279

Chapter Five

The techniques, plasters and pigments ofthe murals and the instruments of the

painters

The canons of painting:

To be successful, a mural painter has to be a master artist, as well as a master

technician jointly in the field. It is not enough that his work only fulfil and satisfy a certain

standard of aesthetic judgement, it is necessary further that the work itself should be long

lasting and durable. This clearly infers that both excellence and durability are the two

fundamental qualities concomitant to each other in a work of mural painting of any period

or tradition. Of these, the former is observable from their superficial appearances

themselves and the latter quality is essentially dependent on effective technique.

According to the description given in the previous chapters, it is evident that both these

qualities have been well represented by the ancient Buddhist mural painters of India and

Sri Lanka though natural elements and human vandalism have damaged their c.reations in

the subsequent periods to a great extent Of these, in fact a perfect technique is, no doubt,

the result of a long period of experimentation in various methods and means, particularly

in the ancient days. 1 Hence this aspect has to be examined very carefully and

systematically though the relevant data is rare and often contradictory.

Thus, when considering the technical aspect of the early Buddhist mural paintings

of India and Sri Lanka, it is evident that it has been studied from two perspectives: first, on
 280

the basis of literary evidences available in the silpa texts; 2and second, by a scientific

examination of specimens of wall paintings. 3 According to both these accounts, it is

obvious that different, incompatible and sometimes even unacceptable opinions have

been expressed on the techniques of paintings of both countries. For instance, as already

mentioned in the previous chapter though there are no extant texts to provide information

on Buddhist canons of paintings during the ancient period, some critics have applied the

details given in the later Hindu canon to the Buddhist paintings of Ajanta too. 4Although it

is difficult to ascertain the religious affiliation of these painters, it is certainly appropriate

to believe that these painters were Buddhist upasakas5 since they were deeply rooted in

1
Asok K Bhattacluuya, Technique of Indian painting: A study chiefly made on the basis of sijpa texts,
Saraswat Library, Calcutta, 1976, Introduction, p.l.
2
Ananda Coomaraswamy, "The technique and theoxy oflndian painting," Technical studies in the field of
fine arts, Fogg Art Museum, Hmvard University, Vol Ill, No. 2, 1934; Siri Gunasinghe, La technique de Ia
peinture Indienne d' Apres les textes du siJpa, Presses Universitaires de France, Paris, 1957; Abanindranath
Tagore, Some notes on Indian aesthetic anatomy and Sadanga or the six limbs of painting; Indian Society of
Oriental Art, Calcutta, (reprinted) 1968; Asok k Bhattachaxya, Technique of Indian painting: A study
chiefly made on the basis of the silpa texts, Saraswat Library, Calcutta, 1976; Jayanta Chakrabarti,
Techniques in Indian mural painting, K.P. Bagchi & co, Calcutta, 1980; Stella Kramrisch, "Introduction to
the Visnudharmottara," Exploring India's sacred art: Selected writings of Stella Kramrisch, ed. Barbara
Stoler Miller, Indira Gandhi National Centre for the Arts, New Delhi, Motilal Banarsidass Publishers, New
Delhi, 1983; The Vishnudharmottara (part ITD: A treatise on Indian painting and image making, Calcutta
University Press, 1928.
3
S. Paramasivan, Annual Reoort of the Archaeological Dtmartment of His Exalted Highness the Nizam's
Dominions, 1936-37, 1939, p.25; Journal of Indian Society of Oriental Art, Vol.XVIll; RB Khan, "Fresco
paintings of Ajanta," Journal of Oil and Colour Chemist Assn, Januaxy 1949, p.24; R. H. de Silva, "The
technique of ancient Sinhalese wall painting- Sigiri," Paranavitana felicitation volume, ed. N.A.
Jayawickrama, Gunasena & co, Colombo, 1965; "Rock painting in Sri Lanka, " Conservation in
archaeology and the I!PPlied arts, (Reprinted of the con1n'bution of Stockholm Congress), June 1975; Nimal
de Silva, "Studies in the techniques of ancient paintings in Sri Lanka," Persj>ectives in archaeology:
Leelananda Prematillaka festschrift, ed. S Senaviratna and others, Department of Archaeology, University
ofPeradeniya, 1990.
4
OM Datta Upadhyaya, "The art of Ajanta: Inner light expressed through ksayavrddhi principle in spatial
dimensions," The art of Ajanta: New per8J>ectives, ed. Ratan Parimoo and others, Books and Books, New
Delhi, Vol. II, 1991, pp.299-311; Pravina Mehta, "Architecture in Ajanta painting," The art of Ajanta: New
perspectives, ed. Ratan Parimoo and others, Books and Books, New Delhi, Vol.II, 1991, pp.312-314; Rai
Anand Krishna, "Certain ateliers, sub-schools and allied schools of the Vakataka period Ajanta paintings,"
The art of Ajanta: New pet'Sj>ectives, ed. Ratan Parimoo and others, Books and Books, New Delhi, Vol.II,
199l,p.258.
5
The term 'upasaka' is generally applied to people who spend considerable time and energy on Buddhist
activities, whether in private rituals or organising support to the sangha RF Gombrich, Teravada Buddhism.
Routledge and Kegan PauL London, 1988, p.75; HP Ray, "Buddhism and trade in early historical India,"
 281

Buddhist legends and teachings as discussed.in the next two chapters in detail. But it is

noteworthy that some critics have gone to the extent to state that the famous six limbs or

Sadanga rules given in the (13th century AD) Commentary ofYasodhara on Kamasiitra, 6

the 'eight limbs' or Astangani presumed in the (11th century AD) Samarangana
7
Siitradhara and the other 'eight limbs' or Guniistaka given in the Citrasutra section of the

prominent Sanskrit text (of the 7th century AD) Visnudharmottarapurana8all ·of which

were non-Bud~ist canons as their names themselves suggest, have been followed not

only by the painters of Ajanta but by the artists ofBagh caves too.~ence, at this point, it

Explorations in art and archaeology of South Asia: Essays dedicated toNG Majumdar, ed. Debala Mitra,
. Directorate of archaeology and museums, Government ofWest Bengal, Calcutta, 1996, p .5 51.
6
Ananda Coomaraswamy, Histozy of Indian and Indonesian art, Munshiram Manoharlal, New Delhi,
(reprinted) 1972, p.88; Jyotsna Saxena, Early Indian paintings in Sanskrit literature, Bharatiya Kala
Pra.kashan, Delhi, 1998, pp. 24 & 70; Asok K. Bhattachacya, Technique of Indian painting: A study chiefly
made on the basis of the silpa texts, Saraswat Library, Calcutta, 1976, p.14; Percy Brown, Indian painting,
YMCA Publishing House, London, 1927, pp.21-22; C. Sivaramamurti, Indian painting, National Book
Trust, New Delhi, 1970, p.37; Sources of history illuminated by literature, Kanak Publications, New Delhi,
1979, p.l38; Karl Khandalavala, The development of Indian paintings, pp.l 0-11; EB Havell, A handbook of
Indian art, ludic Academy, Varanasi, 1972, pp.195, 199; K Bharatha Iyer, Indian art: A short introductio!!,
Asia Publication House, Bombay, 1958, p.66.
7
Asok K Bhattachacya, Technique of Indian painting: A study chiefly made based on silpa texts, Saraswat
Libracy, Calcutta, 1976, Introduction, pp. 96-97; Madanjeeth Singh, The cave paintings of Aj anta, Thames
and Hudson, London, 1965, p.62.
8
See DN Shukla, Silpa sast:ra: Hindu achievements in aeronautics and fine arts, Shri Darshan Lal
Vastuvanmaya Prakasha-sala, Lucknow, 1967, pp.56-60.
9
Percy Brown, Indian painting, YMCA Publishing House, London, 1927, pp.21-22; C. Sivaramamurti,
Indian painting, National Book Trust, New Delhi, 1970, p.37; The art of India, Hany N Abrams PublisherS,
New York, 1977, p.176; Asok K Bhattachacya, Technique of Indian painting: A study chiefly made based
on silpa texts, Saraswat Library, Calcutta, 1976, Introduction, pp.92-97 and 136; Moti Chandra, Studies in
early Indian painting, Asia Publishing House, London, 1970, pp.22-23; Siri Gunasinghe, "Ajanta's shadow
on Sigiriya," The art of Ajanta: New pers,pectives, ed. Ratan Parimoo and others, Books and Books, New
Delhi, Volll, 1991, p.503; DN Shukla, Silpa sastra: Hindu achievements in aeronautics and fine arts, Shri
Darshan Lal Vastuvanmaya Prakasha-sala, Lucknow, 1967, p.I77; Madanjeeth Singh, The cave paintings
of Ai anta. Thames and Hudson, London, 1965, pp.61-62; Jyotsna Saxena, Early Indian paintings in Sanskrit
literature, Bharatiya Kala Pra.kashan, Delhi, 1998, p.70; See also MN Deshpande, "Ajanta: The
Kaleidoscopic enigma of art historians," The art of Ajanta: New Peoo>ectives, ed. Ratan Parimoo and
others, Books and Books, New Delhi, Voll, 1991, pp.22-23; "Cave paintings in India," Wall paintings of
India: A historical persj>ective, ed. OP Agrawal, INTACH Conservation Centre, Lucknow, 1989, p.20; G
Yazdani, Ajanta: Monochrome re;productions of the Ajanta frescoes based on photography, Swati
Publications, Delhi, Vol.III, 1942, p.S; Benoy K Beh~ The Ajanta caves: Ancient paintings of Buddhist
In~ Themes and Hudson, 1998, pp.34-35; KK Chakravarti, "A histoty of wall painting in Madya
Pradesh," Wall paintings of India: A historical persj>ective, ed. OP Agrawala, INT ACH Conservation
Centre, Lucknow, 1989, pp.52 and 63.
 282

is necessary to focus attention on this matter and on the general contents of these

non-Buddhist canonical texts.

As already stated, of these texts, the Commentary ofYasodhara on Kamasutra 10is

undoubtedly not a Buddhist text in any sense. It contains information of erotic nature,

which is incompatible with the ideals of Buddhist teachings. In addition, it is obvious that

the dates of all the Buddhist mural paintings of both India and Sri Lanka of the ancient

period are older than Yasodhara' s commentary. However, it is to be noted iliat it has been

generally accepted that the Sadangas or the six limbs11 ofthe paintings thus presented in

the commentary are; riipabheda (variety of form or the knowledge of appearances);

pramiinam (appropriate proportion or correct perception); bhiiva (description of emotion

or action of feelings . on forms); liivanya yojanam (infusion of grace or artistic

10
According to some scholars, Vatsyayana and Canakya happen to be one and the same person. Therefore
Vatsyayana, the author ofKamasutra should be a person of the fourth century BC. Nevertheless, according
to other scholars the Kamasutra was written in the third century AD. Certain Indian scholars have however
tried to establish that the Kamasutra was written probably in the frrst century AD .. See Anil Baran Ganguly,
Fine arts in ancient India, Abhinav Publications, New Delhi, 1979,p.20. Thus it is evident that Vatsyayana's
date is uncertain though he appears to have been earlier than Kalidasa Hence he is generally assigned to
different periods from 3m century AD to the sixth. See Sures Chandra Banerji, A companion to Sanskrit
literature, Motilal Banarsidass, Delhi, Second edition, 1989, p.l 04. However, it is certain that Y asodhara
wrote the Jayamangala commentmy on the Kamasutra ofVatsyayana under the Chiihamiina king
Visaladeva, middle of the 13th century AD. Ibid, p.lll.
11
The verse reads: Rupabhediih pramanani bhiivalavanyayojanam/
SadrllSyam varnikabbangam iti citram sadangakam II Yasodhara' s commentmy on the Kiimasfitra, part I,
chapter 3: as quoted by Asok K Bhattacharya, Technique of Indian painting: A study chiefly made based Oil
silpa te~ Saraswat Library, Calcutta, 1976, Introduction, pp.14 and 29. Although Tagore bas presented a
detailed account of the six limbs given in the Kamasutra (Abanindranath Tagore, "Sadanga or the six limb~
of painting," Journal of the Indian Society of Oriental Art, Calcutta, 1961; Some notes on Indian aesthetic
anatomy and Sadanga or the six limbs of painting, Indian Society of Oriental Art, Calcutta, (reprinted) 1968,
pp.l-18), Coomaraswamy pointed out that it is impossible to accept his subjective interpretation of these
terms. According to him, these can be far better understood in a purely practical sense as distinction of
types; ideal proportion, expression of mode (with reference to rasa) and preparation of colours (grinding and
levitation etc). See AK Coomaraswamy, History of Indian and Indonesian art, MunshiramManoharlal, N<m
Delhi, (reprinted) 1972, p.88.
 283

representation); siidriSyam (likeness or similitude); and vamikiibhanga (artistic manner of

12
using the brush and colours or mixing of colours to produce an effect ofmodelling).

Accordingly, it is obvious that these six limbs indicate an insight into what should

be the qualities of a good painting and influenced the formation of style. The description

is, however, too general and therefore it can be applicable to any tradition of painting

including the Buddhist tradition, except perhaps the monochrome and modem idealistic

painting traditions. This is further substantiated by the fact that the ancient Chinese artists
13
have also used a similar Sadanga cannon. As some scholars have pointed out, these six

canons of Chinese painting ofHsieh-Ho (479-501) closely follow the six limbs oflndian
14
painting, though the order of classification is somewhat different These are: (1) mental

revolution gives birth to the life rhythm; (2) to bring out the anatomical structure with the

help of the brush; (3) to draw forms in conformity with nature; (4) to make the colours

correspond to the nature of the object; (5) to distribute the lines in their proper places; (6)
15
to propagate the forms by passing them on into the pictures. As also stated by

Coomaraswamy, any parallel to the Chinese six canons of Hsieh Ho cannot be traced

though a likeness to the Chinese idea, but can probably be recognised in connection with

12
Jyotsna Saxena, Early Indian paintings in Sanskrit literature, Bharatiya Kala Prakashan, Delhi, 1998, pp.
24 & 70; Radhakamal Mukerjee, The culture and art of India, George Allen and Unwin, London, 1959,
p.242; Percy Brown, Indian painting, YMCA Publishing House, London, 1927, pp.21-22; C.
Sivaramamurti, Indian painting, National Book Trust, New Delhi, 1970, p.37; Karl Kbandalavala, The
development of Indian paintings, pp.1 0-11; K Krishnamurti, "Tiaditional Indian aes1hetics in theory and
practice: A restatement with special reference to Natya sastra and the graphic art," Indian aesthetics and art
activity, Proceedings of seminar, Indian Institute of Advanced Study, Simla, 1968, pp.48-49; Prithvi Kumar
Agrawal a, On the sadanga canons of painting, Indian Civilisation Series, Prithvi Prakash an, V aranasi, 1981,
introduction, pp.l-7; Anil Baran Ganguly, Fine arts in ancient India, Abhinav Publications, New Delhi,
1979,p.96.
13
Percy Brown, Indian painting, YMCA Publishing House, London, 1927, p.23. For the detailed account of
Chinese six limbs see: Balram Srivatava, Notes of Indian aesthetics (with s.pecial reference to silpa),
Chaukhambha Oriental Research Studies, No.33, Chaukhambha Orientalia, Varanasi, 1985, pp.128-129; Sri
Aurobindo, The significance of Indian art, Sri Aurobindo Circle, Bombay, 1947, p. 79; Mulk Raj Anand,
Album of Indian paintings, National Book Trust, New Delhi, 1973, pp.11-12.
 284

what is said about cetana, the movement of life, in the Visnudharmottarnpurana. 16

Accordingly, it is evident that these six limbs of painting in Chinese seem to have

similarities with the Brnhmanical tradition but have been reinterpreted to suit Chinese
17
exigencies. Hence it is obvious that an overall comparison of the Jndian concept of

'Sadanga' with the six limbs theory of the Chinese pictorial art is not possible.

Of the Indian silpa texts, the most famous one Qn_paintit!g is tl1e Visnudharmottara
' I

of the seventh century AD, which was a supplement or appendix to the

V isnupurana. 18 This manuscript gives the.fullest_acconntJrnown. of the various branches,

methods and ideals of Indian painting. It deals not only with its religious aspects but also

and largely with its secular employment.1:lt.proclaims_the Joy_that colours, forms and the

representation of thmgs seen and imagined .produce. 20 It is noteworthy that though

Madhusudan and Madhava Prasad Sharma edited this Visnudharmottara Mahapurana in

21
1912, discussions on painting proper, with.reference io_Indian silpa .texts was, for all

practical purposes, started by Gopinath Rao. 22 This was followed by the English

translation of the Citrasutra and Pratima/aksana sections of the Visnudharmottara by

14
Radhakamal MukeJjee, The culture and art of India, George Allen and Unwin, London, 1959, p.242.
IS Ibid, p.242. _
16 AK Coomaraswamy, Hist01y of Indian and Indonesian art, Munshiram Manoharlal, New Delhi,
(reprinted) 1972, p.88.
17
Abanindranath Tagore, Some notes on Indian aesthetic anatomy and Sadanga or the six limbs ofpainting.
Indian Society of Oriental Art, Calcutta, (reprinted) 1968, pp.1-18; MuJk Raj Anand, The Hindu views of
.m, Amold-Heinemann, New Delhi, 1957, p.1 08; Phanindra Nath Bose, Principles of Indian silpa sastra
with the text ofMayasastra, The Panjab Sanskrit Book Department, Lahore, 1926, pp.84-85.
18
Stella Kram.risch, Exploring India's sacred art: Selected writings of Stella Kram.risch, ed. Barbara Stoler
Miller, Indira Gandhi National Centre for the Arts, New Delhi, 1983, p.263; C Sivaramamurti, Chitrasutra
of the Vishnudharmottara, Kanak Publications, New Delhi, chap .I, pp.l-16.
19
See Kusum Kumari Gupta, A socio-religious study ofVisnudharmottarapuran~ Harman Publishing
House, New Delhi, 1994.
20
Stella Kramrisch, Exploring India's sacred art: Selected writings of Stella Kram.risch, ed. Barbara Stoler
Miller, Indira Gandhi National Centre for the Arts, New Delhi, 1983, p.263.
21
Madhusudan and Madhava Prasad Sharma, Visnudharmottara Mahapuran~ Venkateshwara Press,
Bombay, 1912.
22
Gopinath Rao, "Painting in ancient India," Modern Review, Vol.XXV, December 1918, p.557.
 285

23
Stella Kramrisch. This translation along with an illuminating introductory note attracted

the attention of many scholars working on Indian art. Consequently, Coomaraswamy also

published an annotated version of only chapter 41 of the Visnudharmottar!!, 24in order to

improve Kramrisch' s translation and to provide an overall understanding of the treatise

with the help of other references and documents. Nevertheless, he experienced difficulties

in the study of the text and he frankly admitted that in spite of his best efforts he could not

solve all the problems related to the technical terms even with the help of the other

experts.

