Journal of Archaeological Science: Reports 14 (2017) 332–339

Contents lists available at ScienceDirect

Journal of Archaeological Science: Reports

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

Chemical analyses of Roman wall paintings recently found in Paphos, MARK

Cyprus: The complementarity of archaeological and chemical studies
C. Balandiera, C. Joliotb, M. Ménagerb, F. Vouvec, C. Vieillescazesb,⁎
a
Avignon University, UMR 8210 AnHiMA, Site Hannah Arendt/Sainte-Marthe, 74 rue Louis Pasteur, 84000 Avignon, France
b
Avignon University, UMR IMBE CNRS 7263/IRD 237/AMU, Campus Jean-Henri Fabre, 301 rue Baruch de Spinoza, BP 21239, 84000 Avignon, France
c
LERM, 23 rue de la Madeleine, CS 60 136, 13631 Arles Cedex, France

A R T I C L E I N F O A B S T R A C T

Keywords: The ancient city of Nea (New) Paphos lies on the south-western coast of Cyprus. It has been a UNESCO World
Archaeometry Heritage site since 1980. Archaeological surveys and excavations led on Fabrika Hill, on the north-east acropolis
Hellenistic and Roman Paphos of the ancient city of Hellenistic and Roman Paphos revealed the existence of important Hellenistic quarries,
Cyprus water supplies and houses of the same period. An important Roman building was found, with traces of wall
Wall paintings
paintings in different spaces. In order to confirm the relative chronology of this Roman building given by ar-
chaeological stratigraphic and stylistic analyses, as well as technical observations, twenty-five samples of dif-
ferent kinds of wall paintings were collected for chemical analyses of inorganic and/or organic constitutive
materials. Frescoes have been identified and all the pigments were of mineral origin. At least two different
phases in the Roman building on the northern part of the site and another probably older kind of wall painting
on the southern part have been successfully differentiated.

1. Introduction. A Roman building found by the French
archaeological expedition at Paphos (Fabrika Hill). Archaeological
context and aims of the study
1.1. Ancient Paphos: site and archaeological research
The ancient city of Nea (New) Paphos lies on the south-western coast
of Cyprus, 15 km to the west of the famous sanctuary of Aphrodite in
Palae (Old) Paphos. It has been a UNESCO World Heritage site since
1980. If Agapenor is said to have created Paphos on his way back from
the Trojan War, no historical evidence can provide the name of its true
creator. Could it be Nikokles, the last Cypriot king of Paphos, or
Ptolemy, Alexander's general? The beginnings of New Paphos are still
being debated (Daszewski, 1987; Młynarczyk, 1990; Bekker-Nielsen,
2000; Balandier, 2014; Vitas, 2016). Despite the fact that different ar-
chaeological expeditions (Australian, Cypriot, Italian, and Polish ones)
had been led in Paphos (for a synthesis on the archaeological results see
Nicolaou, 1966 and Młynarczyk, 1990), the hill on the north-east side
of the ancient city of Hellenistic and Roman Paphos (Fig. 1), called
Fabrika Hill, remained unexcavated until 2008: only the area of the
theatre underwent research, being first excavated by Prof. Richard
Green and then by Dr. Craig Barker and Dr. Smadar Gabrieli for the
University of Sydney. However, the top of the hill and its surroundings
were not researched and remained outside the archaeological preserved
area (Green et al., 2004; Barker, 2016). Dr. Claire Balandier was first
approached by the former Director of the Department of Antiquities of
Cyprus, Dr. Pavlos Flourentzos, to work specifically on Fabrika Hill. The
French archaeological expedition at Paphos (MafaP: Mission arché-
ologique française à Paphos) was thus created under the auspices of the
French Ministry of Foreign Affairs and Avignon University, and the first
excavation season on Fabrika Hill started on 15th April 2008. The first
aim of the expedition was to make an inventory of the archaeological
remains and to identify the different occupation layers of the hill.
During the first archaeological season, a field survey and some ar-
chaeological soundings brought the fact to light that the hill was oc-
cupied from the Hellenistic period until modern times (a cemetery,
probably dating from the 19th century, was identified on the top of the
northern part of the hill).
1.2. New excavations on Fabrika Hill by the French archaeological
expedition at Paphos: Roman building and wall paintings: archaeological
description
Archaeological surveys and excavations revealed the existence of
important Hellenistic quarries, of houses of this same period and of
what was probably once part of the city wall. An underground aqueduct

