See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/301956774

The composition and technology of the 3–4th century CE decorative earthen

plaster of Pithalkhora caves, India

Article  in  Journal of Archaeological Science: Reports · May 2016

CITATIONS READS

5 826

3 authors:

M. Singh S Vinodh Kumar

National Museum Institute Archaeological Survey of India
102 PUBLICATIONS   623 CITATIONS    28 PUBLICATIONS   228 CITATIONS   

SEE PROFILE SEE PROFILE

Sujata Waghmare
National Environmental Engineering Research Institute
15 PUBLICATIONS   156 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

CHARACTERIZATION OF GLAZE TILES View project

The carbon negative uses of Cannabis Sativa L. View project

All content following this page was uploaded by S Vinodh Kumar on 06 May 2016.

The user has requested enhancement of the downloaded file.

 Journal of Archaeological Science: Reports 7 (2016) 224–237

Contents lists available at ScienceDirect

Journal of Archaeological Science: Reports

journal homepage: www.elsevier.com/locate/jasrep

The composition and technology of the 3–4th century CE decorative

earthen plaster of Pithalkhora caves, India
M. Singh ⁎, S. Vinodh Kumar, Sujata A. Waghmare
Archaeological Survey of India, Science Branch, Bibi-ka-Maqbara, Aurangabad 431 004, India

a r t i c l e i n f o a b s t r a c t

Article history: This research focuses on the characterization of ancient decorative earthen plaster of India's Pithalkhora group of
Received 9 October 2015 caves. The restoration strategies being planned for the earthen plaster demand detailed scientific investigation to
Received in revised form 23 March 2016 understand technology and composition so as to prepare matching plaster. The mineralogical, micro-structural
Accepted 26 April 2016
and chemical characterization of the earthen plasters have been investigated by analytical tools like petrological
Available online xxxx
microscope, XRF, CHN analysis, laser scattering particle size analyzer, FTIR, XRD, Scanning Electron Microscopy,
Keywords:
thermal analyzer (DGA/DTA) as well as on-site observations. A comparative study with nearby sites portrays
Earthen plaster the earthen plaster of Pithalkhora coarser (silt: 62–100%, sand: 2–2.2%) with traces of clay (0–2%); hence it
Particle size may be prone to damage by grain breakages. The chemical composition confirms that the aggregates mixed
Aggregate with the earthen plaster belong to that of locally available basaltic origin. SEM studies show addition of quartz
Illite and traces of clay (illite and montmorillonite) to augment the flexibility and performance of the plaster. The res-
Calcite toration plaster should be synthesized with aggregates of locally available basaltic rock mixed with slaked lime in
Quartz addition with rice husk and other vegetal fibres as vegetal additive to overcome the shrinkage of the plaster.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction
The analyses and characterization of ancient earthen plaster in India is
still in primitive stage despite notable progress in last decade worldwide
(Sharma et al., 1995; Singh and Arbad, 2014, 2015). A critical understand-
ing of earthen support layer is very essential as many problems associated
with conservation of ancient decorative surface originate from structural
or support layers. Due to vulnerability of earthen support layers, the liter-
ature in decorative layers on earthen support is very limited (Warran,
1990; Mora et al., 1984; Houben and Guilland, 1984; Griffin, 1999). An-
other major restriction is the availability of samples as research cannot
be completed at the cost of damage to the historic surfaces. To tailor ap-
propriate conservation strategies, the art conservators are required to be
well versed about the problems of earthen architecture. Identification
and characterization of original materials is a vital step for proper under-
standing of the deterioration process of the original fabric on earthen sup-
port layer. Such characterization allows development and synthesis of
novel materials whose composition can be matched with the original ma-
terials. The materials used for conservation of earthen supports should be
chosen in accordance with accepted conservation criteria (compatibility
with original materials: equivalent or weaker strength, similar porosity
and density etc.). The appropriate knowledge about composition and
characteristics of original materials is an essential requirement for identi-
fying the repair materials which fulfil the criteria.
⁎ Corresponding author.
E-mail address: m_singh_asi@yahoo.com (M. Singh).
Pitalkhora Caves (20° 15′ N; 75° 15′ E) located in remotely wildlife
valley in the Satamala range of western Ghats of Maharshtra, India con-
sists of 14 Buddhist Caves and forms one of the earliest centres of the
rock-cut architecture. The Pithalkhora caves represent an early Hinaya
rock cut architecture but the earthen plaster and its painting are of Ma-
hayana period as per inscription found dating back from 250 BCE to 3–
4th century CE (Fergussons and Burgess, 2000; Nagaraju, 1981). This
place was along an ancient trade route (Fig. 1) that connected the Dec-
can tableland with the port of Bharuch on the west coast and the ancient
city of Ujjain to the north. Fig. 2 shows the general view of Pithalkhora
caves. The Chaitya and Monastery Caves in this group have traces of
very beautiful paintings of which some are surviving in the former.
Fig. 3 show the beautiful painting on the earthen plaster of main
Chaityagriha.
In contrast to other rock cut caves of western India, the Pithalkhora
caves were excavated from variety of basalt stone which weathers
fast. As the caves are carved in somewhat softer, fragile basaltic rock
marked by thick clayey veins, here one can also see examples of ‘ancient
conservation’. A feature need special mention is ingenious arrangement
diverting rain water from its way into the caves through cracks by mak-
ing long tunnel like opening in the ceiling and allowing the concealed
drain to lead the water outside.
Based on the ancient Indian painting recipe and recipe for the prep-
aration of earthen plaster written in Sanskrit language in 6th century CE
(Sivaramamurti, 1978), just after the painting works of Pithalkhora/
Ajanta caves, it will be interesting to explore any addition of proteina-
ceous materials in to the earthen plaster of Pithalkhora, documented

http://dx.doi.org/10.1016/j.jasrep.2016.04.017
2352-409X/© 2016 Elsevier Ltd. All rights reserved.
 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237 225

Fig. 1. (a) The map showing the location of Pithalkhora caves. (b) The map shows the location of Pithalkhora caves along with Bhuddhist caves of Maharashtra.

