You are on page 1of 13

THE WALL PAINTINGS IN THE BAGH CAVES--AN

INVESTIGATION INTO THEIR METHODS


BY S. PARAlVŸ
(Arch~ological Chemist, GoveŸ Museum, Madra~)
Received August, 1, 1939

I. Introduction
TH~ Bagh caves (22 ~ 22' N. & 74 ~ 48' E.) contain some of the finest examples
of I n d i a n wall paintings. Mutilated and faded as these paintings ale, t h e y
still constitute a priceless treasure comparable to those at Ajanta. T h e y
have a t t r a c t e d considerable attention.1
The caves, which are nine in number, are situated on t h e southern
slopes of the Vindhya hills in the Amjhera District of t h e Gwalior State.
T h e y are 70 miles from Mhow, the nearest Railway station, on the Raj-
putana-Malwa section of the B. B. & C. I. Railway. T h e cliff side on which
the caves are e x c a v a t e d rises to a height of about 150 feet above the Bagh
river and is remarkable as being only the outcrop of sandstone Ÿ an other-
wise basaltic region. These sandstone caves are s u r m o u n t e d by a deep
b a n d of claystone.
The interior of these caves was at one time fully decoratedwith paintings.
But the caves have crumbled due to t h e excessive weight of the super-
imposed band of claystone. Moisture percolating t h r o u g h it has also
c o n t r i b u t e d to their destruction. Thus, with the crumbling of t h e caves
m a n y of. the paintings have disappeared except a few in caves I I I and IV. ~
T h e paintings p r o b a b l y date Ÿ the early seventh c e n t u r y A.D.,
being contemporaneous with the paintings in caves X V I and X V I I at
Aianta?
The paintings of Bagh and Aianta form a distinct class .by themselves,
and belong to the golden age of Indian classical art, which inspired more
t h a n hall the art of Asia.* T h e y will bear comparison with the best t h a t
E u r o p e could produce down to the time of Michael Angelo. 4

* "As faras their artistryis concerned,thereis little to choos~ b3~we~n th~ pic~ures
of Bagh and Ajanta. Both exhibit the same broad handling oftheir subjects, the s~me
poetry of motion, the same wonderful divers[ty in the poses of their figures, the same
feeling for colour and the same strong yet subtle line-work. In both, decorativebeauty

85
86 S. Paramasivan
II. Experimental Investigations
I n reconstructingthe technique of the painting process adopted by the
ancient classical artists at Bagh, one has to consider the four principal
factors5 that go to make up the paintings. They will be considered here
in order. Samples of three kinds of painted stuccoes that were available were
collected from damaged portions of the paintings in cave IV at Bagh, and
expe¡ conducted with them. Two of them had rough plaster of deep
red and light red ferruginousearth, while the third had rough plaster of lime.
Over them was a layer of fine plaster of some white material, the latter
supporting the painted layer.

(I) The Carrier.


The walls of the cave temples, and probably also its ceilings, serve as
the mechnical supports for the paintings, and function as the carriers. They
are of sandstone. The surface of the carrier has been specially left rough
so that the rough plaster might hold well. B The sandstone caves were unable
to support the excessive weight of the superimposed band of claystone.
Consequently, the walls and ceilings, which are otherwise fairly firm and
compact, have crumbled and fallen to pieces, thus destroyingthe paintings.
Moisture has been percolatinginto the cave from the top, and it has softened
and damaged the earth plaster and the paintings attached to them. I n
several samples, the paintings are covered with white ei¡ caused by
the presence of gypsum, sodium sulphate, magnesium sulphate, e t c .
The effiorescence can be traced to salts car¡ over by the percolating
water from outside and deposited over the surface of the paintings, and not
to the presence of salts in the plaster (vide results of analysis below).