⁎
Corresponding author: Campus Jean-Henri Fabre, 301 rue Baruch de Spinoza, BP 21239, 84000 Avignon, France.
E-mail address: cathy.vieillescazes@univ-avignon.fr (C. Vieillescazes).

http://dx.doi.org/10.1016/j.jasrep.2017.06.016
Received 3 January 2017; Received in revised form 2 June 2017; Accepted 11 June 2017
Available online 17 June 2017
2352-409X/ © 2017 Elsevier Ltd. All rights reserved.
C. Balandier et al. Journal of Archaeological Science: Reports 14 (2017) 332–339

Fig. 1. Map of the ancient city of Paphos with location of Fabrika Hill (after Balandier and Guintrand, 2016).

opening onto a large cistern used in the 2nd century AD, together with
an important Roman building, were also found (Balandier, 2015;
Balandier and Guintrand, 2016). In 2009 the southern wall of this
Roman building was unearthed at the foot of the northern edge of the
hill (Balandier 2011–2012).
A trench opened against the northern face of this wall revealed
traces of wall paintings. Many pieces of red lines on a white background
were found (Fig. 2).
In 2010, another wall was found 3.50 m in front of this first wall
with the remains of a window and a doorway (Fig. 3).
The north face of this wall still bore wall paintings but of a different
style (Figs. 4 and 5): red square panels separated by green and blue
frames (colors not visible on the picture, only on detail).
The continuation of the excavations (2010 to 2016, except 2013)
brought to light the fact that these remains belonged to a large building,
circa 20 m long east/west but whose eastern and northern limits have
not yet been accurately defined. At least nine spaces (rooms and
courtyards) were identified (Fig. 6).
The excavation of Space 1 reached as far as the foundation of the
building and indicated that it had been completely filled after the col-
lapse of the south wall (Balandier, 2013–2014; Balandier and
Guintrand, 2016). It appears that the building was built at the begin-
ning of the 1st century AD and that the fill of the room dates from the
Tiberian times.1
It is probably at that period that the building was partly damaged by
an earthquake or by the pressure of the ground soil. Space 1 was Fig. 2. Samples of fragments of wall paintings from Space 1 (©MafaP).
completely filled with debris and abandoned: the window and the
doorway were closed and the space situated to the north of it displayed and then became a closed space. This new room (Space 0) opened onto
wall paintings of a new kind: red panels with green frames. It seems that another space (Space 2) through its eastern wall, giving access to other
this space was first an open space (given that there was the window) spaces (Spaces 3, 5, 6, 7) to the south of an open space (main court-
yard). No floor remained but its level to just above the bedrock in Space
0 could be restituted. The excavation of the doorway showed that it had
1
Ceramic evidence was studied by Sandrine Elaigne (CNRS, Lyon). probably been narrowed in order to strengthen it after the damage to