in ancient text to enhance the binding properties of the mortar. This will
give additional data on the nature and characteristic of Pithalkhora plas-
ter and help us in synthesis of compatible materials for restoration
work. Adding vegetal additives is the common feature of all earthen
plasterworks to enhance its performance (Miller, 1934, Orazi, 1994).
The types of vegetal additives incorporated in the plaster vary according
to the place and availability of local materials surrounding the site
(Singh and Arbad, 2014). It is essential to investigate and characterize
vegetal remains in the original plaster to synthesize a matching plaster
mix.
226 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237
Fig. 2. Photograph showing the general view of Pithalkhora caves.
1.1. Development of earthen plasters technology
Traditional mud plaster is made with soil composed of sand, silt and
clay with rice husk, straw and other vegetal fibres added sometimes to
prevent excessive cracking during drying. An equal distribution of silt,
sand and clay in plaster mix is desirable to improve the performance
of earthen support. Excessive silt which is neither a good binder nor
an aggregate and produces the plaster that is prone to shrinkage and
cracking. Clay (phyllosilicate) is a term related to grain size (b 2 μm),
with grain shaped like a sheet much thinner than wider, and the grains
are attracted to one another by electrolytic interparticle forces (Basma
et al., 1996). The non-clays are of grain size greater than clays and are
divided into the grain size categories of (2–50 μm in diameter) and
sand (50 μm to 2 mm in diameter). Due to their small surface area,
non-clays show less attraction for water and are non-plastic. The shrink-
age of clay masses can significantly be influenced by the common prac-
tice of mixing other materials – such as vegetal matter, aggregates, lime
or calcite – which may serve as a binder. The common aggregates gen-
erally reported in the earthen plaster mix are sand and straw. As earthen
plaster mix is prepared by preferably exploiting local materials, differ-
ent kinds of aggregates are not very uncommon in plaster. To prepare
a matching fresh plaster for replication, it is essential to focus on the
types and nature of aggregates in the plaster works.
In case of Ajanta caves, the aggregates present in the earthen plaster
are belong to silt loam soil (soil composed of sand, silt and smaller quan-
tity of clay) that are surviving for years and years. The source of aggre-
gates for making earthen plaster is probably from the Waghura river
bed (Singh and Arbad, 2014). Grain size analysis of earthen plaster of El-
lora caves showed although most of the materials is silt size (between 66
and 81%), yet in some sample silt size particles are between 25 and 43%.
Similarly, in the majority of samples sand size grains are between 50
and 75%, except in some mortar between 9 and 15%. As coarse grains
are present in Ellora mud mortar, the plaster showed a slight rigid behav-
iour with tendency to deform by grain breakage. Due to presence of great-
er proportion of higher grain size particles, the earthen plaster of Ellora
was also showing higher porosity (about 20–28%), less compactness
and prone to damage easily (Singh and Arbad, 2015). Silt loam to sandy
loam soils were mostly exploited as raw material for Ellora mud mortar.
As the earthen mortar of Ellora is characterized for its high silt or some-
times sand and very low clay, to ensure maximum adaptability with the
original raw material the particle size analysis test was performed on
the alluvial dry bed soil of now seasonal waterfall points in the cave com-
plex and the result confirmed the source for the a soil is same.
The Pithalkhora is probably the earliest ancient Bhuddhist caves in
western India where earthen plaster technology was initially practiced
in 3rd to 4th century CE (Mithra, 1971). As there is no ancient text deal-
ing about the techniques used by ancient Indian in the preparation of
earthen plaster of Pithalkhora, scientific investigation is only tool to ex-
plore the nature and characteristics of the plaster. It is also interesting to
correlate the nature and the composition of Pithalkhora plaster to that
of earthen plasters of other nearby Buddhist sites like, Ajanta, Ellora, Au-
rangabad caves etc. situated within the radius of about 100 km. This will
help to further enrich our knowledge about the development of earthen
plaster technology vis-á-vis improvement of methodology in later caves
as well as to gain further knowledge about the historical development of
that period.
In the present study, the chemical and mineralogical compositions of
the earthen plasters were investigated with X-ray fluorescence, petro-
logical microscope, XRD analysis and SEM analysis. The optical micros-
copy gave information about the binder, typical shape of the
aggregates and also about the structure of entire components. Sieve
analysis was also carried for the aggregate after chemical separation of
the binder and aggregate. The samples of plasters were also extensively
examined under FTIR, TGA–DTA and CHN analysis for inorganic and or-
ganic additives so as to gain information about true nature of the plaster.
2. Materials and methods
Out of 14 caves at Pithalkhora, remains of earthen plaster with paint-
ings are observed in only four caves namely 3, 4, 5 and 6. In the
 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237
Fig. 3. Photograph shows the beautiful painting on the mud plaster of main Chaityagriha,
Pithalkhora caves.
remaining caves, the earthen plasters with paintings have either lost or
few of the caves could not be painted owing to highly hygroscopic clay
veins frequently encountered within the basaltic stone matrix. On site
visual examination under magnifying lens played a primary role in the
understanding of the plastering technique at Pithalkhora caves. The ex-
amination revealed that two layers of earthen plaster have been applied
on a roughly cut basaltic rock from which the caves were excavated. At
some places within the caves, breccias with a clayey matrix was noticed
which was devoid of any plaster or paintings. The first plaster layer is a
levelling layer with thickness ranging from 10 mm to 40 mm in accor-
dance to the topography of the wall. This layer composed of earth
(black in colour) mixed with organic additives and far coarser in texture
levelling the roughness of the rock surface. Above this layer, a smoother
second plaster layer with thickness 10–15 mm is superimposed. The
preliminary function of the second layer is to get uniform surface to re-
ceive the pigment layer. This layer is little smoother and reddish in col-
our. Above the second plaster layer, a lime wash layer was noticed on
which the painting work have been executed. The photographs of the
earthen plaster layers along with schematic diagram of the plastering
technique employed at Pithalkhora caves are shown in Fig. 4.
One of the major drawbacks in the examination of the decorated
earthen plasters is the availability of samples. As the research cannot
be carried out at the expense of damage to the decorated surfaces, few
grams of samples extracted from the surface where the painting layers
are lost from cave no 5 and 6 of Pithalkhora (5 numbers of samples)
were subjected to scientific examination and the details of the layer
and location of the samples collected are mentioned in Table 1. In the
227
Fig. 4. (a) Photograph showing the plastering technique of Pithalkhora caves. (b) Schematic
diagram of the plastering technique of Pithalkhora caves.
present study, particle size analysis, XRD, petrological microscopy, ste-
reo microscopy were used to ascertain the mineralogical composition
of the earthen plaster.
The earthen plaster samples were allowed to dissolve in 15% dil. HCl
to remove the calcareous aggregates present in and the remnants were
sieved followed by complete drying. The aggregate components were
subjected to petrological analysis to confirm their composition and
grain size. The chemical composition of earthen plasters in the form of
major oxides was determined using an X-ray fluorescence set up (Phil-
lips 1410, Holland) as per standard procedure (Colao et al., 2010), by
mounting a compressed pellet in the sample holder. Accordingly loss
on ignition (LOI) has also been derived.
The mineralogical composition of the earthen plaster of sample nos.
1, 2 and 3 was determined by an X-ray Diffraction (XRD) spectrometer
(Philips 2404, Holland) using a graphite monochromator and Cu Kα ra-
diation. The sample was scanned from 2θ ranging from 5 to 120°. The
presence of minerals was confirmed with the help of data files present-
ed by JCPDS (Joint Committee on Powder Diffraction Standards) 1994.
The XRD data supported the mineralogical findings obtained through
petrological analysis and scanning electron photomicrograph of the
plaster. The percentage of organics in sample nos. 1–5 was determined
by using a Carbon, Hydrogen, Nitrogen (CHN) analyzer (FLASH EA 1112
series, Thermo Finnegan, Italy). The mode of operation is of Absolute
CHN mode. The accuracy of the measurements are ± 0.3%. following
the ‘Dumas Method’. The ‘Dumas Method’ requires complete and in-
stantaneous oxidation of the sample by flash combustion. The combus-
tion products (NO2, CO2, SO2) were separated by the chromatographic
column which was detected by a Thermal Conductivity Detector
(TCD) that provided an output signal proportional to the concentration
of the individual components.
228 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237

Table 1
Details of the earthen plaster samples of Pithalkhora.