Both these scholars, Kramrisch and Coomaraswamy worked mainly on the early

edition of the text done by Madhusudan and Madhava Prasad Sharma. A scrutiny of this

printed work apparently reveals frequent corrupt readings. Certainly, for this corrupt

condition of the text and for a general difficulty in understanding the technical terms and

the conceptions of the ancient Indian painters, many problems relating to the proper

appreciation of the subject dealt with in the text had to be left unsolved by them?~ence,

the importance of the subsequent Priyabala Shah edition lies in the fact that the author

prepared the text with the help of a collated study of at least seven manuscripts, in addition
2
to the above-mentioned earlier text published by the Venkateswara Press. ~evertheless,

23
Stella Kramrisch, Introduction to the Visnudharmottara (pt. llD: A.treatise on Indian painting and image
making, Calcutta University Press, 1924: This has been reprinted as Exploring India's sacred art: Selected
writings of Stella Kramrisch, ed. Barbara Stoler Miller, Indira Gandhi National Centre for the Arts, New
Delhi, Motilal Banarsidass Publishers, Delhi, 1983, pp.263-307.
24
AK Coomaraswamy, "Visnudharmottara (Ill) chapter 41," Journal of the American Oriental Society.
VolLTI, 1932,pp.l-13.
25
Asok K Bhattachaxya, Technique of Indian painting: A study chiefly made based on silpa texts, Saraswat
Library, Calcutta, 1976, Introduction, p.8.
26
Priyabala Shah, Visnudharmottara-purana (Khanda lli1 Gaekwad's Oriental Series, No.l30, Baroda, 2
vols, (VolJ text) 1958 and (Vol.TI trans) 1961.
 286

by the very nature of the study this editor too frankly admits that in spite of the efforts

made by her to present a correct text, it cannot be said that it gives complete satisfaction?7

The Visnudharmottara is generally divided into three major parts. Of these, the

first part discusses the creation of the world, geography, astronomy etc. The second part

deals with Dharma, Rajaniti and polity. The third part or the Citrasutra contains short

treatises on Sanskrit and Prakrit grammar, lexicology, metrics, poetics, dancing, singing,

instrumental music and long treatises on painting, icono-plastic art and architecture. This

Citrasutra has been discussed in nine chapters from 35 to 43 of the Visnudharmottara. Of

these, chapter 40, named as 'Rangavyatikiira,' specially deals with the techniques of

preparing the ground for the wall painting, the first sketching and the preparation of

various shades of colours etc. There is ample evidence to show that the author of the text

had an exceptional insight into the intricate problems involved in the making of a painting

of high order. This is evident from the discussion of the Gunastakas. The

Visnudharmottarapurana says that proper position, proportion and spacing, gracefulness

and articulation, resemblance, decrease and increase are the eight good qualities of a

painting.28It not only prescribes these laws regarding the human form of various types and

categories and their proportions, but also deals with the problem of foreshortening or

ksayavruddhi. Certainly, this discussion on the law of foreshortening is a unique

contribution of the text, for this topic has not been specifically treated in any other Silpa

27
1bid, p.xviii.
28
Phanindra Nath Bose, Principles of Indian silpa sastra with the text ofMayasastra, The Panjab Sanskrit
Book Department, Lahore, 1926, p.83.
29
Asok K Bhattacharya, Technique of Indian painting: A study chiefly made. based on silpa texts, Saraswat
Library, Calcutta, 1976, Introduction, p.ll.
 287

It is however difficult to say how far the theory of painting preserved in the

Citrasutra ofVisnudharmottara served as a practical guide to even Hindu painters. It is

also not easy to correlate the theory with practice as the surviving examples of paintings

except in some cases hardly follow the rules laid down by the author of the
3
Visnudharmottara. <i'or instance, the difference between theory and practice is apparent

in the elaborate formula for preparing the wall surface for painting in the

Visnudharmottara31 as opposed to the simple formula of earth mixed with eow dung and

chaff pressed into the surface at Ajanta as discussed below. It is also evident that the

Visnudharmottara formula was meant for the smooth surface of a structural temple and

not for the rough surface of a cave like Ajanta or Bagh. 32In addition, in the colour shades

too the theory is much more advanced than the practice. For instance, the use of gold,

silver, copper, mica, ultramarine, orpiment, minium, cinnabar etc33 shows an advanced

knowledge of additional colours that contrasts with the limited palette of Ajanta and Bagh

paintings.

Apart from these evidences, it is generally accepted that the chapters of the

Visnudharmottara dealing with painting must have been compiled in the seventh century

AD, most probably later than the latest paintings of Ajanta; and so we are acquainted with
34
the theories prevalent at the time of the full maturity of their practice. Similarly, it is

30
.Moti Chandra, Studies in early Indian painting, Asia Publishing House, London, 1970, p.20.
31
Priyabala Shah, Visnudhannottara-purana CKhanda IID. Gaekwad's Oriental Series, 1958, chap 40,
25-29vv.
32
Moti Chandra, Studies in early Indian painting, Asia Publishing House, London, 1970, p.21.
33
Priyabala Shah, Visnudhannottara-purana CKhanda lll), Gaekwad's Oriental Series, 1958, chap 40,
25-29vv; See also Moti Chandra, Studies in early Indian painting, Asia Publishing House, London, 1970,
r,.21.
4
Stella Kramrisch, Exploring India's sacred art: Selected writings of Stella Kramrisch, ed. Barbara Stoler
Miller, Indira Gandhi National Centre for the Arts, New Delhi, 1983, p.264. Coomaraswamy and Mulk Raj
also believe that the manuscript ofVisnudhannottara belong to the seventh centruy AD. See Ananda K
Coomaraswamy, "The technique and theory of Indian painting," Technical studies in the field of fine arts,
 288

obvious that while enumeratin_g the defects and virtues of _painting the author of the

Visnudharmottara acted as a critic of the contemporary style of painters. In this process,

- '

apparently, the author shows his_partiality for the older tradition survivil}.g in th~_paiQtin_gs

of Ajanta and Bagh in which right stance, right proportion, vertical composition,

modelling, and foreshortening are pronounced features. 35Accor.ding to all these facts, it is

very doubtful whether the Buddhist painters of Ajanta and Bagh caves have used this text

as their guide or handbook as believed ~ some scholars.

Together with the Visnudharmottarapurana, mentioned above, some scholars

suggest that the painters of ~anta andB'!,gh caves used some of the rules of_paintmgs

given in the Samarangana Siitradhara also. In any case, first, it has to be admitted that

among the other canons besides th~ Visnudharmottarapuranl:!-, th~ Samarangana

Siitradhara is acclaimed as one ofthemost valuedsi/pa texts. 36This canonis.primarily a

work on architecture and painting and icono-plastic art are dealt with in chapters 71 to 82.

Ganapati Sastri edited the text on the basis of three available manuscripts in the

1920s.37But only one of them, which belongs to the Baroda central library, contains a

discussion on painting. Apparently, the section on paintings, which is also mutilated and

corrupt in its present state, has thus been edited on the basis of a single manuscript,
'
leaving no scope for improvement by a comparison ofvarious readings as in other cases.

Fogg Art Museum, Hmvard University, Vol.III, No.2, October 1934; Mulk Raj Anand, The Hindu views of
!!rt, Amold-Heinemann, New Delhi, 1957, p.l07; Priyabala shah also believes that the text should be placed
between 450-650 AD. See Priyabala Shah, Visnudhannottara-purana (Khanda III), Vol.I, Gaekwad's
Oriental Series, No.130, Baroda, 1958, p.xxvi; Nevertheless Nagpall believes that this is belong to the sixth
centwy AD. JC Nagpall, Mural paintings in India, Gain Publishing House, Delhi, 1988, p.3.
35
Moti Chandra, Studies in early Indian painting, Asia Publishing House, London, 1970, p.22.
36
T Ganapati Sastri, Samarangana Sutradhara, Geakwad's Oriental Series, Nos.XXX and XXXII, 2 vols,
1924-25; VS Agrawala, Sam.arangana Sutradhara, Geakwad's Oriental Series, 1966.
37
Samarangana Sutradh~ ed. T Ganapati Sastri, Gaekwade's Oriental Series Nos. XXX and XXXII,
Baroda, Vol.I 1924 and Vol.II, 1925.
 289

However, it is said to have been written Q)' king BhQja of Dhara (c. 1000-1055

AD) of the Paramara dynasty. 38Therefore, .it is _clear that the .descriptions __of painting

methods and formulas of ingredients given by the author of Samarangana Siitradhara

cannot be applied to the murals of Ajanta or B~h or an_y Buddhist walLPaintigg site of

ancient India, since these paintings were earlier than the canon by a few centuries, as in

the case of the descriptio~_given in Yasodhara's commen~o_n Kamasutra. Altho_l.!gh it

is needless to discuss this. matter further, it is noteworthy that the most valuable

contribution of the Samarangana Siitradhara~ears to be the discussion of '_eight limbs__,'

which are eight successive stages in the making of a meritorious painting. In fact, the

importance of these eight limbs cannot be over-estimated__, for with their he_!p the entire

procedure involved in producing a -mural in Indian technique can be explained.

Accordingly, it seems that while the well-known six-limbs or Sada_fJJja of_pain~igg

referred to by Yasodhara,. the _commentator of 1he.Ka.masutra39and the eight.qualities or

'gunastaka' of the Visnudharmottara40_enumerate .the aesthetic virtues _of _painting, these

eight limbs or the 'astangani' oftheSamarB.ngana Siitradhara41.enjoinlhe technicaLsteps

essential in particular for wall painting.

38
Asok K Bhattacharya, Technigue of Indian painting: A study chiefly made based on silpa texts, Saraswat
Librlll)', Calcutta, 1976, Introduction, p.l2; Jayanta Chakrabarti, Techniques in Indian mural painting, KP
Bagchi and co, Calcutta, 1980, p.31. ·
39
For the detailed account of these six limbs refer to Abanindranath Tagore, "Sadanga or the six limbs of
painting," Journal of the Indian Society of Oriental Art, Calcutta, 1961; Some notes on Indian aesthetic
anatomy and Sadanga or the six limbs of painting. Indian Society of Oriental Art, Calcutta, (reprinted) 1968,
pp.l-18; See also Phanindra Nath Bose, Principles of Indian silpasastra with the text ofMayasastr!!, The
Panjab Sanskrit Book Department, Lahore, 1926, p.84; Sri Aurobindo, The significance of Indian art, Sri
Aurobindo Circle, Bombay, 1947, pp.84-95.
40
The verse reads: Sthiinam pramiinam bhulambho madhuratvam vtbhaktata I
SadraSyam kSayavrddhi ca gunastakamidam smrtam II Visnudharmottara am, chap. 43; 19v. For the
meaning of eight limbs please refer to AK Coomaraswamy, "Visnudharmottara (Ill), chapter 41 ,"Journal of
American Oriental Society, Vol. LIT, 193 2, pp.1-17.
41
The verses read; Vartika (pra) thamam tesam dvitiyam bhtimibandhanam/
Lekhyam trtiyam syad rekhiikarmlini (vartatemiha laksanam ?)
 290

According to the short descrjption thus_given above_, it is <mparent that these

canons of paintings42contain rules ofIndian paintings__co'lering a long period, at least from

Pancamam (karsakamnacca?) sastham syiid vartanlikramahl

Saptamam (lekhanam lekhakaranam dvicakarma?) tathiistamam// Please refer to AK Coomaraswamy,
"Visnudharmottara (III), chapter 41 ," Journal of American Oriental Society. Vol. LII, 1932, p.l7.
Nevertheless, Bhattacharya reasonably rejected these and has given another interpretation. Pleas refer to
Asok K Bhattacharya, Technique of Indian painting: A study chiefly made based on silpa texts, Saraswat
Library, Calcutta, 1976, Introduction, pp.92-97.
42
Besides Samarangana Sutradhara. regarding few other cannons of paintings, which also belong to ,
non-Buddhist tradition, please refer to PA Mankad, Aparajitapracca, Gaekwad's Oriental Series, No.CXV,
Baroda, 1950. Although this is an exposition of the science of architecture, other arts like sstrology and
music etc have also been discussed. Among them, few chapters on the icono-plastic art and painting also
included. Abhilasitiirtha Cintiimani or Miinasollasa, ed. GK Shrigondekar, Gaekwad's Oriental Series,
NolXXXIV, Baroda, 2 vols. 1939; Abhilasitiirtha Cintamani, ed. R Shama Sastry, University ofMysore
Sanskrit Series, No.69, 1926; See also AK Coomaraswamy, "The technique and theory of Indian painting,"
Technical studies in the :field of :fine arts, Fogg Art Museum, Harvard University, Vol.III, No.2, 1934. The
authorship of the book has been attributed to the Western Chalukyaking Someswara ostensibly Somadeva
III (:first part of the 12th century AD) in the colophon of the text. The third part, which contains a group of20
chapters, deals with architecture, painting, iconography etc. It appears that the Miinasolliisa was the earliest
among the Silpa texts describing the pictorial tradition in South India. (Asok K Bhattacharya, Technique of
Indian painting: A study chiefly made on the basis of si]pa texts, Saraswat Library, Calcutta, 1976,
Introduction, p.18); See DN Shukla, Silpa sastra: Hindu achievements in aeronautics and :fine arts, Shri
Darshan lal VastuvanmayaPrakasha-sala, Lucknow, 1967, pp.63-65. Silparatna is also another important
text on painting. Chapter 46 of the Silparatna is entitled Citralaksana or the characteristics of painting. The
manuscript has been edited by T Ganapati Sastry, Trivendrum Sanskrit Series, No.7 5, 2 vols. 1922.
Subsequently an introductory note of the Citralaksana section i.e. chapter 46 of the Silparatna by KP
Jayaswal appeared only with the terminological meanin_$ of certain words used in the chapter. KP Jl!)'aswal,
"A Hindu text on painting," Journal of the Bihar and Orissa Research Society, Patna, VoliX, 1923,
pp.30-39. Thereafter a translation of chap. 46 of the text by Coomaraswamy wasyublished in sir Asutosh
Memorial volume. See AK Coomaraswamy, "Citralaksana," Sir Asutosh Memorial Volume, Patna,
1926-28, pp.49-61. The author of this manuscript was Sri kumara. However, the Silparatna being the latest
among the major Silpa texts, offers a comparatively dependable reading and from the comprehensive
character of the book it would not be an exaggeration to say that this chapter of the Si]paratna could have
been alone a useful source for the stutiy of the technique of Indian _painting, so far as the tradition of South
India is concerned. Asok K Bhattacharya, Technique of Indian painting: A study chiefly made based on
silpa texts, Saraswat Library, Calcutta, 1976, Introduction, p.20; Jlo/anta Chakrabarti, Techni:9ues in ln~!'ll
mural paintings, Calcutta, 1980, p.4; AK Coomaraswamy, "Citralaksana," Sir Asutosh Memorial Volume,
Patna, 1926-27, p.54. The manuscript ofNaradaSilyaSiistra is also anoth~ important work. V ~van,
"Two chapters on paintings in the NaradaSilpaSastra," Journal of the Indian Society of Oriental Arts,
Calcutta, Vol.IV, 1936. In chapters 60 to 66 ofNaradaSiloaSiistra manuscript discusses the construction of
some special types of painting galleries. Moreover, it d~bes the laws of beautification ofbuildings with
paintings. Nevertheless, compared to the Visnudharmottara and the other former manuscr_9>ts
NiiradaSilpaSii:stra contains scanty information on painting. From the external evidence it is almost certain
that the NaradaSilpaSiistra represents the tradition of South India and belongs to a later date. However, it is
evident that the reading of the text is difficult and full of problems. See Asok K Bhattacharya, Technique of
Indian painting: A study chiefly made based on silpa texts, Saraswat LibJ11!Y, Calcutta, 1976, Introduction,
p.23; See also Jayanta Chakrabarti, Techniques in Indian mural paintings, Calcutta, 1980, p.21: Balram
Srivatava, Notes of Indian aesthetics (with special reference to silpa), Chaukhambha Oriental Research
Studies, No.33, Chaukhambha Orientalia, Varanasi, 1985, pp.l 07-131. The Sudhale.pavidhanD, another
booklet of only 46 slokas, dealing with mural painting, is also found to be one containing useful
information. See VV Sharma, "Method of plastering wall for painting (Sudhalepavidhana)," The Indian
 291

the sixth or seventh century AD onwards. Within this lo!_!g_period, it is obvious that there

were different traditions of the silpa sastra works in India of which the two well known

were the traditions of the south and the north. Of such various canons_, without doubt the

Visnudharmottarapurana and the Samariingana Siitradhara belong to the north Indian

tradition while others beloQg to the south Indian institution. Bu~, in a wider context a stu<!Y

of these texts, particularly the sections on paintings, reveals that in spite of their apparent

differences, there is a common ~_proach to the su:hject. Nevertheless_, it is evident that the

northern tradition of painting, as represented.in .the si/pa. texts. has not .survive.d43.though a

few examples of the southern tradition can still be found in the south Indian region.

Although it is obvious that the Buddhistpaintin_gs of~anta and~~ are differ~nt

from the descriptions given in the silpa cannon so far discussed, it is worth indicating the

magnitude of the problems involved in the stugy of extant s!Jpa texts. Da~e and

mutilations of the manuscripts, sometimes in important sections, constitute a setious

handicap. In addition, copyists' errors are particul~!y_problematic in the identification Qf

the technical words in their original forms. It is also obvious that the texts come from

different regions of the vast Indian sub-continent and as such contain certain technical

terms coined by regional.artists.~Ano.ther.difficulty with si]pa texts.is.that.some of .the

technical instructions were put down in sutra form, which without any explanatory

commentary by one who knows~ remain vague, _obscure and inadequate. ~Therefore, it is
4

Historical Quarterly, Vol.IIL No.1, 1927. It seems that the concise text was written after the Silparatna and
most of the instructions of the Silparatna related to the method of plastering are briefly but systematically
stated in this text. Jayanta Chakrabarti, Techniques in Indian mural painting, KP Bagchi and co, Calcutta,
1980, p.ll.
43
Asok K Bhattacharya, Technique of Indian painting: A study chiefly made based on si1pa texts, Saraswat
Library, Calcutta, 1976, Introduction, p.25.
44
1bid, Introduction, p.S.
45
T tnta Chakrabarti, Techniques in Indian mural painting, KP Bagchi and co, Calcutta, 1980, p.4.
 292

not without reason, that Coomaraswam__y observes that the maste!Y of si]pa texts will

have to be acquired gradually: existing dictionaries offer little aid and often worse than

none. 46

In addition, it is highly probable that practising or professional artists did not write

most of the Indian silpa texts. The authorsll!ps of Samarangana Sutradhara and

Abhilasitartha Cintamani or ManasolHisa, which have been assigned to king Bhoja of

Paramara dynasty and Western Chalu}sya ki~ Someswara re~ectiveJy_, are best

examples in this regard. Similarly, poets who possibly had some association with the

artists and were commissioned.by ihe .rich. patrons to write _such texts. 4~, it is

obvious that mere observation and enquiry by such non-professionals were bound to

have certain limitations. It was_ irn_possible for them to_go <\~ into me minute details of

the technical processes in each and every phase of execution of a painting. This is

particularly evident from the fact that a trite inade_guate formula is often re_peated!Y

prescribed for different categories of painting, as if it was equally applicable to all. For

example, there are no s~arate lists of colours for tem_pera or fresc.o_painting on canvas..

palm leaf and for mural painting etc. In fact, it is obvious that all the colours mentioned in

the list can never be used in a mural_painting alone. No doubt onJy limited colours.. like

yellow ochre, red ochre, lapis lazuli, terre verte, black (carbon), lime or kaolin etc which

are lime resistant and not sensitive to alkalis~ can be used in mu~l_paintin_gs_, since .lime

46
AK Coomaraswamy, "Visnudharmottara (ill) Chapter 41," Journal of the American Oriental Society.
VoLLII, 1932,p.l3.
47
Jayanta Chakrabarti, Techniques in Indian mural painting, KP Bagchi and co, Calcutta, 1980, p.4.
 293

is used as an ingredient for plaster and for either_prime coat or both_prime and subsequent

48
coats in Indian murals.