333
C. Balandier et al.
Fig. 3. The wall with a window and a doorway and wall paintings in course of pre-
servation before being removed. In the background, the back (South) wall of the Roman
house (©MafaP).
Fig. 4. In situ wall paintings on the southern wall of Space 0 (East of the doorway)
(©MafaP). (For interpretation of the references to color in this figure legend, the reader is
referred to the web version of this article.)
Fig. 5. In situ wall paintings on the southern wall of Space 0 (West of the doorway)
(©MafaP). (For interpretation of the references to color in this figure legend, the reader is
referred to the web version of this article.)
Space 1's back wall: traces of wall paintings were observed inside the
wall, behind the northern pillar of the doorway, abutted against the
pillar of the first phase which still bore stucco (Balandier 2013–2014;
Balandier and Guintrand, 2016 - 132). This new pillar still bears some
plaster with a herringbone incision called arriccio, ready to receive the
different thin layers of the fresco: the same herringbone incision was
observed by E. Hadjistephanou in the House of Orpheus excavated by
334
Journal of Archaeological Science: Reports 14 (2017) 332–339
Prof. D. Michaelides for the University of Cyprus: “These frescos were
executed using the pontata technique. On the arriccio layer on the
plaster the same technique of applying the herringbone was evident.
This herringbone on arriccio can could also be seen on the walls of the
Orpheus site” (Hadjistephanou, 2016 - 390). This observation is an
interesting indication, since the House of Orpheus is dated late 2nd or
early 3rd century AD (Michaelides, 1991 - 4), meaning Septimius Se-
verus' reign (Fig. 7).
The Roman building on Fabrika Hill was seriously damaged on its
northern part but the excavation showed that another space (Space 4)
existed to the north of Space 0: small fragments of wall paintings were
still in situ on the north face of the first course of the northern wall.
1.3. Stylistic and technical analyses of the wall paintings
The first stylistic analyses were made by Dr. Fr. Alabe2 where dif-
ferent types of decoration models and styles could be distinguished and
where the conclusion was drawn that only the fragments coming from
the fill of the south-east slope of the hill, below the Roman cistern,
could belong to the Hellenistic period. It was also concluded, however,
that the two or three main different styles of wall painting observed on
the fragments taken from the destruction levels of the building un-
covered on the northern side of the hill, all belong to the Roman im-
perial period, clearly to the so-called second Pompeian style for one
particular kind, and perhaps also to the 1st and eventually to the 3rd
Pompeian style for the others.
Practical technical observations on the making of the wall paintings
were made by E. Hadjistephanou: “These Roman mural paintings use
the fresco technique of the wet on wet application. Plaster layering is a
more complicated process in houses, compared with that of tombs. It is
possible that the construction of the walls or the flat polishing of the
surface, sometimes allowed for the application of multiple thin layers
over an arriccio. The wall paintings on the sites in Nea Paphos generally
consist of a mix of calcareous aggregates, rock fragments and straw
fibers. The technique of the herringbone application was also used on
arriccio, where the composition of plaster was usually 3:1 and 2:1 in
intonaco. Prior to the pigments, is the final application layer, consisting
of a layer of impasto lime followed by a polishing carried out at a very
precise moment. When the plaster is too wet, the colors tend to mix
with the lime surface, aided by the movements of the brush and, as a
result, they change tone. When too dry, they are imperfectly fixed and
in this case, considerable loss in adhesion can be expected. The calcium
hydroxide fails to mix in the water with which the pigments are ap-
plied. This operation was carried out quickly and the artist usually
finished the job or the decorations in the same day. It is not easy to
distinguish the plaster joints, perhaps because of the artist's diligence
when polishing the plaster. Here one can easily observe the deep
brushstrokes, proving that the paint surface was wet. As far as the
colors are concerned, Cyprus is very rich in minerals, with areas such as
Troodos full of umber, sienna, red-yellow ochre. Carbon black is present
too. Copper is the metal of Cyprus and the main constituent in the
production of Egyptian blue, an important color frequently found in
tombs, despite its rarity” (Hadjistephanou, 2016 - 386).
In order to approve the relative chronology of this Roman building
given by archaeological stratigraphic and stylistic analyses, as well as
technical observations, samples of the different kinds of wall paintings
were taken for chemical analyses, in June 2011. The archaeological
question behind the decision to carry out scientific analyses on samples
of wall paintings recovered from the excavation site was to verify if the
results of the stratigraphic and stylistic analyses could be confirmed by
the chemical ones: are there different types of wall paintings?
2
Françoise Alabe (Tours University) made a first inventory and study of the wall
paintings in July 2011.
C. Balandier et al.
Fig. 6. Plan of the Roman house: Location of the different spaces in 2016 (conception: C. Balandier and drawing: E. Rossignol).
Fig. 7. Remains of plaster layer with the herringbone incision (arriccio) (“House of
Orpheus”, © C. Balandier.
2. Materials and methods
2.1. Sampling
In order to analyze inorganic and/or organic constitutive materials,
twenty-five samples of the wall paintings were collected following the
methodology for sampling from materials of cultural property (i.e.
European Committee for Standardization EN 16085:2011–12). Due to
the destructive nature of sampling, the samples were carefully chosen
from areas that had no aesthetic or iconographic value for future re-
construction.
2.2. Optical and scanning electron microscopy
The polished cross-section of paint fragments was elaborated by
335
Journal of Archaeological Science: Reports 14 (2017) 332–339
using a polyester resin Sody 33 and flexible silicone molds TS (ESCIL,
Chassieu, France). The polishing was done on a variable speed polishing
apparatus ESC 200 GTL (ESCIL). The stratigraphy of the different
samples was examined with a microscope Leica DMLB 100 equipped
with a digital photo camera Nikon Coolpix S.
The examination of cross sections were carried out by scanning
electron microscopy (JEOL JSM 6490 LV) and analyzed by electron
probe microanalysis SAM. The backscattered electron technique (High
Vacuum, 20 kV acceleration voltage) on polished sections was used
previous to elemental chemical analysis using EDXS. Secondary elec-
tron imaging on coated carbon (High Vacuum, 20 kV acceleration
voltage) provided high resolution images of the cross sections.
2.3. Microchemical tests
The identification tests were carried out directly on the untreated
samples. By means of these qualitative tests, cations and anions can be
detected in a single pigment particle, a particle as small as is possible to
isolate microscopically. The method has good sensitivity for mineral
pigments and fillers, using predominantly acid-base, redox and com-
plexation chemical reactions (Charlot, 1980; Ashok, 1993; West
Fitzhugh, 1997; Delamare and Guineau, 1999). Solvents and reagents
were purchased from Merck (Darmstadt, Germany). The analyses and
observations were carried out with a binocular microscope and by
testing a positive blank for each chemical reaction. More particularly,
cations as Fe2 +, Fe3 +, Pb2 +, Ca2 +, Al3 +, Mn2 + and anions as CO32 −,
SO42 −, PO42 − and S2 − were looked after.
2.4. FT-IR
The FT-IR spectrometer was a Thermo-Nicolet AVATAR 360 FT-IR
in transmission mode with OMNIC 32 software. FT-IR spectra were
collected in the 400 to 4000 cm− 1 range recording 64 scans per
C. Balandier et al. Journal of Archaeological Science: Reports 14 (2017) 332–339