Sample no. Description Layer Location

1 Earthen plaster 1st layer Cave no. 5

Side wall
2 Earthen plaster 2nd layer Cave no. 5
Side wall
3 Earthen plaster 1st layer Cave no. 6
Ceiling
4 Earthen plaster 2nd layer Cave no. 6
Ceiling
5 Earthen plaster 1st layer Cave no. 6
Upper wall

Particle size analysis is the backbone of materials research, providing properties modified. It is difficult to understand the composition and
distribution curves by which soils can be characterized (Shea and characteristics of original material without any early documentary evi-
Watta, 1939). The sand, silt and clay percentages derived from particle dence on building practices applied at Pithalkhora, it has become an ob-
size analysis provided our major findings during the study of the earth- vious requirement. Local soils are likely to have been used for
en support. Due to scarcity of samples, the traditional hydrometer/sieve construction, and may serve as a good source of repair materials. The ba-
analysis was replaced by an electronic particle size analyzer for faster saltic stone samples collected from the site have also been analyzed find
and more precise analysis. A laser light scattering instrument, which out the chemical composition. The results of our analytical examination
operates on the principle of interaction of laser light with soil particles are presented below.
suspended in water in a highly agitated state (Vitten and Sadler,
1997), was used. We used an Eye Tech particle size and shape analyzer
(Ambivalue Netherlands) that gives accurate results for both size and 3. Results
shape with very small samples. Because only a limited number of sam-
ples could be taken, the analyses were carried out on only a few (04 3.1. Chemical composition of earthen plaster
Nos.) representative samples. For each sample, three measurements
were taken and the value averaged. The chemical composition of the earthen plaster in the form of major
For FTIR spectroscopic analysis, a small quantity of plaster sample oxides was determined using an X-ray fluorescence spectroscopy. From
was mixed with KBr and the pellet was freshly prepared and used. the obtained value shown in Table 2, it appears that slaked lime having
The FTIR spectra were obtained in the region of 4000–550 cm−1 using minor content of magnesium carbonate may be resulted from the basal-
Agilent 600 series FTIR Spectroscopy equipped with nitrogen cooled tic origin aggregates which influence shrinkage of plaster. The lime
MCT detector. The spectrum was taken in 20 scans with the data inter- might have acted as binder for earthen plaster. The addition of lime
val 0.5 nm, and the resolution was 2 cm−1. The FTIR spectra were ob- into plaster mix was also confirmed by reaction with dilute hydrochloric
tained from different areas of the samples. FTIR spectra database with acid which gave effervescence. There is also a possibility of raw mate-
inorganic and organic materials in the sample were used to characterize rials containing the lime. The slaked lime probably might have also
the unknown components in the sample. been added to overcome the inadaptability of local resource materials
Olympus Magnus and LABOMED CSM2 stereomicroscope was used as it acted as cementing materials for earthen mixture. The remarkable
to analyze the aggregates of the earthen plaster samples. The SEM pho- colour of earthen plaster is due to presence of 12–15% iron oxide. The
tomicrographs of the earthen plasters were obtained using a FEI ESEM observed percentage of quartz in the plaster is around 45% indicates
Quant 200, USA scanning electron microscope at a range of magnifica- that approximately half of the content of earthen plaster is of quartz
tions to identify the components of the plasters. Thermo-gravimetric which are the best aggregates of the earthen plaster. As per the results
(TG) and differential thermal analysis (DTA) of the plaster samples obtained from the analysis of local basaltic stones (Table 3) i.e., the per-
were performed simultaneously on ground sample using Diamond centage of quartz present is around 50%, it is confirmed that the aggre-
Model TG/DTA (Perkin Elmer, USA) instrument. The analysis was done gates used for the preparation of the earthen plaster belong to the local
in Air/N2 atmosphere at a flow rate of 40 ml/min from 30° to 1000 °C basaltic stone origin.
using Pt/Rh crucibles and alumina as reference material. In a thermo
gravimetric analysis (TGA), a sample is heated according to a certain
Table 2
temperature program and the changes in sample weight are plotted as Chemical composition (% by weight) of earthen plaster of Pithalkhora.
a function of temperature; thus the weight lost by a sample material
Major oxides Sample
can be calculated in % according to the temperature range. This method
is useful in determining the qualities and the thermal behaviour of a ma- 1 2 3 4 5
Cave no. 5 Cave no. 5 Cave no. 6 Cave no. 6 Cave no. 6
terial. It also gives us an idea about the behaviour of other substances in
Side wall Side wall Ceiling Ceiling Upper wall
the composition of a material. For example, TGA indicates the existence (1st layer) (2nd layer) (1st layer) (2nd layer) (1st layer)
of calcite at 800–900 °C, clay at 560–650 °C and quartz at 573 °C (Gurdal,
SiO2 44.15 45.20 43.28 44.35 45.10
1976). DTA is a sophisticated analytical technique which records the
Al2O3 19.30 15.69 18.71 18.20 17.53
temperature at which the components a mortar decompose. It is partic- Fe2O3 14.85 12.29 14.77 13.50 14.50
ularly good at distinguishing between calcium compounds, and is there- TiO2 2.67 1.95 2.69 2.22 2.52
fore used primarily to determine binder type. Hence, it is used as CaO 6.10 7.96 6.16 6.75 7.20
MgO 2.47 3.63 2.52 3.10 3.20
recognised technique for analyzing the historic mortars.
K2O 0.56 0.76 0.52 0.58 0.65
Microscopic examination of plaster samples was carried out under Na2O 0.83 1.29 0.90 1.10 0.98
an Olympus Magnus MS2 and LABOMED CSM2 Stereo-microscope to MnO 0.14 0.14 0.16 0.15 0.15
identify vegetable fibres used as organic binders in the plasterworks. P2O5 0.30 0.30 0.24 0.32 0.35
A very common question in earthen construction technology has to L.O.I.a 7.65 9.67 8.32 8.12 6.13

do with whether the support materials were excavated locally and their a
L.O.I.: loss on ignition.
 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237 229

