At present the Mukkuruni Vinayaka shrine is a modern building constructed of polished granite. It can be dated with all probability to the late 19th century. This Vinayaka represents the Trimurti. Therefore this shrine has three stupi on the top of the shikharam or cupola. Mukkuruni means three (munru) times four measures (kuruni). On special occasions an offer consisting of three modaka made of four measures (kuruni) of rice are given as nivedya or food offer.



The Early Chola Mukkuruni Vinayaka temple consists of the shrine’s garbhagriha and ardhamandapam with three pavilions or shrines added at a later date.

The shrine can be identified as an Early Chola construction on the basis of
the following characteristics.
(1) The ground-plan and lay-out.
(2) The profile of the kapota.
(3) The shape and decoration of the kudus.
(4) The shape of the podigai or corbel.
(5) The shape and decoration of the shikhara.

But it also confronts us with several unique features.
(1) A rectangular garbhagriha and shikhara rounded off at the corners and topped by three stupi.
(2) An ekatala building with several characteristics of a dvitala: a hara with shala and karnakutis.
(3) The ekatala and dvitala characteristics synthesized by integrating the griva niche into the hara.



The vimana is the part of the structure which includes the sanctum or garbhagriha together with its superstructure, the sikhara. On the basis of what is visible in the photo we can classify the ground-plan of this vimana as belonging to the I-2-b class. [2]

– I indicates an ekatala or one storey building.
– 2 indicates a protruding section in the centre of each vimana wall.
– b means the vimana has six pilasters in each wall.

The protruding section in the wall of the sanctum is clearly visible. The kudus that decorate the kapota indicate the presence of pilasters in the structure of the wall. Based on our observations we can draw the ground plan as in this picture. The pairs of pilasters that flank the central vimana niche on both sides are so close together there was probably no space for additional niches besides the main central niche. The space between these would most likely have been a ‘blind niche’, a niche without sculpture.

The general pattern is that a shrine with six pilasters has two levels or storeys, and would therefore be a dvitala. Here we have a ground plan with six pilasters, but no second tala. From the kudus on the kapota of the roof of the ardhamandapa it can be concluded the ardhamandapa walls had two pilasters each. The presence of the shrine in front of the niche indicates there was a devakoshta or niche with a murti or deity placed in it.

From the spacing of the pilasters, and the projection of the vimana wall, it can be inferred this shrine would have had one niche in each wall of the garbhagriha and one niche in each of the ardhamandapa walls, with a total of five niches. It is not possible to know which deities would have found a place in the niches of this shrine.



The Mukkuruni Vinayaka temple in the photo shows an ardhamandapa directly connected to the garbhagriha, but slightly narrower. A faint shadow on the kapota above the first tala seen in the photo indicates where and how the ardhamandapa is attached to the garbhagriha. The ardhamandapa is also rather short. Maybe only half the length of the visible vimana wall. With its niche being almost as wide as the ardhamandapa wall itself. This is a 1-A-1 lay-out. [3]

– 1 indicates one niche in each vimana wall
– A indicates a garbhagriha directly connected to the sanctum and slightly narrower.
– 1 indicates one niche in each wall of the ardhamandapa

1-A-1 temples generally belong to the older phases of construction. Later temples show a straight wall with a false antarala [4] where the sanctum and the porch connect. ‘Koyils‘ shows that this lay-out was applied in the period between 885-910 in the area of the Caveri Delta. [5] And in South Arcot between 945-965. [6]



The pilasters are octagonal as far as is visible. From what can be seen on the photo the podigais or corbels of the temple were angular and possibly throated. This would mean they belong to the period before the construction of the Rajarajesvara temple in Tanjore. T-shaped podigais (for instance as the ones belonging to the mandapa in front of the shrine) were applied for the first time in the Rajarajesvara temple around the year 1000 CE and consequently systematically applied in the construction of stone
temples by Chola architects.



The kapota or cornice of the first tala is typical for all South Indian temple structures.Under the Pallavas this cornice was completely straight. Under the Cholas it developed a more and more pronounced bell-shape. This kapota has a relatively straight profile with only a very slight curve. The kudus are wide and horseshoe-shaped without circles. The top of the kudus are decorated
with simhamukhas, also a feature of Early Chola temple architecture.