spectrum. 5 mg of each sample was mixed with 150 mg of KBr,
homogenized and pressed under 10 T·cm− 2 in order to form a thick KBr
pellet. These samples were directly analyzed by infrared spectroscopy.
All spectra were baseline corrected and smoothed. The principal
Component Analyses (PCA) were performed on the first derivative of
whole FT-IR spectrum using TQ analysis software with mean centering
data and a covariance algorithm.
3. Results and discussion
3.1. Stratigraphy
A stratigraphic study of the twenty-five samples revealed groups
with common values, and, each group was well differentiated, showing
various artistic and technical implementations (Vieillescazes et al.,
2016). An evolution of the techniques clearly appeared, ranging from
directly applying a paint layer on lime mortar/sand, grey (House, Space
1) or white with sometimes a brown lower mortar (House, Space 2 and
3), to a more elaborate construction of the pictorial matter. In this last
case, an intermediate ground layer whitewash, grey (House, Space 4
and courtyard) or white (Area -Cistern), was added, including many
fragments of shells. It is important to note that the future of a painting
and especially its degradation through time will usually depend on
which technique of execution was applied. In our case, frescoes have
been identified, and they are fortunately quite strong and time-re-
sistant. Table 1 summarizes the characteristic observations.
3.2. Microchemistry
Microchemical tests led to the conclusion that all the pigments were
of mineral origin (see Table 1). Iron oxides, whose coloring power is
very strong, are those which were mainly identified. Iron oxides were
present either as pure oxides for red colors or mixed with impurities for
yellow ochres. These results are consistent with the pale appearance
observed. Blacks were more difficult to identify. In accordance with
their dark or very dark intensity, they also contained oxides: dark
sienna or black iron oxide pigments (Mora and Philippot, 1977;
Stefanaggi, 1997; Barbet, 2009).
3.3. FT-IR spectroscopy
All the plasters and mortars were analyzed by FT-IR spectroscopy.
The different infrared spectra exhibited the same chemical fingerprint
as the one given in Fig. 8. The bands centered on 712, 876, 1432, 1797
and 2514 cm− 1 are characteristic of calcite (Al-Hosney and Grassian,
2005; Zhang et al., 2012; Derrick et al., 1999). This indicates the use of
lime in the final plaster and in the preparatory layer. Furthermore, the
weak band at 1082 cm− 1 indicates the occurrence of quartz, owing to
the presence of sand in the sample. Furthermore, we found faint signals
of clays with an absorption band centered on 1028 cm− 1 (SieOeSi)
and, sometimes, the two peaks characteristic of iron oxide at 521 and
454 cm− 1 (Bellamy, 1975; Hein et al., 2009; Miller and Wilkins, 1952).
All the infrared absorption of plasters and mortars differ only by the
global intensity of the chromophore absorbance, pointing notably to
differences in the quartz/calcite and clay/calcite ratios.
In order to differentiate between the formulations of plasters and
mortars, a statistical approach based on principal component analysis
was applied. The most significant principal component scores were PC1
and PC2, which show up to 66,7% of the total variability of the sample.
The results are given in Fig. 9.
Such a statistical approach allows the samples to be classified into
five different groups corresponding to different plaster or mortar for-
mulations. This result, with its five different groups: the fine white and
grey plasters, and the white, grey and brown mortars, is fully consistent
with the layers previously described in the stratigraphic study. The
intraspecific similarities in the infrared spectrum of each group validate
the sampling protocol and the groups of layers carried out by micro-
scopy.
Furthermore, the different excavation locations share many simila-
rities: (i) All the plasters had substantially different compositions from
those of mortars, (ii) The cistern area, Space 2, Space 3 and restored
spaces showed similar white mortar, (iii) Space 1, Space 4 and the
courtyard presented analogous grey mortar. The differences between
different locations are essentially due to the evolution of the coating
and painting techniques previously discussed in the stratigraphic study.
The identification of the pigment was obtained for all colors and
samples. In all the different excavations and locations, soil-based ma-
terial, such as ochre or clay containing a high proportion of iron oxide

Table 1
Stratigraphic study: characteristic observations.