Table 3 decorative earthen plaster, the hygrometer/sieve analysis methods

Chemical composition (% by weight) of basaltic stone samples of Pithalkhora caves. were discarded. As the electronic particle size analyzer needs minimum
Major oxides Sample quantity of sample and yields results which are more precise and reli-
1 2
able, the laser particle size analyzer was used in this investigation.
Four earthen plaster samples extracted from cave nos. 5 and 6 of
SiO2 47.87 51.47
Pithalkhora were analyzed for particle size distribution and graph of
Al2O3 16.72 15.94
FeO 3.25 2.99 the particle size of Pithalkhora is shown in Fig. 5. From the particle
Fe2O3 5.08 5.33 size analysis, it is observed that earthen plaster of Pithalkhora is far
TiO2 2.38 2.30 coarser than what was expected and contains too less the clay particles.
CaO 6.49 6.05 Grain size analysis showed that of most of the materials are silt size (be-
MgO 3.12 2.88
K2O 5.18 6.05
tween 62 and 100%) followed by sand size (between 2 and 22%). How-
Na2O 4.49 3.61 ever, in all the samples proportion of clay size particles are far too less
MnO 0.14 0.18 between 0 and 2%. The earthen plaster of Pithalkhora is showing highly
P2O5 0.83 0.88 rigid behaviour with greater tendency to deform or damage by grain
L.O.I.a 4.45 2.32
breakage due to presence of large proportion of coarse size grains.
a
L.O.I.: loss on ignition. The earthen plaster composition of Pithalkhora, based on the grain
size distribution is plotted on sand–silt–clay ternary diagram, Fig. 6.
3.2. CHN analysis From the ternary diagram we observed that silt loam soil (soil com-
posed of more silt, less sand and negligible quantity of clay) was proba-
The most common form of elemental analysis, CHN analysis, is ac- bly sourced in the preparation of earthen plaster of Pithalkhora caves.
complished by combustion analysis. In this technique, a sample is
burned in an excess of oxygen, and various traps collect the combustion 3.4. Petrological analysis
products—carbon dioxide, water, and nitric oxide. The masses of these
combustion products can be used to calculate the composition of the The earthen plaster sample nos. 1–5 of Pithalkhora was observed
unknown sample (Shedlovsky, 1960; Spiro, 1960). The carbon, hydro- through petrological microscope and principal mineral compounds
gen and nitrogen analysis of earthen plaster sample nos. 1–5 was car- identified are shown in Table 5. From the petrological data, it is observed
ried out and the results of CHN analysis are shown in Table 4. The that quartz, calcite and feldspar are the major minerals identified in all
earthen plaster of Pithalkhora is marked by addition of more percentage the samples. The white creamy colour lime is found mixed in fine grains
of vegetal matter probably to compensate the shrinkage property of of euhedral transparent to honey red coloured quartz. The dark green to
plaster owing to low content of slaked lime. Besides, proteinaceous ad- black coloured pyroxene was also observed in the plaster sample. The
ditives in form of vegetal fibres are generally found added in Indian sample also shows the presence of organic fibres. In sample no. 2, dark
earthen plaster works (Singh and Arbad, 2014, 2015), the peak for the green to black coloured pyroxene was observed along with feldspar
same can be observed under FTIR spectra of the plaster. and organic fibres. The presence of plagioclase in the earthen plaster as-
sociated with pyroxene indicates composition of eutectoid, attributable
to original basaltic formation, from which the majority of the starting
3.3. Particle size analysis materials have been derived. The presence of iron oxide and iron silicate
in various samples is due to preparatory materials made up of crystal-
Particle size analysis provide distribution curve through which the line and amorphous phase rich in ferrous substances. The petrological
type of soil can be characterized. The particle size distribution of soil data is in agreement to the XRD analysis of the plaster used for determi-
structure also precisely influence the humidity transfer (Kathleen, nation of mineralogical composition.
1997) in the earthen plaster. The sand, silt and clay derived from particle Two thin sections of Pithalkhora earthen plaster were examined
size analysis determine the type of earthen support. As discussed earlier, under petrographic microscope to gain knowledge about mineralogical
obtaining large quantity of sample for analysis is a major problem for composition. The characteristics observed under the microscope

Table 4
Organic content in the soil samples by CHN analysis.

Sample and location Carbon (%) Hydrogen (%) Nitrogen (%) Total (%)

1 6.96 1.38 0.56 8.90

Cave no. 5
Sidewall
(1st layer)
2 7.49 1.25 0.46 9.2
Cave no. 5
Sidewall
(2nd layer)
3 4.82 1.45 0.47 6.75
Cave no. 6
Ceiling
(1st layer)
4 6.25 1.59 0.45 8.29
Cave no. 6
Ceiling
(2nd layer)
5 6.83 1.48 0.50 9.1
Cave no. 6
Upper wall
(1st layer)

% mentioned here is w/w.

230 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237
Fig. 5. Particle size analysis graph of earthen plaster samples 1–4, Pithalkhora caves.
include colour, colour variation under plane polarized light and crossed
polarizars. In Fig. 7a and 7b, the thin sections showed the presence or-
ganic matter such as vegetal fibres along with quartz. From Fig. 7c and
7d it was noticed that the existence of rice husk in the earthen plaster.
The occurrence of clay minerals such as illite, montmorillonite along
with calcite, quartz and organic matter has been seen from images in
Fig. 8a–d. The presence of above mentioned minerals and organic mat-
ter have also been confirmed by XRD, SEM-EDX and FTIR analysis of
Pithalkhora earthen plaster samples.
3.5. Aggregate analysis
The earthen plaster samples of Pithalkhora were dissolved in 15%
hydrochloric acid followed by subsequent washing of the aggregates
to neutral pH. The aggregates were sieved after drying and separated
as per grain size. Fig. 9 shows the grain size of the aggregates found
mixed in the earthen plaster. It is observed that fine to medium grained
aggregates of basaltic origin might have been mixed in the earthen plas-
ter for strength and durability. Further, the shrinkage of soil can signifi-
cantly be influenced by common practice of mixing other materials such
as vegetal matter, aggregates, lime or calcite all of which will serve as
binder.
Fig. 6. Ternary diagram of Pithalkhora earthen plaster samples.
3.6. XRD analysis
XRD analysis is one of the versatile techniques that can be used to
identify any crystalline substances like most of the minerals. X-ray dif-
fraction pattern of sample nos. 1 and 3 is shown Figs. 10 and 11. From
the spectra, it is observed that major component in both the samples
are quartz, calcite and feldspar. Quartz coming from fillers predomi-
nates in XRD pattern in coarse grained first preparatory layer. Calcite
predominate XRD pattern in fine particle size fraction. The peak corre-
sponds to Orthoclase feldspar has also been also noticed in the XRD.
This is the mixed layer peak of illite and montmorillonite clay. Hence
the result demonstrates the presence of clay minerals in sample no. 3
which is in consistent to SEM images of the plaster as well as data ob-
tained through particle size analysis. However, in sample no. 1 peak
for clay minerals at 9.90 is not seen denoting lack of any clay like illite
in few earthen plaster of Pithalkhora in agreement to particle size distri-
bution graph of the plaster.
3.7. FTIR analysis
A few fragments of earthen plaster sample nos. 1–5 of Pithalkhora
caves were observed under Fourier Transform Infrared (FTIR) spectros-
copy. For FTIR spectroscopic measurement, a small quantity of plaster
sample was mixed with KBr and using pelletizer a fresh pellet was pre-
pared and used for FTIR analysis. Infrared (IR) was subsequently used
for getting the information on some main components present in the
earthen plaster and for determination of presence of organic compo-
nents and colourants. The FTIR spectrum of sample no.3 is shown in
Fig. 12. It reveals the most characteristic SiO2 tetrahedral band at the
spectral region 1037 cm− 1 and 470 cm− 1. The absorption band at
540 cm−1 is probably due to the presence of clay minerals especially
from the kaolinite family (Van der Marel and Beutelspacher, 1976).
The peak at 748 cm−1 might have been caused either by Feldspar or
clay minerals. IR spectra also show a characteristic band at 3518 cm−1
due to free hydroxyl ions of the water. The peak at the frequency of
1627 cm−1 is due to the presence of N\\H group or organic matter pres-
ent in the earthen plaster such as in the form of rice husk and other or-
ganic fibres. Two shallow peaks of methyl group at ca. 2920–2850 cm−1
may be also the presence of organic matter present in the earthen
plaster.
3.8. SEM analysis
The scanning electron photomicrograph of the earthen plaster of
Pithalkhora is shown in Fig.13a–d at different magnifications. The earth-
en plaster shows coarse black inclusion of ferruginous silicate (pyrox-
ene) along with quartz probably bound together with calcite. Some
micro granules indicating the presence of feldspar as well as quartz
and iron minerals are observed. Besides, pore lining and pore filling cal-
cite along with clay minerals like illite/montmorillonite are also seen at
higher magnification. SEM image at 1000× shows the clusters of illite
present as elongated disk like crystal filling the micro porous structure
of the earthen plaster. The presence of illite/montmorillonite gives flex-
ibility to the plaster and enhances the mechanical and physical property
of the earthen plaster. The SEM results correlate with XRD and particle
size analysis data of earthen plaster of Pithalkhora.
3.9. Analysis of vegetal matter
There is a common practice of mixing of vegetal matter in the earth-
en plaster work since early time to avoid shrinkage and to gain strength
(Miller, 1934, Orazi, 1994). Vegetal additives found mixed in the plaster
are generally of local origin plant remains. Whereas rice husk has been
reported mixed in the earthen plaster of Ajnata, mixing of hemp was de-
tected at Ellora (Singh and Arbad, 2014). The organic additives of
Pithalkhora were observed under magnifying lenses and also under
 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237 231