is the key-note to whicb_all else is attuned, and both ate as free from realism as t h e y are
from stereotyped convention. The artists, to be sure, have p o rt ra y ed their subjects
direct from life--of t h a t there can be no doubt, but however fresh and vital t h e i r port-
rayal m a y be, i~ never misses t h a t q u a l i t y of abstractionwhieh is indispensable to mural
decoration, as ir is, indeed, t o all t r u l y great painting. True, there is n o t hi ng left a t
Bagh to equal some of the surprisinglymajestic figures, such as the famous Boclhisattva,
a t Ajanta, but in one respect at a n y tate the paintings of Bagh have an a d v a n t a g e o v e r
those of Ajanta. F o r whereas at A j a n t a most of the paintings appear to have been done
p i e c e m e a l - - a t Bagh they give the impression of having been conceived and executed at
one and the same time, of at a n y tate in conformity with a single well-thought out
scheme . . . . To be adequately appreciated t h e y (the paintings) must be seen to their full
scale and in the archi~ec~ural set~ing for which t h e y were d e s i g n e d . " - - S m J o H ~
~~/IARsttALL.
TZide Sir ffohn Marshall & o~hers~ The Bagh Caves ( L o n d o n : I n d i a Socieby),
1927, pp. 17-18.
S. Paramasiva~z Proc. I1zd. A cad. Sci., A , roL X, PL V i l

3. Cave No. 4 at Bagh, Facade (Frescoes are inside the s h u t t e r s )


(With the kind perrnisMon of the Dir~etoe of Arehmology, Gwalioe State)
The H/all Paintings in t/ze Bagh Caves 87

(2) The Ground.


The following experiments were conducted to s t u d y the nature of the
ground which h a d ' b e e n prepared by the Bagh artists to lay the coloured
designs on.

Study of the microsection.--Microsections of the painted stuccoes containing


all the different layers were prepared.7 Since the earth stuccoes were very
soft, considerable care was bestowed in preparing them. Particles of dust
adhering to the prepared surface were removed very carefully with a soft
brush. AII examinatioI1 of the surface under the travelling microscope
revealed three lines of cleavage~ or junctions in the case of the earth
stuccoes and two in the case of the lime stucco thus showing the presence of
four and three distinct layers respectively in them. T h e y collsisted of
the top layer of paint with a layer of s o m e white material serving as fine
plaster just underneath, with two layers of rough plaster ill the case of the
earth stuccoes and olle layer in the case of lime stucco. The thicknesses of
t h e different layers were as follows :

Earth Stuccoes
Lime Stucco
Deep red Light red

lnH1, mm. mm.

PMnted Stucco 8-2O .2 8.1~20.2 3.8-6 -9

Rough P l a s t e r - -
First layer 6.8-19.9 6.9-19.0
1 3.4-6.5
Second layer 1.0 1.0

Fine PIaster 0-1 0.1 0.2

PMnt Film 0.1 0.1 0.2

Average thickness of
of the painted
stucflo 17.2 17.2 4.9

On account of the sof~ness of the earth stucco i~ was difficult to separate the
differen~ layers along the junctions. In the case of the lime stucco, suzh separation
could be effected with greater case,
88 S. Paramasivan

The earth stuccoes are very much thicker than the lime ones. The paint
film in the lime stuccoes is twice as thick as t h a t on the earth stuccoes. I n
other words, the paint film on the lime stucco shows a certain lack of
delicacy in the artists' handling of the brush. Ir is, therefore, probable
t h a t the Iime stucco does not belong to the p a l m y days of Bagh, or t h a t the
paintings on the earth and lime stuccoes were done b y different groups of
artists.
Size of the Particles in the Plaster.--The rough plaster was carefully
separated from the layer offine plaster and the layer of paint film and gently
crushed between fingers so as to avoid breaking the larger particles into
smaller ones. The particles composing ir were mechanically separated and
graded according to size by Robinson's method.S The proportion and
sizes of the particles were as follows : - -

< 200/x, 200/x-700/x >700/z


(Per cent.j (Per cent.) (Per cent.)

Earth Plaster :
Bright red 17 50 33
Light Red 20 40 40
Lime Plaster 6O 40 nil

I n the case of the earth plaster, particles smaller than 200/, are mostly
composed of laterite, silica and clay, while the larger particles ate mostly
of sand and laterite. In the case of the lime plaster, particles smaller than
200 q are composed of lime, clay and sand, while the iarger particles ate
mostly composed of sand.