Sample Pictural layer Fine grey plaster Fine white plaster White mortar Grey mortar Brown mortar

H. S. 1.1 Iron oxide (red ochre) Lime, sand

H. S. 1.2 Iron oxide (red/yellow ochre) Lime, sand
H. S. 2.1 Iron oxide (brown-earth) Lime, sand +
H. S. 2.2 Iron oxide (red ochre) Lime, sand +(little)
H. S. 2.3 Iron oxide (red ochre) Lime, sand Indeterminate
H. S. 3.1 Iron oxide (red ochre) Lime, sand +
H. S. 3.2 Iron oxide (red ochre) Lime, sand +
H. S. 4.1 Iron oxide (red ochre) Indeterminate Lime, sand
H. S. 4.2 Black iron oxide with chalk white Lime, shells Lime, sand
H. S. 4.3 Iron oxide (red ochre) Lime, shells Lime, sand
H. S. 0.1 Iron oxide (red ochre) Lime, shells Lime, sand
H. S. 0.2 Chalk white Lime, shells Lime, sand
H. S. 0.3 Iron oxide (red ochre) Lime, shells Lime, sand
H. S. 0.4 Black iron oxide Lime, shells Lime, sand
H. S. 0.5 Iron oxide (red/yellow ochre) Indeterminate Lime, sand
H. S. 0.6 Iron oxide (red/yellow ochre/dark sienna) Lime, shells Lime, sand
A.C. 1 Black iron oxide Lime, shells Lime, sand
A.C. 2 Iron oxide (red ochre) Lime, shells Lime, sand
A.C. 3 Iron oxide (red ochre) Lime, shells Lime, sand
A.C. 4 Black iron oxide Lime, shells Lime, sand
R. S. 1 Iron oxide (red ochre) Lime, shells
R. S. 2 Iron oxide (red/yellow ochre) Lime, shells
R. S. 3 Iron oxide (red ochre) Lime, shells
R. S. 4 Iron oxide (red/yellow ochre) Lime, shells

Keys: H.S. X: House (North), Space (1, 2, 3, 4), sample n° X; H. S. 0: courtyard; A.C.: Area Cistern (South East); R. S.: Restored Space.

336
C. Balandier et al. Journal of Archaeological Science: Reports 14 (2017) 332–339

Fig. 8. Infrared spectrum of a grey plaster from Space 1.

Two main techniques are commonly observed in mural paintings: (i)

al secco, where the pigments, dissolved in an organic binder, are applied
on a dry plaster, (ii) al fresco, where the pigments, dissolved in a sa-
turated aqueous solution of lime, are applied on a wet lime plaster
(Perego, 2005). As noted in Fig. 10, the painting layer contains a high
proportion of calcite and no trace of any organic binder. On the basis of
these results, it is concluded that the pigments in this case were applied
using the al fresco technique.

