The Colas c35.s2: Painting

The Cōḷas c35.s2

Language: English

Painting

c35.s2.1

Cōḷa painting, like all other forms of Cōḷa art, was a continuation and development of Pallava-Pāṇḍya work in the field. Literary evidence on the extent and excellence of early Tamil murals is both authentic and extensive; but no specimens have survived, and we have no means of forming any close idea of the nature of the art as it was practised then. In fact paintings are delicate products of art and are the first to suffer by the action of time and weather, not to speak of chemical changes in rocks or in the materials used. Often excellent early paintings have been overlaid, as at Tanjore, by later work not so good. In the midst of many gaps and much uncertainty, there remains no doubt about the continuity of a painting tradition in the Tamil country.

c35.s2.2

The fragmentary Pallava paintings still traceable in the 'Cave temples' of Tirumayam and Māmạndur, and in the structural temples of Panamalai and Kāñcī and Māmallapuram as also in the rathas of Māmallapuram date from the seventh and eighth centuries, while the outer layer of paintings in Śittannavāśal, often mistaken to be Pallava, and those in the Tirumalaipuram cave temple in the Tinnevelly district are both Pāṇḍya in origin and belong to the ninth century.⁶⁰ Of the Cōḷa paintings, the most important are those in the *pradakṣiṇa* passage round the sanctum of the Tanjore temple, most probably coeval with the temple and belonging to the time of Rājarāja I and Rājēndra I.

c35.s2.3

From the standpoint of the technique employed, the South Indian paintings stand apart from the better known paintings of Ajaṇṭā, Sigiriya, Bāgh, Bādāmi and Ellora. Here the ground-plaster is of coarser lime mortar below with a finer coat of lime wash above it on which the pigments have been laid in accordance with the 'fresco' or 'lime medium' technique. The absence of any adhesive such as gum or glue in the laying of the pigments is the important feature that distinguishes them from the painting in the Deccan and Ceylon. There the ground-plaster is of mud—clay or alluvium, and sand mixed with very small quantities of lime and containing organic materials like cowdung and husks or vegetable fibres acting as 'binding medium'; this ground-plaster is covered by a lime wash of egg-shell thickness and painted with mineral colours mixed with glue or gum as a fixitive; the technique is 'tempera' and the painting could be done after the lime wash had dried up. In Śittannavāśal on the other hand, the ground-plaster is a mixture of lime and sand forming a coarse ground with a thin layer of fine lime-plaster over it; on this the paintings have been executed in mineral colours without the mixture of any adhesive. Since on examination the colours appear to have been mixed in lime water and applied on the dried up surface, this is strictly 'fresco-secco', different from true fresco work where the paint is applied on the still wet lime surface, a very good example of which we find in the Tanjore Cōḻa paintings. All the other paintings in the South Indian series found in the temples of Mamandur, Tirumayam, Kāñcī, Nārttāmalai, Somapālayam, Lepākṣi, Tirugōkarṇam, Malaiyaḍipaṭṭi, Travancore, Cochin and elsewhere belong to the class of fresco-secco.⁶¹