However, besides these Indian texts, no indication whatsoever exists that the

ancient Sri Lankan artists followed detailed corp._plex prescriPtions similar to those

followed by the non~uddhist painters oflndia.~ut, it is_notewort:hy that there is .a large

volume of unpublished information mainly confined to ola-leaf manuscripts, having as

their subject matter architecture, scu_!pture~aint4!.g and other minor arts and· crafts tho~

scholars in Sri Lanka have rarely been inclined to pursue studies connected with these .

aspects. 5<Many such handbooks of medieval _and late mediey_al periods_hav_e_been found

particularly in various family collections and libraries but have little to contribute on the

ancient tradition. 51 How.ev.er~_as inthe_.case_oftheJ:ndian_.c.anons it.is...evidentJhatmany of

these works onsilpa sastras ofthe island too are in Sanskrit, of a corrupt nature, followed

by paraphrases in Sinhalese. For instance, the art canon ofthe Sri Lankan__painters ofthe

medieval period is a work named Rupavaliya, which is also in the same form and

language. According to Coomaraswarp._y it is a work of Brahmanical origin datable to

about 11th century AD52 and its_contents were J;ummarised by: him_ as e_onsisting in J.he

main of the iconographic characteristics of various gods and some ornamental devices,

which were known to sculptors . and painters. 53Jn .addition, there are. _a few other

48
1bid, p.l7.
49
Siri Gunasinghe, "Ajanta's shadow on Sigiriya," The art of Ajanta: New pers.pectiyes, ed. Ratan Parimoo
and others, Books and Books, New Delhi, Vol.II, 1991, p.503.
50
Nandasena, Mudiyanse, "Silpasastra works in Sri Lanka," Journal of the Royal Asiatic Society of Great
Britain and Ireland, London, No.I, 1978, pp.69-73.
51
Nandadeva Wijesekara, Early Sinhalese painting, Saman Press, Maharagama, 1959, p.77.
52
AK Coomaraswamy, Medieval Sinhalese art, New York, Second edition, 1956, pp.111-113.
53
Ibid, pp.111-113.
 294

unpublished manuscripts dealin_g direct!Y with the su~ect ofpaintin_g, but they too belon_g

to the medieval and late .me.dieval perio.ds. 54

According to this short description, though it is clear that nothing remains of

precisesilpa canons of the Buddhistpainters of Sri Lanka of the ancient period, there is no

evidence to suggest as some scholars have done that the ancient painters of the island have

also used some of the Indian canons lik~- Visnudharmottarapuran~- Samarangana

Sutradhara and some other still.later. non~uddhist .regulations. 55 Some ·writers have

incorrectly gone to the extent to state that the Buddhist painters of Sri Lanka have used the

six limbs given in the Commentary .on the Kamasutra ioo. 5 ~ence, it .is obvious .1:hat when

examining the techniques of ancient Buddhist mural painting traditions of India and Sri

Lanka the most suitable way is to reJy _u_pon the information obtainable from scientific

researches and the close examination of the extant paintings rather than on details given in

the canonical texts. Nevertheless, it should not be forgotten that there are some limitations

in relating to the scientific researches too. For instance, if the plasters of paintings

survived for a period of more than thousanc!_yeat:s, there m!!J have been deGQm_position of

the organic material and occurrence of chemical and physical changes in the ingredients.

54
For further details of these manuscripts please refer toM. Somathilake, A historical study of mural
paintings in Buddhist temples during the 18th and 19th centuries of Sri Lanka, Unpublished MA dissertation,
University ofPeradeniya., 1996, fifth chapter.
55
Nimal de Silva, "Studies in the techniques of ancient paintings in Sri Lanka," Pet'Sj>ectives in
archaeology: Leelananda Premati.laka festschrift 1990, ed. S Senaviratna and others, Department of
Archaeology, University ofPeradeniya, 1990, pp.l27-132; Raja de Silva, "Painting: Early period 247 BC to
800 AD," Archaeological ~artment centenary 0890-1990) commemorative series, Volume five-
Painting, ed. Nandadeva Wijesekara, State Printing Corporation, Colombo, 1990, pp.9, 16-17 and 30; R. H
de Silva, "The technique of ancient Sinhalese wall painting- Sigiri," Paranavitana felicitation volume, ed.
N.A. Jayawickrama., Gunasena& co, Colombo, 1965, p.l05.
56
Phanindra Nath Bose, Principles of Indian silpasastra with the text ofMayasastr!!, The Panjab Sanskrit
Book Department, Lahore, 1926, p.84; Nanda Wickramasinghe, "Mural paintings: 800 AD and 1200 AD,"
Archaeological Department centenary (1890-1990) commemorative series, Volume five - Painting, ed.
Nandadeva Wijesekara., State Printing Corporation, Colombo, 1990, p.65. The later states that "at this stage,
 295

Therefore, even modem science would also not be able to reveal the. qriginal ii_!gredients

correctly as in the case.oftheformulas__given_in the si/pa canons. 57

The support and the ground of the paintings:

As discussed in the third and fourth ch~ters in detail_, most of the Buddhist mural

paintings of India and Sri Lanka were either in cave temples, brick-built temples or

structural stupas. Hence, it is essential to focus attention at this_point at lea.St bri~fly, on

architectural forms and the special characteristics of these monumental structures also.

For example, the caves of 1\janta fall into two distinct__grol!Ps s~parated from each other

by a fairly long interval, the first belonging to the second century BC and the second, to

the fifth or sixth century AD. There are altQgether at Ajant!, includiQg the unfinished

ones, a total of thirty-one caves, ofwhich cave nos 9, IO, I9, 26 and 29 are caityagrihas

and the rest of the caves are monasteries. Of these_. the earliest caves at ~anta are the

caitya halls no 9 and I 0 and the viharas no I2 and I3. Of the caitya halls at Ajanta, cave no

I 0 is considered the oldest and chronolQgicallY it follows the caiJYa halls ofBh~a and

Kondane. Hence, its architectural forms are more highly developed than those of the

earlier caitya halls; particularly .the .spring_of!he horse.shoe.arch.on the ~de. 5SWhereas

the fa~de screen of cave no I 0 is carved in wood and the pillars of the nave slant inward,

as they would if the structure were of wood and thrust of the roofhad to be counteracted,

he (the artist) converts his imaginary picture into a material picture, in accordance with the six-fold
r,nnciples, ~ch as iconography, size, colours etc." Ibid, p.65.
7
Nimal de Silva, "Studies in the techniques of ancient paintings in Sri Lanka," Perspectives in
archaeology: Leelananda Prematilaka festschrift 1990, ed. S Senaviratna and others, Department of
Archaeology, University ofPeradeniya, 1990,p.l30.
58
For further details of architectural evolution of these early caitya balls and viharas see Percy Brown,
Indian architecture. <Buddhist and Hindu periods), DB Taraporevala and sons; Bombay, 3rd ed, (n.d),
pp.25-30.
 296

the fayade of cave no 9 was executed in stone and_pillars are_peiJ>endicular. An interesting

feature of cave no 9, reflected in the later plan of the Vihara shrine, is that the end is flat

· rather than apsidal. However., these catrya caves are intemal!Y divided b_y colonnades into

a central nave, an apse and side-aisles, the latter continuing behind the apse and thus,

providing for circumambulation. In the second__phase of excavatiol!, th~enerall~_yout of

the monasteries was standardized, though each one of them continued to present some

individual features. At this st:Me a shrine with a colossal im~e ofBuddha Wa.s introduced

in the middle of the back wall of the pillared hall opposite the entrance-door. The shrine

has in many caves an antechamber in front. Sometlnies there are subsidia_!Y shrines in the

back and sidewalls. 59 It js __generally .believed that the .rock .cut viharas at Ajanta as

elsewhere were intended as residential quarters only, since they consisted of an astylar

hall surrounded by a series of cells, small sguare or rectan_gular chambers cut further into

the rock. 60

The method adopted in Ajanta rock-hewn temples was followed in the caves of

Bagh too although differentlY d~ed in detail. Con$e_guen~~ th~ener:aJ _plan of the

caves at Bagh was the same as the other caves, but these are on a far bigger scale than the

caves of 1\janta. Th~__general scheme. .consists of a main central hall with cells for

meditation all round and a main shrine with a stupa as the chief place of

worship. 61 Another similar set of caves. are at Ellom~Like. the.Buddhist.caves of Ajanta,

these follow certain layout patterns that are common to caityas and viharas, but show

certain definite stylistic differences in the architectural elements. This is reflected ~ the

59
Deb ala Mitra, Buddhist monuments, Sahitya Samsad, Calcutta, 1971, p.176.
60
Sheila L Weiner, Ajanta: Its place in Buddhist art, University of California Press, London, 1977, p.37.
 297

pillar types, niche forms and prevalent _plinth mouldin_gs etc. This difference is at once

corroborated by the iconography of the two cave groups, since Ellora is demonstrably

shown as quite evolved .and tending to he nearer the V.ajrayana. 62

It would be of interest to discuss the process of excavating these caves as evident

from the upfinished floOI and .columns of cav..e nos24 aruL25_ at Ajanta.63 This process,

commenced after the outline had been marked and excavation started from the ceili!!g that

was finished first. The work then continued downwards by the cutting of deep alleys with

sharp and heary instruments like theyickaxe followed _try the breakigg of the intervenin_g

ridges, leaving solid blocks, for pillars where necessary, till the floor was reached which

was the last to receive attention. Th~_prelimin~ work of excavation was thus done b_y

pickaxe and the rest, ·including the finish and carving was entirely executed by hammer

and chisel. According to the method of excavation, after fmishigg the ~yade and the

veranda the excavator went deep into the interior, attending first to the hall and next taking

up the antechamber, shrine or cells as the case m~ be_, the_procedure_ofthe excavation

being the same.64 Thus, it is clear. that the caves wer_e not_ only hewn out from top to

bottom vertically but also from front to back. Accordingly, one can observe that the

rock-cut architecture of the Indian_paintigg sites is considerab_!y different from the method

of construction of structural monuments. In addition, it is interesting to note that the work

61
S Fyzee Rahamin, The Indian art of drawing pictures in the caves ofBagh, The Indian Daily Mail,
Bombay, 1928, p.7.
62
KV Saundara Raj an, Cave temples of the Deccan, Architectural Survey ofTemples, No.3, Archaeological
Survey of India, New Delhi, 1981, p.l7.
63
Debala Mitra, Ajanta, Archaeological Survey of India, Eleventh edition, 1996, p.8; Benoy K Behl, The
Ajanta caves: Ancient paintings of Buddhist India, Themes and Hudson, 1998, p.30.
64
Deb ala Mitra, Ajanta, Archaeological Survey of India, Eleventh edition, 1996, pp.8-.9; See also Walter M
Spink, "The achievement of Ajanta," The age of the Vakatakas, ed. Ajay Mitra Shastri, Harman Publishing
House, New Delhi, 1992, p. 194; Himanshu Prabha Ray, Monastety and guild: Commerce under the
Satavahanas, Oxford University Press, Delhi, 1986, p.l84.
 298

of quarrying, dressing and finishing of Buddhist caves presumably went hand in hand.

This is particularly evident from the fact that even the unfinished caves of Ajanta bear

traces of paintings. Hence, it would seem that_plasterin__g and _painting was also taken in

hand immediately and simultaneously, with the excavation of the caves. For example, in

cave no 4 at Ajanta, the carviQgS on th~_pillar are unfinished, but the_paintiggs 9n the
65
ceilings are complete. Therefore, as evident from Ajanta, it may he presumed that the

chisel-work and brushwork of paintings at the mural sites of Indi~ proceeded

simultaneously.

According to the method of excavation thus described, the outer surface of the

inner walls of the caves, cut into the hard and compact volcanic tr:a_p-rock or basalt,

characterised by vesicle and amygdaloidal cavities, constituted the carrier for the plaster

on which the pigments were laid. Thus it is obvious that the surface of the carrier was

rough and uneven, with deep furrows and chisel-marks produced in the course of

excavation of the caves b_y the slow process of hammer and chisel strokes. ~d~d~ this

itself was an advantage, as a tooth was provided to the plaster laid on the surface.66_Thus,

at Ajanta, at least three types of rock surfaces were made, i.e. the most uneven, moderately

uneven and even surfaces. Of these, the most uneven surface can be seen on the ~il4t_gs

and walls of the cells in the viharas. Particularly, the rough surface of the ceiling with

deep chisel-marks was made to hold fi~b'in strQI_!g bond theh~_g!n_g_plas~re~_grounds,

so that it did not crumble easily. Moderately uneven surfaces were made on the vertical

sidewalls and front walls of most of the vihara halls of the fifth or the sixth century AD.

65
Ganesh Halo~ "Ajanta painting: Its technique and execution," The art of Ajanta: New pers,pective, ed.
Ratan Parimoo and others, Books and Books, New De~ Vol.II, 1991, p.377. See also Debala Mitra,
Ajanta, Archaeological Swvey of India, Eleventh edition, 1996, p.9.
 299

The third type was used to bring out details of ornamental pillars, for which a very high

degree of evenness of surface was felt necessary. Besides these later caves where at least

three kinds of rock surfaces can be seen, it is noteworthy that the wall surfaces of the early

cave nos 9 and 10 at Ajanta were more evenly chiselled.67 These general descriptions of

the method of excavation and also the making of walls particularly of the later caves of

Ajanta are directly applicable to the method of digging out the caves ofBagh, Ellora and

Aurangabad too.

In contrast to India where Buddhist paintings now survive only in the rock-hewn

cave temples, at least two categories of places are known where the painters of ancient Sri

Lanka drew Buddhist murals. These are the natural rock caves or rock surfaces and the

constructed walls ofthe shrines or the walls of relic chambers ofthe st11_pas. In the case of

a cave or a rock, a technique of deepening a rock shelter or cave was developed by

'peeling off' the weathered su~ce and striated ll!)'ers of the rock to form a de~p cavity or

declivity below a large natural boulder or projecting cliff-face. Meanwhile, a deeply

incised groove or drip-ledge, running in a horizontal line as far as permissible marked the

upper limits of the peeled surface, which further protected the cavity thus formed. In fact,

this drip-ledge quite effectively prevented rainwater flowing down the sides ofthe boulder

or cliff, into the depression. Consequently, this device helped to keep the surface where

the paintings were executed dry, the water beiqg directed to flow down the drjp led_ge

vertically beyond the limits of the mouth of the cave. 68

66
BB Lal, "The murals: Their composition and technique," Ajanta murals: An album of eighty-five
re.productions in colour, ed. A Ghosh, Archaeological SUIVey of India, New Delhi, 1967, p.53.
67
· Ganesh Haloi, "Ajanta painting: Its technique and execution," The art of Ajanta: New perspective. ed.

Ratan Parimoo and others, Books and Books, New Delhi, Vol.II, 1991, p.368.
68
Nandadeva Wijesekara, Early Sinhalese painting, Saman Press, Maharagama, 1959, pp.77-78.
 300

Besides, in most cases, the area to be utilised for paintings having beeri so selected

m the cave was enclosed and enlarged by the addition of frontal screens of

walls-originally made of mud and wattle or rubble, later of brick or stone masonry and

surmounted by a lean to roof to form a cave shrine. ~owever, an even surface was
6

ultimately obtained by plaining away the irregularities on the rock face. Subsequently, as

in the case of the rock cut caves of India, the surface was converted into a series of tiny

pittings by means of a pointed iron instrument, to ensure substantial adhesion of the

overlying plaster to become one composite structure with the rock. 70The rock face thus

plastered and smoothed was finally covered with paintings.

According to this short description, one major difference is clearly evident

between the cave temples of ancient India and Sri Lanka where the paintings still remain.

Primarily, the Indian paintings have been located in edifices chiselled out of the living

body of the rock. This method of carving out buildings from massive rocks is lacking in

Sri Lankan painting sites. Hence, the question arises at this point: why did the Sri Lankans

deviate from the practice thus accepted in India? It is obvious that the quality of rock in Sri

Lanka differed to make it a burden to excavate buildin_gs out of it, since it was too hard and

strong unlike the case of Ajanta, Bagh, Ellora etc. Another aspect of this issue is the

suggestion that the Indian paintings served the needs of a few residents within an area or a

temple itself whereas in _sri LankaJhese were presented to a wider congregation?1 For

instance, Dehejia concludes that it is a distinct probability that Ajanta's painted narratives

69
Senaka Bandaranayake, The rock and wall paintings of Sri Lanka, Lake House Bookshop, Colombo,
1986,p.ll. 0

70
Nandadeva Wijesekara, Early Sinhalese painting, Saman Press, Maharagama, 1959, p.78.
71
Vidya Dehejia, Discourse in early Buddhist art: Visual narratives oflndilb Munshiram Manoharlal
Publishers, New Delhi, 1997, p.34; Nandadeva Wijesekara, Early Sinhalese painting, Saman Press,
Maharagama, 1959, pp.95-96.
 301

were not intended exclusivelY or even __primari!Y for __pul;>lic viewing. In view of their

location within residential temples, it is plausible that the murals were to be experienced

by the monastic communi!)' of Ajanta and further stressed that the narratives m'!Y have

been addressed to the resident monastics with the intention that they benefit from the

Buddhist message of the tales. For exam_ple~ conteii_lplatigg on the fortitude d~l~ed ~

ascetic Khsantivadin who, when his limbs were tom apart, only blessed the king who

passed the order, may have served as an 4I~iration to the .1janta monks and facilitated

their spiritual advancement 72In this context, it is to b.e noted that some scholars hav.e
73
concluded that Ajanta was a.centre .ofBuddhist education as discussed in the third .and

seventh chapters in detail. This may be equally applicable to the painting sites ofBagh and

Ellom too as their massive architectural elements s1:1_ggest.

Besides these rock surfaces, the second variety of surfaces where paintings are

available in Sri Lanka, but not in India are the brick walls of a shrine as in the case of

Tivamka image house and the walls of the relic chambers of various stupas. It was found

that these painted walls of Sri Lanka were constructed us~ different materials and

techniques. Consolidated mud walls, stone walls in mud mortar, laterite in mud mortar

and burnt bricks in mud mortar are some of the cl~-based techn!gues used in

constructing walls, while brick in lime mortar, stone in lime mortar are some of the lime

bindings used in wall constructions. It is evident that all these walls w~r~_plastered usin_g

72
Vidya Dehejia, Discourse in early Buddhist art: Visual narratives of India, Munshiram Manoharlal
Publishers, New Dellii, 1997, p.34.
73
Shobhana Gokhale, "Ajanta: The centre of monastic education," The art of Ajanta: New perspectives, ed.
Ratan Parimoo and others, Books and Books, New Delhi, Vol.I, 1991, p.149.
 302

74
either clay mortar or lime mortar before decorating them with paintings. lt is noteworthy

that unlike the ordinary brick walls of a shrine thus described, there were several

enclosures of the relic chambers of the la,r.ger ~tu_pas. Usual!Y three compartments were

located one above the other in the centre-of a solid mass of brick masonry, which formed

the dome and basal terraces ofthe stu_pa. These chambe~_genera~]y served as repositories

for the relics, which signified the presence of the Buddha within the monument along

with other valuable thin_gs. The chambers were embedded in the masonry and total!Y
75
inaccessible after completion _ofthe construction. Therefore, as. also mentioned in the

third chapter, the paintings thus executed were actually never viewed once the monument

was completed. Hence no doubt that these wer~_part of the ritual ;w_paratus that activated

the symbolic meaning of the chambers and their contents.

It is evident tha:tplasteri!J._g the walls of such buildiggs or stt1pas or caves for

executing paintings during the ancient period of both countries was not an ordinary

method, generally adopted for the.cus.tomary houses. 7~ .the..Buddhistwall paintings of

India, the process of execution of the murals appear to have been the same in all the

examples that have survived, . accoiding . to ..Brown~ interpretation. 77 Nevertheless,

74
Nimal de Silva, "Studies in the techniques of ancient paintings in Sri Lanka," Perspectives in
archaeology: Leelananda Prematilaka festschrift 1990, ed. S Senaviratna and others, Deparlment of
Archaeology, University ofPeradeniya, 1990, p.l28.
75
Senaka Bandaranayake, The rock and wall paintings of Sri Lanka, Lake House Bookshop, Colombo,
1986,p.74.
76
For the description of methods of plastering buildings in ancient India, see Stella Kramrisch, "Plaster,"
BC Law Volume, ed. DR Bhandat:kar and others, The Indian Research Institute, Calcutta, Part I, 1945,
pp.611-616; VV Sharma, ''Method of plastering walls for painting (Sudhalepavidhanam)," The Indian
Historical Quarterly, Vol.lli, No.I, 1927, pp.53-59. This article discuses the descriptions of plastering
methods given in an ancient Indian manuscript that belong to the Hindu tradition.
77
Percy Brown, Indian painting, YMCA Publishing House, London, 1927, p.112. He describes that over the
surface of the rough excavated wall of rock, a mixture of clay, cow-dung and pulverised trap rock was
applied, to the thickness of on~ighth to ~quarters of an inch. Sometimes this first dressing also
contained finely chopped straw or rice husks. This ground was then coated with an exceedingly thin layer of
 303

different opinions have been eJgJressed on the subject of the_pr~aration of_plaster and the

contents or ingredients that were used by the Buddhist painters of India during the ancient
78
period, (especially in relation to Ajanta) by other scholars like Coomaraswamy,
79 8 81 82
Nagpall, Craven, 'bebala Mitra and Rowland etc.

