Journal of Archaeological Science: Reports 57 (2024) 104595

Contents lists available at ScienceDirect

Journal of Archaeological Science: Reports

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

Plasters and mortars from the theatre in Nea Paphos (Cyprus): A

multidisciplinary study
Paola Pizzo a, b, *, Jan Válek a, Petr Kozlovcev a, Dita Frankeová a, Alberto Viani a
a
Institute of Theoretical and Applied Mechanics of the Czech Academy of Sciences, Czech Republic
b
Institute of Classical Archaeology, Faculty of Arts, Charles University, Czech Republic

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

Keywords: The study of plasters and mortars constitutes a relevant tool in reconstructing the technological knowledge of the
Plaster workforces at play. Through a multidisciplinary and multiscale approach, this project characterizes selected
Hydraulic lime samples from the Hellenistic-Roman theatre in Nea Paphos (Cyprus), highlighting the characteristics of the raw
Gypsum
materials, the productive processes, and the functional properties. Petrographic analysis combined with x-ray
Mortar analysis
Cypriot archaeology
diffraction, thermal analysis and scanning electron microscopy allowed to collect a complete set of elemental,
Theatre architecture mineralogical and chemical data, all essential for the correct characterization of the samples.
According to our results, it appears clear that both Hellenistic and Roman workforces were able to exploit all
the raw materials locally available, displaying a remarkable knowledgeability on how to enhance the perfor­
mance of the mortar mixtures creating a durable and resistant material that survived to this day. Our results
display how a conscient choice was made in the selection of the typologies of plasters and mortars for the
fulfilment of different purposes. Furthermore, natural hydraulic lime (NHL) was identified in association with a
water-related structure.

1. Introduction of our first case study, the theatre of Nea Paphos. The profuse avail­
ability of material in this specific site allowed us to use it as a trial and a
Plasters and mortars have regained a prominent role in the Cypriot stand-alone case study (see Table 1).
research field during the last few decades (see the works of Amadio, The foundation of this research consisted in agreeing in a univocal
2018; Balandier et al., 2017; Theodoridou et al., 2013; Philokyprou, definition of what plasters and mortars are. De facto, different science
2012a, 2012b). Although interest in understanding this widely available fields have different definitions. The main components of plaster and
material has risen, studies concretely focusing on them are still scarce mortar mixtures are binders – the choice of which largely depends on the
and are mostly interconnected with the field of conservation rather than state of the art, fashion and the geological availability (Marinowitz et al.,
archaeology. The techniques used to characterize the samples are 2012; Philokyprou, 2012a) – and aggregates, whose function is to
similar in both approaches, but the differentiation in aims and objectives enhance the properties of the mixture (Weiner, 2010, 185, 189). Lime is
between the two disciplines have resulted in limited intercommunica­ among the most commonly historically employed binders (Theodoridou
tion and exchange of data between the two fields. The lack of a et al., 2013, 3263; Weiner, 2010, 185, 186); thus, unexceptionally, lime
comparative database, and in some cases of any available published binders are widespread in Cyprus, geologically rich in limestone de­
data, has pushed this research team to start investigating the plaster posits (Philokyprou, 2012a, 2012b). Other binders commonly found in
production in Cyprus in a wide area and across a protracted chrono­ the archaeological record include gypsum and mud. For this research,
logical span. In particular, our research has been focusing on the wide mud plasters have been not taken into account, as our main aim is to
region of modern-day Paphos, an area interested in intense settling ac­ answer open questions concerning the study of the lime and gypsum
tivities since the Early Bronze Age. Five archaeological sites, across the plaster production in Cyprus. In particular, our attention was focused on
Late Bronze Age to the Early Roman periods, have been studied in order the characterization of the binders, with the specific purpose of recog­
to identify synchronic and diachronic developments or changes in the nizing whether or not there would be compatibility with the raw ma­
industry of plaster production. In this article, we will present the results terial locally geologically available. Considering the long life of the

* Corresponding author at: Institute of Theoretical and Applied Mechanics of the Czech Academy of Sciences, Czech Republic
E-mail address: pizzo@itam.cas.cz (P. Pizzo).

https://doi.org/10.1016/j.jasrep.2024.104595
Received 7 March 2024; Received in revised form 9 May 2024; Accepted 13 May 2024
Available online 22 May 2024
2352-409X/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
P. Pizzo et al. Journal of Archaeological Science: Reports 57 (2024) 104595