c35.s2.4

In Tanjore, as elsewhere, the coat of fine plaster seems to have been laid when the coarse plaster below was still wet. The painting (in true fresco method) was also applied when the surface was still wet. In this process the pigment penetrates the surface, and soon, the wet calcium hydroxide in the lime, as the water-content evaporates, reacts with the carbon dioxide of the air forming a thin protective and transparent film over the surface. The same thing happens even where the pigments are laid in lime medium on the dry surface; though there is no penetration of the pigment, the glassy protective film is still there. This is the merit of the fresco process. In Tanjore old and well-slaked lime burnt over a wood fire and derived perhaps from shell or lime-stone seems to have been used; no marble dust has been found in analysis. The pigments with the exception of black are found to be of mineral origin, the principal colours used being black, yellow, brown, red, blue, green, yellowish green and light blue; the materials employed are lime for white, wood charcoal or lamp black for black, ultra-marine for blue, toned down with lime or fine sand for light blue, ochres for yellow, brown and red, terreverte for green, ultramarine and yellow ochre for yellowish green, ultramarine and terreverte for bluish green. The technique employed has restricted the palette to material which will not react and change when mixed with lime—a limitation of the fresco process. Since again in the fresco all the pigments dry out lighter and that gradually in a few week's time, the artist must have a good idea of the relative warmth of the tones to be ultimately expected. The Tanjore artists have understood this so well as to achieve a fine harmony of colour. In some places the colour films do not adhere so firmly as in general they do, and here the pigments seem to have been applied in lime medium. Apparently the paintings were begun in fresco, and were finished or retouched in lime medium. True fresco work demands greater swiftness and precision of execution than the lime medium technique. This means the execution of the work panel by panel and the careful jointing of the panels to match in tone and colour. In the Cōḷa frescoes at Tanjore the joints are so well concealed that it is not quite possible to estimate the rate at which the painting was done and the area covered by an artist in a day. The plaster coats are so thin that they could not have retained their moisture for more than a very short time. The artists seem to have completed one wall per day or during such time as it remained wet, in which case no joint would be visible. This would not have been impossible, since the wall is conveniently divided into horizontal panels by patches of colours and each panel may have been worked by a single artist or a group. As a matter of fact the themes are made up of many small scenes on each panel; each scene may have been painted by one artist, a number of them working simultaneously. The area of the panels varies from 24 square feet to 60. It is not also impossible that the joints are concealed by the horizontal patches of colour. To put in so much detail of dress and ornament, and to achieve so much general decorative effect even in the disposition of the figures under such limitations of time as the fresco process imposed required great skill of the artists indeed.

c35.s2.5

The paintings in the Tanjore temple in the walls and ceilings of the *pradakṣiṇa* underneath the *vimāna* were originally lighted by three doorways in the middle of the exterior walls of the *garbhagṛha* on the south, west and north; these openings were closed by Vijayarāghava Nayak between 1653 and 1659 by rubble walls which have now been demolished by the archaeological department. The openings correspond in position to the *devagoṣṭhas* (niches) in the central bay on each side on the inner walls enshrining large sculptures appropriate to them. Vertical pilasters corresponding to the system of bays and recesses of the outer wall divide the *pradakṣiṇa* passage into fifteen chambers each separated from its neighbour by a constriction which is really an opening composed of joists, lintel and sill. There are two chambers on either side of the main entrance on the east, five each on the north and south, three behind on the west. The first three chambers, starting in *pradakṣiṇa* order from the south of the main entrance, are entirely covered, walls and ceilings, with the Nāyak series of paintings of the sixteenth and seventeenth centuries which do not seem to cover any Cōḷa paintings underneath. The fourth, eighth and twelfth chambers, those with the openings on the outer wall, are dominated by the large sculptures on the inner wall, and the last three chambers (13-15) have again only Nāyak paintings with no Cōḷa layer underneath. The remaining six chambers have excellent Cōḷa paintings, visible where the Nāyak layer overlaying them has pealed off, and the biggest problem facing the archaeologist is to find a successful way of stripping the Nāyak layer elsewhere without damaging the underlying Cōḷa paintings.⁶²

c35.s2.6

The theme of the Tanjore Cōḷa paintings is religious, drawn mostly from the hagiology that was later worked up into the *Periya-Purāṇam*. The episodes of the life of Sundaramūrti form the subject of some of the best panels. On the top of the panel on the west wall (chamber 7) is the Kailāsa scene—Śiva seated in *Yogāsana* on a tiger skin with the couchant Nandi (bull) in front and a group *ṛṣis* with a couple of *apsara* maidens dancing at the opposite end; Śiva is painted red and one of the *ṛṣis* blue. Just below is the picture of Sundaramūrti and his friend Cēramān Perumāḷ, another *nāyanār*. The scene depicts the journey of the two friends to Kailāsa at the invitation of Śiva; the youthful Sundara is seated on the fast moving elephant sent to fetch him, and Cēramān on his horse speeded by his whispering the *pañcākṣara* (Śaiva credo) into the ears of his horse which, as a result, soon mounted the sky and overtook the divine elephant ridden by Sundara! The white elephant of Sundara is the centre of the picture; and right in front is the prancing steed, its rider looking back and beckoning to Sundara. The king wears a long moustache, a close beard, and a large tuft of hair behind his head. He is bare-bodied except for the *dhoti* round his waist. Round the neck he has a tight necklet and a dangling cord with a *rudrākṣa* bead strung in it. The harness and trappings are well worked out on the body of the white horse. At the right and left top corners are a group of celestial beings half hidden by the clouds—the right group consisting of *apsara* maidens and *gandharvas* showering lotus petals, dancing, and playing on musical instruments. The left group seems to comprise *ṛṣis*. Right in front of Cēramān’s steed is the most elegant of the dancing *apsara* maidens, her body bent in rhythmic curves and her right arm stretched out in front and left bent in *abhaya* pose, her face turned back. Her diaphanous clothing and the elaborate coiffure, jingling anklets and wristlets add charm to her graceful form (frontispiece).