But, unlike these earlier scholars, Lal outlined the actual ground of the painting of

Ajanta in detail in 1960s. Accordin_g to him the l'!Y!Iill <>_(.ground cat}_genef(llfy be

described as constituting ofthree layers: the pigments; fine plaster; and coarse plaster with

two distinct lines ofjunctions which have been confirmed_ey microscopic examination of

the painted stucco. Of these, the paint layer is generally about 0.1 mm in thickness. The

white plaster, about the thickness of, and in some senses reseplbling, an eggshell On this polished shell-like
surface, the frescos were painted in watercolour. See Ibid, p.ll2.
78
Coomaraswamy describes that the technique of paintings at Ajanta and of Indian wall painting as follows:
The surface of the hard porous rock was spread over with a layer of clay, cow-dung and pulverised
trap-rock, sometimes also mixed with rice-husks, to a thickness of from three to twenty millimetres (118 to
%inch). Over this was laid a thin coat of fine white p~Iished lim~plaster, which was kept moist while the
colours were applied and afterward slightly burnished. AK Coomaraswamy, The arts and crafts ofindia and
~ Today and Tomorrow's Printers and Publishers, New Delhi, 1913, p.87; Histozy of Indian and
Indonesian art, Munshiram Manoharlal, New Delhi, (reprinted) 1972, p.89.
79
Nagpall descnbes that the extensive Buddhist wall paintings in Ajanta and Bagh are made on mud plaster
prepared from local resources. These clay aggregates have their distinctive components depending on the
rock formation of the area and vaty in tint In any case, the plaster on the walls is generally built up in several
layers, the uppermost being the least coarse. Various kinds of filling materials have been used to minimise
the cracking of the mud aggregate, like straw, husk, hair, coarse, fibres etc, which also strengthen the mix.
Dung is added in the mix to impart a degree of plasticity. JC Nagpall, Mural paintings in India, Gain
Publishing House, Delhi, 1988, pp.23-24. .
80
Craven believes that the surface of the rock walls of the caves of Ajanta was itrst prepared by a coating of
potter's clay, mixed variously with cow dung, straw and animal hair. Once this was levelled to a thickness of
half an inch to two inches, it was coated with a smooth, fine white lime plaster, which became the actual
painting surface. Roy C Craven, A concise histozy of Indian art, Themes and Hudson, London, 1976, p.l27.
81
Debala Mitra believes that the ground was first prepared by laying a rough layer offerruginous earth,
mixed with rock-grit or sand, vegetable fibres, paddy husk, grass and other fibrous material of organic
origin, on the swfaces of the walls and ceilings, which were left rough so that they might act as 'tooth' for
the rough plaster to set firmly upon them. Next, a second coat of mud and ferruginous earth mixed with !me
rock-powder or sand and !me fibrous vegetable-materiel, was applied to it. The surface was finally finished
with a thin coat of lime-wash. Debala Mitra, Ajan~ Archaeological Survey of India, Eleventh edition,
1996, p.l2.
82
Rowland thinks that the rough surface of the wall or vault was first covered with a layer of clay or cow
dung mixed with chopped straw or animal hair. When this had been smoothed and levelled, it was given a
coating of gesso (fine white clay of Gypsum) and it is on this ground that the actual painting was done.
Benjamin Rowland, The art and architecture of India Buddhist Hindu, Jain, Penguin Books, USA, 1953,
p.243.
 304

underlying layer of fine _plaster var!~s in thickness from c~v~ to cave and ral!ges in

thickness between 2 and 3 mm. The thickness of the lowest coarse of plaster below varies

over a wide range, bein_..g_primari!Y determined on a__particular ~ot 12J the unevenness in

the rock surface. 83

Thus, it is obvious that altogether Ajanta paintings are composed of three major

layers; the support or the carrier.; th~__groun~_; and the l4:!Yer of_IDgments. Of these_, -~

already stated the support or the carrier is the surface of the rock, which is rough in nature.

Hence, it is <;ertain that the SJ!Pj)Ort is the actual base on which th~aintiJ!&s are executed.

Generally, these surfaces of walls and ceilings at Ajanta are full of uneven

spherical-cavities created due to the ev.ru>oration oftra._m>eg_gases duri_n_g the ef1!Ption. But

it is obvious that while excavating the caves; the surface of the rocky-carrier was

originally ro._l!gh~ned with chisel-m~ks in different d~rees to make better bonqin_g and to

suit the ground to be _covered by painting as..already.noted above. 84

It was also mentioned above that the ground was mainly prepared by the

application of two coats o(plaster on the carrier or the S!!m>Ort. Of thes~ the first coat w..as

coarse in texture with a considerable amount of fibrous vegetable material and rock-grit or

sand. The unevenness of th~ chiselled rock surface was___pr9_pedy rectified llY th~

application of this coat. This was then made smooth by another layer of mud and

ferruginous earth, again mixed with fin~ rock__powder or sand and fine fibrous ~etable

85
material. This fairly rough surface was .finally rendered smooth. by the application .of a

83
BB Lal, "The murals: Their composition and technique," Ajanta murals: An album of eighty-five
n;productions in colour, ed. A Ghosh, Arohaeological Survey of India, New Delhi, 1967, p.54.
84
Ganesh Haloi, "Ajanta painting: Its technique and execution," The art of Ajanta: New pe!Sj)ective, ed.
RatanParimoo and others, Books and Books, New Delhi, Vol.II, 1991, p.368.
85
BB Lal, "The murals: Their composition and technique," Ajanta murals: An album of eighty-five
n;productions in colour, ed. A Ghosh, Arohaeological Survey of India, New Delhi, 1967, p.54;
 305

thin layer ofJime-wash, which was then painted_over. 86Jt has been revealed that this whole

ground of the painting is composed of mud-plaster containing about 10 to 12 percent of

combined w~ter and organic matter_, such as y~etable fibres,_padQy hus~_,_grass and other

fibrous material of organic origin and rock-grit or sand. Silica is present to the extent of

about 60 percent and iron ~d alumina account for 27 _percent. In addition_, lime and

magnesia are also presenuoJ:he extent.of2to3_percent. 87

It is obvious that roughened or grainy surface (more or less according io the nature

of the work) mention~_d aboye ·is alvve1ys considered as a kind of bond and h~Jps the

application of any kind of ground to the support and subsequent coat of paint to the

ground. Consequel}f!Y, J()_!!g}mess of a ~!!J>_po_rt o_Lground is called the 'tooth' or the 'ke_y'

of the swface. ~owever, itis notewDrthy.thatbe.ing.mainly ofmud,_the _ground _is .soft

8

and porous and it does not possess the natural strength and durability of common lime

plaster. ~evertheless; it seems that._great care was exercised by the artist in Jaying the
8

ground, for although almost completely free from lime, it still had a degree of

compactness and hardnes~ which rendered it fit to receive the brushwork. Particularly__, a

study of the grain-size or the particles constituting the plaster has shown that the mixture

of ferruginous mud a.Q<lgri!!J ro_~k_ppw_der and sand was very carefu!!Y_ga!l_ged and tha.t

the grains of silica possessed marked angularity, which has considerably contributed to

"Presetvation of mural paintings," Ancient India: Bulletin of the Archaeological Swvey of India, No.22,
1966, (reprinted) 1973,p.90.
86
SK Bhattacharya, The stozy of Indian art, Atma Ram and sons, Delhi, 1966, p.49; BB Lal, "The murals:
Their composition and technique," Ajanta murals: An album of eighty-five reproductions in colour, ed. A
Ghosh, Archaeological Swvey of India, New Delhi, 1967, p.54; "PresetVation of mural paintings," Ancient
India: Bulletin of the Archaeological Survey of India, No.22, 1966, (reprinted) 1973, p.90.
!rl BB Lal, "The murals: Their composition and technique," Ajanta murals: An album of eighty-five
reproductions in colour, ed. A Ghosh, Archaeological Swvey of India, New Delhi, 1967, p.53.
88
Ganesh Haloi, "Ajanta painting: Its technique and execution," The art of Ajanta: New pers_pective, ed.
Ratan Parimoo and others, Books and Books, New Delhi, Vol.ll, 1991, p.368.
 306

the firmness and compactness .Of the_plasier. 90The_fibrous .material incorporated in the

plaster for imparting to it a degree of strength has, however proved to be a serious

drawback, for it has de~ed__paJ1!y on account of d~redatjons lD' insect.:Pest and parfur
on account of the natural decomposition of the vegetable-matter under conditions of high

humidity and temperature. Accord~ to all these observations it is clear that the lo~ev_ity

or deterioration of a painting mainly depends on the nature and preparation of the ground
. 91
m most cas;es.

In this context, it is noteworthy that though the main ingredient of the Ajanta

ground plaster is clay mixed with bindi!!g materials as discussed above_, the i!!gredients

and the quality of ground differs due to the difference in the evenness or not of the support

base. In fact, it is for this reason that one notices severliJJYpes Qf_ground_plasters in_Ajanta.

For instance, it has been observed that for a very uneven support the painters had used

three layers of ground; for modera~~!y uneven Sl!Ql)Ort two !~ers of.ground_; and for an

uneven support just one layer of ground. It has been revealed that in each layer, binding

ingredients_ofthe cliD' _gr_ound also differ~<t but as alrea~_y no_ted lil!l__e was not used alQ!!g

with the plaster. 92

Thus, to fill in the voids of the chiselled furrows of the most uneven rock surface,

the first layer was made thick t()__even_m> the surface for sub~~ent cqats of!ID'~rs. It W::t.S

found that in this layer, a kind of grass seeds, paddy-skin, fibre and stone chips of uneven

sizes were mixed with cliD'· BJ.lt.it should not be fQI.gottJm that all these i_!illredients haye

89
BB Lal, "The murals: Their composition and technique," Ajanta murals: An album of eighty-five
re.productions in colour, ed. A Ghosh, Archaeological Survey of India, New Delhi, 1967, p.53.
90
Ibid, p.54.
91
See BB Lal, "Preservation of mural paintings," Ancient India: Bulletin of the Archaeological Swvey of
India, No.22, 1966, (reprinted) 1973, p.90.
 307

not always been found together. 93 The second layer was made of fine clay and sand mixed

with fifle fibre. At some places cotton and long grass have also been found from which it is

evident that the early scholars did not identify this stratum ofplaster s~arate!Y. However_,

it is to be noted that this layer was applied between the first and the third layers, on the

most uneven rock surface. But_, sometimes this_plaster was used directlY as the first l'!Yer

on another variety of wall surface i.e. moderately uneven support, for adorning with

paintings. It is evident that on this moderat~ uneven surface_, a thick liD'er of_ground

preparation was not required unlike the case of most uneven surface. In addition, it is

noteworthy that this moderateJy uneven Sl!P.POrt does not have tooth eno_!l_gh to hold a

heavy ground, like the uppermost layer. Hence, this finer and lighter second layer is the
94
more a:ppropriate one for this kind of uneven _support

The third layer is made of clay and very fme sand. It is evident that in most cases

the paintinps of .t\jan__ta caves were executed on this l~_yer. In this_process_, it is common

practice that after a clay model has been made and before colours are applied, a clay paste

comprising of clay mixed with waterjs awlie__d on the moclel wiJh ~ tbi.ck brush w_obtain
a smooth surface. The base becomes clean and smooth by this method and helps in

colouring as well as brin_gin_g ou.tthe__eff~ts of the colours, N__ev_ertheless_, it is ~ifi_cant

that the surfaces of the earliest caves no 9 and 10 are more whitish than those of a later

period, since lime plaster w_as U$e__d in the_prQCess_of_pl3Sterin,gJ1:!e_two cayes_.. Of_th~e_,

particularly on the left wall of cave no 9, it is clearly evident that the paintings were

originally .made on lime_-pla,stered_ground. But as mentioned in th~_previous ch~_pters

92
Ganesh Haloi, "Ajanta painting: Its technique and execution," The art of Ajanta: New pers,pective, ed.
Ratan Parimoo and others, Books and Books, New Delhi, Vol.II, 1991, p:370.
93
Ibid, pp.371-37 4.
 308

these were later covered u_p Q)' cl~_plaster and painted anew around the fifth century

Apart from these details obtainable from close observations of the paintings, it

would be interesting to note the laboratq.!Y investigations and scientific analysis of the

paintings of Ajanta done in the 1930s, which reveal at least two stages of plastering, i.e.

rough and fine, in t,h~ _p_r~aration of the_ground for the murals. Thus, as in the case of the

early identifications, as recognized by Paramasivan too, clay and sand played·a significant

role in the paintings of the caves of Ajanta for th~r~aration of the ro_!!gh,_plaster of the

ground, while lime was present in a very low percentage. The presence ofvegetable fibres

in large quantity, alo~ with other v~etable_products~ m!J be __presumed_, based on

chemical analysis. It is also revealed that fine plaster was prepared out of a mixture oflime
96
with a small percentage of Gypsum, i.e. Calcium Sulphate.

As in the case of Ajanta, at Bagh too a close examination of broken portions,

suggests th~:tt 1:4~ paintilJ.g seems to have been done subseguent to the final smooth4!g Qf

the surface. According to the process, perpendicularity of the wall does not appear to have

been attended to in. the _original and un_d~ound cb).sell!!!g, for .th_e ro.!!_gh_plaster is in

places several inches thick, in others quite superficial, but over both is an even and finer

94
Ibid, pp.371-37 4.
95
1bid, pp.371-37 4.
96
S Paramasivan, "Technique of the painting process in the cave temples at Ajanta," Annual Re,ports of the
Archaeological De_partment of His Exalted Highness the Nizam's Dominions, 1936-37, pp.25-36. He
further says that the rough plaster of lime was common with south Indian paintings.
 309

finish of l,ime. 97Scientific investigations have revealed that this rough plaster was .made
98
out of mud.

Besides Bagh, though the paintings are not extensively available today, at Ellora

too chiselling first rol!ghened the rock surface so that the._plaster could stick to it_p_ro_per!y

as in the case of the walls of Ajanta. It has been revealed that the plaster was composed of

clay mixed with cow-dlJ.Q& chQQPed husk, fme sand etc and its thickness varied from

about 1 to 1.5 em. Though it does. not belong to the caves of the Buddhist gro~, on

examination the plaster from Kailasa, the elaborate Hindu cave at the site has been found

to be made of ferruginous eart:h,. containing large quantities of silica, iron and alumina.

The percentage oflime is ve.ry low and could have been_present in the mud as an im_purizy.

Over this rough plaster were given one or two more coats of lime p~ which was made

of lime and a small pr:oportion of calcium su]phate. This__plaster was reinforced with

vegetable fibres. Thus,. the ma.in._ difference between the technique of paintings at Ajanta

and that ofEllora is that there is no evidence of lime_plaster ~lied over mud_plaster in

the later caves at Ajanta as examined by Paramasivan~u:t, it has to be .admitted that

there were many imperfections in the process of laboratory research followed by

1
Paramasivan and the .conclusions that he.made. 0<Jience, when considering the techniques
of Buddhist mural paintings of Indi~ it is more reasonable to ac:U:!ere to the information

obtainable from close observations of the paintings themselves.

97
E Impey, "Description of the caves ofBagh in Rath," Journal of the Bombay Branch of the Royal Asiatic
Society, Vol.V, July 1856, p.560.
98
S Paramasivan, "Technique of the painting process in the cave temples at Ajanta," Annual Rq>orts of the
Archaeological ~artment ofHis Exalted Highness the Nizam's Dominions. 1936-37, pp.25-36.
99
S Paramasivan, "Technique of painting process in the cave temples at Ellora," Annual Rq>ort of the
Archaeological Department ofHEH the Nizam's Dominions for 1938-39, Calcutta, 1939, pp.31-38.
 310

As mentioned above_, unlike in Indi(b the Buddhist wall _pain~J!&S of Sri Lanka_,

that are now extant are to be found on two durable supports: rock and brick walls of

shrines and the relic vaults of the stu_pas. It is obvious that in case the Sl!PIJOft was a brick

wall, laying an external plaster on which the paintings could be executed was. th~ only

requirement. In case the surface of the S\!I)_port was a roc~ firs!, it was artificial]y

roughened to allow for the adherence of the lowest layer of ground as already noted, This

technique is cl~rly di~l~ed -~the SUQPOrt at ~it}ya as also at Ajanta_, B~h and

elsewhere. It is evident that at all these places, the chiselling work was done with al) iron

point but in Sri Lanka the chjps removed were smaller than in India as is evidenced b_y the

marks. Undoubtedly, the quality of the rock mainly decided. this procedure and

consequently, smaller_pitti!!gs were _r~uired to be _plastered at the_paintjn_g sites of Sri

Lanka.lOl

Besides support or the carrier thus prepared, when_considering the ground of the

paintings of the Sri Lankan mural site~ Bel~ the earliest scholar who worked on Sjgiri_ya

had concluded that this was about half an inch thick and comprised of tempered earth and

kaolin of a reddish brown hu~ strengthened with rice-husks and._per4_~s shr_eds of

coconut fibre. He further concluded that upon this were laid at least two coatings of white

lime, a qUJU1er to half an inch _thick. 102..But, it has been found by later _schola.rs that the

ground pl~ter ofSigiriy_a was .applied in tbre.e..distinct layers.as in the case of Ajanta. 103

100
RH de Silva, "The evolution of the techniques of Sinhalese wall painting and comparison with Indian
painting methods," Ancient Ceylon, Journal of the Archaeological Survey of Ceylon, No.1, 1971,
pg.98-103.
Nandadeva Wijesekara, Early Sinhalese painting, Saman Press, Maharagama, 1959, p.96.
102
HCP Bell, "Interim report on the operations of the archaeological survey at Sigiriya 1897 ,"Journal of the
Royal Asiatic Socie1y (Ceylon Branch), Vo1.XV, No.48, 1897, p.ll4.
103
RH de Silva, Sigil"!ya, Department of Archaeology, Sri Lanka, -1976, p.l9.
 311

Of them, as correctl_y observed l;ly Khan Bahadur Sana Ullah, an archaeological

chemist of India who also participated in several restoration works of the wall paintings of

Sri Lanka, the rock surface_, which is rol_!gh and uneveJ1_, has at the outset been__plastered

with local red alluvium reinforced with vegetable fibres and rice husks. This mateJjal is

very gritty and contains an ~_p_preciable amount of ferric oxide_, which accounts for its red

colour. 1 ~e next layer consists of a buff coloured composition containing sand..as .the

principal ingredient besides some clay, lime and vegetable fibres. The material is quite

soft and can be readi!Y _powdered between the f~ers. It also contains 5_percent calcium

carbonate, corresponding to 2.75 percent lime, which is insufficient to confer necessary

strength to the plaster. It is certainly inconceivable that the ancient__painters of the island

could have selected such a weak material for preparing the foundation for their paintings

and the on?' reasonable e'SJ)lanation is that or_iginal_!y this leyer had set sufficientJy hard

but has been reduced to the present weak state through natural deterioration, in the course

of so many centuries. ProbabJy some adhesive substances such as cow-dung, cereals_,

starch, glue, gum etc were employed and these have also perished by decomposition. The

third or the uppermost ll!)'er, about 1/8 inch in thickness, is com_posed of sand and lime

mortar to the value of 17.5 percent and is in a fair state of preservation. It has been worked

up to a smooth and even surface with sm>erficial ~lication o:(_pure lim~ in order to

receive the brushwork. The total thickness of the composite plaster varies between 3/8
105
inch and 1 inch according to the contour ofthe rock surface.