Table 5 For the thermal characterization of materials, weight loss noticed in

Minerals present in the samples identified from petrological microscope. the region 31 °C to 380 °C is attributed to the absorbed water (hygro-
Sample Minerals Major minerals scopic water) and chemically bound water of hydrated salts and hy-
1 Hematite iron oxide (Fe2O3) Calcite, quartz,
draulic components such as calcium silico allumminated hydrates
Cave no. 5 Olivine (magnesium silicate) hematite detected through petrological analysis. A second moss loss was detected
Sidewall Kutnahorite (calcium manganese carbonate, at about 380–600 °C, which can be attributed to the transformation of
(1st layer) CaMn(CO3)2) phase. For instance, it could be caused by the heat alteration of clay min-
Albite Na(AlSi3O8)
erals or the oxidation of organic matter. The small mass loss observed
Cancrinite Na6Si6Al6O24
Calcite Ca(CO3)2 between 600 and 950 °C is indicative of the presence of calcium carbon-
Clinoferrosilite Fe(SiO3) ate in the plaster. The continuous decomposition of carbonate without
Cristobalite synthetic SiO2 steps suggests absence of sharply defined different degree of crystallin-
Quartz (SiO2) ity of the lime. The DTA analysis showed a broad peak connected with a
2 Kutnahorite (calcium manganese carbonate,
Cave no. 6 CaMn(CO3)2)
small weight loss at about 405 °C which could be attributed to combus-
Ceiling Albite Na(AlSi3O8) tion of organic substances. The mass loss occurred at around 1000 °C is
(1st layer) Cancrinite Na6Si6Al6O24 due to the decomposition of calcite (CaCO3). The presence of slight mass
Calcite Ca(CO3)2 loss around 450–500 °C may be the presence of thermal action of hy-
Clinoferrosilite Fe(SiO3)
drated lime.
Cristobalite synthetic (SiO2)
Hematite iron oxide (Fe2O3)
Quartz (SiO2) 4. Discussions
3 Kutnahorite (calcium manganese carbonate,
Cave no. 6 CaMn(CO3)2) The Chemical composition of the component values obtained which
Upper wall Albite Na(AlSi3O8)
is also in close consonance to the composition of earthen plasters of
(1st layer) Calcite Ca(CO3)2
Clinoferrosilite Fe(SiO3) nearby sites Ajanta/Ellora. However, distinct characteristic difference
Cristobalite synthetic SiO2 in the composition of earthen plaster of Pithalkhora to that of Ajanta/El-
Hematite iron oxide (Fe2O3) lora (Singh and Arbad, 2014, 2015) was observed in the compositional
Quartz (SiO2)
content ratio of carbonates of calcium and magnesium. These compo-
4 Kutnahorite (calcium manganese carbonate,
Cave no. 6 CaMn(CO3)2) nents together showed the value between 8 and 11% in Pithalkhora
Ceiling Albite Na(AlSi3O8) earthen plaster which may be the impurities on account of weathering
(2nd layer) Calcite Ca(CO3)2 of basaltic rock. The calcite component along with magnesium observed
Cancrinite Na6Si6Al6O24 in the earthen plaster works of Ajanta caves is between 12 and 15% and
Clinoferrosilite Fe(SiO3)
that of Ellora caves between 15 and 30%. As the lime acts as cementing
Cristobalite synthetic SiO2
Hematite iron oxide (Fe2O3) agent to the earthen mix, low percentage of calcitic component might
Quartz (SiO2) may be attributed for looseness and early fall of plaster at Pithalkhora
5 Hematite iron oxide (Fe2O3) caves at high speed. The low percentage of calcitic component may be
Cave no. 6 Olivine (magnesium silicate)
resulted from the aggregates used that probably from that locally avail-
Upper wall Kutnahorite (calcium manganese carbonate,
(1st layer) CaMn(CO3)2)
able basaltic origin or there may be the possibility of result of diagenesis
Albite Na(AlSi3O8) if the water in the cave is acidic (Karkanas et al., 2000). Since, the tech-
Cancrinite Na6Si6Al6O24 nology for preparation of earthen plaster for decorative works probably
Calcite Ca(CO3)2 first time practiced at Pithalkhora, the chemical analysis data also repre-
Clinoferrosilite Fe(SiO3)
sent historical/technological development status of that period and
Cristobalite synthetic SiO2
Quartz (SiO2) about the improvement of methodology in later caves in western India.
The CHN data for the plaster also confirms the presence of organic
additive in the plasterworks of Pithalkhora as represented by percent-
microscope. The vegetal additives detected were rice husk of local ori- age of nitrogen, carbon and hydrogen. The percentage of carbon is rela-
gin. Fig. 14 shows the microscopic image of ancient rice husk extracted tively more comparing to hydrogen and nitrogen confirms that the
from the plaster and to that of new rice husk. The microscopic image presence of organic matter in the earthen plaster. The relative high per-
show closer similarity in the cellulose structure of rice husk. The pres- centage of nitrogen suggests the presence of proteins in the plaster as
ence of rice husk in the plaster has also been confirmed by the petro- the samples were not contaminated with bats urine. The possibility of
graphic study of the thin section images of the earthen plaster. It is contamination has been ruled out due to the observation during the col-
essential to use rice husk in required proportion during the preparation lection of samples. This is in agreement to FTIR spectra of the earthen
of new plaster for restoration being planned for Pithalkhora caves. plaster showing the peaks for either amides or other organic matter
pointing about mixing of organic materials.
The presence of larger grain size particles is the major reason for
3.10. Thermal analysis (DGA-DTA) greater porosity (about 25–35%) and less compactness of the
Pithalkhora plaster which is also highly prone to damage easily. This is
Thermogravimetry (TG) and differential thermal analysis (DTA) was in contrast to particle size data of earthen plaster of Ajanta and Ellora
performed simultaneously on ground sample at a flow rate of 40 ml/min caves wherein clay proportions noticed is about 15% and between 15
from 30° to 1000 °C using Pt/Rh crucible and alumina as reference ma- and 30%, respectively giving better compactness to the plaster (Singh
terial. As the boundaries between thermal events which occur in mor- and Arbad, 2014, 2015). Our investigation concludes about better selec-
tars during controlled heating are strictly arbitrary, it makes difficult tion of raw materials for preparation of earthen plaster of Ajanta/Ellora
to compare the results of thermal analysis given by various authors and improvement of technology after the initiation of first plastering
(Martens et al., 2005; Moropoulou et al., 1995; Bakolas et al., 1998). work at Pithalkhora. The result also clearly points reasons for damage
Therefore, we decided to apply commonly used temperature range to the Pithalkhora plaster on account of slight pressure. This is also in
which will best document thermal behaviour of ancient plaster. For agreement with utilization of silt loam soil for the earthen plaster
the examined sample no. 2, the typical thermal curves obtained by works of Ajanta and Ellora mostly sourced from the bank of nearby
DTA/DG analysis are shown in Fig. 15. river and its characteristic modified by adding other ingredients
232 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237