3.4. SEM

The examination of seven cross sections with five samples from the
house on the north and two samples from the courtyard, allowed the
following to be determined:
Fig. 9. Principal component analysis of infrared spectrum of plasters and mortars.
⁎ Three layers are well differentiated on each sample, with one layer
of mortar and two paint layers,
with very similar infrared profiles were found (cf. Table 1). Such a
⁎ The mortar matrix is composed of a binder of lime including a large
result gives a very clear idea of the diversity of natural earth pigments
amount of distinct aggregates, principally limestone, shelly fossils
available in Cyprus at that time. All materials exhibit almost the same
and plants, and more rarely, siliceous inclusions. The matrix is co-
infrared fingerprint leading to the same identification.
herent on the top of the layer and the contact with aggregates is
Fig. 10 gives: (i) the infrared spectrum of a red pigment of one
successfully done. The mortar is more friable at the bottom of the
sample from Space 1 (ii) its value with the subtraction of the above
layer. Most of the aggregates in the samples are between 200 μm
plaster layer done in order to avoid the calcite absorptions. The char-
and 1 mm and correspond to limestone and fossil fragments. Other
acteristic bands of kaolinite and ochre with the bonded OH group
aggregates are composed of fine particles (< 15 μm) and are silic-
stretching at 3616, 3692 and 914 cm− 1 and the main absorption band
eous in nature (Fig. 11).
at 1032 cm− 1 can be thus clearly seen (Bikiaris et al., 2000; Chukanov,
⁎ The paint layers were applied to fresh (wet) plaster, with a mixture
2014; Miller and Wilkins, 1952). The free OH group of kaolinite absorbs
of lime, pigments and filler. Paint-colored layers were most fre-
around 3436 cm− 1, and the water bonded signal in the kaolinite sheet
quently directly applied on the substrate. In only one case, a white
at 1637 cm− 1 (Miller and Wilkins, 1952). The presence of ochre is also
layer was applied first, probably a preparatory wash with lime.
shown by the band at 794 cm− 1, and the two strong peaks at 524 and
These colored layers are very thin (< 20 μm).
468 cm− 1 characteristic of Fe2O3 (Hein et al., 2009). Such a result
⁎ The pigments employed are: lime, colored clays, red iron oxide, and
clearly indicates the presence of an ochre, a natural earth material,
Egyptian blue, mixed with a crushed limestone filler (Fig. 12).
composed of varying amounts of octahedral iron oxides, alumino-sili-
⁎ The fresco painting technique seems to be the one applied.
cate clays and calcium compounds (Gettens and Stout, 2011). More-
over, kaolinite is present in various types of ochre (Bikiaris et al., 2000)
from 0 to 26,75% (Elias et al., 2006).