104
Khan Bahadur Sana Ullah, Rt;port on the treatment of the Sigiriya frescoes and suggestions for the
preservation of paintings in the various shrines and old monuments in Ceylon, Ceylon Sessional Paper,
No.xxi, September 1943, Government Press, Colombo, 1943, p.3.
105
Ibid, p.3.
 312

Thus, it is evident that in the_process ofla_yi~_g the__ground, there was a tendency to

use lime (mixed with clay and sand) on thicker grounds and to retain the admixture of

vegetable fibres in both the fer~inous c~~l~ers as well as on lime-based_grounds. It is

interesting to note that more than one layer of lime-based ground has been used in a
1
painting uplike the murals of Ajanta. ~wever, the surface of.each layer of .ground was

left irregular to provide a key for the next layer. In addition, on the surface of th~ clay

ground, there was a concentration of__paddy husks toj:>rovide further key or tooth for the

next layer of lime-based ground. Finally, the application of a lime wash to receive the
107
paints was retained. Accordingly, it is clear that the.murals of Sigiriya are.painted upon

a thin layer of finely ground sand and lime, probably produced by burning aquatic
10
molluscan shells. ~t is thus evident that as in the.case.of Ajantathe paintingsofS~riya
109
were also ,executed on a carefully prepared .surfa.c.e though the process considerably

differed from that used at Ajan~.

Though the above applies to the paintings- of Sigiriya, the main principl~ and

techniques hold well for other sites of the island as well. Of thes~ in the_painti__gg of

Hindagala, two types of ground are laid. The first type of ground is found in the upper part

of the painting and this is finished in a better techni_gue than was used in liD'ing the second

type in the lower part of the painting. No straw or-other fibres except for paddy bus~ are

106
This has been described by Gunasinghe that "At Sigiriya two more layers of plaster were appli~, this
time with a quantity of lime added to the mix. It would appear that the second layer was specially meant to
fill out the undulations of the rock surface which, unlike at Ajanta, was not completely chiselled into an even
plane. Over the second layer was applied a third, containing a very large proportion of lime (33 .3 percent
according to de Silva), mixed with sand and vegetable fibre. It is significant that at Ajanta neither of these
intermediate layers of plaster has been noticed." Siri Gunasinghe, "Ajanta's shadow on Sigiriya," The art of
~anta: New: pers,pectives, ed: Ratan Parimo~ and other:', Books and B~o~s, New Delhi, ~ol.II, ~ 991, ~.502.
1
RH de Silva, "The evolution of the techniques of Sinhalese wall pamting and companson wtth Ind1an
ftainting methods," Ancient Ceylon, Journal of the Archaeological Swvey of Ceylon, No.1, 1971, p.93.
00
PEP Deraniyagala, "Some side lights on the Sinhla monastery-fortress of Sihagiri," Spolia Zeylanica,
Vol:XXVI,Partl, 1973,p.71.
 313

present and the grains of sand in the mix <!J)_pear to have been sieved_., for they are finer than

in the ground of type two. Flakes of mica are also present in this ground._ The average

thickness of the cor;n_plet~__ground is 6.1 mm and the surface is trowelled. The _ground of

type two in the lower areas of the painting is more compact than grO\lild_one._ The sq,rface

is also not trowelled and does not bear an overall pr~aration for the_paintip_g. Th~_ground

is in two layers, the lower layer being 5 S mm and the upper layer 7.3 mm in thickn~ss; it

is buff in colour and consists of a cl~ sand mixture containigg no o_!ganic fiores oryadd__y
110
husks.

Though it is not clear at both the sites of Sigiriya and Hindagala, anthill cla:y may

have been used. The saliva content exuded Q)' the ants im_parts to the clay an adhesive

quality. The plaster mixed with this. clay affords better internal circulation of air owi~g to

the minute ppepiQgS created within and h,ejps then~Q)r better__preservation of the_plaster

from desiccation and consequent disintegratiofr. The uppermost surface re~ins

comparativ.ely rougher and the whole composition has _hardened like cement. 1HUnlike at

Sigiriya, at Hindagala, a very thin white coating composed of clay instead e£ the- usual
112
lime is l3;id on the ferruginous clay ground. Apart from .the composition of-this coat of

109
VA Smith, A histoty of fine arts in India and Ceylon, DB Taraporevala and sons, Bombay, 1969, p.IOO.
110
Raja de Silva, "Painting: Early period 247 BC to 800 AD," Archaeological De.partment centenaxy
( 1890-1990) commemorative series, Volume five - Painting, ed. Nandadeva Wijesekara, State Printing
Corporation, Colombo, 1990, p.31.
111
Nandadeva Wijesekara, Early Sinhalese painting, Saman Press, Maharagama, 1959, pp.78-79.
112
The lime required for the preparation of plaster would have been taken from two sources, either from the
sea as shells or as corals of geological origin. Analysis of the painting ground from various sites shows a
very small ratio ofMgO: CaO and it seems improbable that the mineral limestone was the source oflime for
these ground. The very small ratio of Magnesium Oxide to Calcium Oxide in Sri Lankan lime plaster can be
accounted for if coral and other aragonitic forms of shells were the source of Calcinm carbonate for burning
lime. Nimal de Silva, "Studies in the techniques of ancient paintings in Sri Lanka," Perspectives in
archaeology: Leelananda Prematilaka festschrift 1990, ed. S Senaviratna and othe1"S, Department of
Archaeology, University ofPeradeniya, 1990, p.l30. For further details of obtaining the lime for grounds
see Raja de Silva, "Painting: Early period 247 BC to 800 AD," Archaeological De.partment centenmy
 314

lime wash, the method of layii_!g the whole _ground is similar to that found at Dam bulla

which is now in an exposed condition. 113

At V essagiriya, the first layer of ground is white lime sand mix with a trace of clay

varying from 0.4 to 1.0-mm thickness accordin_g to the contours ofthe rock surface: This

coat was suppressed by a thin white coating of lime in thickness of0.8 to 1.0 rnm._ with a

very small p_ro_portion ofcll!)'. Unlike other site~ mal!)' areas of__paintin_g do not have the

first ground layer, but only a thin coating oflime laid direct on the rock. Nevertheless, it is

evident tha~ in certain areas, there ..are .three layers ofthin coating. 11 ~esides, at Gonagolla,

the ground consisted of two layers based on lime. The first coat varied from 2.0 to 2.~ mm

in thickness and consisted of limeJ sand and a little cl~. The second l~yer ~_purer in

lime 0.8 to 1.0 mm in thickn~containing a little sandand trowelled smooth to receive

the paints. 115

The Pulligoda paintings were done on a thin coating oflime with an average of0.8

mm spread on a layer of brown cley-san<t chQID>ed straw and_pad<!y husks in thickness of

3.0 to 7.9 mm. There was a greater proportion of paddy husks. and straw on the s~ce of

the ground than in its body and flakes of mica were__present The interface between the

brown coloured ground and the white coating_ was. irregular_ It is evident that the ~rface

of the coating was gene~llY worked smooth but was uneven to a ~light extent co,;n_pared

0890-1990) commemorative series, Volume five- Painting, ed. Nandadeva Wijesekara, State Printing
Corporation, Colombo, 1990, pp.35-36.
113
RH de Silva, "The evolution of the techniques of Sinhalese wall painting and comparison wi~ Indian
painting methods," Ancient Ceylo!!, Journal of the Archaeological Swvey of Ceylon, No.I, 1971, p.93;
"Rock painting in Sri Lanka," Conservation in archaeology and the applied arts, (Reprinted of the
contribution to the Stockholm Congress), June 1975, London, pp.69-73.
114
Raja de Silva, "Painting: Early period 247 BC to 800 AD," Archaeological Department centenmy
( 1890-1990) commemorative series, Volume five - Painting, ed. Nand adeva Wijesekara, State Printing
Corporation, Colombo, 1990, p.32.
m Ibid, p.32.
 315

with other places. Hence, this shows a reluctance or lack of ex_perience·in usi:Q.g a thicket

coat oflim~ ground.ll 6At Mahiyangana, the .ground of the paintin.g _is .in three layers, the

first is a brown mixture of clay and sand admixed with paddy husks, varying between 3. 7

and 6.2 mm in thickness. The second_l~yer was a_l)ale buff coloured lime-cll!Y and sand

layer with a few plant fibres included. This was 0.41 to 1.0 mm in thickness. The third

layer comprised of a white co;:tJ;ing of lime sand with a trace of clay 0.8 to 1.2 mm thic~

the surface being trowelled smooth and hard so as to expose occasionally the grains of the

lower layer oflime plaster. 117

In contrast, it was found that at early sites like Sigiriya, Dambulla and

Maraveediya, paintings were mainlY on cll!Y-based_plaster as in the case of ~anta and

Bagh. In places like Mahiyanganaya and Pulligoda the first ground layer was of clay with

the outer layers containing more lime. It is interestin_g to note that the_plasters at the caves

of boulder garden and Cobra-hood cave of Sigiriya, Gonagolla, Mihintale, vestibules oj

Ruvanvalisaya, Tivamka im!:!Be house and Lankatilaka were main!y lime-based_plasters.

However, it is to be noted that though clay plaster was in continuous use, special

significance may be attached to the use of mud plaster in early samples 11pto _about 8tll

century AD, due to the presence of paddy. husk and chopped straw, not found in mud I

plaster applied after the 12th century AD. 118

Apart from these rock surfaces, on the stone pillars and other covered surfa~es oj

stone monuments and in vestibules of the stt,Ip~ a ve_ry thin coating of__plaster was

applied. In fact very little of this composition remainS- except in a few fragme9ts a1

116
Ibid, p.32.
117
Ibid, pp33-34.
 316

Galvihara and Ruvanvalisrora. It is evident that in all these cases on_!y one layer of fine

lime and sand plaster was used. Over this was laid a very thin upper surface and the whole

layer was not thicker than half an inch at the.most. 11 9_N"ev.ertheless, the plas.ters applied on

the brick built walls at Polonnaruva differ from those on stone pillars, rock and cave walls

so far discussed .. It is obvious that i~ _guali!Y~ it is inferior and has nowhere received the

same minute degree of attention as in the earlier stone surfaces. The first layer covering

the brick wall consisted oflime_plaster madeu_p of sand_, clay and lime mixed to a suitable

consistency. The clay is a greyish alluvium deposit. The plaster is thick in places but

varies between half to three _guarter inches. The surface th~ _pr~ared has been covered

with another thin layer of fme lime not thicker than one sixteenth of an inch. This final

coating was smoothed over with the trowe1. 120

Accordingly, it is clear that the outermost layer that received the pigments in all

the ancient plasters ofthe SriLankan murals upto the end of Uth centnry AD was prepared

in lime as. in the case of Ajanta and Bagh, except at Hindagala where instead a layer of

white clay has been used as the outetyigment receivi~ l~er. It is also clear that there is

no essential difference in the methods and materials of the wall paintings even in the later

phase of th~ period concerned thq_1_!8h the tim~.P is considerable since it runs "!:l.P to the

12th century AD. Hence in Sri Lankan context i.t is .obviollS that there is ..a .continuity of

technique from the earli.esttimes of.tbe historical perio.d up to the 12th or 13th .centuries

AD. The only conspicuous difference in this last phase Is that 1n addition to the rock

118
Nimal de Silva, "Studies in the techniques of ancient paintings in Sri Lanka," Pers.pectives in
archaeology: Leelananda Prematilaka festschrift 1990, ed. S Senaviratna and others, Department of
Archaeology, University ofPeradeniya, 1990, p.l29.
119
Nandadeva Wijesekara, Early Sinhalese painting, Saman Press, Maharagama, 1959, p.79.
120
Ibid, p.79.
 317

supports, there survive paintings on brick walls. 121 Besides, it is to be noted that the

paintings on the brick walls of the relic chambers at Dadigama which belong to this last

phase are unusual in technique and do not find a counterpart anywhere in the island. It has

been found that this ground was of crushed brick, a material that has also been used as a

plaster in the contemporary Baddasimapasada or the chapter house in the Alahana

parivena in Polonnaruva. It was a few millimetres thick and the superficial coating

consisted of white clay122as in the case of Hindagala. interestingly enough, in all these

cases, organic materials, which would have served to consolidate the ground and act as a

binding medium were also used, which would dry and impart strength to the whole

ground. This binding medium has been shown to be a mixture of an emulsion of a plant or

vegetable gum and a drying oil. 123

It has to be noted at this point that as mentioned in the third chapter, it is recorded

in the Vinaya Pitaka that the Buddha allowed walls of dwellings to be coloured and

paintings of flowers and creepers were permitted. 124 In addition, chapter six of the

121
RH de Silva, "The evolution of the technique of Sinhalese wall painting and comparison with Indian
painting methods," Ancient Ceylon, Journal of the Archaeological SUIVey of Ceylon, No.1, Janwuy 1971,
p.94; "Rock painting in Sri Lanka," Conservation in archaeology and the applied arts, (Reprinted of the
contribution to the Stockholm Congress), June 197 5, London, pp.69-73. For the detailed accounts of the
techniques of the paintings of Sri Lanka during the ancient period, as revealed by the scientific tests please_
refer to Raja de Silva, "Painting: Early period 24 7 BC to 800 AD," Archaeological Department centenary
( 1890-1990) commemorative series, Volume five - Painting, ed. Nandadeva Wijesekara, State Printing
Corporation, Colombo, 1990, pp.21-25 and 31-34.
122
RH de Silva, "The evolution of the technique of Sinhalese wall painting and comparison with Indian
painting methods," .Ancient Ceylon, Journal of the Archaeological SUIVey of Ceylon, No.1, Janwuy 1971,
p.94; Nanda Wickramasinghe, "Mural paintings: 800 AD and 1200 AD," Archaeological Department
centenazy (1890-1990) commemorative series, Volume five- Painting, ed. Nandadeva Wijesekara, State
Printing Corporation, Colombo, 1990, p.65.
123
RH de Silva, "The evolution of the techniques of Sinhalese wall painting and comparison with Indian
painting methods," Ancient Ceylon, Journal of the Archaeological SUIVey of Ceylon, No .I, 1971, pp.92-93;
"The technique of ancient Sinhalese wall painting - Sigiri," Paranavitana felicitation volume, ed. NA
Jayawickrama, Gunasena, Colombo, 1965, pp.96-99; Raja de Silva, "Painting: Early period 247 BC to 800
AD," Archaeological Department centenary ( 1890-1990) commemorative series, Volume five - Painting,
ed. N andadeva Wijesekara, State Printing Corporation, Colombo, 1990, pp.l-3 9.
124
The Book of Discipline (Cullavagga), ed. LB. Homer, Luzac, London, 1952, Vol. V, p.212.
 318

Cullavagga. an early Buddhist canonical text, which forms a part of the Vinaya, records

some of the materials of construction that may be used for Buddhist temples. In

accordance with these stipulations, the paintings within caves were done on a ground

consisting of a clay mix containing paddy husks. An organic binding medium was also

added for strengthening the earth layers and causing them to adhere to the rock

support. ~ addition, in order to enable pigments to adhere to the surface ofthe plaster, it
1

is recorded in the Vinaya Pitaka that the Buddha said "anujtiniinii bhikkhave

stisapakuddam sitthatelakam. " 1 ~omer has translated this short statement as "I allow,

monks, mustard powder, oil of beeswax." 127 Consequently, this mixture has been

identified as "a paste made of mustard seeds and oil of beeswax." 128 Besides, the

Dammapaddattakatha or the Commentary on the Dhammapada of fifth century AD also

states that the vegetable fibres in the form of rags cut into small fragments were mixed
129
with the ground clay.

Although these ingredients were recommended by the early Buddhist texts,

according to the above analysis it is evident that the ancient Buddhist painters of both

countries do not seem to have strictly adhered to these ingredients or the rules. But, it has

to be admitted that with the exception of the binding medium of the plasters, which is

made of mustard seeds and oil of beeswax other details of the preparation of grounds

adopted by the Buddhist painters of both countries were quite consistent with these

125
The Vinaya Pitakam, ed. H Oldenberg, Williams and Norgate, London, Vol.II, 1880, pp. 4 and 151;
Vinaya Texts, ed. TW Rhys Davids and H Oldenberg, Sacred Books of the East, Vo1.XX, Oxford, part Ill,
1885,p.l70.
126
The Vinaya Pitakam, ed. H Oldenberg, Williams and Norgate, London, Vol.II, 1880, p.151.
127
ill Homer, The Book of the Discipline. (Cullavagga) Luzac and co, London, Vol V, 1952, p.212.
128
Vinaya Texts, ed. TW Rhys Davids and HOldenberg, Sacred books of the east, YoLXX, Oxford, part Ill,
1885, p.171.
 319

Vinaya rules. This may also indicate usage of separate canons of paintings belonging to

the Buddhist tradition which are now lost and were most probably different fr~m the

Hindu canons of painting that present different processes, formulae and ingredients.

The technique:

It is evident that the French term technique is derived from the Greek word

Takhnikos,130which gives the meaning of the method of doing or performing something

especially in art or science. 131In relation to the sphere of art, the technique of an artist may

apply either to the medium or method in which he works or of his particular treatment of

the material he is using, his lighting or general approach to the work of art. 131'hus it is
133
evident that the technique of a painting is almost dependent on its support or ground i.e.

processing of the plasters and the method of application of the colours on them. In this

context, it is significant that though terms such as "Ajanta frescoes" or "frescoes of

Ajanta" and "Sigiri frescoes" or "frescoes of Sigiri" 134 and so on have been used by

129
Dhammapadattakatha, ed. HC Norman, Pali Text Society, London, 1912, p.219; EW Burlingame,
Buddhist legends, HaiVard University Press, 1921, Volll, p.25.
130
The Pergamon dictiomuy of art, ed. John Fi1zmaurice Mills, Pergamon Press, Oxford, London, 1965,
~.162.
31
Joyce M HawKins and others, The Oxford study dictionaxy, Penerbit F ajar Bakti, SDN, BHD, Kuala
Lumpur, 1994,p.713: ·
132
The Pergamon dictionaxy of art. ed. John Fi1zmaurice Mills, Pergamon Press, Oxford, London, 1965,
f.l62.
33
Kurt Herberts, The complete book of artists' techniques, Themes and Hudson, London, 1958, p.23.
134
See Portfolio of Buddhist art: Historical and modern, Panl Carus, The Open Court Publishing co,
Chicago, 1906, p.2; VA Smith, "Indian paintings at the festival of empire, 1911," The Indian Antiquaxy, ed.
Richard Carnac and others, VoLXL, 1911, p.297; G Yazdani, "The wall paintings of Ajanta," The Journal of
the Bihar and Orissa Research Society, Vol.XXVII, 1941, pp.6-33; Mukul Chandra Dey, My pilgrimage to
Ajanta and Bagh. Gian Publishing House, Delhi, (reprinted) 1986, pp. 27, 236,237 etc; TN Mukhmji, Art
manufactures of India, Navrang, New Delhi, 1974, p.l3; Moni Bagchee, Ananda Coomaraswamy: A study.
Bharatha Manisha, Varanasi, 1977, pp.27 -28; WE Gladstone Solomon, Jottings at Ajanta, The Times Press,
Bombay, 1923, p.26; Karl Khandalavala, Indian sculpture and painting: An introducto:ty study, DB
Taraporevala sons, Bombay, (n.d), p.51; Charls Fabri, ''Frescoes of Ajanta: An essay," Marg, ed. Mulk Raj
Anand, Vol.IX, No.I, December 1955, pp. 61-76; Marguerite Marie Deneck, Indian art, Hamlyan
Publishing Group, Middlesex, 1967, pp.24-25; Romila Thapar, A history of India, Penguin Books India ltd,
 320

almost all scholars without any explanation there has been much discussion regarding the

exact process of murals of these sites and much difference of opinion has been expressed

by the experts.