c35.s2.7

Below this panel is the scene of dispute between Sundara and Śiva on the eve of his marriage (*ante*, p. 474). On one side we see Śiva as a bearded old man, sitting in front of Sundara, clutching the palm-leaf document, and speaking to the *sabhā* with raised arm and uplifted face. Again on the other side is the old man producing the palm-leaf deed executed by Sundara’s grandfather, the bewildered Sundara standing before him in eager expectation of the verdict. On the faces of the members of the *sabhā* is depicted a varying conflict of emotions. The right side of the picture depicts the culmination of the episode—a temple entered by the whole lot of them. The lower panel showing some women engaged in cooking may be part of the portrayal of preparations for Sundara’s marriage.

c35.s2.8

Elsewhere on the west wall (chamber 9) we get a grand scale painting of Naṭarāja and his devotees, and the subject is apparent though much of the original painting is still overlaid with later work. There are many women of high rank included in the scene.

c35.s2.9

But the grandest composition in the whole series is the Tripurāntaka panel on the north wall (eleventh chamber). It is a battle scene. Śiva is standing on the deck of the chariot in *ālīḍha* pose, his left knee bent, and the whole weight of his body thrown on the right leg which is placed forward; his eight arms carry different weapons, one of them, holding a long bow in front. His vibrant frame and defiant expression suggest vigorous action. On the driver's seat is the four-headed Brahmā holding the reins and the whip. This is the centre of scene. In front are the horses of the *asuras* facing Śiva and his *gaṇas*, the whole foreground depicting fights with different weapons between the two groups. On the top is seen Durgā on her lion thrusting her spear into the body of an *asura* while her lion is holding another by the neck.

c35.s2.10

Another panel on the western wall of the shrine recently uncovered depicts Naṭarāja in Kanakasabhā worshipped by a royal devotee attended by his many queens and retinue. There is no difficulty in identifying the royal figure as Rājarāja I, the founder of the temple, who called himself *Śivapādaśēkhara* and named every unit of measurement Āḍavallān, indicating his intense devotion to the Lord of the Dance—Naṭarāja. His principal queens and consorts are portrayed on a large scale as standing behind him, while the other queens and also the retinue are smaller.

c35.s2.11

There is a second circumambulatory over the one in which the Cōḷa paintings have been discovered. The inner walls of this circumambulatory contain traces of old Cōḷa plaster and painting showing that it was also painted as the one below. Running round the inner wall at almost eye-level are bas-relief sculptures depicting the *tāṇḍava* poses of Śiva. Though there is provision for 108 sculptured panels, only 82 have been completed and the rest are blank. These rather crude sculptures were perhaps covered by painted stucco and illustrate the poses described by Bharata, The lines of the figures are drawn in light red or brown and deepened by forcible blacks and reddish browns. The other colours used for flesh and drapery are defined by delicate brush work in different colours. There is a conscious attempt at modelling here. Though there is not much diversity in the poses of figures, yet they are far from being stereotyped. The celestial beings, *apsaras* and *gandharvas* 'have a certain bend of the body as if they had floated into shapes on waves of an invisible sea.' The lines of the seated figures of women have more grace and charm than the standing ones. The dancing forms are full of action and expression. The faces are drawn in three quarters front and profile visible and square in outline with pronounced chin. The hair is done into elaborate coiffures of different patterns with small ringlets falling in front on the face, and decked with flowers, buds, and ornaments shaped like crescent and star. The eye brows are set low in human forms and high in celestial beings; the eyes themselves are linear and pisciform and the eyelids are not pronounced, but none the less expressive of emotion. The noses are long, straight, and sensitive, very rarely curved, the nostrils wide and mobile. The variety of ornaments worn by women is a study by itself. The drapery consists of a *sārī* of diaphanous muslin worn round the waist and covering down to the ankles and thrown into graceful folds decorated by floral patterns or horizontal lines, and held in position round the waist by sashes of different colours, their ends hanging in folds. The bust is generally bare except for a piece of cloth worn over the left shoulder and passing between the ample bosoms under the right arm. The men are of strong build, with beard, moustache and knots of hair on the head.