Fig. 7. a) and c) Thin section image of earthen plaster under petrological microscope with plane polarized light. b) and d) Thin section image of earthen plaster under petrological
microscope with crossed polar.

Fig. 8. a) and c) Thin section image of earthen plaster under petrological microscope with plane polarized light. b) and d) Thin section image of earthen plaster under petrological
microscope with crossed polar.
 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237 233

Fig. 9. Aggregate components in the earthen plaster of Pithalkhora separated by sieve.

(Singh and Arbad, 2014, 2015). There is only disagreement in clay per-
centage of Pithalkhora plaster (0–2%) with Ajanta and Ellora plasters
(5–15%) that may one of the reason for the failure of plaster. The com-
mon aggregates generally reported in the earthen plaster mixture are
sand and straw. As preparation of earthen plaster exploits local mate-
rials, different kinds of aggregates are not uncommon in the plaster of
varied geological set-up. To prepare a matching fresh plaster for restora-
tion, it is essential to focus on types and sizes of aggregates in the
plasterworks.
There may be possibility of inclusion of proteinaceous adhesive in
preparation the earthen plaster of Pithalkhora. The presence of
proteinaceous matter might have occurred from the proteic adhesive
or from the rice husk. This is in consistent to the FTIR results of Ajanta
and Ellora earthen plaster wherein additive proteic adhesives have
been reported (Singh and Arbad, 2014, 2015). However, no peak of cal-
cium oxalate was detected in the spectrum of Pithalkhora plaster
denoting absence of any oxidative product of proteic binding materials.
The FTIR spectra confirms the addition of proteic binding material in the
earthen plaster mix of Pithalkhora in agreement to the data reported
through CHN analysis of the plaster. It is noted that oxidation of proteic
material into calcium oxalate was detected in the earthen plaster works
of Ajanta cave (Singh and Arbad, 2015) through FTIR.

Fig. 10. XRD spectrum of earthen plaster sample no. 2.

234 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237

Fig. 11. XRD spectrum of earthen plaster sample no. 3, Pithalkhora caves.

Traditionally, the earthen plaster is prepared by mixing sand, silt and
clay. Straw is some time added in to the mix to prevent excessive cracking
during drying. In the earthen plaster with high aggregate (sand + silt) to
binder (clay) ratio (4:1), it has been observed that adding significant
quantity of straw helps to reduce undesirable shrinkages (Prost et al.,
1998). Whereas clay on drying forms a hard compact mass, this is not
the case with non-clay minerals because of their larger grain size. Due
to their small surface area, non-clays are much more irregular with re-
duced grain to grain contact surface and contact cohesion much less
than clay (Prost et al., 1998). However, for the earthen support to function
well, it is desirable to have an equitable distribution of sand, silt and clay.
More silt or sand in the earthen plaster is neither a good binder nor an ag-
gregate and that is prone to shrinkage/cracking and damage with small
pressure. In recent years the special clay such as sepiolite are being
added to mortars and cements to improve the engineering characteristics
(Maritinez-Ramirez et al., 1995; Andrejkovicˇová et al., 2011, 2012). A re-
cent study reported addition of sepiolite in the preparation of mortar for
earthen based wall restoration (Andrejkovicˇová et al., 2013). Sepiolite
clays are also used to absorb chemicals which can be dusted or sprayed
on plants or on the surface of the grounds (Murrey, 1999). The addition
of 10% sepiolite fibres enhances mechanical and physical properties of
the plaster (Kavas et al., 2004).

Fig. 12. FTIR spectrum of mud sample no. 3, Pithalkhora caves.

 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237 235

Fig. 13. SEM images of plaster samples of Pithalkhora caves at various magnifications (a. presence of calcite and pyroxene; b. presence of quartz and illite; c. presence of feldspar and calcite;
d. presence of montmorillonite, illite, pyroxenes and iron minerals).

