directly the pilasters and is not cut by any of the niches.

This adhiÀ¶h¡na can be understood as the direct support of the pillars in a wooden construction. The crossed and radiating beams, the extremities of which are seen in the lower moulding, could be the representation of a tight wooden base for a superstructure. This feature can be seen in modern wooden temple cars. The second band or prati can be a representation of a connecting beam passing through the pillars. Because it is continuous, it ensures the fixation of the pillars and the tightness of the building.

South facade, Central niche.

The wall (p¡da) shows a series of small pilasters (ku·yastambha) carved in the body of the massive stones. This is a representation of thin wooden pillars of a hypostyle pavilion, the space between the pillars being filled with some material. The other structures carved between the pilasters are secondary structures, which will be described below. The represented pillar consists of a base, a shaft of square section and an upper part made of three elements, a bulb slightly protruding, a chamfered band and a large abacus, with thin transition mouldings. The middle element and the upper one are much larger than the shaft. They are the most conspicuous elements and may be the kala¿a and ma¸·i or (phalaka) of áilpa¿¡stra.

Between the abacus of the pilaster and the first architrave there seems to be an intermediary system of corbels. We call corbel the protruding element (potik¡ or bodhik¡) which supports an architrave (uttara) and which is the usual intermediary between a pillar and any superstructure in Indian wooden or stone architecture. The corbel does not rest directly on the abacus. In the space between them appears a square element, which has the breadth of the body of the pillar. That may be a representation of the square tenon engaged in the abacus and the corbel, visible in the space between them, and which is of square section to prevent the abacus and corbel to turn. In áilpa¿¡stra it is called v¢raka¸¶ha or (v¢raka¸·a) and is given as the terminal part of the pillar below the potik¡. In the angles of the monument, where orthogonal architraves are crossed, there is a system of crossed orthogonal corbels.

The entablature is the representation of two superposed architraves (uttara). Where there is a representation of an angle, orthogonal architraves are crossed and their crossed extremities are shown. These extremities appear above the corbels, so that we see a series of three corbelling elements above the pilasters. Such a superposition of corbels and architraves is a feature of wooden architecture, which can be still observed in recently built houses in North Karn¡¶aka.

The pair of architraves bears a long, non-curved eave. The angles have a turned up point. Occasionally there are attic windows called n¡sik¡. The eave is topped by one more protruding moulding, which is a kapota after a small recess. It has the shape of a shorter eave with n¡sik¡. This is not a duplication of the previous eave. The role of the kapota is to mark the termination of an architectural element. Here it indicates the termination of the first level of the building. It has also a functional role. The very long eave rests on the architraves in cantilever position and a strong counterweight must be placed above in order to prevent it to topple down.

Architraves, eave and kapota are given in áilpa¿¡stra as one unit called prastara.

The prastara described so far supports, not one superstructure, but a frieze of figured superstructures of reduced sizes.

Each one comprises an adhiÀ¶h¡na similar to that of the wall, i. e. the moulding made of beam extremities and the upper band (prati), then, after a small recess, a dome-shaped roof (¿ikhara). Because of the division and composition of the side into multiple representations of secondary structures, the frieze is made of several types of ¿ikhara. Those placed in the angles are of square plan and called kar¸ak£¶a. Those in the axis are of oblong plan and called madhyakoÀ¶ha or (bhadra¿¡l¡). Those in the intermediary left spaces are square, are still more reduced in size and called paµjara.

With this frieze of ¿ikhara we can consider that the first level (tala) of the monument is completed. It tops not only the first level of the vim¡na, but also the rangama¸·apa and porches, going around the whole monument, in the same way the base and wall do. The vim¡na portion is topped by two more levels, the structure of which is quite similar to the first one. These two upper levels are of diminishing size and less elaborate than the first one. Because of their reduced sizes they look like attics. But we can recognise in their structure the wall (the adhiÀ¶h¡na of which is hidden in the terrace of the lower level behind the frieze of (¿ikhara) with pilasters, the prastara (without the longer eave), the superstructure with its adhiÀ¶h¡na and frieze of ¿ikhara.