337
C. Balandier et al.
Fig. 10. (a) Infrared spectrum of a red pigment layer of Space 1 (b) Result of the subtraction of the red pigment spectra by the above plaster layer. (For interpretation of the references to
color in this figure legend, the reader is referred to the web version of this article.)
Fig. 11. Sample R.S.2: view of a cross section under SEM - image of the atomic contrast
(©LERM).
4. Conclusion
Chemical analyses confirmed the stratigraphic and stylistic studies
of the different wall paintings found on Fabrika Hill. They showed that
the two different kinds of decoration observed in the Roman building,
in Space 0 on one hand, and Spaces 1, 2 and 3 on the other, were not
made with the same materials and techniques. This is therefore further
evidence to support the chronological proposal that they belonged to
two different phases of the Roman building. Moreover, chemical
338
Journal of Archaeological Science: Reports 14 (2017) 332–339
Fig. 12. Identification of Egyptian blue pigment used into a yellow pictural layer (sample
R.S.2.)-Elementary analysis (©LERM).
analyses made on samples taken on the eastern slope of the same hill,
below the large Roman cistern, helped to determine that these frag-
ments were of another kind: they do not belong to the same building;
they were found in a fill and probably come from destroyed Hellenistic
buildings, first destroyed by an earthquake and then filled to rebuild the
area at the very end of the 1st century BC or the very beginning of the
1st century AD. To resume, Hellenistic quarries were filled at the very
end of the 1st century BC, perhaps after the well-known earthquake of
15 BCE when Emperor Augustus gave funding to rebuild the city. The
whole area was then covered by a thick layer – circa 4 m - of earth
C. Balandier et al.
mixed with fragments of Hellenistic artefacts. At the beginning of the
1st century AD, a large Roman building was built. It was damaged soon
after its construction, probably by an earthquake (under Tiberius'
realm?), and it appears that its southern rooms had been filled and
abandoned. It was then restored and new wall paintings were decorated
in Space 0.
The chemical analyses confirmed both the stratigraphic and stylistic
analyses of the wall paintings made by the archaeologists and the
technical observations made by the restorers3: we have at least two
different phases in the Roman building on the northern part of the site
and another probably older kind of wall painting on the southern part.
Archaeological excavations showed that the first phase of the Roman
building, on the North side of the hill, decorated with a first kind of wall
paintings, was destroyed by an earthquake and two rooms of the
southern wing were totally filled and abandoned. The new phase of the
house was decorated with wall paintings of another kind. On the
southern part of the hill, the fragments of wall paintings present inci-
sions showing that they belong to a previous period (probably late
Hellenistic).
The comparison of the archaeological results with the chemical ones
is significant, and it confirms the need for archaeologists and chemical
researchers to work together (Vieillescazes et al., 2016).
Finally, the components analyzed are known for this period and
they are common in Cyprus, a country rich in raw materials. The pro-
duction of lime had been known since the Neolithic period and the
colored clays continue to be extracted to this day (Philokyprou, 2010).
Given that there were copper deposits on the island, the Egyptian blue
was probably made on site. Two types of sand were used: sand from
quarries and sand extracted from the sea.
Conflicts of interest
The authors declare that they have no conflicts of interest.
Acknowledgment
This work was supported by a financial grant from the French
Ministry of Foreign Affairs and by the University of Avignon. We want
to thank the Department of Antiquities of Cyprus and the Museum of
Paphos for their support, especially the restorer of wall paintings,
Vangelis Hadjistephanou, and its team.
References
Al-Hosney, H.A., Grassian, V.H., 2005. Water, sulfur dioxide and nitric acid adsorption on
calcium carbonate: a transmission and ATR-FTIR study. Phys. Chem. Chem. Phys. 7,
1266. http://dx.doi.org/10.1039/B417872F.
Ashok, R., 1993. Artists' Pigments: A Handbook of Their History and Characteristics. vol.
2 National Gallery of Art and Oxford University Press, Washington and New York.
Balandier, C., 2011–2012. La mission archéologique française à Paphos: résultats de la
deuxième campagne de fouilles sur la colline de Fabrika (2009). In: Report of the
Department of Antiquities, Cyprus, (in press).
Balandier, C., 2013–2014. Découverte d'une maison romaine sur la colline de Fabrika.
Résultats des travaux de la mission archéologique française à Paphos en 2010. In:
Report of the Department of Antiquities, Cyprus, (in press).
Balandier, C., 2014. Des anciennes capitales de royaumes aux nouvelles villes portuaires:
réflexions sur l'évolution du réseau urbain de Chypre à l'époque hellénistique. In:
Balandier, Cl., Chandezon, Chr. (Eds.), Institutions, sociétés et cultes de la
Méditerranée antique. Mélanges d'histoire ancienne rassemblés en l'honneur de
Claude Vial, Scripta Antiqua 58, Éditions Ausonius, Bordeaux, pp. 179–209.
Balandier, C., 2015. Hellenistic and Roman Paphos. The French archaeological expedition
on Fabrika hill. In: Kyprios Character. History, Archaeology and Numismatic of
3
The wall paintings have been remove for restoration and conservation by Vangelis