The shikhara or cupola is bulbous, ribbed, and rectangular with rounded corners. [7] The bulbous profile is a typical Chola feature. This sikhara is decorated with ribs. Ribbed shikharas are found on 8 temples that were build in the central Caveri Delta by one of the earliest Chola kings, Aditya I. These are the shrines of the (seven) Sapta Sthana temples, all situated around Tanjore, and the Nagesvara temple in Kumbakonam. These 8 temples are stylistically related and form a core of architectural development
and experiment, according to Dr.Hoekveld [8]. Other temples build in the same period, that is before 910 CE, have smooth shikharas. Also throughout the 10th and 11th century most shikharas were smooth.



A hara or roof decoration consisting of karnakutis and shalas is highly unusual for an eka-tala shrine. The only other building that is an ekatala with a complete hara is the Bhima Ratha in Mammalapuram, as far as we have been able to establish. And only very few Early Chola temples are eka-tala with karnakutis, the small miniature shrines on the corner. But this hara also has a shala, a rectangular pavillion at the centre of the haraShalas often have a niche with a deity. Here this niche is very large and functions
at the same time as the griva niche. The deity in this niche is Dakshinamurti, Shiva as a teacher, who is usually placed facing south. He is flanked by rishis on both sides of the niche. This combination of ekatala and dvitala, with 6 pilasters, a hara, and a griva niche synthesized with the shala in this way, is unique.

In the karnakutis we see small sculptures of standing figures. This kind of decoration is already found in early Chalukya (6th to 9th century) and Pallava (6th to 9th century) temples. It is also a feature of the Sapta Sthana temples, the even earlier Vijayalaya Cholishvara temple in Narttamalai, the related and contemporaneous Nagesvara temple in Kumbakonam, and most other dvi– and tri-tala Chola temples dated to before 910 CE. From the photographic evidence at our disposal we conclude that this feature
was much less common in later temples. All these points together make it entirely probable to argue for a date of construction during the earliest phase of Chola architectural development.

This is another reason to argue for an early date for the Mukkuruni Vinayaka temple depicted in the photo.



Alas in the photo not enough is visible of the base to be useful for analysis.


In the photo we can see three secondary buildings that must have been added to the original shrine some time after its original construction. In front of the ardhamandapam is a mukhamandapam. Its base, pillars, roof and general lay-out show it to be of a much later date, very possibly Later Chola. The base is decorated with a lotus moulding and with sculpted panels of which we cannot see the details. The pattika is shaped like a kapota with kudus. Above the pattika is another layer of sculpture the details of which cannot be made out. The steps leading into the mandapam and the shrine are positioned on the south side of the platform in such a way devotees would enter from the side and would not directly approach the deity in a straight line. They are decorated with a curved railing in the form of the tongue of a makara sculpted at the top of the railing. The structure is reminiscent of a ratha mandapam, a hall in the form of a chariot, although there is no evidence of wheels or horses. The pillars are polygonal, and divided in three equal sections from above the low base. The kapota has a strong curve. We would expect the space under the mandapam to be dark, but it is possible to make out details of the dvarapala niches. It may be that there was no roof cover at the time of this photo.

A closed shrine has been erected in front of the southern ardhamandapam niche. This indicates the deity positioned in this
niche was under worship and the devotees had felt the need to convert the niche into a shrine. The jagati of the base base has a pronounced lotus moulding, a round kumuda, and a kapota in the place of the pattika. The kapota of the roof shows a reasonable pronounced bell-shape. The base could be dated to the later Chola period (compare for instance the base of the Airavatesvara temple in Darasuram).

The third attached edifice is an open mandapam with possibly eight pillars. It shows a base that has almost completely
disappeared in the surrounding pavement and the remains of a flight of steps, also partly covered by the pavement. Five of the visible pillars have square and polygonal mouldings. The fifth is round with a pronounced kumbha. Faintly visible behind the pillars there seems to be a kind of structure with mouldings. As the mandapam is open it would most likely have been used to perform
certain rituals, probable during festivals. There may have been a small pedestal or platform for such occasions. We can see it has been ritually painted with alternate white and a dark color. Probably red as is the custom. This confirms the concept that it was used for ritual purposes. Of the base only the kapota-shaped pattika with several decorative kudus are visible. Under this kapota
there seem to have been sculpted panels. The kapota of the roof shows a strong wave-form pointing to a relatively late date.

From the evidence we may conclude a considerable period of time elapsed between the construction of the original shrine and the
building of the secondary structures. The base of all three structures actually points to the time of the later Cholas in the 12th century. They seem to be related to the ratha type structures popular in that period. The curve of the kapota of the three attached shrines shows a much stronger slant in comparison with the kapota of the original Mukkuruni Vinayaka shrine. Compare for instance the kapota of the Raja Sabha build in the first half of the 12th century during the reigns Vikrama Chola (1118-35) and Kulottunga II (1133-50) [9]. And the kapota of the mandapam originally situated in the inner prakara. It may therefore be argued a
considerable time elapsed between the original construction and the building of the secondary shrines.