It is evident that among the types of mural painting, Indian silpa texts refer to only

two types of mural painting, tempera and fresco secco and in actual execution of the

subcontinent too, it was restricted mainly to these two varieties of work. 135 Consequently,

art technologists have argued whether the murals of Ajanta and Bagh are· fresco (i.e. al

fresco; fresh or wet ground), fresco buono (i.e. huon fresco; genuine wet ground) or fresco

secco (i.e. dry fresco; process of painting on dry plaster) and tempera (painting with

pigments mixed with chalk or clay and diluted with weak glue or size) etc. The opinions

expressed by scholars like Griffiths, 136 Coomaraswamy, 137 Goloubew, 138 Rowland, 139

New Delhi, Vol.I, 1966, p.190; Krishna Chaitanya, A history of Indian painting: The mural tradition.
Abhinav Publications, New Delhi, 1976, p.50; HCP Bell, "Interim report on the operations of the
archaeological swvey at Sigiriya in 1895," Journal of the Royal Asiatic Society (Ceylon Branch), VolXIV,
No.46, 1895, pp.56-60; "Interim report on the operations of the archaeological swvey at Sigiriya (Second
season) 1896," Journal of the Royal Asiatic Society (Ceylon Branch), VolXIV, No.47, 1896, pp.254-258;
"Interim report on the operations of the archaeological swvey at Sigiriya, 1897 ," Journal of the Royal
Asiatic Society. (Ceylon Branch), Vol:XV, No.48, 1897, pp.98, 107-113; "Paintings," Archaeological
Swvey of Ceylon Annual Reports, 1905, p.l6; "Demalamahasaya paintings," Journal of the Royal Asiatic
Society (Ceylon Branch), VolXXVI, No. 71, 1918, pp.199-20 1; MD Raghavan, "The Sigiriya frescoes,"
Spolia Zeylanica. ed. PEP Deraniyagala, Government fuss, Colombo, 1948, pp.65-72; William E Ward,
"Recently discovered Mahiyangana paintings," Arbbus Asiae, Vol.XV, Nos. 1&2, p.109; WB Marcus
Fernando, Sigirjya. Archaeological Department, Colombo, 1967, p.4; Argus John Tresidder, Ceylon: An
introduction to the resplendent land. D van Nostrand co, Canada, 1960, p.200; Roland Silva and others,
Cultural treasures of Sri Lanka. Minisby of Cultural Affairs, 1973, (no page numbers), See painting section;
Ravindra Pandya, Cave paintings of Sri Lanka, Art Centre, Pragati Society, Ahmadabad, 1981, p.37;
Nandana Chutivongs and others, Paintings of Sri Lanka: Sigiriya. Archaeological Swvey of Sri Lanka,
Centenary publication, Central Cultural Fund, 1990, p.44; Rhys Davids, "Sigiri, the lion rock near
Pulastipura, Ceylon and the thirty-ninth .chapter of the Mahavamsa," Journal of the Royal Asiatic Society of
Great Britain and Ireland. New Series, Vol. VII, pp.209-211; K Krishna Murthy, Buddhism in South Asill,
Classical Publishing co, New Delhi, 1990, p.63; Nadun, "Mural painting," Viskam (Creativity), Souvenir on
the occasion of the 5lh Non-aligned Summit Conference held in Colombo, Sri Lanka, 9-20 August 1976,
f:.56; Sushila Pant, The art of Ceylon, (no publication data), 1970, pp.20-26.
35
Jayanta Chakrabarti, Techniques in Indian mural painting, KP Bagchi and co, Calcutta, 1980, p.l4.
136
Griffiths believes that the Indian practice of wall painting at Ajanta as elsewhere is in fact a combination
of tempera with fresco. He concludes that the hydraulic nature of Indian lime makes it possible to keep a
surface moist for a longer time than in Europe-produces a closer and more intimate liaison between the
colour and the lime and a more durable and damp-resisting face than the open texture of European fresco.
 321

142 d
Have11, 140.-..h
n avnam,· 141Cha1tanya
·
an HaI01·143are conspicuous
.
among such arguments.

Of these opinions, it is obvious that the most popular one was that the ancient paintings of

He further believes that this art has been practised all over India since the time of the Ajanta paintings. John
Griffiths, The paintings in the Buddhist cave temples of Ajanta Khandesh, 1896, Caxton Publishers, Delhi,
(reprinted) 1983, Voll,pp.l7-18.
137
Coomaraswamy believes that though the method was fresco, processing of painting at Ajanta differed
chiefly from Italian fresco in the greater length of time the Indian lime remains damp and in the fact that the
surface, after colouring, is burnished with a small trowel, by which process the colour is deeply ingrained.
AK Coomaraswamy, The arts and crafts of India and Ceylon, Today and Tomorrow's Printers and
Publishers, New Delhi, 1913, p.87.
138
Goloubew in his monograph on Ajanta says that the paintings are true frescoes, though some of them
have been finished or retouched by a process analogous to tempera. See V Goloubew, "Documents pour
servir a I' etude d'Ajanta, Les peintures de la premiere grotte" Ars Asiati9!, Vol. X, Paris and Brussels,
1927.
139
Rowland believes that although Indian wall paintings can never be described as frescoes, in the true sense
of the word, it is notable that the plaster ground was kept moist during the application of the pigments.
Benjamin Rowland, The art and architecture of India Buddhist. Hindu. Jain, Penguin Books, USA, 1953,

f,w ~~ell thinks that the process employed was usually that which is known in Italy as fresco buono, in
2

which colours mixed with lime-water are applied to a prepared surface of the finest plaster, while it is still
wet, so that they are chemically united with the ground. Indian fresco buono when the wall is a suitable one
is an exceedingly permanent process for interior decoration and much more durable in a tropical climate
than oil painting. Nevertheless, as it was largely used in exposed situations or in buildings that were not in
themselves of a permanent kind. In fact, very few of the early Indian fresco paintings have survived and
Ajanta paintings seem to be a combination of fresco buono and fresco secco. According to him this view is
supported by very good historical evidence, the hereditary craft-tradition of Indian mural painting. The chief
difference between Indian fresco and the Italian process of fresco buono is that the colours are united to the
plaster ground by mechanical action (beating with trowel and polishing) as well as chemically, by the action
of the lime. The Indian fresco is given a highly polished surface, which in the dust-laden atmosphere of
India is a great advantage, as it prevents accumulations ofdirt and enables the wall to be cleaned with a dty
duster or by syringing it with water. EB Havell, The art heritage of India comprising Indian sculpture and
painting and ideals of Indian art, DB Taraporevala and sons, Bombay, 1908 and 1964, pp.65, 72 and 110.
For the account of another variety of fresco painting of India, see Jayanta Chakrabarti, "Fresco-secco
painting in India: A fresh technical scrutiny," As.pects of Indian art and culture, SK Saraswati
commemoration volume, ed. Jayanta Chakrabarti and DC Bhattacharya, School ofHistorical and Cultural
Studies, Calcutta, 1983, pp.l54-161.
141
Bhavnani believes that ancient painters of India practised both techniques of tempera and fresco paintili.g
methods. Enakshi Bhavnani, Decorative designs and craftsmanship of India, DB T8raporevala sons,
Bombay, 1969, p.15.
142
Chaitanya believes that Ajanta painting is not true fresco, as the colours were not applied when the lime
was wet Krishna Chaitanya, Arts of India, Abhinav Publications, New Delhi, 1987, p.48.
143
Haloi states that on close observation of the murals at Ajanta, it becomes obvious that there are no traces
of joining of plaster-grounds, as in the case of fresco buono and he further describes that the colours used
were opaque or semi-opaque in nature and mostly of mineral origin. For making different shades, two or
more colours were mixed together. For tonal variations and opacity, white was often mixed with other
colours. In the Ajanta paintings, for obtaining white colour, lime has been used as an inert pigment and
mixed with other colours when found necessary, this has contributed to the resulting brilliancy.
Accordingly, it is clear that Ajanta murals as that of gouache medium painted on dried mud plaster ground,
instead of tempera, fresco buono or secco mentioned in many case studies. Ganesh Haloi, "Ajanta painting:
Its technique and execution," The art of Ajanta: New pers.pective, ed. Ratan Parimoo and others, Books and
Books, New Delhi, Vol.II, 1991, pp.366-367.
 322

India have been done according to the fresco method or following some kind of fresco

technique. 1M-under this impression, some scholars have gone to the extent to state that

Indian painting, at least what remains of it, was till the 13th century AD an art of fresco

painting or variation of fresco process. 145 But, others have concluded that in the long

history of extant Indian mural paintings, from the time of Ajanta to the 19th century AD

the most common medium has been tempera. 1 ~ence, it is essential to examine at this

point what the conspicuous characteristics of a fresco painting or the fresco process are in

contrast to the tempera technique.

It is obvious that the term 'fresco' is derived from the Italian term connoting
147
fresh. Although this method has been practised in Minoan and other ancient
148
civilisations the technique that we now call 'true fresco' that is 'huon fresco' was first

perfected as a practical theory in Italy in about 1300. 14~t denotes the most durable form of

art known to Europe where a piece of painting must be completed on a wet and freshly laid

144
Rama Sivaram, "Jnflnence of Ajanta in the Deccan and in south Indian painting," The art of Ajanta: New
persJ>ectives, ed. Ratan Parimoo and o1hers, Books and Books, New Delhi, Vol.II, 1991, p.468; VA Smith,
A histmy of fine arts in India and Ceylon, DB Taraporevala and sons, Bombay, 1969, p.89; Phanindra Na1h
Bose, Principles of Indian silpasastra with the text ofMayasastra, The Panjab Sanskrit Book Department,
Lahore, 1926, p.86; Lamence Binyon, Painting in the F areast: An introduction to 1he histoty of pictorial art
in Asia, Edward Arnold, London, 1913, p.36; A Foucher, The beginnings of Buddhist art and o1her essays in
Indian and Central Asian archaeology, (tr. LA Thomas and FW Thomas), Indological Book House
Varanasi, 1914, (reprinted) 1972, p.I60.
145
Karl Khandalavala, Indian sculpture and painting: An introductozy study, DB Taraporevala sons,
Bombay, (n.d), p.5l.
146
Jayanta Chakrabarti, Techniques in Indian mural painting, KP Bagchi and co, Calcutta, 1980, p.l4.
147
Joyce M HawKins and o1hers, The Oxford study diction my, Penerbit F ajar Bakti, SDN, BHD, Kuala
Lumpur, 1994, p.27l.
148
The Pergamon dictionruy of art, ed. John Fitzmaurice Mills, Pergamon Press, Oxford, London, 1965,
p,.70.
49
See Mrs Merrifield, The art of fresco painting as practiced by the old Italian Spanish masters with a
preliminruy inquity into the nature of the colours used in fresco paintings with observations and notes, Alee
Tiranti ltd, London, 1952, See introduction; For a discussion of fresco paintings in Italy during the late
middle ages and 1he renaissance please refer to Millard Meiss, The great age of :fresco: Discoveries.
recoveries and survivals, Phaidon Press, London, 1970, See introduction, pp.l3-21.
 323

stucco ground before another piece of plaster is prepared. 15~t consists of painting with

lime-resistant colours (not chemically affected by lime) on damp lime-plaster, i.e. plaster

that has not yet set. In this process as the plaster can only be painted on in this state the

painter divides his work into so-called 'day-pieces,' each day piece being the area, which

he can finish in one day. Thus, in 'huon fresco' nothing can be added or altered after the

plaster has set This time factor gives fresco painting an extraordinary vitality because it

means that the brushwork must be quick, the forms monumental and the range of colours
151
limited. In addition, when the plaster sets, the particles of colour crystallise into the wall

and remain permanently fused with it They cannot flake off as they can in secco method;

the fresco can only be damaged if the wall decays in some way, by efflorescence, for

instance, or if the plaster flakes off. The relevant plaster layers must therefore be very

carefully built up. 152

The term fresco-secco means painting on lime plaster, which has set unlike the

genuine fresco. In fact, it is a technique of great antiquity the pigments being ground in an

aqueous binding medium. In this technique, the completely dry lime plaster wall is

thoroughly saturated with lime water (or baryta water) and left overnight The painting

then takes place on a moist surface as in fresco, except that the colours are mixed with a

solution of casein glue or egg yolk, instead of being ground in water only. This is the

traditional meaning of secco, but in a wider sense, as in its original Italian meaning, it can

150
For further details of process of fresco see Encyclopaedia of paintings, ed. Bernard SMyers, Hutchinson,
London, 1956, pp.l85-186; The Pergamon dictionazy of art, ed. John Fitzmaurice Mills, Pergamon Press,
Oxford, London, 1965, p.70.
151
Kurt Herberts, The complete book of artists' techniques, Themes and Hudson, London, 1958, pp.42 and
47.
152
Olle Nordamark, Fresco painting: Modern methods and techniques for painting in fresco and secco,
American Artists Group, New York, 1947, pp.9-31; Kurt Herberts, The complete book of artists'
techniques, Themes and Hudson, London, 1958, p.44.
 324

indicate any dry (as opposed to fresco or fresh plaster) method of wall painting, including

for instance, work such as tempera painting upon a dry wall. 153Nevertheless, it has to be

admitted that the so-called secco technique or painting on dry plaster has always existed

side by side with fresco method. It is evident that in this process, to ensure that the

pigments will fuse completely with the ground as they do in 'buon fresco,' the surface is

thoroughly soaked and washed down several times with slaked lime to which some fine

river sand has been added. The paint must be applied while the slaked liine is in this

viscous state. The execution should be free, dashing and a good deal of skill is required

not only in the handling of the paintbrush but also in making the best of a restricted palette

and equally restricted range of expression. 154

Accordingly it is clear that though the process is somewhat similar fresco secco is

in some ways easier to paint than genuine fresco. In this technique, the artist is neither

limited to colours which resist the action of lime, nor is he tied to the "day-piece."

However, the various methods of secco painting depend on whether size, casein or

distemper is used for the ground for the vehicle: on these in tum depend the quality and

intensity of the colours. 15 sm fact, innumerable variations are possible according to the
combination of ingredients: the grain ofthe ground, the nature ofthe vehicle (which can

make the colours transparent or opaque), the colours themselves and all these combined to

153
The Pergamon dictionruy of art, ed. John Fi1zmaurice Mills, Pergamon Press, Oxford, London, 1965,
f.l54.
54
See Mrs Merrifield, The art of fresco painting as practiced by the old Italian Spanish masters with a
preliminazy inquity into the nature of the colours used in fresco paintings with observations and notes, Alee
Tiranti ltd, London, 1952, p.xi; Kurt Herberts, The complete book of artists' techniques, Themes and
Hudson, London, 1958, p.40; See also JW Mollett, Illustrated dictionruy of words used in art and
archaeology, Sampson Low co, London, 1882, p.l46.
155
See Olle Nordamark, Fresco painting: Modern methods and techniques for painting in fresco and secco,
American Artists Group, New York, 1947. Specially refer to the details of fresco ground, plastering the
fresco ground and painting the fresco etc, pp.21-37 ,.67-86.
 325

enhance or subdue each other's effect. Consequently mural paintings executed iri secco

method may look very diffe~ent; the technique chosen would depend on the immediate

purpose ofthe work to be carried out. 156

However, it is evident that another, quite dissimilar branch of secco is practised

with opaque distemper. In this technique it is not the grain of the plaster ground, which

constitutes the surface but a soft, matt, almost cloth-like layer of paint, which completely

conceals the texture of the wall. The colours are rich and deep. The influence of the

necked plaster is eliminated in the finished work, so that the whole picture surface must be

painted. Consequently, the painting is isolated from the rest of the wall, assuming an

independence, which may entirely change the character of the room. 157In addition, there is

still another secco techrlique that is particularly conducive to free, large-scale work. With

this method distemper and colours are glazed on to a smooth ground, which has

previously been soaped. The brush glides easily over the surface; there is no roughness to

resist its progress. The brush strokes do not entirely disappear and delicate effects similar

to watercolours can be obtained. 158

Of these few variations, it is interesting that the greatest masters always had a

special predilection for the technique of fresco. Certainly, this process of painting gave

them greater inspiration than any other technique. The question that arises at this point is

why this predilection for fresco? What distinguishes fresco from other methods of

painting? Basically, it is clear that there are advantages to adhering to the technique since

156
Kurt Herberts, The complete book of artists' techniques, Themes and Hudson, London, 1958, p.34; See
also Olle Nordmark, Fresco painting: Modern methods and techniques for painting in fresco and secco,
American Artists Group, New York, 1947, pp. 10-13; AP Laurie, Materials of the painter's craft. TN Folis,
London, 1910,p.86.
157
Kurt Herberts, The complete book of artists' techniques, Themes and Hudson, London, 1958, p.38.
1
~ Ibid, p.38; See also WG Constable, Painter's workshop, New York, 1953, p.60.
 326

fresco is very durable and the colours are extraordinarily luminous, which makes them so

suitable a medium for wall painting. In addition, there is a subjective reason from the

painter's point of view since in fact fresco painting is a challenge to the artist: a 'thrilling

tension persists as long as he paints:' he has a constant struggle with the wall, he t;nust

work quickly and accurately, without doubt he must be absolute master of the technique
159
and every brush stroke must be right the first time.

However, in the context of this study it is obvious that the term fresco is often

incorrectly applied to what are best described as mural paintings of ancient India and Sri

Lanka. lndeed, these Buddhist murals at Ajanta and elsewhere now it seems beyond

doubt, particularly from the evidence of restoration works and the scientific tests, w~re

executed in tempera technique on dry plaster, the layers of plaster being as ~in as
egg-shell in some places. In addition, polished sections of the painted stucco also clearly

show that the pigments stand out as a thin but distinct layer on the ground unlike in the

case of the fresco method. Obviously, no diffusion of colours into the body of the plaster

is noticed; the pigments are neatly superimposed on the surface without permeating into

the body of the plaster. Thus, the form of the pigments, the flaking of colours, the lifting

up of paint-ftlm in the form of cups and the blistering effect, all these show that the

technique of paintings at Ajanta was not that of fresco buono, which is not affected by

160
water since the pigments are not softened by it This conclusion is further confirmed by

the fact that in most cases the insides of the cave surfaces can hardly be smoothed

159
Kurt Herberts, The complete book of artists' techniques, Themes and Hudson, London, 1958, p.43.
160
BB Lal, "The murals: Their composition and technique," Ajanta murals: An album of eigh1y-:five
rg>roductions in colour, ed. A Ghosh, Archaeological Survey of India, New Delhi, 1967, p.55.
 327

sufficiently to take wet fresco. In addition, they sweat a great deal which is not suitable

for the technique of wet fresco either. 161

It may be emphasised at this point that in the process of preparing the ground and

then fastening colours on that ground, the binding medium plays a very significant role in

painting. In fact, as in the case of plaster in the chamcterisation of the technique of a

painting, the nature of the medium is always taken into considemtion and accordingly,

the universally accepted classification such as, oil, watercolour, tempera, ·fresco etc is

generally formulated based on the medium. 162But, unfortunately the binding medium of

Ajanta has now largely perished partly because of autoxidation and partly due to the

depredations by insect-pests. Consequently, the bond between the pigments and plaster

has now become weak and this has resulted in the flaking off of pigments. Under these
163
conditions, the problem of identifying the binding medium even through chemical

analysis is rather difficult and there are almost insurmountable factors that stand in the

way of arriving at a defmite conclusion. However, the results of such scientific

researches also indicate that in the Ajanta mumls glue was the only organic binding

medium164that held the pigments firmly to the ground unlike in the fresco method. Thus,

all these facts would definitely point to a tempem technique at Ajanta. ~evertheless, it
16

is to be noted that the word tempem is loosely used for any method of painting in which

the colours are mixed with such substances as egg white, egg yolk, casein, glue or

161
Asok Mitra, "The Ajanta and Bagh styles," Panorama of Indian painting, Publication Division, Ministry
of Information and Broadcasting, Government of India, 1968, pp.S-1 0.
162
Asok K Bhattachatya, Technique of Indian painting: A study chiefly made based on silpa texts, Saraswat
Libnuy, Calcutta, 1976, p.87.
163
Encyclopaedia of paintings, ed. BernardS Myers, Hutchinson, London, 1956, p.468.
164
S Paramasivan, "Technique of the painting process in cave temples at Ajanta," Annual Report of the
Archaeological Department of His Exalted Highness the Nizam's Dominions, 1936-37, p.30.
 328

gelatine. But, in reality, true tempera is when the colours are ground with egg yolk
166
only. According to this clarification, it is obvious that the technique used at Ajanta was

a kind of tempera, which amounted almost to fresco secco that is to say the pigment was

applied to dried plaster. 167

It has been revealed that the technique ofBagh is also tempera. '168Halder and Khan

pointed out that gum was used in this process as the medium169as in the case of the murals

of Ajanta. In addition, there does not seem to be much difference in the technique of wall

paintings at Ellora and Ajanta too,170though it has been concluded by some scholars that

the paintings of Ellora have neither been executed in the tempera technique nor are they

true frescoes for there was no other binding medium used except lime. 171 Thus, according

to all these descriptions, it is clear that though the term "fresco" is in use even today for the

Buddhist mural paintings of India, undoubtedly this usage is incorrect since the technique

of all these paintings is tempera rather than fresco.