c35.s2.12

The panels in the Cōḷa layers exposed so far do not suggest any grouping except in the Tripurāntaka and Sundaramūrti panels, and till the entire Cōḷa layer is exposed, one cannot say definitely that the figures in the other groups are not inter-related.

c35.s2.13

The fragments of paintings in the Vijayālaya-Cōḷēśvaram are much faded. The two large ones are on the north wall of the *ardhamaṇḍapa*. The two large figures are those of Bhairava and Naṭarāja; the stiff pose of the figures indicates a late age when the mural art was decadent; but the fragments of plaster revealing a few pretty cherub-like faces on the south wall are strongly reminiscent of the Tanjore School, and these paintings may tentatively be assigned to the late Cōḷa period—late twelfth or early thirteenth centurȳ.

c35.s2.14

2. *Studies.*
3. *Archeologie du sud de l'Inde*, i. p. 116.
4. 167 of 1894; *JMU.* xiv p. 28.
5. 392-4 of 1924.
6. 36 of 1931; 91 and 92 of 1895.
7. See drawings reproduced from J. Dubreuil at the end of the chapter for an idea of the technical terms of South Indian architecture.
8. *Indian Art and Architecture (Buddhist and Hindu)*, p. 98.
9. Name in 282 of *Pudukkōṭṭai Inscriptions (PSI)* of Yr. 11 of Māṟavarman Sundara Pāṇḍya I (1227). Original construction by Śāttan pūdi alias Iḷaṅgoḍi-araiyar and restoration by Mallan Viḍuman alias Tennavan Tamiḻadi Araiyan after damage to temple (*kaṟṟaḷi*) in storm and rain—both mentioned in a recently found inscription on the basement of the central shrine underneath the *dvārapālaka* to the north of the entrance. *JOR.* viii pp. 208-9. See also *JISOA.* v (Coomaraswami vol.) p. 85.
10. For all the Pudukkottah temples discussed here see generally K. R. Venkataraman—*Manual of Pudukkōṭṭai*, the reports of the State Museum for faslis 1349-51, *JOR*, vii-xii and *JISOA* (Coomaraswami volume)—articles by Venkararanga Raju and S. R. Balasubrahmanyan.
11. 310-11; and 316 and 319-320 of 1903.
12. *JISOA.* vii pp. 113-5.
13. 586, 589 and 605 of 1904.
14. 148 and 155-8 of 1919.
15. 141 of 1904 (*SII.* vii 154) of Yr. 5 of Rājakesari (Sundara Cōḷa).
16. 104 of 1914, *EI.* xix p. 86.
17. 558 of 1921.
18. 199 of 1907; *'SII.* iii. No. 124.
19. 364 and 378-80 of 1924.
20. 335 of 1902, *EI.* vii p. 133.
21. 348, 359 of 1903.
22. *JISOA*, vii, pp. 113-5.
23. 192 of 1925.
24. *PSI.* 24; trn. pp. 24 and 30 ff.
25. *PSI.* 14.
26. Such instances, among others, are, the Uttarakailāsa in the Pañcanadeśvara temple of Tiruvadi (Tanjore) built by Rājarāja's queen Dantiśakti (219 of 1894); the Vaidyanātha of Tirumalavāḍi (Trichinopoly) rebuilt in the last years of Rājarāja I and completed under his son Rājendra I (91 and 92 of 1805); the twin shrines of Śiva and Viṣṇu at Dadapuram (S. Arcot) built by Rājarāja's sister Kundavai about 1016 (8 of 1899); the Ariñjigai-Iśvara, now called Cōḷēśvara at Mēlpāḍi (Chittoor) built by Rājarāja I as paḷḷipaḍai in memory of Ariñjaya who fell at Āṟṟūr (ante p. 187); the Śiva Devāle No. 2 at Polonnaruva in Ceylon (ASC. Report 1906 pp. 17-22); the Irungōḷeśvara, now Nīlakaṇṭheśvara of Laddigam (N. Arcot) which bears an inscription of the ninth year of Rājarāja (551 of 1906); the Haratīrtheśvara at Tiruvaranguḷam (Pudukkottah state); the Gangaikoṇḍa-cōḷēśvara of Kūḷambandal (S. Arcot) built before 1034 (414 of 1902); and Tripurāntakēśvara of Kūvam (Chingleput district) built about 1050 (328 of 1909).