The sand grains and particles of laterite are larger in size in the rough
Ÿ than in the fine plaster, the latter containing minute particles whose
diameters vary from 14/~ to 28 q The use of coarse sand grains for the
underlayers and finer ones for the top is recommended b y artists in the
West2
Analysis of the Plaster.--In order to determine the method ofiginally
employed b y the Bagh artists for preparing the rough plaster, its chemical
composition was ascertained, the results of analyses of representative
specimens being as follows : - -
The W'all Paingi~zgs in the .Bagh Caves 89

E~rth Plaster Lime

Deep Red I L i g h t Red Plaster


(Per cent.) (Per cent.) (Per cent.)

~Ioisture . . . . 0.27 0.42 0.91

Carbon dioxide, COs .. 1.25 0.65 12.68

Combined W a t e r and Organic


Matter . . . . 0 -55 1.03 9.34

Sllica, SiO 2 .. 87 -22 87.10 51.18

[ron, F%Oa .. 9 1 7 6 6 -12 3.11 l


Alumina, A120 a 0.44 5-29
2 -23*
Phosphoric Acid~ PzO s 0.12 0-10 I
i
]:itanic Acid, TiO~ . . . . 0.09 0.07 J

Lime, CaO . . . . . . 2.11 0.75 22.41


~Iagnesia, MgO . . . . 0 -56 0.44

Manganese, ]KnO . . . . 0.75 0.39

5alphuric A n h y d r i d e , SO a .. 9 * 9 0.04
Mkalies . . . . . . . 0.44 0.62 1.21
Kitrogen . . . . . . 0.13 0-09

TOTAL . . 100-05 100.06 I 100.00

* Only iron and alumina preclominate, phosphoric acid and ~itanic acid occurring
in negligible quantities.

S o m e of t h e scrapings of t h e fine plaster were c o m p l e t e l y f r e e d f rt ohme


p a i n t film a n d t h e r o u g h plaster. On t r e a t i n g t h e m w i t h dilute h y d r o c h l o f i c
acid, t h e y dissolved w i t h effervescence a n d e v o l u t i o n of c a r b o n dioxide w i t h
t h e s e p a r a t i o n of fine particles of silica. T h e acid s o l u t i 0 n g a v e r e a c t i o n s
for c a l c i u m a n d for a sulphate, t h e l a t t e r occurring o n l y in r e l a t i v e l y s m a l l
p r o p o r t i o n s . T h u s t h e fine plaster was a m i x t u r e of lime a n d c a l c i u m
s u l p h a t e w h i c h h a d b e e n a p p l i e d to t h e surface of t h e rough plaster t o s e r v e
90 S. Paramasivan

as fine Ÿ ~~ In the lime stucco, the fine plaster was composed of lime
alone without any trace of calcium sulphate. Since ir was impossible to get
samples of the fine plaster, unaccompaniedby rough plaster, the proportion
of lime and calcium sulphate was not determined.
Thus the ground was prepared out of naturally occurring ferruginous
earth of of artificially prepared lime plaster. The principal components
of the earth plasters ate silica, iron, alumina and lime. In the case of the
lime plaster, the principal components ate lime and silica.
The proportions of silica in the earth plasters are almost identical,
while their colour seems to have been inŸ by the proportion of iron.
The proportionof lime present in the earth plasters is too low to have been
purposely add'ed to the original material. Hence ir should have been,
present in the original materials as ah impurity. The earth plasters are
fairly firm and strong,t Ir wi11 be shown below that the rough plasters do
not contain any organic binding medium. Hence the consotidation of the
earth plaster may be attributed to the plasticity of clay11 present in it,
and this is evident from the proportionoŸiron and alumina. In the case of
the lime plaster, lime and sand contributeto the firmness and strength of the
plaster.12 The plasters have also been reinforcedand bound with vegetable
fibres,13 the lime plaster containingconsiderablequantities of them.
The percentage of alkalies and snlphuric anhydride in the earth
plasters is low. In other words, presence of soluble salts and of gypsum
which gives rise to effiorescence on the surface of the paintings can be
ignored. ~ The other impurities in the earth plasters are not of much signi-
ficance in reconstructing the technique of the painting process or in
affecting the permanence of the paintings.
From the small proportionof iron and alumina in the lime plaster, ir
is clear that the percentageof clay in i t i s very low. Thus a pure rich lime
having no hydraulic properties has been used. From the low percentage
of sulphuric anhydride, one can ignore the presence of gypsum, which