Table 1 theatre, we attempted a sampling strategy covering the different phases

Briefly summarises the major features of the paphian theatre from the founda­ of the theatre’s main structures, with the aim of identifying eventual
tion, around 300 BCE, to the final abandonment. All data and information chronological changes in fashion, or technological developments.
contained in the table are reported in the website of the paphos theatre In order to achieve our aim, we employed a multidisciplinary
archaeological project (www.paphostheatre.org). analytical approach involving archaeological and geological sampling,
Period Major features optical microscopy, chemical and mineralogical analyses. With this
Early Hellenistic (4th-2nd The foundation date of the theatre is around 300 system, we were able to investigate the raw materials, the production
century BCE) BCE. Few remains date to this chronological processes and the technological tools.
period. The orientation of the cavea is believed to
be roughly the same as today. It is not possible to
2. The context: Site and geology of the area
concretely define the dimension and shape of the
theatre for this earlier phase. Several post-holes in
the area of the stage led archaeologists to believe Nea Paphos was founded during the 4th century BCE by King
the stage building must have been a removable Nikokles as the new central harbour town of the region. The pre-existing
wooden structure. In the area of the orchestra, seat of Nikokles’ dynasty was the town of Paphos, located in the modern-
scant remains of mud-packed floor are visible
beneath the later levels.
day village of Kouklia. Nea Paphos served as main trading port for the
Late Hellenistic (2nd-1st During the 2nd century BCE, the theatre underwent easily and readily available geological and mineralogical resources of
century BCE) its first major refurbishment, following the the region: copper outcrops on the Troodos foothills, and lavish timber
traditional Alexandrian implant, with particular forests (Papuci-Władyka, 2020, 73). Spanning between the Hellenistic
resemblances to the theatre in Alexandria, Egypt.
period and modern day, Nea Paphos is one of the most long-lasting sites
The stage building was constructed in stone.
To this phase also belongs the construction of the in Cyprus.
so-called “Charonian” tunnel, which runs below
the stage and orchestra, at the time built with 2.1. The theatre
perishable materials on an elevated level.
Early Roman Age (1st century Augustan phase
BCE-2nd century CE) After a devastating earthquake that hit the island,
The Theatre is located in the area of Fabrika Hill, the North-Eastern
and particularly the region of Paphos, either in 15 quarter of the ancient city of Nea Paphos, near to where the N-E entrance
or 17 BCE, a major reconstruction phase happened gate to the city is supposed to have been located. The theatre has been in
during the reign of Augustus. The western function as a venue for spectacles for almost seven centuries (Barker,
containment wall of the cavea was completely
2015; Green et al., 2015; more information and documentation avail­
replaced, after the original collapsed.
Antonine phase able at https://www.paphostheatre.org) and was used continuously for
Another devastating earthquake seemingly hit different purposes for at least three more.
Paphos around the year 126 CE. Areas across the The first excavations on the site of the theatre date back to the 60 s,
city of Nea Paphos were subjected by major when archaeologist K. Nicolaou discovered the top rows of the seating
reconstructions under the auspices of Emperor
Antoninus Pius. The renovations in the theatre
(Nicolaou, 1966). The presence of the theatre, however, remained
were commemorated in a partially preserved merely a hypothesis until 1987, when the University of Trier opened a
inscription thanking the Emperor and his adoptive test trench in the orchestra area (Barker, 2015). The Paphos Theatre
successor, Marcus Aurelius. Archaeological Project (PTAP), an Australian research team, resumed
During this phase, the cavea was cut back into the
excavations in 1995, and has since directed over two decades of suc­
hill, while the rest of the theatre was profusely
embellished and brought up to date with the cessful working seasons. The methodology adopted by the PTAP project
Roman standards. The parodoi were vaulted and involves stratigraphic excavations in trenches, although the whole area
the inner walls frescoed, with part of the painted is nowadays mostly uncovered.
decorations being still visible today. The orchestra The plan of the theatre differs from the standard Hellenistic models –
was paved with imported marble slabs, and the
stage building – whose foundations are still visible
characterized by a horseshoe shape – by displaying a semi-circular
– was decorated with marble or marble-covered implant; it is likely that the original shape followed the Hellenistic
superstructures. The stage was a double storey tradition and was modified under development during the 1st century
building with colonnaded facade. BCE, in order to follow the Roman standards (on the theatre’s Hellenistic
Late Roman Age (3rd century During the 3rd century CE the theatre underwent
phases see Green et al., 2015). The date of foundation is still under
CE) its last major reconstruction. The reason for the
transformation was due to a change in the purpose discussion. Pottery findings in the area have been dated as early as the
of the building, which was now altered to allow the 4th century BCE, suggesting that the theatre was built for a settlement
flooding of the orchestra for water spectacles. For pre-dating the foundation of Nea Paphos, the presence of which has been
this reason, a 1 m high parapet wall with netting long hypothesised and discussed by archaeologists and historians (Green
above was built between the orchestra and the
proedria (the first rows of seats) for combat and
et al., 2015; Hayes, 1991). Furthermore, inscribed letters unearthed
venationes. The marble slabs paving the towards the top of the cavea have been dated around 300 BCE (Green
orchestra’s floor were overlayed by a pink et al., 2015). More definitive data concern the abandonment phase of
waterproof cement revealed by excavation. The the structure. After several major and minor refurbishments and re­
parodoi, which would have normally granted
works, the theatre was definitively abandoned during the 4th century
access to the proedria were partially closed,
probably as a water-containment measure. CE, most likely due to heavy damages caused by a devastating earth­
After the 4th century CE The theatre was officially abandoned during the quake (Barker, 2015). The increasing spread of Christian religion might
late 4th century CE. After the abandonment, the have contributed to the decline of the theatre venue (Barker, 2015;
site became a quarry for materials, employed – Green et al., 2015), as the difficult relationship between these two in­
among others – in the construction of the
Chrysopolitissa Basilica (located not far South from
stitutions is well documented in several areas of the Empire (Davis,
Fabrika Hill), which took place over centuries 2010). After the abandonment, the theatre was used as a source for
beginning during the 5th century. There are traces building materials to be used in other locations, including the nearby
of Medieval activities tied to metalworking and Chrysopolitissa Basilica. Quarrying activities ceased around the 7th
ceramic production in the whole area of southern
century CE. During the Medieval period, the site against Fabrika Hill was
Fabrika, as well as inside the theatre perimeter.
utilised for extensive productive activities, with the presence of struc­
tures related to metallurgic activities and a major workshop for sgraffito

2
P. Pizzo et al. Journal of Archaeological Science: Reports 57 (2024) 104595

Table 2 Table 2 (continued )

Displays the samples object of this study divided by functional category. For Functional Sample Sub- Location General characteristics
each sample, the identification code, number of sub-samples, number of frag­ group ID Samples
ments, sampling location, and general characteristics are indicated.
shows no traces of
Functional Sample Sub- Location General characteristics pigments.
group ID Samples NPT 6 NPT 6a Cavea – 6th Fragments of plaster
Masonry NPT 2 — W parodos Binding mortar joining the row coating the seatings in the
mortars masonry of the vaulting cavea. The surface is
wall of the Western smoothed and evened, but
parodos. Thick and shows no traces of
compact, no visible pigments.
surface finishing. Stone NPT 6b Cavea – 9th Fragments of plaster
impressions on all the row coating the seatings in the
surfaces cavea. The surface is
NPT 8 — Vomitorium Binding mortar originally smoothed and evened, but
joining the masonry of the shows no traces of
vomitorium wall. The pigments.
samples were collected NPT 6c Cavea – 11th Fragments of plaster
from the crest of the wall, row coating the seatings in the
thus the exposed surface cavea. The surface is
presented traces of smoothed and evened, but
weathering. shows no traces of
Wall NPT 1 — South of W Thin fragments of wall pigments.
plasters parodos plaster with traces of red NPT 6d Cavea – 9th Fragments of plaster
pigment, and possibly row coating the seatings in the
green. Collected from a cavea. The surface is
small niche in the wall. smoothed and evened,
NPT 9 — W parodos Fragment of historical and shows traces of a dark
wall wall plaster encapsulated reddish brown pigment.
in modern repair mortar. NPT 7 NPT 7a Displaced Fragments of plaster
The surface appears coating the seatings in the
polished. cavea. The surface is
NPT 10 — W parodos Fragment of thick wall smoothed and evened, but
wall plaster with traces of shows no traces of
pigment applied on a pigments.
polished surface. NPT 7b Displaced Fragments of plaster
NPT 11 NPT 11a E parodos Fragment of very thin wall coating the seatings in the
wall plaster with traces of cavea. The sample is
brushstrokes. brittle and crumbly.
NPT 11b E parodos Fragment of very thin wall Floors NPT 5 — Corridor Soft but compact fragment
wall plaster with traces of of stone topped by a bright
brushstrokes. white layer, originally
NPT 11c E parodos Fragment of very thin wall interpreted as a plaster
wall plaster with traces of coating or a lime wash.
brushstrokes. These three NPT Orchestra See above.
samples were collected 3a-3b*
from different parts of the *
Sample NPT 3b and NPT 3a are contemporarily waterproofing and flooring
same wall.
mortars.
Hydraulic NPT 3 NPT 3a* Orchestra Pink mortar covering the
floor of the orchestra.
Ceramic aggregates are pottery.
visible at the naked eye. Despite the theatre is still occasionally being referred to as “Helle­
The surface appears
nistic”, not much of this first phase is currently visible (Barker, 2015;
smooth and even.
NPT 3b* Orchestra Bulk sample of floor Green et al., 2015), and most of the samples analysed in this study
substratum with concern indeed the Roman phases. When attributing our samples to
waterproof layer in specific chronological periods, we decided to refer to the dating of the
between. The sample was structures associated, fully aware that there might be discrepancies
the understructure of NPT
3a.
between the construction phase and the decoration, other than super­
NPT 12 — E parodos Fragments of pink imposing refurbishment of different periods.
waterproof cement
collected from the wall
and the floor of a structure 2.2. The geological context
associated with water
containment purposes.
Ceramic aggregates are The Paphian region on the west coast of Cyprus is located near the
well visible at the naked geological interface between the Circum Troodos Sedimentary Succes­
eye. The surface appears sion and offshore Facies of Paleogene, Neogene and quaternary deposits.
smooth and even with
The outcrops of the Circum Troodos Sedimentary Succession are located
evident traces of
weathering. approximately 5 km northwest of the central part of Paphos. A typical
Seat sealing NPT 4 — Cavea – 2nd Fragments of plaster representative of these rocks are the carbonate sediments of the Miocene
row coating the seatings in the Pakhna formation (Eaton and Robertson, 1993). Is characterised by
cavea. The surface is alternating of marls, chalks and limestone deposits (Scirè-Calabrisotto
smoothed and evened, but
et al., 2017). The local limestone types are referred as the Havara and
Kafkalla deposits. Havara is a surficial deposit of porous limestone type
with presence of chalk or marl. As Kafkalla are referred the hardened