c35.s2.15

27. *JISOA*, ii, p. 4.
28. Ante, p. 184.
29. *JISOA*, ii, p. 4.
30. Ante, p. 235.
31. This is the reason why the ground plan of this temple given by Percy Brown can no longer be held to be accurate or complete; there is no ground plan or section in existence for Tanjore—a state of affairs that calls for the early and serious attention of the Archaeological Department.
32. *JISOA*, ii, p. 5.
33. *SII*. ii, No. 61; Intr. p. (13).
34. 22 of 1895 (*SII*. v. No. 578).
35. 335 of 1917.
36. 65 of 1890 (*SII*. iii, No. 22) from Karūr (Rājēndra II); 307-8 of 1901, from Dharmapuri (Kulōttunga I); 70 of 1911 from Śrīvāñjiyam (Rājarāja II) and so on.
37. See 429 of 1912 (Vallūru, Tanjore), 577 of 1905 (Vijayaman-galam Coi.), 504 of 1912 (Ūṭṭattūr, Trichi), etc.
38. *Ācāryapuṣpāñjali*, p. 6.
39. 47 of 1918.
40. 227 of 1921.
41. The account of these temples that follows is based on K. R. Srinivasan's paper 'The last of the great Cōḷa temples', *JISOA*, xvi. (1948), pp. 11-33.
42. *ARE*. 1920, pp. 102-7 and pls. I-VI; also 1908, II, 66-7.
43. *Hindu Iconography*, T. A. G. Rao, ii, pp. 171-4.
44. *ARE*. 1908, p. 81, para 68.
45. *OZ*. (N. F. 29), 1933, p. 5.
46. H. K. Sastri in his *South Indian Images of Gods and Goddesses* draws attention to several instances of iconographic variations in images of the same deity.
47. IV. ll. 157-8.
48. *OZ*., 1933, p. 165.
49. 132 and 136 of 1925; *ARE*. II, 10.

c35.s2.16

51. Tirukkarrali-piccan (the mad man of the stone temple) is a title rather than a name, and was borne later by Parāntakan Śiriya-vēḷār, the general of Parāntaka II (291 of 1908), and Ārūran Kamban of Tirumaṇañjēri (A.D. 991, No. 9 of 1914). Other portraits in Tiruvā-ḍutuṟai, similar to that of Piccan, are that of Ambalavan Tiruviśalūran Tirunāvukkaraiyan (133 of 1925) and some others of less certain date (106 and 141 of 1925). See *SII.* iii, plate xi for images (artistically not important now) of Gaṇḍarāditya worshipping the deity of Tirunallam-uḍaiyār at Kōnērirājapuram (Tanjore district), built by his queen Śem-biyan-mahādēvi (*SII.* iii, Nos. 146 and 147 = 450 of 1908 and 626 of 1909); the king's head-dress, simple and elegant, with strings of pearls, and his ornaments on the arms, neck, and chest, in a way recall the similar decoration of the so-called *Cōḷa-pratima* of the Bhōganandīśvara temple in Nandi (*MAR.* 1914 and 1915), which is a good example of the type of idealised portrait mentioned above in the text.

c35.s2.17

Epigraphical references are found in inscriptions to portrait bronzes of (1) Śembiyan Mahādēvi (A.D. 1020) in a village of the same name—*ARE.* 1926, II, 24, (2) Parāntaka II and his queen Vānavan Mahādēvi, mother of Rājarāja I, and of Rājarāja I himself and his queen Lōka-mahādēvi, all solidly made, in the Tanjore temple—the inscribed image of Rājarāja I now in use there is obviously a much later substitute—*ARE.* 1925, II, 12, *ante.* pp. 168, 189 n. 3.