* F i n e plaster of gypsum has been used in t h e wall paintings of B a m i y a n in Afghan-


istan and Kizil in Chinese Turkestan, which had cultural contact with India.
t In T h e B a g h Caves (London: Indi a Soeiety), 1927, pp. 16-17, S i r ffohn
Russel remarkst h"at th, e rinzaffois less tenacious t h a n at A j a n t a - - a fact t h a t has contri-
buted in no small measure to the deteriorationof the paintings." This is to be a t t r i b u t e d
to the larger proportion of clay in the A j a n t a plaster t h a n in the Bagh ones. F u r t h e r ,
t h e A j a n t a plaster has a larger proportion of vegetable fibres in ir t h a n t h e B a g h plaster
has. This has also contributed Lo the b e t t e r consolidation of t he A j a n t a plaster.
Proc. lnd. Acad Sci., A , roL X, PI. V I I I

4. Cave No. 4 at Bagh, fresco-painting. A p a r t of an elephant procession


The Wall Paintings in the Bagk Caves 91

prevents the setting of mortar and causes efflorescence to appear on the


surface of the paintings. The small proportion of impurities such as
alkalies and other soluble salts, t h e firm adherente of the plaster and the
absence of slaking on the ground indieate t h a t special attention should
tlave been paid to the preparation of lime. x4

Inert Materials in the Plaster.--Since the consolidation of the earth


plasters is due to t h e plasticity of clay present in ir, silica, iron, manganese
and lime ( p r e s e n t a s the carbonate) are the inert materials. Silica alone
occurs to the extent of 87%. I n the case of lime plaster, the results of
chemical analyses show t h a t sand alone is the inert material, f o r a pure r i e l
lime without the admixture of an inert material would not have yielded a
satisfactory plaster.1~ But furtherexperimentsls conducted show t h a t marble
dust (which was invariably used with the Italian frescoes and which gire
the same chemical reactions as earbonated lime) is not present in the
plaster. But the' plaster contained particles with deasity 2-72 I". The
siliea present in the plaster was associated with h y d r a t e dyellow oxide and
anhydrous red oxide of iron, thereby showing t h a t unburnt limestone has
been added to the lime for preparing the plaster or t h a t partially burnt
limestone has been used for the purpose,le The sand grains serving as the
inert materials looked sharp and angular and have contributed to the
¡ and strength of the plaster.17

Technique of laying the ground.--The percentages of eombined water


and organic m a t t e r in the two earth plasters and in the lime plaster are
0.55, 1.03 and 9-34 respectively. T h a t this is not due to the presence of
any organic binding medium like drying oil, glue, albumin of casein is
proved by the absence of any stain 18 when the plaster is aeted upon by
methylene blue, m e t h y l violet, acid green or iodoeosin. Thus t h e organic
m a t t e r present in the plaster must be the vegetable fibres and some organic
impurities. I n the case of the rough plaster of lime, a sample of it softened
considerably under the action of water. On treating ir with dilute hydro-
chloric acid, it disintegratedwith evolution of earbon dioxide and separation
of silica, the acid solution giving tests only for calcium. Thus the lime
plaster has been composed of a mixture of burnt and unburnt limestone or
of limestone partially burnt, and hence lacking in consolidation. The
approximate,probablecomposition of the limestone used for the preparation
of the plaster sectas to be as follows :--
92 S. P a r a m a s i v a n

(Per cent.)
C~lcium carbcnute, CzCO8 .. 44.17
S"lica, ~iO2 . . . . . . 53.30
Ircn and Alumina,, Fe203 -b A]aO~ 2.45
Calcium sulphate, CaSO4 .. 0.08
TOTAL .. 100.00