3
P. Pizzo et al. Journal of Archaeological Science: Reports 57 (2024) 104595

Fig. 1. Displays the exact locations in which samples were collected. Plan of the theatre courtesy of the Paphos Theatre Archaeological Project archives, created by
Geoff Stennett, altered by author (addition of the sampling locations).

upper portion of crust of Havara, which is the result of soil forming summarized in the following five categories: masonry mortars,
process due to lime enrichment (Schirmer, 1998). comprising mortar mixtures used as binding material; wall plasters,
The central part of the Paphos region is formed by geologically encompassing the samples applied on wall surfaces with coating and/or
younger formations of Cenozoic age. These are the outcrops of Pliocene decorative functions; waterproofing mortars, consisting of samples
Nicosia formation (silty or sand marl deposits), Pleistocene marine employed as waterproofing coatings; seat sealing, the plasters specif­
terrace deposits and aeolian deposits (calcarenite formed by cementa­ ically coating the seats on the cavea; and flooring surfaces. At a first
tion of fine-grained sit deposited in shallow marine environment or macroscopic evaluation, the samples appeared well preserved with no
coastal sand dunes) and Pleistocene/Holocene gravels, sands and silty relevant issues of conservation. Moss and plant traces were visible,
sediments or clays fluvial deltaic, marine, aeolian or lacustrine deposits especially in the samples collected in the cavea area. The masonry
(Amadio, 2018; Zomeni, 2012). mortar samples, collected respectively from the W parodos area and the
Due to the geological and topographical conditions in the drift area vomitorium, presented traces of superficial recrystallization, possibly due
of Troodos Mountains, the fragments of igneous rocks from the central to the exposition to weathering agents. Quite a few samples preserved
part of the island may be found in the sedimentary rocks around Paphos. traces of pigmentation on the surface, especially in the area of the W
These can be mainly rocks from ophiolite complexes, pillow lavas, parodos. With the exception of the two masonry mortars, all the samples
gabbro or serpentinite. These fragments occur mainly as components of collected in the theatre are lime-based plasters.
river sediments that are brought to the Paphos area by the Ezousa River The analytical procedure adopted is based on previous studies
or Agriokalami stream. (Kozlovcev and Válek, 2021; Philokyprou, 2012b; Válek et al., 2020;
Kozlovcev et al., 2023; Letourneux and Feneuille, 2012; Válek, 2015; for
3. Materials and methods characterization of the material also see Groot et al., 2004; Middendorf
et al., 2004) and combines multidisciplinary and multiscalar ap­
A total of 45 fragments have been collected from the archaeological proaches. After the preliminary, macroscopic assessment – which aimed
site of Nea Paphos Theatre, under the auspices and with the permission at recording generic features, including weight, dimensions, surface
of the Department of Antiquities of Cyprus. The collected samples have appearance, and conservation status – fifteen samples and sub-samples
been subdivided into 12 groups and attributed a specific code and were selected for the production of blue-stained, uncovered, polished
number (see Table 2). The sampling strategy followed a pre-existing, thin sections. Two sub-samples were not suitable for thin section prep­
and still unpublished, study – carried out by the PTAP mission – aration due to lack of material (NPT 6c, NPT 7b), while one sample (NPT
focusing on dating the theatre’s plasters. The aim of this sampling was to 9) presents extensive modern repair, which would invalidate the char­
cover all the available plastered features, in order to obtain a holistic acterization of the historic mortar. The selected samples were first
overview (Fig. 1). Sampling was minimally invasive, while all the per­ analysed by means of petrography (optical microscopy OM) in order to
formed analyses were destructive. gain information about composition and technological features; subse­
We sampled plasters with different functions, which can be quently, high resolution images of the structure, in addition to the

4
P. Pizzo et al. Journal of Archaeological Science: Reports 57 (2024) 104595

Table 3
Lists all the samples and the analysis performed on them.
Sample ID OM SEM-EDS XRD-QPA TA-DSC