Hadjistephanou (Restorer at the Department of Antiquities of Cyprus) and his team. See
Hadjistephanou, 2016. A programme of preservation and improvement of Fabrika hill is
under study: a shelter will be built upon the remains of the Roman building.
339
Journal of Archaeological Science: Reports 14 (2017) 332–339
Ancient Cyprus, . http://kyprioscharacter.eie.gr/en.html.
Balandier, C., Guintrand, M., 2016. Fabrika, un quartier résidentiel à Paphos? Résultats
archéologiques et réflexion historique sur l'évolution urbaine du secteur Nord-Est de
la ville antique. In: Balandier, C. (Ed.), Nea Paphos. Fondation et développement
urbanistique d'une ville chypriote de l'antiquité à nos jours. Etudes archéologiques,
historiques et patrimoniales, Ausonius Editions, Mémoires 43, Bordeaux, pp.
121–143.
Barbet, A., 2009. La peinture murale romaine, les styles décoratifs pompéiens, 2nd ed.
(Picard, Paris).
Barker, C., 2016. The theatre precinct of Nea Paphos (Fabrika): recent excavations by the
Australian mission. In: Balandier, C. (Ed.), Nea Paphos. Fondation et développement
urbanistique d'une ville chypriote de l'antiquité à nos jours. Etudes archéologiques,
historiques et patrimoniales, Ausonius Editions, Mémoires 43, Bordeaux, pp. 91–104.
Bekker-Nielsen, T., 2000. The Foundation of Nea Paphos. In: Isager, S., Nielsen, I. (Eds.),
Proceedings of the Danish Institute at Athens III, Athens, pp. 195–207.
Bellamy, L.J., 1975. The Infra-red Spectra of Complex Molecules. Methuen and Co Ltd,
London.
Bikiaris, D., Daniilia, S., Sotiropoulou, S., Katsimbiri, O., Pavlidou, E., Moutsatsou, A.P.,
Chryssoulakis, Y., 2000. Ochre-differentiation through micro-Raman and micro-FTIR
spectroscopies: application on wall paintings at Meteora and Mount Athos, Greece.
Spectrochim. Acta A 56, 3–18. http://dx.doi.org/10.1016/S1386-1425(99)00134-1.
Charlot, G., 1980. Analyse qualitative rapide des cations et des anions. Dunod Université.
Chukanov, N.V., 2014. Infrared Spectra of Mineral Species. Springer Netherlands,
Dordrecht.
Daszewski, W.A., 1987. Nicocles and Ptolemy: remarks on the early History of Nea
Paphos. In: Report of the Department of Antiquities, Cyprus, pp. 171–175.
Delamare, F., Guineau, B., 1999. Les matériaux de la couleur. Gallimard, Paris.
Derrick, M.R., Stulik, D., Landry, J.M., 1999. Infrared Spectroscopy in Conservation
Science. Getty Conservation Institute, Los Angeles.
Elias, M., Chartier, C., Prévot, G., Garay, H., Vignaud, C., 2006. The color of ochres ex-
plained by their composition. Mater. Sci. Eng. B 127, 70–80. http://dx.doi.org/10.
1016/j.mseb.2005.09.061.
Gettens, R.J., Stout, G.L., 2011. Painting Materials: A Short Encyclopedia. Dover
Publications, New York.
Green, G.R., Barker, C.D., Gabrieli, R.S., 2004. Fabrika: An Ancient Theatre of Nea
Paphos, Nicosia.
Hadjistephanou, E., 2016. Practical approaches in the conservation of wall paintings in
Nea Paphos. In: Balandier, C. (Ed.), Nea Paphos. Fondation et développement urba-
nistique d'une ville chypriote de l'antiquité à nos jours. Etudes archéologiques, his-
toriques et patrimoniales, Ausonius Editions, Mémoires 43, Bordeaux, pp. 375–391.
Hein, A., Karatasios, I., Mourelatos, D., 2009. Byzantine wall paintings from Mani
(Greece): microanalytical investigation of pigments and plasters. Anal. Bioanal.
Chem. 395, 2061–2071. http://dx.doi.org/10.1007/s00216-009-2967-6.
Michaelides, D., 1991. The house of Orpheus. In: Stanley-Price, N. (Ed.), The
Conservation of the Orpheus Mosaic at Paphos, Cyprus. The Getty Conservation
Institute, Burbank, California, pp. 3–10.
Miller, F.A., Wilkins, C.H., 1952. Infrared spectra and characteristic frequencies of in-
organic ions. Anal. Chem. 24, 1253–1294. http://dx.doi.org/10.1021/ac60068a007.
Młynarczyk, J., 1990. Nea Paphos III Nea Paphos in the Hellenistic Period. Centre
d'archéologie méditerranéenne de l'Académie polonaise des Sciences et Centre po-
lonais d'archéologie méditerranéenne de l'Université de Varsovie au Caire, Varsovie.
Mora, P.L., Philippot, P., 1977. La conservation des peintures murales, Ed. Compositori,
Bologne.
Nicolaou, K., 1966. The Topography of Nea Paphos. In: Bernhard, M.-L. (Ed.), Mélanges
offerts à Kazimierz Michalowski, Varsovie, pp. 561–601.
Perego, F., 2005. Dictionnaire des matériaux du peintre. Belin, Paris.
Philokyprou, M., 2010. The earliest use of lime mortar and gypsum mortars in Cyprus. In:
Second Historic Mortars Conference HMC 2010 and Rilem 202-RHM Final Workshop,
pp. 309–316.
Stefanaggi, M., 1997. Les techniques de la peinture murale, cours international.
Université Paris XII.
Vieillescazes, C., Joliot, C., Ménager, M., 2016. Les peintures murales de la maison de
Fabrika: analyses chimiques. In: Balandier, C. (Ed.), Nea Paphos. Fondation et
développement urbanistique d'une ville chypriote de l'antiquité à nos jours. Etudes
archéologiques, historiques et patrimoniales, Ausonius Editions, Mémoires 43,
Bordeaux, pp. 145–151.
Vitas, D., 2016. The foundation of Nea Paphos: a new Cypriot city or a Ptolemaic ka-
toikia? In: Balandier, C. (Ed.), Nea Paphos. Fondation et développement urbanistique
d'une ville chypriote de l'antiquité à nos jours. Etudes archéologiques, historiques et
patrimoniales, Ausonius Editions, Mémoires 43, Bordeaux, pp. 241–248.
West Fitzhugh, E., 1997. Artists' Pigments: A Handbook of Their History and
Characteristics. vol. 3 National Gallery of Art and Oxford University Press,
Washington and New York.
Zhang, Z., Xie, Y., Xu, X., Pan, H., Tang, R., 2012. Transformation of amorphous calcium
carbonate into aragonite. J. Cryst. Growth 343, 62–67. http://dx.doi.org/10.1016/j.
jcrysgro.2012.01.025.