As in the case of the Buddhist mural paintings of India, various views have been

expressed concerning the. technique of wall paintings of Sri Lanka too. A century ago,

Bell had stated that like the so-called frescoes at Ajanta those of Sigiriya are strictly

165
BB Lal, "The murals: Their composition and ~que," Ajanta murals: An album of eighty-five
~roductions in colour, ed. A Ghosh, Archaeological Smvey of India, New Delhi, 1967, p.55.
1 The Pergamon dictionazy of art, ed. John Fitzmaurice Mills, Pergamon Press, Oxford, London, 1965,
Pt.l62.
61
Jeannine Auboyer, Landmarks of1he world's art: The oriental world India and Sou1heast Asia, Paul
Hamlyan, London, 1967, p.28.
168
SK Dikshit, A .guide to 1he Central Archaeological Museum. Gwalior, Department of Archaeology and
Museums, Government ofMadya Pradesh, 1962, p.l9; Sudhakar Na1h Mishra, Gupta art and architecture,
Agam Kala Prakashan, Delhi, 1992, p.191; KK Chakravarti, "A history of wall painting in Madya Pradesh,"
Wall paintings of India: A historical perspective. ed. OP Agrawala, INTACH Conservation Centre,
Lucknow, 1989, p.53.
169
Asita Kumara Haldar, "The paintings ofBagh caves," RYru!m, Vol. VITI, 1921, pp.l3-14; RB Khan,
"Fresco paintings of Ajanta," Journal of Oil and Colour Chemist Assn, January 1949 p.24.
170
MK Dhavalikar, Masterpieces ofRast:rakuta art at 1he Kailas, Taraporevala and sons, Bombay, 1983,
p.35.
 329

paintings in tempera, i.e. the pigments used were mixed with some liquid vehicle and laid
72
on a dry surface/ though a few years later he changed his mind and stated that the latest

pronouncement is that they are true frescoes. 173 Subsequently, Coomaraswamy also

referred to the paintings of Sigiriya as frescoes and appeared to have been guided by the
174
publications of Bell to which he refers. Meanwhile, based on Bell's first statement

Smith also states that the paintings were wrought in tempera on a dry
175
surface. Nevertheless, Havell stated that there is no known process of tempera or oil

painting, which would stand exposure to tropical weather for nearly fifteen hundred years

as those at Sigiriya have done. Hence, he believes that the process employed both at

Ajanta and Sigiriya was some modification of the present Indian fresco
176
process. Subsequently, Dhanapala177and Wijesekara178also suggested that at Sigiriya at

least, true fresco elements are obtainable to a greater extent than at the subsequent sites of

the island, although it was a modification adapted to suit the local conditions. The latter,

further concludes that it is difficult to deny that the process employed at Sigiriya was a

mixture of true fresco and tempera since a careful examination reveals traits of both

techniques. Therefore, he believes that the paintings were not on the whole true fresco but

171
S Paramasivan, "Technique of painting process in the cave temples at Ellora," Annual Rq>ort of the
Archaeological Dej)art:ment of HEll the Nizam's Dominions for 1938-39, Calcutta, 1939, pp.31-38.
172
HCP Bell, "Interim report on the operations of the archaeological SUIVey at Sigiriya (Third season)
1897 ,"Journal of the Royal Asiatic Society (Ceylon Branch), Vol.XV, No.48, 1897, p.114; Archaeological
SUIVey of Ceylon Annual Rq>orts 1897, George JA Skeen, Government Printer, 1904, p.l4.
173
HCP Bell, Archaeological Survey of Ceylon Annual Rq>orts, 1905, p.16.
174
AK Coomaraswamy, History of Indian and Indonesian art, Edward Goldston, London, 1927, p.163.
175
VA Smith, A history of :fine art in India and Ceylon, Second edition, Oxford, 1930, p.109.
176
EB Havell, Indian sculpture and painting, John Murray, London, 1908, p.171.
177
DB Dhanapala, "A short note on the techniques of the Sigiriya pictures," University of Ceylon Review,
Volll, 1944, pp.64-67; The story of Sinhalese painting, Saman Press, Maharagama, 1957, pp.l1-14;
Buddhist paintings from shrines and temples in Ceylon, Colons, UNESCO, Amlicare Pizzi, APA, Milano,
1964, pp.l4-15.
178
Nandadeva Wijesekara, "Comparative study of Sinhalese paintings with contempor.uy Indian
paintings," Ceylon Historical Journal. Vol.l, No.2, 1951, p.97.
 330

a slight modification of the method devised and put into practice by the artists of the
. d.179
peno

It is evident that the later scholars have also continued this approach with slight

modifications. Of these, the opinion expressed by Maranzi, an expert Italian restorer is

180
conspicuous, since he has stated that the technique of Sigiriya is fresco secco and to

confirm this notion he has wrongly concluded that the paintings of Sigiriya are executed

on a thin lime wash laid direct on the rock surface. 181 Similarly, some writers have gone to

the extent to state that the paintings of Sigiriya show affmity to Italian huon fresco 182

while some others have stated that the murals of Sigiriya were painted on the damp walls
183
and it was natural for the colour to spread on such a wall though an observant eye fails to

find such characteristics on the paintings of the site.

Thus, it is evident that there is a difference of opinion among scholars on the

techniques of Sigiriya paintings as in the case of Ajanta murals and therefore, it is

necessary to discuss the issue in some detail. Particularly, Dhanapala is of the opinion that -

blue and green were omitted from the Sigiriya palette and this would be very difficult to

account for if the technique employed were tempera, but is easily explained if the

possibility of the techniques being fresco is allowed. Only natural earths could be used in

fresco painting, but blue and green have to be obtained from sources other than natural

179
Nandadeva Wijesekara, Early Sinhalese painting, Saman Press, Maharagama, 1959, p.83.
180
L Maranzi, Ceylon presetVation of mural paintings, UNESCO, Paris, July 1972, p.4.
181
Ibid, p.4.
182
PEP Deraniyagala, "Some side lights on the Sinhala monastery-fortress ofSihagiri," Spolia Zeylani~
VolXXVI, Part I, 1973, p.72; Ravindra Pandya, Cave paintings of Sri Lanka, Art Centre, Pragati Society,
Ahmadabad, 1981, p.17; John Lindsay Opie, "The mysteries of Sigiriya,"Island Ceylon, Thames and
Hudson, London, 1970, p.l31.
183
Ravindra Pandya, Cave paintings of Sri Lank!!, Art Centre, Pragati Society, Ahmadabad, 1981, p.23;
Nandana Chutivongs and others, Paintings of Sri Lanka: Sigiriy!!, Archaeological Survey of Sri Lanka,
Centenary publication, Central Cultural Fund, 1990, p.44.
 331

earths. Hence, their alleged absence in the Sigiriya painting. The presence of green at the

site has been proved a later addition. As he further concluded, the presence of an alteration

in the position of a hand in one of the figures, which was mentioned in the previous

chapter, is attributed to the employment of a fresco technique. In the genuine fresco

technique the colours are so thin that the altered hand in the figure is still visible in spite of

the frantic efforts ofthe artist to hide it. The artist changed his mind but before he had time

to erase the original hand the plaster dried. 184

In fact, although the blue pigment has not been used in Sigiriya murals m

abundance, 185 it is evident that the green was originally used and this was not a later

addition. Further, Dhanapala is incorrect in stating that the blue and green have to be

obtained from sources other than natural earths: the blue pigment or lapis lazuli and the

green earth or terre verte are naturally occurring materials as their names themselves

suggest. In addition, regarding the presence of alteration in the position of a hand, an

examination of the painting does not disclose any evidence of scraping, over-painting of

background, overall colour or any other attempt to erase the yellow line of the original

hand. No doubt that the artist changed his mind, painted the contours and coloured another

hand (Plate Xll), but if the plaster had dried before the artist had time to erase the hand as

Dhanapala believes, it is difficult to understand how he found the time to finish the

contours of the hand in the altered position and to paint the colours in a fresco

184
See DB Dhanapala, "A short note on the techniques of the Sigiriya pictures," University of Ceylon
Review, Vol.ll, 1944, pp.64-67; The stoty of Sinhalese painting, Saman Press, Maharagama, 1957,
pp.ll-14; Buddhist paintings from shrines and temples in Ceylon, Colons, UNESCO, Amlicare Pizzi, APA,
Milano, 1964,pp.l4-15.
185
1t is evident that one of the ornaments of the ladies on the main rock ofSigiriya has been painted in blue
colour. In addition, the paintings at boulder gardens of the site are also reveal that the blue pigment has
abundantly used.
 332

186
technique. Hence, it is evident that as Coomaraswamy also observed, the basis and

technique of Sigiriya are practically the same as at Ajanta187and as Sana Ullah concluded,

possibly some adhesive substances were also mixed with the composition, but these have

perished by decomposition. However, since the colours were affected by water the

painting method appears to be tempera. 18~t is interesting to note that it has been revealed

by modern scientific tests also that the characteristics of paintings at Sigiriya are

consistent with a tempera technique. 189

Though in all the other painted sites of the island, the method employed was also ·

entirely tempera as in the case ofSigiriya, some scholars believed that the paintings found

in the relic chamber of Mahiyangana stupa 190 and the murals at the Tivamka image
191
house belonged to the fresco type. lt was argued that particularly the paintings of

Mahiyanganaya are in fresco technique, because being a relic chamber and humid like an

Etruscan tomb, it was painted in tempera on wet plaster and the humidity of the

environment facilitated the formation of calcium carbonate fixing the colours in fresco

over a period of months. 192Thus the technique was described as fresco secco finished in

186
For furlher details of analysis against the view ofDhanapala, please refer to RH de Silva, "The evolution
of the technique of Sinhalese wall painting and comparison with Indian painting methods," Ancient Ceylon,
Journal of the Archaeological Swvey of Ceylon, No.I, Januacy 1971, pp.91-92.
187
AK Coomaraswamy, The arts and crafts of India and Ceylon, Today and Tomorrow's Printers and
Publishers, New Delhi, 1913, p.88.
188
Khan Bahadur Sana U11ah, Re,port on the treatment of the Sigiriya frescoes and suggestions for the
preservation of paintings in the various shrines and old monuments in Ceylon, Ceylon Sessional Paper,
No XXI, September 1943, Government Press, Colombo, 1943, p.l; See also WG Archer and S
Paranavitana, Ceylon: Paintings from temple shrine and rock, New York Graphic Society, UNESCO, Paris,
1957, p.23. .
189
R H. de Silva, "The technique of ancient Sinhalese wall painting- Sigiri," Paranavitana felicitation
volume, ed. N.A Jayawickrama, Gunasena & co, Colombo, 1965, p.l 03; Also William E Ward, ''Recently
discovered Mahiyangana painting," Arttbus Asiae, Vol. XV, 1952, p.l09.
190
Nandadeva Wijesekara, Early Sinhalese painting, Saman Press, Maharagama, 1959, p.58.
191
HCP Bell, Annual Re.port of the Archaeological Swvey of Ceylon, 1909, p.l5; DB Dhanapala, The stmy
of Sinhalese painting. Saman Press, Maharagama, 1951, pp.29-39.
192
L Maranzi, Ceylon preservation of mural paintings, UNESCO, Paris, July 1972, p.5.
 333

tempera method193though some early scholars have rightly observed that as in the other

cases these have also been done on dry plaster194using tempera pigments. 195 Thus, it is

obvious that no evidence is available to support the view expressed by some scholars that

the ancient paintings of Sri Lanka are frescoes. Instead, it has been established that the

general technique of the Buddhist mural paintings of Sri Lanka has always been some

kind of tempera as in the case of the Buddhist mural paintings of India. Therefore, it is to

be stressed here again that though the term "fresco" is in use even today for the

identification of paintings at Sigiriya and elsewhere on the island it is a misnomer in every

sense, as in the case of so-called frescos of Ajanta since these were also painted according

to the method of the tempera technique.

The pigments and the binding medium ofthe paints:

According to the above discussion, it is obvious that the paintings of Ajanta have

been executed after applying a lime wash mixed with gum or glue over the mud-plastered

ground. These paintings clearly display a variety of pigments and the most conspicuous

primary paints are yellow, red, blue, white, black and green along with mixtures of these

in various shades. It has been found that these pigments with the exception of black are of

mineral origin and lime resistant in nature. 1960f these, the red and the yellow pigments are

1931bid, p.3.
194
William E Ward, "Recently discovered Mahiyangana painting," Artibus Asiae, Vol. XV, 1952, p.l 09; S
Paranavitana, "The art and architecture of the Polonnaruva period," The Ceylon Historical Journal, ed. SD
Saparamadu, Special number on tlle Polonnaruva period, Vol.IV, Nos. 1-4, 1954-1955, p.88.
195
James Emerson Tennent, Ceylon: An account oftlle island physical historical and to;pogng>hy, Longman,
London, VolL 2nd edition, 1859, p.475.
196
Ganesh Haloi, "Ajanta painting: Its technique and execution," The art of Ajanta: New pe!Sl>ective, ed.
Ratan Parimoo and others, Books and Books, New Delhi. Volll, 1991, p.375.
 334

red and yellow ochres197respectively and lampblack or carbon was used for black colour

extracted from lamp. For white kaolin, lime and gypsum (calcium sulphate) were used and

for green, glauconite i.e. terra verte. Lapis lazuli, a costly semi-precious mineral was used
198
for blue and this alone was imported usually either from Persia or Badakshan as it has

not been reported from the neighbourhood. In contrast, these blues are still exceptionally

fresh and bright, but conspicuously absent in the earliest paintings. ~t is evident that
1

200
historically this lapis lazuli is known to have been used for ornaments too.

However, there is also hardly any evidence for the use of copper compounds such

as malachite for green and azurite for blue. Terra verte, mineralogically called glauconite

is a green complex white ferrous silicate and is a secondary product of weathering of the

basalt Ochrous clays, such as red and yellow ochres were also evidently obtained from

the clayey product of weathering of the rock. The presence of such other pigments as

verdigris, vermillion, cinnabar, orpiment, realgar or red lead has not been established on
01
the basis of chemical, micro-chemical and spot analysis.Z Accordingly, it is evident that

except lapis blue, all the other pigments viz. white, red, yellow and green were locally

available as residual products of the volcanic rock.~us while there is no clear evidence

for the use of any organic colouring-material, whether or not it was used remains an open

197
S Paramasivan, "Technique of the painting process in cave temples at Ajanta," Annual Report of the
Archaeological De,partment ofHis Exalted Highness the Nizam's Dominions. 1936-37, p.29.
198
VA Smith, A histoty of fine arts in India and Ceylon, DB Taraporevala and sons, Bombay, 1969, p.90.
199
Debala Mitra, Ajanta. Archaeological SlllVey of India, Eleventh edition, 1996, p.l2.
200
G. Yazdani, Ajanta: Monochrome reproductions of the Ajanta frescoes based on photography, London,
1930, Voll,p.2.
201
Nevertheless, of these, Smith describes that the yellow of ancient painters is believed to have been
always orpiment, a natural arsenic sulphide. See VA Smith, A history of fine arts in India and Ceylon, DB
Taraporevala and sons, Bombay, 1969, p .90 Haloi believes that bright red from lead which can only be
found in Vessantara Jataka of cave no 17. See Ganesh Haloi, "Aj anta painting: Its technique and execution,"
The art of Ajanta: New peoo>ectiye, ed. Ratan Parimoo and others, Books and Books, New Delhi, Vol. II,
199l,p.375.
 335

question, since even if it was used it must have perished, leaving behind no evidence of its

possible prior existence?03

The colours of the murals at Bagh are also excessively vivid with marked contrasts

in blue, red, green and yellow as in the case of Ajanta. 204Earthen mineral and stone

pigments, identical with those employed at Ajanta are known to have been used. Besides,

it is evident that the application of lac-red was also known at Bagh. 205Thus, the pigments

of the site were vegetable and mineral dyes of which the former is not clearly observable

at Ajanta. However, in contrast, the painters of Ajanta and Bagh commonly accepted

pigments such as yellow ochre, red ochre, carbon and lime. It seems that there was

definite reason behind the use of such a limited range of pigments in Indian murals. As

pointed out by Paramasivan, only those pigments were selected for the wall paintings that

were not sensitive to some minerals, for the painting processes followed by the painters

consistently involved a predominating use of some minerals like lime as evidenced by the

painting process ofEllora. 2~owever, it is evident that the colours at Ajanta can easily be

softened by plain water, as in the case of the ground plaster at the site. This is an indication

of employment of water-soluble binding medium with the pigments, so that the binder

202
BB Lai. "The murals: Their composition and technique," Ajanta murals: An album of eighty-five
~roductions in colour, ed. A Ghosh, Archaeological SlllVey of India, New Delhi, 1967, p.54.
2 1bid, p.54; See also JC Nagpall, Mural paintings in India. Gain Publishing House, Dellri, 1988, p3. Few
other scholars have also given some short accounts of the pigments used by the painters of Ajanta and Bagh.
For instance, see VA Smith, A history of :fine arts in India and Ceylon, DB Taraporevala and sons, Bombay,
1969, p.90; Debala Mitra, Ajanta, Archaeological SlllVey of India, Eleventh edition, 1996, p.12; Ganesh
Haloi, "Ajanta painting: Its technique and execution," The art of Ajanta: New perspective, ed. Ratan
Parimoo and others, Books and Books, New Delhi, Vol II, 1991, p.375.
204
Impey wrongly concludes that this was mainly due to the influence of Greek paintings. See E Jmpey,
''Description of the caves ofBagh in Rath," Journal of the BomblJY Branch of the Royal Asiatic Society,
Vol V, July 1856, p.562.
2
os Asita Kumara Haldar, "The paintings ofBagh caves," Rupam, Vol.VTII, I 921, p.l5.
206
S Paramasivan, "Technique of the painting process in the cave temples at Ellora," Annual Re,port of the
Archaeological Dta>artment ofHis Exalted Highness the Nizam's Dominions, 1936-37, p.38; See also Asok
 336

holds the particles of the pigments together adheres to the ground in form of a separate

layer at the same time?07It has also been revealed that a clear proof of the existence of

vegetable glue or gum as the binding medium has been furnished in some cases by a

careful chemical and microscopic analysis of pigments. This chemical analysis further

shows that animal glue was probably employed. 208

In contrast to Ajanta and Bagh, the range of colours used at Sigiriya is restricted to

a few colours. The dominant colours thus used in the paintings are yellow~ burnt sienna,

brown green and white. Of these, yellow is used in two different tones, that is the golden

hue of the skin of the noble ladies in general and the light creamy colour seen in the blouse

of one of them. Green is usually copper green and rarely the light leaf-green, which is

sometimes seen in the headdress of the ladies. White is often applied to pick up the

ornaments and the white of the eyes, while the outlines of the figures are usually in a

brown tone. Pink appears occasionally in flowers. Grey is rather rare but used in two

tones, light and dark, as seen in the headdress of the lady with the creamy yellow

blouse.Z~ight blue, which has been claimed as not being used in the series210is also

found, applied sparingly. 211 This colour is reserved for special features such as the eyes,

the gems upon the tiaras and the wristlets and neck ornaments etc. Thus it is evident that

beside these primary colours, a range of mixture of colours has also been used at the site.