NPT 1 X — X X
NPT 2 X X X X
NPT 3a X X X X
NPT 3b X X X X
NPT 4 X X — —
NPT 5 X — — —
NPT 6a X X — —
NPT 6b X X X X
NPT 6c — — — —
NPT 6d X X — —
NPT 7a — — X X
NPT 7b — — — —
NPT 8 X — X X
NPT 9 — — — —
NPT 10 X X X X
Fig. 2. Sample NPT 8, in situ before the collection, on the crest of the wall of
NPT 11a X X — —
NPT 11b X X — — the vomitorium. On the top part it is possible to observe traces of weathering
NPT 11c X X — — and recrystallization.
NPT 12 X X X X
4. Results
elemental composition of the binder and selected aggregates particles,
4.1. Masonry mortars
were obtained with SEM-EDS (scanning electron microscopy coupled
with an energy dispersive spectrometer) analyses. Mineral composition
Masonry mortars were preserved in different areas of the theatre,
data were acquired using XRD-QPA (x-ray powder diffraction with
however it was possible to safely sample – without causing any struc­
quantitative phase analysis by Rietveld (1969) method) and compared
tural or aesthetical damage – in two areas: the southern wall running
with the results of TA-DTG (derivative thermogravimetry) (for a com­
parallel to the western parodos (sample NPT 2), and in the walls of the
plete list of the analysis, see Table 3 below).
vomitorium (sample NPT 8), located on the North-West sector of the
For the petrographic analysis, a standard polarized light microscope
cavea. It is noteworthy to mention that these two samples are likely
(PLM), Olympus BX53M with an Olympus DP27 digital camera, was
belonging to the earliest phases of the theatre’s history.
utilized. A “cold cathode” type Mk 5-2 was associated to the microscope
Preliminary macroscopic observation immediately highlighted
to further investigate binders and binder-related particles; the analysis
strong similarities between the two samples, which appeared remark­
was performed under the following conditions: the electron beam cur­
ably different from the other plasters collected in the theatre. The
rent was at 260 µA at 15 kV voltage. SEM-EDS analyses were performed
mortars display a fine, homogeneous, compact, light-grey binder matrix,
on carbon-coated samples with a Tescan MIRA II LMU scanning electron
with well-sorted and evenly-distributed particles. The outer, exposed
microscope with an energy-dispersive analytical system (Bruker AXS)
surfaces present similar responses to weathering agents, with extensive
under the following conditions: back-scattered electron mode (BSE)
surface recrystallization (Fig. 2). OM analysis (Fig. 3) allowed us to
with electron accelerating voltage corresponding to 15 kV, at a WD of
identify the binder as well-sorted gypsum – as highlighted by the
15 mm, in low vacuum. The chemical composition was quantified by
maximum particle size inferior to 0.5 cm. Silica and carbonate particles
Brucker Esprit 2.5 software, without standardization. The analytical
are distributed in the matrix as a result of the burning process of the raw
results were expressed in oxide form.
material, rather than intentional addition of aggregates. According to
Thermal, as well as XRD analysis, were performed on samples ground
the XRD-QPA and TA (respectively Table 4 and Table 5) analyses on the
to analytical fineness. Each sample was firstly gently crushed in a mortar
binder-rich fractions, the gypsum content is over 90 % in both cases (see
and sieved through a set of different sieves ranging from 90 to 63 µm.
Table 4), while the calcite content is around 7 %.
The fraction above 90 µm was considered to be corresponding mainly to
Conclusively, the petrographic, mineralogical and chemical analyses
the aggregates, while the fraction below 63 µm was considered to be rich
confirm the similarities between the two samples, allowing us to suggest
in binder (Diaz et al., 2022). Before the XRD analysis, an internal stan­
that the gypsum used for their production comes from the same source,
dard (ZnO, 10 wt%) was homogenized with the sample. Data were
and that a similar production process was followed. This statement is
collected on a diffractometer D8 Bruker Advance pro (Cu Ká radiation,
supported by the presence of limestone fragments with natural Si-rich
40 kV and 40 mA) with 0.01 ◦ C step size 2ϴ and counting time 0.4 s/
infills detected in both NPT 2 and NPT 8 (see Fig. 3B). This type of
step. Quantitative phase analysis (QPA) – to determine crystalline and
particles has not been observed in any of the geological samples
amorphous phases – was performed by the Rietveld method (1969)
collected.
using Topas 4.2 from Bruker AXS. Derivative thermogravimetry was
Technologically speaking, the thickness of samples (up to 5 cm)
performed on TA Instruments Discovery SDT 650 under the following
suggests the employment of construction techniques that required slow
conditions: Nitrogen atmosphere, heat rate 20 ◦ C/minute, temperature
setting gypsum. One possible method to achieve it is to produce the
range 50–1000 ◦ C. TG, DTG and heat flow curves were collected and
plaster with heterogeneous calcination temperatures that affect and
analysed with MS Discovery; special focus was directed towards the
modify setting times (Elert et al., 2023). Furthermore, the observation of
regions between 50–250 ◦ C (detection of gypsum), 250–550 ◦ C (esti­
the matrix allowed us to estimate a firing temperature for the raw
mation of hydraulic phases) and between 600–850 ◦ C (carbonate
gypsum stone lower than 900 ◦ C, as demonstrated by the incomplete
decomposition) (Földvári, 2011). When the spectrum displayed possible
burning of limestone and siliceous particles.
traces of organic materials the sample was analysed again with ther­
mogravimetry coupled with a mass spectrometer (TGA-EGA, evolved
gasses analysis), measuring the eventual exothermic processes in Ni­ 4.2. Wall plasters
trogen and air atmospheres. The detection and identification of evolved
gases were realised by quadrupole mass spectrometry MS Discovery. Wall plaster samples were collected from different areas of the
theatre, with focus on the parodoi walls (NPT 9 and 10 for the western

5
P. Pizzo et al. Journal of Archaeological Science: Reports 57 (2024) 104595

Fig. 3. A (left) shows the freshly cut section of sample NPT 2 under the stereomicroscope, with gypsiferous particles (G), limestone fragments (C) and siliceous
particles (S) visible at low magnifications; while (B) (right) shows the same sample under OM-PPL, with the Si-rich limestone fragment in the centre (A).

parodos; NPT 11a, b, c for the eastern one). One additional wall plaster
Table 4 was collected from a wall running parallel to the western parodos (NPT
Summarizes the XRD-QPA results on the binder-rich fractions of each sample, 1). The parapet wall erected between the orchestra and the proedria
taking into consideration the amorphous and the three main compositional preserved traces of a thin wall plaster, impossible to sample. In the area
mineral phases: calcite (CaCO3), gypsum (CaSO4*2H2O) and silica (quartz or
around the theatre, scattered architectural elements – including col­
chert; SiO2). The samples with two layers (NPT 6b and NPT 7a) were mechan­
umns, capitals and blocks of stone – preserved scanty traces of pig­
ically cut and separated to obtain: NPT6b1/NPT7a2 corresponding to the fin­
ishing layers; and NPT6b2/NPT7a1 consisting of the ground layer. Sample NPT mented decorative plasters; however, being their original location,
10 displays a sequence of two layers mechanically detached and analysed function, and date unknown, it was deemed less relevant to sample
separately. XRD-QPA was performed on the superficial part of the sample (10a) them.
and in the lower area (10b), which contained remnants of a waterproofing The wall plaster samples in object of study have different structures,
mortar layer. Amorphous phases can be indicators of possible hydraulicity. compositions and finishing treatments (Fig. 4), making connections and
Sample ID CaCO3% CaSO₄% SiO2% Amorphous% comparisons hazardous. However, it is possible to observe a certain
degree of similarity between the samples collected from the same
NPT 1 66.1 1.9 1.7 23.2
NPT 2 6.7 92.8 <0.5 0.0
structures. Furthermore, the type of aggregates employed in the mortar
NPT 3a 47.6 <0.5 2.1 49.4 mixtures are similar to what has been observed in the wider context of
NPT 3b 71.4 n.d. 1.0 27.0 the theatre: river sand, limestone fragments and crushed shells.
NPT 6b1* 95.0 <0.5 <0.5 4.5
NPT 6b2* 75.0 n.d. 1.1 21.9
4.2.1. Western parodos wall plasters
NPT 7a1* 72.5 n.d. 1.2 23.6
NPT 7a2* 77.4 n.d. 1.5 15.0 The two samples collected from the western parodos wall consist of
NPT 8 7.5 91.4 0.7 0.0 up to 7 cm thick blocks of compact and well-sorted lime plaster
NPT 10a 60.1 0.6 <0.5 37.8 distributed in layers. The sequence of layers appears as follows: a thin
NPT 10b 61.6 1.1 29.9
<0.5
more porous layer was set directly on the stone blocks and let to partially
NPT 12 91.3 <0.5 1.7 6.4
dry before a thicker and more heterogeneous plaster layer was laid on
top of it. The pigment seems to have been applied directly on the fresh
plaster (more evident in sample NPT 1, see Section 4.2.3). The first layer,
Table 5 applied directly on the stone, consists of a mixture of lime binder with
Summarises the TA results on the binder-rich fractions of each lime-based finely crushed clay-based materials (possibly bricks or ceramic frag­
sample, highlighting the percentage of weight loss in the temperature range of ments) and sand. Characteristic of this layer is the presence of a highly
interest for the decomposition of CaCO3. the expected temperature range is set irregular pore structure with traces of deposited salts inside (Fig. 5). The
between 650 ◦ C and 850 ◦ C, however the analysed samples display an earlier second layer consists of a heterogeneous mixture of lime binder with
start of the process, possibly due to the presence of different carbonates (i.e. limestone fragments, river sand, crushed shells, and occasionally
aragonite). Calcium carbonate content was calculated following the following
crushed clay particles. Phenomena of calcite dissolution and precipita­
wt.loss%(600 − 850 C)*100
◦