The Shri Nataraja temple is situated between the two rivers Velar and Coleroon (Kollidam) near the east coast of Tamil Nadu. It is situated in what are the approximate boundary areas  between two traditional regions, Chola Nadu and Nadu Nadu. Chola Nadu or
Land of the Cholas consists of the Caveri Delta. It was where the Chola dynasty had its power base. Nadu Nadu means literally Central Land. It consisted of the areas around modern Cuddalore and Pondicherry and is now generally known as South Arcot.

But although Chidambaram could be considered to belong to South Arcot geographically, historically it has always had strong connections with the Chola dynasty. The presiding deity, Shri Shiva Nataraja was the Kula Devata or family deity of the Cholas. Scenes of Rajaraja Chola worshipping Shri Nataraja in Chidambaram are depicted in the Rajarajesvara temple in Tanjore, build around the year 1000 CE. Generally the deity of Chidambaram was held in high respect by all the kings and dynasties of South India. This is witnessed by the compositions of the Tevaram, by inscriptions, and by legends and oral traditions.

It could be argued that its geographical positioning outside the delta proper would disallow for an early dating, according to the diffusion model as proposed by Dr.Hoeveld. But its spiritual importance for the Chola kings could support the following
argument. Aditya I was the king who commissioned the building of the Sapta Sthana temples between 872 and 885. He is known to have campaigned against the Pallavas and conquered their territory during his reign . He is also known to have spend
considerable time at and near Chidambaram and he and his son Parantaka I are the first historical kings to have gilded the roof of the Chit Sabha (the sanctum sanctorum of the Nataraja temple), according to inscriptions. So this temple was of great personal
significance to them.

Should it therefore surprise us if either Aditya I or Parantaka I build a shrine for Vinayaka as part of the complex? Lets enumerate the arguments. The shrine’s ground-plan (I-2-b) and lay-out (1-A-1) indicate a building date before 910 CE if we treat this temple as belonging to the core-area of the Chola dynasty, the Caveri Delta. The shape of its podigai points to a time of construction before Rajaraja (985 CE). The ribbed shikhara and karnakuti sculptures connect it with the Sapta Sthana shrines and the Nagesvara temple in Kumbakonam, all build by Aditya I before 886. The secondary mandapams seem to have been added at a considerable later date (see especially the differences in the profile of the kapotas) and can be likened to architectural
conventions popular in the late 11th and 12th century.

We would therefore argue it is likely this shrine was build around the year 900 CE under patronage of the Chola king Aditya I. From this we may conclude the following:

(1) The temple complex already had more or less the present lay-out, structure and size at this date, around 900 CE. The positioning of the shrine in this place indicates the prakara lay-out was already established, as was the placement of
the gopurams. The discovery of the photo of this Early Chola shrine calls for a reconsideration of the accepted history of the Nataraja temple in Chidambaram.

(2) Certain academic publications have suggested that before the 10th century the temple and its traditions had a completely different structure and meaning.There were certain supportive arguments for this position. The discovery of proof for the existence of a temple that very likely was constructed during the earliest phase of Chola architecture impels us to reconsider these

(3) This discovery supports the findings of Dr.Sharada Srinivasan. [10] Her metallurgical analysis of Ananda Tandava Murtis suggests some may have been crafted as early as the 9th century under patronage of the Pallava dynasty. Connections between the Pallava dynasty and the Chidambaram temple are also evidenced by the literary tradition.


[1] Hoekveld-Meijer, G. Koyils in the Colamandalam. Typology and Development
of Early Cola Temples. An art-historical study on geographical principles.

PhD Thesis of the Vrije Universiteit te Amsterdam, 1981.

[2] Hoekveld 1981, page 86-88

[3] Hoekveld 1981, page 150-53

[4] Hoekveld 1981, page 150-186

[5] Hoekveld 1981, page 155

[6] Hoekveld 1981, page 155

[7] Hoekveld 1981, page 275

[8] Hoekveld 1981, page 104-111 and 373-377

[9] Younger 1995, page 103-104, Natarajan 1994, page 148-49

[10] Srinivasan, Sharada. Shiva as ‘cosmic dancer’: on Pallava origins of
the Nataraja bronze. in World Archaeology, volume 36, no.3 September

The authors would like to thank the Digital South Asian Library for its kind permission to reproduce the six photos
of the Shri Shiva Nataraja temple in Chidambaram. All other photos by the authors.