K Bhattacharya, Technigye of Indian painting: A study chiefly made based on silpa texts, Saraswat Library,
Calcutta, 1976, p.69.
2
m Ganesh Haloi, "Ajanta painting: Its technique and execution," The art of Ajanta: New pers.pective, ed.
Ratan Parimoo and others, Books and Books, New Delhi. Vol.II, 1991, p.366.
208
BB Lal, "The murals: Their composition and technique," Ajanta murals: An album of eighty-five
~roductions in colour, ed. A Ghosh, Archaeological Swvey of India, New Delhi, 1967, pp.54-55.
2 Nandana Chutivongs and others, Paintings of Sri Lanka: Sigiriya, Archaeological Swvey of Sri Lanka,
centenary publication, Central Cultural Fund, 1990, p.43.
210
VA Smith, A histocy of fine arts in India and Ceylon, DB Taraporevala and sons, Bombay, 1969, p.IOO.
211
PEP Deraniyagala, "Some unreconled frescoes from Sigiriya," Journal of the Royal Asiatic Society,
(Ceylon Branch), Vol.XXXVIII, No.1 06, 1948, p.87.
 337

It is evident that the composition of these pigments is as follows: red-red ochre,

yellow-yellow ochre, green- green earth, terre verte or iron silicate and black-carbon. Of

these, it is believed that powdered terra verte imported from India was the green pigment

used. The buff pigments contain iron (as ochre), white material insoluble in

2N-hydrochloric acids and concentrated s~lphuric acid and are probably yellow ochre and
12
kaolin or silica. However, the white is not calcium sulphate.Z Thus, both at Ajanta and

Sigiriya, the principal colours and the materials used are largely the· same: red, yellow,

green, and black. Nevertheless, it is obvious that red, yellow, green, black, white and blue

are the dominant colours used at Ajanta producing a much more varied palette than at

Sigiriya. Thus, the richness of colour at Ajanta becomes obvious when compared to the

yellow-orange-red spectrum of Sigiriya. 213It is also apparent that there is a vast difference

in the use of colours in the portrayal of the human body and nature, between Ajanta and

Sigiriya. 21 ~is difference is further confirmed by the fact that there is an absence of

artificial highlighting of prominent features by using white as applied at Ajanta as

discussed in the previous chapter in detail. Instead, the subtle application of colour as seen

at Sigiriya makes the figure glow as if with light from within. 215

Although Smither has given quite a different list of colours a century ago,
216
specially with relation to the murals of vestibules of the stupas it has been found that the

212
R H. de Silva, "The technique of ancient Sinhalese wall painting- Sigiri," Paranavitana felicitation
volume, ed. N.A. Jayawickrama, Gunasena & co, Colombo, 1965, pp.99-l 00.
213
Siri Gunasinghe, "Ajanta's shadow on Sigiriya," The art of Ajanta: New persj>eCtiyes, ed. Ratan Parimoo
and others, Books and Books, New Delhi, Vol.II, 1991, p.503.
214
Ravindra Pandya, Cave paintings of Sri Lank!l, Art Centre, Pragati Society, Ahmadabad, 1981, p.l7.
m Nandana Chutivongs and others, Paintings of Sri Lanka: Sigiriy!l. Archaeological Survey of Sri Lanka,
centenary publication, Central Cultural Fund, 1990, p.44.
216
Smither describes that the blue was probably indigo, the red vermilion and the yellow orpiment (a
preparation of arsenic and sulphur, producing a gold-coloured pigment). Yellow from its religious
association, was pre-eminently popular with the Buddhists, it is therefore, the predominating colour in all
their decorations, the red taking the next place and .the blue while subordinate to the .others, being often
 338

pigments commonly used for the other early painting sites of Sri Lanka were lime for

white, red ochre, yellow ochre (orange shades being obtained by mixing the colours), the

green terre verte and carbon for black. Mixtures of red ochre and carbon black produced
217
brownish colour. The rare blue pigment lapis lazuli (natural ultramarine) was found in a

few paintings of the early period. Of these, besides the paintings of Sigiriya mentioned

above one of the fine examples is the depiction of the goose and the floral designs found

on the northern and eastern vestibule of the Jetavana stupa, which is now deposited at the

archaeological museum at the site. As in the case of India, this mineral has not been

reported in Sri Lanka and appears to have been obtained via India from Badakshan in
218 219
northeast Afghanistan where the only deposit in Asia is known or from lake Baikal.

As in the case of the method and the materials of the plasters, it is interesting to

note that the pigments used during the last phase of the period concerned i.e. around

eleventh and twelfth century AD are also not different from those used in the early sites,

but the blue lapis lazuli is apparently absent from the painters' palette. In addition, at

Ajanta, Sigiriya and Hindagala the yellow so largely used at Anuradhapura is very rare. It

was found that the other pigments used in the later phase were black containing iron and

mixed with yellow and sometimes with white as well to form green of various shades; white is also
frequently used alone, but principally as a ground work for the paintings. He further concludes that the
colours used appear to have been tempered with gum, to give them body. JG Smither, Architectural remains
of Anuradhapura, Ceylon Government Press, Colombo, 1894, p.42.
217
RH de Silva, "The evolution of the technique of Sinhalese wall painting and comparison with Indian
painting methods," Ancient Ceylon, Journal of the Archaeological Survey of Ceylon, No.I, Janwuy 1971,
pp.93-94. Nevertheless, Nimal de Silva believes that the use of blue pigment in ancient Sri Lanka, the lapis
lazuli, imported from India or Persia. Nimal de Silva, "Studies in the techniques of ancient paintings in Sri
Lanka," Pers,pectives in archaeology: Leelananda Prematilaka festschrift 1990, ed. S Senaviratna and
others, Department of Archaeology, University ofPeradeniya", 1990, p.131.
218
RH de Silva, "The evolution of the technique of Sinhalese wall painting and comparison with Indian
painting methods," Ancient Ceylon, Journal of the Archaeological Survey of Ceylon, No.I, Janwuy 1971,
pp.93-94; Nimal de Silva, "Studies in the techniques of ancient paintings in Sri Lanka," Pers,pectives in
archaeology: Leelananda Prematilaka festschrift 1990, ed. S Senaviratna and others, Departnient of
Archaeology, University ofPeradeniya, 1990, p.131.
 339

silica and red containing iron ochre. 220Thus, in contrast, the four major colours left in

ancient Sri Lankan paintings drawn before the 13th century AD were red ochre, terre verte,

lapis lazuli and in some of the paintings one can see the traces of black pigments. 0( these,

· both terre verte and lapis lazuli were not available in Sri Lanka and would have been

imported from elsewhere as already mentioned-above. It is obvious that due to the scarcity
221
of these materials Sri Lankan artists have used it sparingly. These investigations also

revealed that two sources of obtaining pigments might have been exploited: One was the

mineral pigments and the other a vegetable source. Of these, it is evident that nat;urally
222
obtainable mineral pigments were utilised wherever possible, except in the rare case of

black, which is obtained from ~bon.

When considering the binding medium of paints in Sri ~itis to be norep that

the pigment layer of Sigiriya contains a water-soluble synthesis that hastens out as a white

fluffy material on the addition of excess ethanol Hence it has been concluded that_ it is a

vegetable gum; the other component ofthe binding medium was identified as a drying oil,
223
which was water insoluble, but soluble in ethanol, carbon tetrachloride and chloroform.

Even at the other painting sites oftheisland,_the binding medium ofpaintas iil_the case of

binding medium in the ground was usually an emulsion of plant gum and a drying oil

though in a few paintings, the binding medium was found to consist of a plant gum

219
Georgina Herramann, "Lapis Lazuli: The early phase ofits trade," Iraq, Vol.XXX, No.1, 1968, pp.21-57.
220
RH de Silva, "The evolution of the technique of Sinhalese wall painting and comparison with Indian
painting methods," Ancient Ceylon, Journal of the Archaeological Survey of Ceylon, No.I, January 1971,
g.94.
1
Nimal de Silva, "Studies in the techniques of ancient paintings in Sri Lanka," Pemx;ctives in
archaeology: Leelananda Prematilaka festschrift 1990, ed. S Senaviratna and others, Department of
Archaeology, University ofPeradeniya, 1990, p.l31.
222
Nandadeva Wijesekara, Early Sinhalese painting. Saman Press, Maharagama, 1959, p.81.
223
R H. de Silva, "The technique of ancient Sinhalese wall painting- Sigiri," Paranavitana felicitation
volume, ed.N.A. Jayawickrama, Gunasena & co, Colombo, 1965, p.IOO.
 340

24
only? It has to be noted at this point that from a study of literary documents too it has

been concluded by some scholars that ancient Sri Lankans were aware of a technique of
25
painting in an oil medium? Certainly, it is obvious that if oil had not been used as the

medium for outdoor work, as in the case of the paintings of late medieval period of Sri

Lanka, it is difficult to understand how these could have lasted. ~ence it is obvious that
2

there was an apparent reason for using an oil medium for the wall paintings of ancient Sri

Lanka since almost all the murals were located outdoor, unlike the painting'sites oflndia.

According to the discussion so far, it is clear that there are some similarities as

well as some significant dissimilarities between the two mural painting traditions of both

qountries, where elements such as the supports, grounds, techniques, pigments and the

binding gum etc are concerned. Of these, the difference between the architectural

elements where the paintings have been executed is conspicuous. In addition, apart from

the prepared rock surfaces designed in somewhat dissimilar technique, it was mentioned

above that the ground of the painting at Ajanta consists of a layer of ferruginous clay

mixed with sand, vegetable fibres and paddy husks over which is laid a superficial coating

of lime in preparation for the paints. Accordingly, the technical process of painting was

found to be tempera and the medium soluble in water was detected. It is seen that these

224
R.H de Silva, "The evolution of the techniques of Sinhalese wall painting and comparison with Indian
painting methods," Ancient Ceylon, Journal of the Archaeological SUIVey of Ceylon, No.I, 1971, pp.93-94;
See also "Rock painting in Sri Lanka," Conservation in archaeology and the aj>plied arts, (Reprinted of the
contnbution to the Stockholm Congress), June 1975, London, pp.69-73; Nimal de Silva, "Studies in the
techniques of ancient paintings in Sri Lanka," Perspectives in archaeology: Leelananda Prematilaka
festschrift 1990, ed. S Senaviratna and others, Department of Archaeology, University ofPeradeniya, 1990,
p.l3; Khan Bahadur Sana Ullah, Re.port on the treatment of the Sigiri:yafrescoes and suggestions for the
preservation of paintings in the various shrines and old monuments in Ceylon, Ceylon Sessional Paper,
No .XXI, September 1943, Government Press, Colombo, 1943, p.1.
225
Raja de Silva, "Painting: Early period 247 BC to 800 AD," Archaeological De.partment centenazy
(1890-1990) commemorative series, Volume five- Painting, ed. Nandadeva Wijesekara, State Printing
Corporation, Colombo, 1990, p.4; James Emerson Tennent, Ceylon: An account of the island physical
historical and topograj>hy, Longman, London, Vol.I, 2nd edition, 1859, p.491.
 341

227
technical methods are fairly similar to those adopted at Sigiriya and most of the other

painting sites of the island.

In addition, two layers of ground clay have also been found both in the caves of

Bagh in India and Dambulla in Sri Lanka. It is also evident that as in the case of Sri

Lankan paintings, in Bagh too the base was .provided wit:h_ a Iime-basecl groupd of

appreciable thickness (from 3.4 to 6.5 mm.) reinforced with vegetable fibres?28Hence, it

is clear that there is a considerable similarity between the technique of laying the ground

in Indian painting sites and the earliest Sri Lankan paintings not only with relation to the

murals of Ajanta and Sigiriya, but also to the murals at other sites as well. Another

similarity is that the common earth pigments, yellow earth (yellow ochre), red earth (red

ochre) and green earth (terre verte) are found in both countries, though the blue pigment,

lapis lazuli was used primarily in the Indian paintings and rarely in Sri Lanka since it is

evident that this pigment is known to be present only at a few places of the island. Besides,

though Gypsum was widely used as a white pigment at Ajanta, it has not been found in Sri

Lankan paintings.229

Instruments of the painters:

:

It has not- been possible -to obtairi any direct evidence regarding the -implexpents

used by the ancient Buddhist mural painters of India and Sri Lanka. These being made of

226
AK Coomaraswamy, Mediaeval Sinhalese art, Pantheon Books, New York, 2nd edition, 1956, p .165.
w lUI de Silva, "The evolution of the techniques of Sinhalese wall painting and comparison with Indian
painting methods," Ancient Ceylon, Journal of the Archaeological Swvey of Ceylon, No.1, 1971, p.l 03;
Raja de Silva, ''Painting: Early period 247 BC to 800 AD," Archaeological Department centenary
(1890-1990) commemorative series, Volume five- Painting, ed. Nandadeva Wijesekara, State Printing
Corporation, Colombo, 1990, p.l5.
228
RH. de Silva, "The evolution of the teChniques of Sinhalese wall painting and comparison with Indian
painting methods," Ancient Ceylon, Journal of the Archaeological Slirvey of Ceylon, No.I, 197t, p.l02.
 342

perishable materials have not outlived the paintings. However, of these, no doubt the

painting brushes had a prominent place. The early silpa texts have described separate

brushes of different thickness, broad, fine or medium to be used for each shade of colour
23
and the brush should be shaped like the sprout of a Banyan tree. ~terestingly enough

the brush strokes as evident in the paintings of Ajanta, Bagh, Tivamka shrine and

elsewhere are quite consistent with these classifications.

Beside these brush strokes, it is noteworthy that two actuaL illustrations in the

extant paintings of Sri Lanka of the period concerned also disclose some information

regarding the brushes of ancient painters. These paintings preserve figures of persons in

the very act of portraying figures with the brush.- These examples are seen in the Sasa

Jataka painting at Dimbulagala and the other at Tivamka image house. ln. these paintings,

the god Sakra is depicted in the very act of painting the figure of the hare on the moon. As

evident by the paintings, the brush is a thin object and the method of holding the brush, the

comparative size and shape can also be judged. From the other evidence especially

relating to the medieval period of Sri Lanka, it can be noticed that three types of
231
brushes, sometimes in accordance with the instructions of painting canons mentioned

above, might have been used~ One was- the rough brush madefrom.the aerial roots of trees

and may have been large and crude. The smaller brushes were fashioned out of cat or

ZJ.9lbid, p.l03.
230
Jayanta Chakrabarti, Techniques in Indian mural painting, K.P. Bagchi & co, Calcutta, 1980, p .71; Asok
K Bhattacharya, Technique of Indian painting: A study chiefly made based on silpa texts, Saraswat Libnuy,
Calcutta, 1976, pp.49-51. It is to be noted that according to 1he descriptions given in 1hese silpa texts,
Coomaraswamy has also presented three kinds ofbrushes, which were used by the ancient painters as
coarse, medium and fine. As he has obseiVed, sweeping on is to be done wi1h the coarse one, using it
crosswise; do 1he modelling with 1he medium one and holding the point sideways. Ananda CoomarllS}Vamy,
"The technique and theory of Indian painting," Technical studies in 1he field of fme arts. Fogg Art Museum,
HIUVard University, Vol. ill, No.2, 1934, p.62.
 343

squirrel's hair, the tuft being fastened to a thin wooden holder. The finest brush with a

trailing point suitable for fine line drawing was made ofTelitana grass.

As in the case of the painting brushes, it has not been possible to obtain any direct

evidence relating to the palettes of the painters too. Nevertheless, from the same scene of

the Sasa Jataka painting at the Tivamka shrine it transpires that the palm of the freehand of

the painter himself was used to contain the pigments immediately in use. But, it is worth

noting that the artist DL Perera who made the line drawings in 1909 reproduced the god

Sakra in this 'Sasa Jataka' painting making use of his left palm as a palette. This was

subsequently contested by the artist SM Senaviratna who-copied the drawings in 1969 and

who reproduced a palette more or less like a coconut shell in Sakra 's left hand.

Nevertheless, due to the sorry condition ofthe actual painting, one cannot now judge from
232
an examination of the same painting, which of these two is correct. In addition, it is

evident that clay pots filled with paint, which belong to the later centuries, have· been

discovered within painted cave temples of Sri Lanka. Hence, it is not unlikely that clay

vessels of various sizes contained the other requisite pigments. 233Thus, probably, the palm

of the hand of the artists and the small clay vessels helped to contain the pigments in use as

the palettes of the painters during the ancient period.

Apart from these two main instruments of the painters, it is. evident that atleast the

painters of Ajanta used compasses to draw the numerous circular and geometrical designs

on the canopies on a large scale. The large circles of designs drawn on the canopies of the

cave nos one, two and others of the later period are the best examples for the indication of

231
M. Somathilake, A historical study of mural paintings in Buddhist temples during the 18th and 19111
centuries of Sri Lanka, Unpublished MA dissertation, University ofPeradeniya, 1996, See fifth chapter,
which deals with the techniques of paintings.
 344

using such compasses. 23'1n addition, it is noteworthy that as in the case of the above

instruments, there is no clear evidence regarding the implements that have been used for

grinding the pigments too. However,__pits and holes in the rock floor are seen in ~anta and

afford substantial evidence of the preparation of the pigments by the Buddhist painter who

required his colours in good quantities. 235por instance, at least nine such smalLholes can

be seen close to each other on the left side, between the sidewall and the line of pillars on

the floor of cave no 10 alone (Plate XXIII) thou_gh Spink has incorrectly decided that these
236
have been made by cowboys for sharpening their axes.

Thus, it is obvious that there are similarities as well as dissimilarities and analysis

shows the use of ingredients in one _place that cannot be traced in the other_, especial!}'

materials used for laying the ground and. the- pigments. In this respect, where Ajanta and

Sigiriya do agree is in the use of a final application of an ~shell thin l~_yer o(pure lime as

the priming_ There are also interesting differences in the manner in which they were

employed. From these findings of differences between the two traditions of__paintin__gs_, it

becomes quite clear that while there is a basic similarity especially between Ajanta and

Sigiriya as compared to other sites of the island in the matter of pre_parit!g the _ground,

there are also significant differences that indicate different traditions._ The most important

fmding, in this regard is the use of oil_ b_y the Sri Lankar1 _painters. Unlike at Ajanta_, at

Sigiriya and elsewhere in Sri Lanka a gum and a drying oil have been incorporated in the

mix. At Ajanta, only glue had been used as a bindin_g medium althou_gh in other sites in

232
CE Godakumbura, Murals at Tivamka pilimage, Archaeological Department, Colombo, 1969, p.l7.
233
Nandadeva Wijesekara, Early Sinhalese painting, Saman Press, Maharagama, 1959, pp.81, 96-97.
234
Of these paintings, one of the best reproductions of a canopy painting is deposited at the National
Museum of India, New Delhi.
235
JC Nagpall, Mural paintings in India, Gain Publishing House, Delhi, 1988, p.26.
 345

India vegetable gums have also been used. However, no trace of any kind of oil has been

recognised in these investigations at Ajanta and other Buddhist sites of peninsular India

where lime seems to .be the principal astringent. Hence, the most important technical

difference between Ajanta and Sigiriya and the other places of the island is in the binding

medium each one favoured. It might be suggested that it is the deviations, such as the use

of oil, rather than similarities, which are intrinsic to the technical process and not
7
necessarily borrowed that are significant.23 But, atthe same time it has to be' admitted that

Indian and Sri Lankan wall paintings, dating from the earliest times till about the end of

the 12th century AD are considerably similar as regards the technique of laying the ground
238
and also the pigments employed. This homogeneity of techniques, between .the

paintings ofthe two countries.may probably be. due to the.adoption.ofa common.silpa text

that we have lost today.

236
See catalogue no 515. 66 ofW Spink collection at the photo archives of American Institute of Indian
Studies, Gurgaon, Haryana
237
Siri Gunasinghe, "Ajanta's shadow on Sigiriya," The art of Ajanta: New penmectives, ed. Ratan Parimoo
and others, Books and Books, New Delhi, Vol.II, 1991, pp.502-504.
238
RHde Silva, "The evolution of the techniques of Sinhalese wall painting and comparison with Indian
painting methods," Ancient Ceylon, Journal of the Archaeological Survey of Ceylon, No.1, 1971, p.1 03.