formula: CaCO3 %content = . tion are well documented (Fig. 6).

CO2 molarmass
Chemical and mineralogical analysis on the binder-rich fractions of
Sample Wt loss % Wt. loss% Wt. loss% CaCO3% the samples highlight the presence of a relatively pure lime binder
ID 600 ◦ C-850 ◦ C 550 ◦ C-850 ◦ C 500 ◦ C-850 ◦ C content
(CaCO3% above 60 %) with traces of gypsum and silica minerals. XRD-
NPT 1 37.6 85.5 QPA detects a considerable percentage of amorphous phase (between 30
NPT 3a 24.6 55.9
and 40 %), possibly linked to the presence hydrated hydraulic phases.
NPT 3b* 32.7 74.3
NPT 41.8 95.0
6b1* 4.2.2. Eastern parodos wall plasters
NPT 39.2 90.5 The samples from the eastern parodos wall are just a few millimetres
6b2* thin, with a bright white finishing and clear brushstrokes marks
NPT 7a1 37.6 85.5
NPT 39.8 90.5
(Fig. 7A). Despite the reduced thickness, all three samples (NPT 11a, b,
7a2* and c) present a double layered structure, remarkably similar to the one
NPT 10a 30.7 69.7 observable in the seat sealing functional category discussed further (see
NPT 10b 36.9 83.9 Section 4.5). The layer directly attached to the wall is coarser, more
NPT 12 35.1 79.8
porous and less well-sorted. It consists of a mixture of highly pure lime
binder with river sand of fine grain (max. grain size 1 mm). On top of it
lays the finishing layer, a thinner stratum consisting of pure lime mixed

6
P. Pizzo et al. Journal of Archaeological Science: Reports 57 (2024) 104595

Fig. 4. The three different types of wall plasters collected in the theatre as photographed during the documentation phase, after a preliminary cleaning. On the left,
sample NPT 10, a thick double layered wall plaster, with a smooth, polished surface with possible traces of pigments; in the centre, NPT 11a a double layered, thin
wall plaster with a bright white finish applied with brush; on the right, sample NPT 1, a single layer of lime plaster topped with bright red and yellow pigments,
applied with a fresco or semi-fresco technique.

elemental composition come from the SEM-EDS analysis (CaO above 86

%; SiO2 around 3.5 %; presence of SO3 and MgO in most cases between
0.5 % and 3 %; traces (below 1 %) of Al2O3 and FeO – this corresponds to
the average of measurements taken on the binders of samples NPT 11a, b
and c; for each sample 5 different points were measured).

4.2.3. Other wall plasters

The only other wall plaster collected is NPT 1, gathered from a
structure with a niche located south of the western parodos wall. This
sample is the only one to have a single layer structure among all the wall
plasters analysed. The exterior appearance is similar to the one of the
eastern parodos wall samples: thin lime plaster with a smooth finish and
traces of brushstrokes. However, the mortar mixture was visibly
different, with NPT 1 displaying a much denser, lime-richer binder with
higher concentration of aggregates. These mostly consist of fine (below
1 mm) river sand, limestone, and crushed shells. The plaster was
Fig. 5. OM-PPL (left) and cathodoluminescence (right) microphotograph of smoothed with the help of a tool, whose action resulted in the horizontal
sample NPT 1; the dark matrix corresponds to the lime binder, while the needle- orientation of the aggregates and in the presence of a crack running
shaped crystals correspond to salts. Cathodoluminescence allowed us to parallel to the surface. The surface was then completed by adding a
completely exclude the identification of the crystals with calcite recrystalliza­ bright red pigment, which was applied directly on the surface as proven
tion, as they would have appeared as bright red in CL. by the colour seeping through the underlying layer (Fig. 8).
In conclusion, it appears evident that the decorative plasters were
not applied at the same time on the two walls of the parodoi and on the
other structure. This might be the result of minor refurbishments
happening during different moments of the long history of the theatre.
At any rate, despite being ascribable to different phases, the overall
composition of the theatre wall plasters is similar, and the materials
employed are all of local sourcing. In fact, the presence of radiolarian
chert from the Dhiarizos river, as well as the limestone fragments
remarkably similar to the geological formations of the Paphian region
are both element in support of a local provisioning of the materials.
Nonetheless, a precise sourcing location has not been identified yet.

4.3. Waterproofing mortars

The definition “hydraulic” is contradictory and often discussed by

scholars of different discipline. The term hydraulic means the ability of a
binder to set and harden in wet conditions and highly hydraulic lime-
based systems can harden under water. These binders are not water
soluble and thus provide a certain water resisting properties to a mortar.
The distinguishing feature of hydraulic mortars is the presence of un­
Fig. 6. OM-XPL microphotograph of sample NPT 10 with focus on the interface
stable silicate phases (Weiner, 2010, 188). Silica minerals can be natu­
between layers 1 and 2. Legend: 1 = first layer; 2 = second layer; arrows =
precipitation and recrystallization traces. rally present in and to produce a natural hydraulic lime suitable raw
material has to be found and burnt to form hydraulic phases that react
with water during the hardening phase of the binder. Similar hydrated
with crushed shells. No traces of pigments are visible. Both layers well
hydraulic phases are formed by a reaction of uncarbonated lime – i.e.
cautiously smoothed and evened, and some time must have passed be­
slaked lime in the form of calcium hydroxide – with certain rocks/sand
tween the applications of the two: in fact, traces of dark, organic
of volcanic origin. The pozzolanic reaction is often manifested by a re­
pollution are visible in the interface (Fig. 7B).
action rim around the reacting aggregate, but its presence is not always
Due to the scarcity of available material, XRD and TA analyses were
clearly identifiable. The absence of a reaction rim does not mean that a
not performed on the sample. All the chemical information about the
pozzolanic reaction has not occurred in the binder (Calzolari et al.,