The fine plaster did not soften when acted upon b y water. Otherwise
its behaviour towards staining agents a n d dilute hydrochloric acid was
exactly similar to that of the rough plaster of lime. Thus lime has brought
a b o u t the consolidation of the fine plaster.
Method of laying the ground.--From experimental results it is clear
t h a t the artists applied the first coat of earth plaster of of lime plaster
containing vegetable fibres. The former received a second coat of rough
plaster of earth to a thickness of. 1 mm. The t o t a l thickness of the rough
plaster varied from about 7.8 mm. to 20 mm. and from 3.4 mm. to 6.5 mm.
respectively depending upon the inequalities of the surface of the carrier.
While the ferruginous earth was naturally occurring, t h e lime plaster had
been prepared either b y mixing b u r n t and unburnt Iimestone or b y partially
burning the limestone, and preparing the plaster out of it. The rough plaster
has been given a coat of fine plaster containing a mixture of lime and calcium
sulphate in the case of the earth plaster and of lime alone in the case of lime
plaster and the surface smootkened.
(3) The Pigments.
The following pigments were identified19 in the paintings a t Bagh : - -
Yell ow ochre.
Red ochre.
Terre verte.
Lapis lazuli.
Carbon.
Lime.
Over the lime plaster, the pigments that were identified were the yellow and
red ochres, carbon and lime. Thus the colour scale is v e r y limited. Evi-
d e a t l y only those colours which were easŸ available locally were employed.
(4) The Binding Medium.
The paint film is fairly adhering to the plaster. I t easily disintegrates
when soaked in cold and boiling water and no gum could be extracted with
The Wal! Paintings in the gag/~ Caves 93

water from it. No vchiele of organic binding medium could be extracted


from the paint film with ether, earbon disnlphide and chloroform used
separately and successively. There was no stain on the paint film w h e n it
was acted upon b y methylene blue and methyl violet, b u t acid green stained
the paint film t h e r e b y showing the presence of glue a s a binding medium, ls
When a small particle of the paint film was covered with a drop of water,
a glue ring with characteristic glue crackle formed on evaporation of the
water. ~o Thus a tempera technique had been employed in the execution of
the paintings. W i t h a few samples of the earth stucco, no stain was dev-
eloped on the paint film either with acid green or iodoeosin. On t r e a t m e n t
with dilute hydrochloric acid, such paint films disintegrated with evolution
of carbon dioxide, the acid solution giving tests for caleium. This reaction
was characteristic of the paint film on all the samples of lime plaster collected.
B u t the pigments had not interfused or spread beneath the plaster ground,
and henee the technique is one of lime medium. ~ This is to be expected
for the layer of lime on the earth plaster is t o o very thin for the ground to
remain wet for sufficient time for true fresco process to b e adopte& I n the
case of the lime stucco, the rough plaster has not well con.solidated on account
of the poor quality of the lime which is mostly unburnt while, for the true
fresco process, well burnt and well slaked lime is used. 22 The fine plaster
on the rough plaster is too thin to retain any moisture for any considerable
time for the paintings to be done in the true fresco style.

Black wets poorly a n d a little gum or glue must be added t o it before


grinding3a A sample of the black paint gives the usual stain for glue
with acid green, ls

Since the paintings have been done either in tempera of in lime medium,
i t i s very difficult to say the extent of the ground t h a t was covered in the
course of a single day from a detailed examination of the joinings of day-to-
day's work. 24 Joints in the plaster are more easi!y visible in true fresco than
in lime medium. F u r t h e r colours employed for the background makes it
difficult to detect any joint.

I n conclusion, the author wishes to express his thanks to Mr. M. B.


Garde, Director of Arehmology of the Gwalior State, who took the trouble
to collect the samples of painted stuccoes on which investigation were carried
out, and for bis generous help, to Mr. J. F. Blakiston, formerly Director-
General of Archmology in India and to Rao Bahadur K. N. Dikshit, Director-
General of Archmology in India, for their great interest in the subject.
94 S. Paramasivan
REFERENCES