c35.s2.18

Some stone portraits, not well preserved, are: (1) a king and queen worshipping a *liṅga* in the Śiva temple of Tiruviśalūr (Tanjore district) below which is a record of the *tulābhāra* and *hiraṇyagarbha* performed there respectively by Rājarāja and his queen Lōkamahādēvi (42 of 1907, *EI.* xii, p. 121, n. 2); (2) a king or noble seated cross-legged on the floor and worshipping a *liṅga* in a ruined Śiva temple at Olagapuram (S. Arcot)—129 of 1919; (3) Anantaśiva worshipping a *liṅga* in a *maṇḍapa* built by him in the temple of Tiruviśalūr (T. G. Ara-vamudan, Portrait Sculpture, fig. 10); and (4) Kundavai dancing before Śiva in the temple built by her at Dādāpuram (S. Arcot)—17 of 1919.

c35.s2.19

T. G. Aravamudan (*op. cit.*, pp. 38-9 and fig. 13) identifies a Cōḷa general Kēttan Ādittan and his sister Kaliyavvai—of 168 (a) of 1922—in a couple of bronze statuettes from the Kālahasti temple. In the Śiva temple of Śrīmuṣṇam (S. Arcot) is a portrait of Tambirān-tōḻan Mānakkañjāṟan who used to recite Tiruppadiyam in the temple (255 of 1916). At Anbil the people set up (c. 1250) an image in the Prēma-purīśvara temple of a certain Paḷudaiyāṇḍān who, at the cost of his life, protested against unjust revenue demands on the villagers (596 of 1902).

c35.s2.20

Popular local tradition identifies a large stone figure in the Śrī-raṅgam temple with poet Kamban; two damaged stone figures at Tēra-ḻundūr with Kamban and his wife; and a large well preserved loose stone image in Ēkāmranātha temple at Kāñcīpuram with Karikāla Cōḷa.

c35.s2.21

⁵² *Rūpam*, 1930, No. 40, p. 1.

c35.s2.22

⁵³ 168 (b) of 1922.

c35.s2.23

⁵⁴ 168 (a) of 1922.

c35.s2.24

*55. H. K. Sastri, *South Indian Images*, p. 162.
56. *Ibid.*, p. 125 and fig. 80.
57. *Annamalai University Journal*, vol. iii, p. 43.
58. *OZ.* (N. F. 10) 1934, pp. 176-86.

c35.s2.25

59. The plates are still unpublished as already noted (p. 15, n. 1 *ante*), and the seal is reproduced by the kind permission of the Government Epigraphist, Mr. N. Lakshminarayana Rao. The details of the seal are described elsewhere.

c35.s2.26

60. *Proceedings of the Indian History Congress*, vii, (1944), pp. 168-76.

c35.s2.27

61. The technique of Indian paintings has been studied by Dr. S. Paramasivan, Archaeological Chemist, and the details are to be gathered from his valuable papers of which the following may be mentioned in particular:
(1) Indian Wall Paintings, *JMU.* xii (1940), pp. 96-128, and xiii, (1941), pp. 1-15;
(2) Annual Reports of Hyderabad Archaeology Department, 1936-7, pp. 25-38 on Ajaṇṭā and Ellorā;
(3) Technical Studies, Harvard; v. 4, (1937) pp. 222-39 on Tanjore, viii, 2, (1939), pp. 83-9 on Śittannavāśal;
(4) Proceedings of the Indian Academy of Sciences: vii, 4, (1938), pp. 282-90 on Vijayālaya-Cōḷēśvaram; x, 2, (1939), pp. 77-84 on Kāñcīpuram; x, 9, (1939), pp. 85-95 on Bāgh.

c35.s2.28

⁶² S. K. Govindaswami described the subject-matter of these paintings first in the *Annamalai University Journal*, ii, (1933), and in *JISOA.* i, (1933), pp. 73-80, but he went wrong over the technique of the paintings. See also C. Sivaramamurti in *Trivēṇi*, vi, (1933), pp. 227-34; O. C. Ganguly in *IAL.* (NS. ix), 1935, p. 86, and Dr. Paramasivan in *JOR.* ix, (1935), p. 363.