7
P. Pizzo et al. Journal of Archaeological Science: Reports 57 (2024) 104595

Fig. 7. A (left) stereo-microphotograph of sample NPT 11a with well visible brushstrokes; 7B (right) OM-PPL microphotograph of sample NPT 11a with focus on the
interface between the two layers.

the waterproof flooring of the orchestra (sample NPT 3a), and a pink
mortar associated with a structure related to the sewage of water (NPT
12), located between the eastern parodos and the Nymphaeum. Both
these features are dated to the Roman phases of the theatre, with the
orchestra floor refurbishment being securely dated not earlier than the
3rd century CE (Barker, 2015). On a macroscopic level, these two
samples appear quite similar: they consist of up to 5 cm thick fine lime
mortar mixed with considerable amount of finely crushed backed-clay
particles. Crushed bricks or pottery fragments had been previously
observed in minor quantities in the other samples from the theatre. In
the case of the waterproofing samples, however, the percentage of clay
content is much more significant.
At the moment of sampling, we were informed that the structures on
which the samples were applied had been associated with water-related
functions (Barker, 2015), however we could not be sure whether they
possess hydraulic properties. Chemical and mineralogical analyses were
fundamental for the understanding of this. According to the XRD-QPA
results, these macroscopically similar samples have little in common
on the mineralogical composition level. While the orchestra floor (NPT
Fig. 8. Sample NPT 1 in thin section under the optical microscope combined 3a) presents a high percentage of amorphous content (amorphous – 49.4
with UV light. The red pigment was applied with a fresco technique, as high­ %, CaCO3 – 47.6 %) – generally associated with hydrated hydraulic
lighted by the pigment seeping through the binder. On top of the pigment, a
phases – sample NPT 12 is a lime-rich binder (CaCO3 – 91.3 %, amor­
new layer of mortar is visible, hinting to possible repairs (arrow).
phous – 6.4 %), with scanty traces of silica mineral phases. Thermal
analysis confirms this dissonance, pointing to the possible use of natural
2023). hydraulic lime (hereinafter NHL) in the production of NPT 3a – the or­
In this paper, we use the term “waterproofing” to refer to this cate­ chestra floor – as hinted by the CO2/H2O ratio being 4.75 (Bakolas
gory of samples, pointing out their function rather than the material et al.,1998). In the case of NPT 12, this same ratio was equal to 27,
property. pointing towards the use of a non-hydraulic lime (Calzolari et al., 2023).
Waterproof features associated with the theatre mainly consisted of To conclude, SEM-EDS mapping further strengthens the idea that the

Fig. 9. SEM-EDS maps of samples NPT 3a (left) and NPT 12 (right). In NPT 3a silica (green) particles are distributed in the calcite (red) binder, while in NPT 12Si is
almost exclusively contained in the pozzoloanic aggregates.

8
P. Pizzo et al. Journal of Archaeological Science: Reports 57 (2024) 104595

Fig. 10. A (top)) Schematic representation of the orchestra floor section. From top to bottom: layer 3a, waterproofing mortar; layer 3b, coarse and thick lime plaster
layer (blue lines indicate the part not sampled and only ideally reconstructed); layer 3c, the irregular patches of waterproofing mortar; layer 3d, the binding and
smoothing mortar; layer 3 s, the sandstone blocks. The three squares represent the thin sections produced from this sample, highlighting the layers visible in each of
them. Drawing by author. 10B (bottom)) scan of the thin sections represented schematically in Fig. 10A (red rectangles); from left to right: TS3A (=layer NPT 3a),
TS3B1 (=layer NPT 3b); TS3B (=layer NPT 3b, 3c, 3d, 3 s).

orchestra floor sample (NPT 3a) was produced using NHL (Fig. 9), while
the water sewage-related sample was crafted employing regular air lime,
and its waterproofing properties were achieved by adding burnt clay
particles. The cementitious index (CI) of both samples was calculated
following this formula (after Arizzi and Cultrone, 2021; Figueiredo
et al., 2016):
1.1Al2 O3 + 2.8SiO2 + (0.7Fe2 O3 )
CI =
CaO + 1.4MgO
The results, however, were surprising similar with NPT 3a having an
average CI of 0.33, and NPT 12 of 0.27. According to the classification
proposed by Holmes and Wingate (2002), sample NPT 3a is slightly
hydraulic, while NPT 12 is on the border between fat lime and slightly
hydraulic – the threshold being at 0.3.
Finally, it is worth mentioning that pozzoloanic plasters and mortars
were by far more widespread in Roman times than earlier (Theodoridou
et al., 2013; Jackson et al., 2012).

4.4. Floors
Fig. 11. Photograph of sample NPT 3a and 3b in situ (next to scale bar).
Apart from the floor of the orchestra (sample NPT 3b), not many

9
P. Pizzo et al. Journal of Archaeological Science: Reports 57 (2024) 104595

Fig. 12. Picture of the detachment between layers 3c and 3d (left: stereomicroscope picture; right: om-ppl picture).

other floor surfaces were visible in the theatre. Scanty traces of lime
plaster or lime wash were visible on the floor of the narrow corridor
between the parapet wall and the proedria but they were not sampled.
Thus, all the information about the floor coatings come from the single
orchestra floor sample. This structure is related to the Roman phases of
the theatre, and more specifically to the 3rd century CE.
Sample NPT 3b consists of a 10 cm thick block, displaying a sequence
of at least three layers, on top of which laid sample NPT 3a (discussed in
Section 4.3). The whole structure was about 20 cm thick, and it was not
possible to collect a cohesive block with all four layers. Sample NPT 3b
alone encompasses four distinguished layers with different binders,
aggregates, porosity, and, possibly, even functions (see Figs. 10 and 11).
Construction starts with small sandstone blocks of similar dimensions
but irregular shapes (layer 3 s in Fig. 10) topped and bound together by a
thin, brown lime mortar (layer 3d). This mixture of lime and sand was
likely applied to smoothen and even the surface before adding the
subsequent layers. Between layer 3d and 4c, a series of flat and elon­
gated pores highlights the presence of adhesion problems, with the
waterproofing layer (3c) having detached from the underlying structure Fig. 13. Stereomicrophotograph of sample NPT 7b section. Finishing layer is
(Fig. 11 and Fig. 12). Layer 3c presents a fine, light-grey lime binder on top.
mixed with clay-based materials, limestone fragments and sand. This
stratum is not evenly distributed throughout the sample and rather ap­
pears in patches of different thickness. The thicker layer (3b) consists of
a mixture of lime binder and coarse, well-rounded limestone fragments.
XRD and TA analysis of the binder-rich fraction of layer 3b show a
lime-rich binder with moderate content of silica minerals. Considerable
amorphous content – 27 % – and an earlier start of the decomposition of
carbonates in TA curves can be ascribed to the presence of hydraulic
phases.
This sample raised a series of questions concerning the construction
techniques of floors structures. At the moment of publication, we lack
any comparative data either from other theatres or from other public
structures in Nea Paphos. Samples gathered from the agora of Nea
Paphos are currently under examinations, and the results will be pub­
lished soon. Until then, this sample remains a unicum and its interpre­
tation requires cautiousness.