1. Transactions of the J5iterary Society of Bombay, 2, 206-14.


2. Journal of the Bombay Braneh of the tr Asiatie
Society, 4, 5 4 3 - 5 7 3 .
Indian Antiquary, 39, 225.
Sir J o h n M a r s h a l l a n d The Bagh Caves in the Gwalior State ( L o n d o n : I n d i a
others S o c i e t y ) , 1927.
Burgess, J. .. Notes on Bauddha l~ock Temples of Ajanta, their Paintings
and Sculptures and on the Paintings of the Bagh
Caves ( B o m b a y G o v e r n m e n t C e n t r a l P r e s s ) , 1879,
94-95.
Benjamin Rowland and The Wall Paintings of India, Central Asia and Ceylon
others (Boston: M e r r y m o u n t Press), 1938, p p . 79-81.
Zoc. cit., p p . 79-81.
3. Sir J o h n l~Iarshall a n d Zoc. c/t., p. 22.
others
4. Zoc. cit., p. 5.
5. D a n , e l V. T h o m p s o n . . The Materials of Medieval Paint~ng (i~ew t t a v e n : Y a l e
U n i v e r s i t y P r e s s ) , 1936, p p . 4 3 - 4 7 .
6. o. Zoc. c/t., p. 39.
7. S. P a r a m a s i v a n 99 " T e c h n i q u e of t h e P a i n t i n g P r o c e s s i n t h e T e m p l e of
Vijayalaya Cholisvaram in the Pudukkot~ai State,"
Proc. Ind. Acad. Sci., 1938, 7, 284.
8. N o r m a n G. C o m b e r .. A n Introduetion to the Scientific Study of the SoiI
( L o n d o n : A r n o l d ) , 1927, 4 8 - 4 9 .
G. W . R o b i n s o n .. Soils: Their Or$gin, Constitution and Classification
( L o n d o n : M u r b y ) , 1932, 12-13.
9. A.P. Laurie .. The Painter's Methods and Materials ( L o n d o n : SeeleF
& Co.), 1926, 197.
10. R u t h e r f o r d , :L G e t t e n s . . " T h e M a t e r i a l s i n t h e W a l l P a i n t i n g s of B a m i y a n , "
Technical Studies, 1938, 6, 186.
" T h e i~Iate¡ i n t h e *Wall P a i n t i n g s f r o m Kizil i n
C h i n e s e T u r k e s t a n , " Technical Studies, 1938, 6, 283.
11. Searle The Chemistry and Physics of Clays and other Ceram~c
Materials ( L o n d o n : E r n e s t B e n n ) , 1924, 262-63.
12. J . W . Mellor A Comprehensive Treatise on Inorganic and Theoretical
Chemistry ( L o n d o n : L o n g m a n s , G r e e n & Co.,
1923, 3, 677.
13. 8 @ Encyclopcedia Britannica, 1 4 t h Edi~ion, 18, 39-50.
14. Max Doerner Ths Materials of the _4rtist and their use in Painting,
translated (New Y o r k : H a r c o u r t , B r a c e & Co;),
1934, 269.
Tke Wal! Painting's in t/le Bagh Caves 95
15. S. P a r a m a s i v a n .. Loa. ci~., p. 2869
16. a 9 Zoc. cit., p . 286-87.
16. b .. ]5oc. cit., p . 287.
16. c ,. Zoc. c/t., p. 287.
17. 9. Z o c . cit., 2 8 i .

18. William Ostwal4 .. " I c o n o s c o p i c S t u d i e s (Microscopic I 4 e n t i f i c a t i o n of


Homogeneous Binding l~Iediums)," t r a n s l a t e d ,
T e c h n i c a l Stuclies, 1935-36, 4, 1 3 8 4 2 .

19. M a r t i n de W i l d .. T h e Scient 91 E x a m ~ n a t i o n of Pict 91 ( L o n 4 o n : G. B e l l


& Sons, L t d . ) , 1929.
Rutherford J. Gettens " The S t a g e Microscope i n t h e R o u t i n e E x a m i n a t i o n of
a n d G. L . S t o u t P a i n t i n g s , " T e c h n i c a l S t u d i e s , 1935-36, 4, 224-28.

20. .. " T h e M a t e r i a l s i n t h e W a l i P a i n t i n g s of B a m i y a n ,
A f g h a n i s t a n , " T e c h n i c a 9 1 S $ u d i e s , 1938, p. 190.

21. l~ary H a m i l t o n A n c i e n ~ P a ~ n t ~ n j (New H a c e n : Y a l e U n i v e r s i t y P r e s s ) ,


Swindler 1929, p. 418.

22. A. P . L a u ¡ .. Greek and Rom~n Patnt$ny ( C a m b r i d g e: U a i v e r s i t y


Press), p. 83.

23. S. P a x a m a s i v a n .. " T h e M u r a l P a i n t i n g s in t h e B r i h a d i s v a r a T e m p l e a t
Tanjore--An I n v e s t i g a t i o n i n t o t h e ~r
T e c h n i c a l S t u d i e s , 1936-37, 5, p. 283.

You might also like