Besides, in order to overcome inadaptability of local raw materials,
other inorganic materials such as calcite or slaked lime are often
added to earth mortar by just mixing during preparation of mortar
(Jerome and Stackley, 1993). It has been observed that adding lime to
the earthen materials; significantly influence its shrinkage properties
as lime works as cementing agent for the earthen mix (Austin, 1990).
However, the possibility of using reactive fillers in the earthen mix has
still to be fully explored. It seems that materials such as calcite, silica
and ferric oxide act like cementing agent forming chemical bonds
with clay minerals that may reduce swelling behaviour (Forth, 1990).
Protein may react chemically with clay of the earthen plaster by an ex-
change of inorganic cations in the clay with the organic, a mechanism
relating to ability of amino acid to encourage clay flocculation (Sikels,
1981).
The deterioration of wall paintings on earthen support is most often
due to loss of cohesion and adhesion of the support layers and its subse-
quent detachment. The earthen plaster begins to absorb moisture at
around 67% relative humidity (Fiero, 1997). The main mechanism for
deterioration of earthen plaster is the loss of mechanical strength, caus-
ing degradation of the binder which results the paint layer to lose

Fig. 14. Microscopic image of ancient rice husk along with the new one.
236 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237
Fig. 15. a) DG-DTA diagram of earthen plaster sample no. 2, cave no. 5. b) DG-DTA diagram of earthen plaster sample no. 3, cave no. 6, Pithalkhora.
adhesion and fall. In-situ climatic condition and addition of surface coat-
ings can be detrimental to paint layer as well as to the earthen fabric due
to differential dimensional change and difference in water vapour per-
meability (Chiavari et al., 2002). The surface films will inhibit transmis-
sion of water vapour through various substrate layers to the outer
surface. At Pithalkhora, the single most deterioration factor observed
is the seepages of rain water, retention of rainwater in clayey veins
within the rock matrix and high humidity noticed inside the cave.
When active deterioration is observed, any repair or restoration effort
without understanding the causes and mechanism of deterioration are
only temporary measures and often cause more damage in the long
run (Cather, 1999).
The deterioration of earthen plaster must have been occurred due to
its coarser nature and lesser percentage of clay present in it. From the
analytical studies it is found that the aggregates used for the preparation
of earthen plaster belong to the origin of basaltic rock. There may be loss
in calcium carbonate percentage due to diagenesis of the plaster with
water present in the caves. It is also identified that the minerals present
in the aggregates are quartz, calcite, plagioclase, illite/montmorillonite
and Iron minerals from petrographic studies, XRD, thin section analysis
and SEM analysis. The presence of rice husk and vegetal fibres has also
been confirmed by thin section studies, FTIR, CHN analysis and thermo
gravimetric analysis. The data sets will definitely be crucial in prepara-
tion of compatible restoration plaster.
The chemical composition of the component values obtained which
is also in close consonance to the composition of earthen plasters of
nearby sites of Ajanta/Ellora. Due to presence of large proportion of
coarse size grains, the earthen plaster of Pithalkhora is showing highly
rigid behaviour with greater tendency to deform or damage by grain
breakage. Based on the grain size distribution is plotted on sand–silt–
clay ternary diagram, it is confirmed that silt loam soil (soil composed
of more silt, less sand and negligible quantity of clay) was probably
sourced in the preparation of earthen plaster of Pithalkhora caves. This
may be the main reason for the early decay of the earthen plaster in
compared with the earthen plasters of nearby sites Ajanta/Ellora caves.
5. Research on conservation of earthen plasters
The published literatures available on the earthen plasters are in lim-
ited numbers. Dedicated investigations of material composition, struc-
tural support, plaster and paint layer etc. is needed for better
understanding to tailor the proper conservation strategies. The earthen
plaster samples may be required of few grams of at least 7 to 10 g for the
proper scientific investigation. The visual analysis also plays on vital role
to select the sample extraction and condition of the samples. The appro-
priate scientific analysis of may be carried in systematic way. The
analytical techniques such as XRF or Electron Microprobe Analyzer
(EMPA) may be employed for identification of chemical constituents.
The mineralogical investigation can be studied using X-ray diffractions
(XRD), petrological analysis, thin section studies, microscopic studies,
SEM. The chemical constituents may also be confirmed using FTIR
spectroscopy. The presence of lime binder may be confirmed with
thermo gravimetric analysis. The identification of sizes and shape
of the aggregates are most important in case of preparation of the
restoration plaster. The study of nearby soil and rocks will also give
the idea regarding the technology and constituents of the ancient
plasters.
6. Conclusion
From the scientific examination, it is revealed that the earthen plas-
ters of Pithalkhora caves are too coarser showing predominantly rigid
behaviour and prone to damage with slight pressure due to presence
of only traces of clay minerals for compactness of the plaster. As the
preparation for earthen plaster for decorative purpose might have
been first initiated at Pithalkhora, this research also throws light about
the technological development and improvement of methodology
while preparing earthen plaster for decorative works in later caves.
The traces of clay minerals encountered in some samples might
have given flexibility and strength to those plasters. The result
show utilization of silt loam raw soil for Pithalkhora plaster works.
As the aggregates are having the similar composition of local basaltic
origin, the restoration plaster will be synthesized using the aggre-
gates of the local basaltic rock. In addition, the vegetal remains in
the form of rice husk and vegetal fibres (around 10% weight) should
invariably be mixed in the plaster for compatibility and strength. The
required quantity of the slaked lime may be added to prepare the
restoration plaster. The result also teaches about the development
of earthen plaster technology in western India.
Acknowledgement
The authors are grateful to Prof. D.N. Singh, Dr. Kannan Iyer and Dr.
Somenath Mondal of Indian Institute of Technology, Mumbai, Prof.
M.M. Sardesai of Dr. Babasaheb Ambedkar Marathwada University, Au-
rangabad and S.N. Deshpande of Govt. Institute of Science, Auranagabad
for their kind help and support during investigations. Authors wish to
acknowledge the help of Dr. Samrat Mukherjee of BIT, Ranchi for ren-
dering help in analysis. We are thankful to our Director General and Di-
rector (Science) at Archaeological Survey of India, New Delhi for their
constant support during the course of this research.