4.5. Seat sealings

A unique feature of the Nea Paphos theatre is the presence of plasters Fig. 14. OM-PPL microphotograph of the top surface of sample NPT 6a, with
coating the seating rows of the cavea. While this feature is known in possible traces of a dark pigment or coating (in red polygon).
several other Hellenistic and Roman theatres – including the above-
mentioned one in Alexandria of Egypt – comprehensive studies of seat Roman times. In the Paphian theatre traces of seat coating are visible
sealing plasters have never been completed in Cyprus. Functionally throughout all the cavea, with a remarkable status of preservation. The
speaking, these plasters serve contemporarily an aesthetical purpose and sampling procedure aimed at avoiding any damage to the preserved
a practical one, as they cover the seats from weathering agents (Barker, layers, thus we collected detached fragments visibly belonging to still-
2015, 35). Considering their function, these plasters were likely cleaned in-place features. In total, seven samples were collected from the cen­
and even replaced in short intervals of time. Therefore, it is impossible to tral sector of the cavea between the proedria and the 11th row (NPT 4,
offer a precise chronological window, and it is only logical to assume 6a, 6b, 6c, 6d, 7a, and 7b).
that they must belong to the latest phases of use of the theatre, during Observing the various samples collected, it seems plausible to retrace

10
P. Pizzo et al. Journal of Archaeological Science: Reports 57 (2024) 104595

Fig. 15. OM-PPL (left) and XPL (right) of sample NPT 6b with the layer bond well visible in XPL due to the different density of the binder.

a production model consisting of a double-layered system (Figs. 13 and consisted in smoothing the surface, possibly with polished tools that left
15) of lime plasters with a well-smoothed, polished, and occasionally no marks; in several cases inorganic pigments, mostly containing iron,
painted surface (Fig. 14). The first layer, in direct contact with the stone have been applied on the surface with the aid of brushes (as revealed by
blocks of the seats, consists of a coarser mortar mixture of lime and well the clear brushstrokes marks, see Fig. 7 above). It was not possible to
sorted river sand. The presence of radiolarian chert and magmatic rocks identify particular or clear changes in the diachronic perspective, as
– including olivine, serpentinite and basalt – allowed us to identify the overall the samples presented relevant similarities. The only notable
source of this sand in the hydrogeological basin of the Diarhizos river, exception, in this regard, is the appearance of waterproof samples in the
which collects magmatic rocks and minerals from the Circum Troodos later Roman phases. In regard to the hydraulic mortars, it is noteworthy
catch. The second, thinner, layer, also defined as “finishing layer”, to underline again the use of NHL in one sample, although whether it
consists of a much finer mixture of lime binder and shells’ fragments. was casual or intentional cannot be established.
Through the observation of the thin sections it was possible to deduce From the analytical data gathered from this site, it is also possible to
that the first layer was levelled and partially dried before the application hypothesise that masonry mortars were preferably produced out of
of the finishing one (Fig. 15). gypsum. This preliminary assumption seems to be further corroborated
Chemical analyses confirm the higher content of calcite in the fin­ by the analysis of the materials sampled in the Agora of Nea Paphos,
ishing layer, indicating richer and more refined binders were used to currently under study.
produce these specific plasters. SEM and OM further highlight the fact In conclusion, this study might not provide extensive information on
that both layers were levelled, in fact, it is possible to observe a denser the diachronic developments and evolutions of the plaster industry from
binder in the upper surfaces of both layers. Pigment analysis have yet to Hellenistic to Roman times; however, it contributes to the knowledge of
be carried; however, preliminary observations in OM coupled with UV the productive processes and the raw materials selection in use at the
light highlighted the possible presence of organic material – either from times. Up to now, little was known about Cypriot plaster industry; this
the pigment or its binder. Additional RAMAN spectroscopic analysis are research aimed at answering prompt and urgent questions such as the
in order for the identification of the pigments, which – at the naked eye – raw material sourcing, the mastery of stone burning and any relevant
appear as faded away in all the samples except NPT 6d. trend in the choice of plaster and mortar making. Further analysis on
Below are summarizing tables of the compositional analyses carried samples from the Agora of Nea Paphos, from Kissonerga-Skalia, Palae­
out on the theatre samples. paphos and Yeronisos Island – currently under study – will allow to
outline a clear overview of the industry from the end of the Bronze Age
5. Conclusions to the Roman times.

At the conclusion of this trial study, it could be safely supposed that CRediT authorship contribution statement
the binder used in the theatre sample was the result of processing of local
stones. Limestone is a widely available resource in the area surrounding Paola Pizzo: Writing – review & editing, Writing – original draft,
the archaeological site of Nea Paphos. Quarrying sites have been iden­ Methodology, Investigation, Data curation, Conceptualization. Jan
tified even within the premises of the ancient theatre itself (Green et al., Válek: Writing – review & editing, Validation, Supervision, Funding
2015), signifying that the materials for the construction were possibly acquisition, Conceptualization. Petr Kozlovcev: Writing – original
acquired directly in situ. Lime production could have not been an draft, Data curation. Dita Frankeová: Formal analysis, Data curation.
exception. Nonetheless, all the limestones collected from the catchment Alberto Viani: Formal analysis, Data curation.
of the modern district of Paphos displayed chemical and morphological
similarities to a relevant degree, making the identification of a specific Declaration of competing interest
source not possible. In order to obtain such an information, it would
have been relevant to analyse partially burnt lime particles. However, no The authors declare that they have no known competing financial
features as such were identified in the theatre samples, suggesting that interests or personal relationships that could have appeared to influence
the firing process occurred under carefully monitored conditions and the work reported in this paper.
that the burnt stones were attentively selected before the slaking. Lime
and gypsum lumps, on the other hand, were identified in relevant Data availability
quantities. The analysis of these binder-related particles, however, did
not allow any assumption on the type of slaking or mixing utilised. Available data are included in the manuscript.
The study of the aggregates reveals the prevalent use of readily
available materials. In particular, the substantial presence of seashells Acknowledgments
suggests the use of sea sand, which was easily accessible in the imme­
diate proximity of the building site. The predominant surface treatment We would like to acknowledge the Department of Antiquities of