 M. Singh et al. / Journal of Archaeological Science: Reports 7 (2016) 224–237
References
Andrejkovicˇová, S., Ferraz, E., Velosa, A.L., Silva, A.S., Rocha, A.S., 2011. Fine sepiolite ad-
dition to air lime–metakaolin mortars. Clay Miner. 46, 621–635.
Andrejkovicˇová, S., Ferraz, E., Velosa, A.L., Silva, A.S., Rocha, A.S., 2012. Air lime mortars
with incorporation of sepiolite and synthetic zeolite pellets. Acta Geodyn. Geomater.
9, 79–91.
Andrejkovicˇová, S., Ferraz, E., Velosa, A.L., Silva, A.S., Rocha, A.S., 2013. Air lime–
metakaolin–sepiolite mortars for earth based walls. Constr. Build. Mater. 44,
133–141.
Austin, George S., 1990. Adobe and related Building materials in New Mexico USA in 6th
International conference on the conservation of Earthen Architecture. In: Grimstad,
Kirstem (Ed.), Adobe Preprints, Las Cruces. New Mexico. USA. Getty Conservation In-
stitute, Los Angeles, pp. 417–423 (Oct 14–19).
Bakolas, A., Biscontin, G., Moropoulou, A., Palazzi, D., Zendri, E., 1998. Characterization of
structural byzantine mortars by thermogravimetric analysis. Thermochim. Acta 321
(1–2), 151–160.
Basma, Azm, S., Homound, Al., Abdallah, I., Husein, M., Mahamed, A., 1996. A swelling
shrinkage behavior of natural expansive clays. Appl. Clay Sci. 11 (2–4), 211–227.
Cather, S., 1999. Assessing causes and mechanism of detrimental changes to wall painting.
In: Gowing, R., Heritage, A. (Eds.), Conserving the Pained Past. Post prints of a confer-
ence organized by the English Heritage, London, pp. 67–74.
Chiavari, G., Ioelle, M., Prati, S., Santo, Padre, P., 2002. PY–GC–MS of the synthetic poly-
mers used in past restoration on Giotto's mud painting at the Scrovegni Chapel
(Pedova). Chromatographia 56 (11−12), 763–767.
Colao, F., Fantoni, R., Ortiz, B., Vazquez, c.M.A., Martin, J.M., Ortiz, R., Idris, N., 2010. Quarry
identification of historical building materials by means of laser induced breakdown
spectroscopy, X-ray fluorescence and chemometric analysis. Spectrochim. Acta B
65, 688–694.
Fergussons, James, Burgess, James, 2000. The Cave Temple of India. third ed. Munshiram
Manoharlal Publishers Pvt. Ltd, New Delhi, pp. 242–246.
Fiero, K., 1997. Cliff dwelling walls - the earthen plaster project at Mesa Verde. Cult. Res.
Manag. Bull. 20 (10), 38–39.
Forth, H.D., 1990. Fundamental of Soil Science. eighth ed. Wiley, New York.
Griffin, T.J., 1999. Earth grounds in wall painting conservation. An Investigation of their
Working Properties and Performance Characteristics. Courtauld Institute of Arts
(M.A. thesis).
Gurdal, E., 1976. A Research about Northern and Middle Anatolia Gypsums (Doctorate
thesis) ITU Faculty of Architechture, Istanbul, p. 32.
Houben, Hugo, Guilland, Hubert, 1984. Earth Conservation, A Comprehensive Guide. In-
termediate Technology Publication, London.
Jerome., Pamela, Stackley, 1993. Analysis of Bronze Age mud brick from Palaikastro Crete.
In: Alcada, Margarida (Ed.), 7th International Conference on the Study and Conserva-
tion of Earthen Architectures, Silves, Portugal, 24–29 October, Lisbon, pp. 381–386.
Karkanas, P., Yosef, O., Goldberg, P., Weiner, S., 2000. Diagenesis in prehistoric caves: the
use of minerals that form in situ to assess the completeness of the archaeological re-
cord. J. Archeol. Sci. 27 (10), 915–929.
Kavas, T., Sabah, E., Celik, M.S., 2004. Structural properties of sepiolite-reinforced cement
composite. Cem. Concr. Res. 34 (11), 2135–2139.
Kathleen, Fiero, 1997. Cliff dwelling walls – the earthen plaster project at Mesa Verde.
CRM Bull. 20 (10), 38–39.
Maritinez-Ramirez, S., Puertas, F., Blanco-Varela, M.T., 1995. Carbonation process and
properties of a new lime mortar with sepiolite. Cem. Concr. Res. 25, 39–50.
View publication stats
237
Martens, G., Elsen, J., Brutsaert, A., Deekers, M., Brulet, R., 2005. Physical and chemical
evaluation of lime mortars from Tournai (Belgium). International Building Lime Sym-
posium, Orlando, Florida.
Miller, T., 1934. Adobe or sun-dried brick for farm building. Farmer's Bulletin No.1720. De-
partment of Agriculture, Washington D.C., U.S.
Mitra, Debala, 1971. Buddhist Monument, Shri Mahendra Nath Dutt. first ed. Shishu
Sahitya Samsad Private Limited, Culcutta.
Mora, Paola, Mora, Laura, Phillippot, Paul, 1984. Conservation of Wall Painting.
Butterworth, London.
Moropoulou, A., Bakolas, A., Bisbikou, K., 1995. Characterization of ancient, byzantine and
later historic mortars by thermal and X-ray diffraction techniques. Thermochim. Acta
269-270, 779–795.
Murray, H.H., 1999. Applied clay mineralogy today and tomorrow. Clay Miner. 34, 39–49.
Nagaraju, S., 1981. Buddhist architecture of Western India (C.250 B.C.–C.A.D. 300). Agam
Kala Prakashan, 34, New Delhi.
Orazi, R., 1994. Mud-brick architecture in Sardinia, in Ceramics in architecture. In:
Vincenzini, P. (Ed.), Proceeding of the International Symposium in Architecture of
the 8th CIMTEC World Ceramic Congress and Forum on New Materials, Florence,
pp. 219–229 (June-28–July 1, Italy).
Prost, R., Koutit, T., Benchare, A., Haurd, E., 1998. State and location of water absorbed in
clay minerals-consequence of the hydration and swelling–shrinkage phenomenon.
Clay Clay Miner. 14 (2), 117–131.
Sickels, L.B., 1981. Organics vs. synthetics, their use as additives in the mortar. In Mortars,
Cements and Grouts Iron Is Used in the Conservation of Historic Building Symposium.
Ed. Rome. ICCROM, 25–52.
Sivaramamurti, C., 1978. Chitrasutra of the Vishnudharamottara. Kanal Publication, New
Delhi.
Sharma, R.K., Gupta, H.O., Kanotra, Y.K., 1995. Evaluation of physical characteristics of
‘mercula’ admixed clay for sealing of cracks in earthen plaster of monasteries, Ladakh
region. In: International Centre for the Study of the Preservation and the Restoration
of Cultural Property (Ed.), Methods of Evaluating Products for the Conservation of Po-
rous Building Materials in Monuments: International Colloquium, Rome, 19–21 June
1995: Preprints. ICCROM, Rome, pp. 41–49.
Shea, Fred, Watta, C.E., 1939. Dumas method for organic nitrogen. Ind. Eng. Chem. Anal.
Ed. 11 (6), 333–334. http://dx.doi.org/10.1021/ac50134a013.
Shedlovsky, T., 1960. Conductometry. In: Weissberger, A. (Ed.), Physical Methods of Or-
ganic Chemistry, 3d ed. Technique of Organic Chemistry vol. 1. Interscience Pub-
lishers, Inc., New York (pt. 4, chap. 45).
Spiro, M., 1960. Determination of transference numbers. In: Weissberger, A. (Ed.), Physi-
cal Methods of Organic Chemistry, 3d ed. Technique of Organic Chemistry vol. 1.
Interscience Publishers, Inc., New York (pt. 4, chap 46).
Singh, M., Arbad, B.R., 2014. Characterization of traditional mud mortar of the decorated
wall surfaces of Ellora caves. Constr. Build. Mater. 65, 384–395.
Singh, M., Arbad, B.R., 2015. Characterization of 4th–5th century A.D. earthen plaster sup-
port layers of Ajanta mural paintings. Constr. Build. Mater. 82, 142–154.
Van der Marel, H.W., Beutelspacher, H., 1976. Atlas of infrared spectroscopy of clay min-
erals and their admixtures. Elsevier Scientific Pub. Co., pp. 1–396 (Science).
Vitten, J.S., Sadler, Y.L., 1997. Particle size analysis of soils using laser light scattering and
X-ray absorption technology. Geotech. Test. J. 20, 63–73.
Warran, J., 1990. Conservation of Earth Structures. Butterworth, Heinemann, Oxford.