11
P. Pizzo et al. Journal of Archaeological Science: Reports 57 (2024) 104595

Cyprus, with particular mention to the conservator Ms. Eve Hadji­ Jackson, M.D., Vola, G., Všianský, D., Oleson, J.P., Scheetz, B.E., Brandon, C., Hohlfelder,
R.L. 2012, Cement microstructures and durability in ancient Roman seawater
georgiou. In addition, we would like to thank Dr. Craig Barker and
concretes. In: Válek, J., Hughes, J.J., Groot, C.J.W.P. (Eds.), Historic Mortars.
Professor J.R. Green for consultation on the archaeological site. Characterisation, Assessment and Repair, RILEM Bookseries (7), pp. 49–76. doi:
This research is part of the PlaCe-ITN network, funded by the Eu­ 10.1007/978-94-007-4635-0_5.
ropean Union’s Horizon 2020 Research and Innovation Programme Kozlovcev, P., Kotková, K., Frankeová, D., Válek, J., Viani, A., Mařiková-Kubková, J.
2023, Characterisation of historic mortars related to the possibility of their
under the Marie Sklodowska-Curie GA no. 956410. radiocarbon dating, Mikulčice and Pohansko Archaeological sites. In: Bokan
Bosiljkov, V., Padovnik, A., Turk, T. (Eds.), Conservation and Restoration of Historic
References Mortars and Masonry Structures. HMC 2022. RILEM Bookseries (42) doi: 10.1007/
978-3-031-31472-8_14.
Kozlovcev, P., Válek, J., 2021. The micro-structural character of limestone and its
Amadio, M., 2018. From deposits to social practices: Integrated micromorphological influence on the formation of phases in calcined products: natural hydraulic limes
analysis of floor sequences at Middle Bronze Age Erimi-Laonin tou Porakou, Cyprus. and cements. Mater. Struct. 54, 217. https://doi.org/10.1617/s11527-021-01814-7.
J. Archaeol. Sci. Rep. 21, 433–449. https://doi.org/10.1016/j.jasrep.2018.07.023. Letourneux, J. and Feneuille, S. 2012, Mineralogical and Microstructural Analysis of
Arizzi, A., Cultrone, G., 2021. Mortars and plasters—how to characterise hydraulic Mortars from Kushite Archaeological Sites, in J. Válek, J.J. Hughes, C.J.W.P. Groot
mortars. Archaeol. Anthropol. Sci. 13, 144. https://doi.org/10.1007/s12520-021- (eds.), Historic Mortars. Characterisation, Assessment and Repair, RILEM Bookseries
01404-2. (7), pp. 37-48. doi: 10.1007/978-94-007-4635-0_4.
Bakolas, A., Biscontin, G., Moropoulou, A., Zendri, E., 1998. Characterization of Marinowitz, C., Neuwald-Burg, C., Pfeifer, M. 2012, Historic documents in understanding
structural Byzantine mortars by thermogravimetric analysis. Thermochim. Acta 321, and evaluation of historic lime mortars. In: Válek, J., Hughes, J.J., Groot, C.J.W.P.
151–160. https://doi.org/10.1016/S0040-6031(98)00454-7. (Eds.), Historic Mortars. Characterisation, Assessment and Repair, RILEM Bookseries
Balandier, C., Joliot, C., Ménager, M., Vouve, F., Vieillescazes, C., 2017. Chemical (7), pp. 15–24. doi: 10.1007/978-94-007-4635-0_2.
analyses of Roman wall paintings recently found in Paphos, Cyprus: the Middendorf, B., Hughes, J.J., Callebaut, K., Baronio, G., Papayanni, I. 2004,
complementarity of archaeological and chemical studies. J. Archaeol. Sci. Rep. 14, Mineralogical characterization of historic mortars. In: C. Groot, G. Ashall, J. Hughes
332–339. https://doi.org/10.1016/j.jasrep.2017.06.016. (Eds.), Characterization of Old Mortars with Respect to their Repair, RILEM TC 167-
Barker, C., 2015. Twenty years of the University of Sydney excavations of the theatre COM, pp. 21–36.
precinct in Nea Paphos in Cyprus. In: Australian Archaeological Fieldwork Abroad Nicolaou, K., 1966. The Topography of Nea Paphos. In: Mélanges Offert á Kazimierz
III, Ancient History Resources for Teachers, pp. 26–63. Michalowski, pp. 561–601.
Calzolari, L., Medeghini, L., Baiocchi, I., Zanzi, G.L., Mignardi, S., 2023. Aqua Papuci-Władika, E., 2020. Paphos Agora Project (PAP): Interdisciplinary Research of the
Alexandrina and Fragole cistern: characterization of mortars from Roman Jagiellonian University in NEA PAPHOS UNESCO World Heritage Site (2011–2015)
constructions, Rome (Italy). Archaeol. Anthropol. Sci. 15 (183) https://doi.org/ First Results (1). Historia Iagellonica, Krakow.
10.1007/s12520-023-01885-3. Philokyprou, M., 2012b. The beginnings of pyrotechnology in Cyprus. Int. J. Architect.
Davis, T.W., 2010. Earthquakes and the crisis of faith: social transformation in Late Herit. 6, 172–199. https://doi.org/10.1080/15583058.2010.528145.
Antique Cyprus. Buried History 46, 5–16. Philokyprou, M., 2012a, The earliest use of lime and gypsum mortars in Cyprus. In:
Diaz, J., Ševčík, R., Mácová, P., Menéndez, B., Frankeová, D., Slížková, Z., 2022. Impact Válek, J., Hughes, J.J., Groot, C.J.W.P. (Eds.), Historic Mortars. Characterisation,
of nanosilica on lime restoration mortars properties. J. Cult. Herit. 55, 210–220. Assessment and Repair, RILEM Bookseries (7), pp. 25–36. doi: 10.1007/978-94-007-
https://doi.org/10.1016/j.culher.2022.03.014. 4635-0_3.
Eaton, S., Robertson, A.H.F., 1993. The Miocene Pakhna Formation, Cyprus, and its Rietveld, H.M., 1969. A profile refinement method for nuclear and magnetic structures.
relationship to the Neogene tectonic evolution of the Eastern Mediterranean. Sed. J. Appl. Cryst. 2, 65–71. https://doi.org/10.1107/S0021889869006558.
Geol. 86, 273–296. https://doi.org/10.1016/0037-0738(93)90026-2. Schirmer, W., 1998. Havara on Cyprus: a surficial calcareous deposit. Eiszeit. Gegenw.
Elert, K., Bel-Anzué, P., Burgos-Ruiz, M., 2023. Influence of calcination temperature on 48, 110–117. https://doi.org/10.23689/fidgeo-1441.
hydration behavior, strength, and weathering resistance of traditional gypsum Scirè-Calabrisotto, C., Amadio, M., Fedi, M.E., Liccioli, L., Bombardieri, L., 2017.
plaster. Constr. Build. Mater. 367, 130361 https://doi.org/10.1016/j. Strategies for sampling difficult archaeological contexts and improving the quality of
conbuildmat.2023.130361. radiocarbon data. The Case of Erimi-Laonin Tou Porakou, Cyprus. Radiocarbon 59
Figueiredo, C., Lawrence, M., Ball, R.J., 2016. Chemical and physical characterisation of (6), 1919–1930. https://doi.org/10.1017/RDC.2017.92.
three nhl 2 binders and the relationship with the mortar properties. In: Conference Theodoridou, M., Ioannou, I., Philokyprou, M., 2013. New evidence of early use of
Paper, REHABEND 2016, May 24–27, Burgos, Spain. artificial pozzolanic material in mortars. J. Archaeol. Sci. 40, 3263–3269. https://
Földvári, M., 2011. Handbook of Thermogravimetric System of Minerals and Its Use in doi.org/10.1016/j.jas.2013.03.027.
Geological Practice. Geological Institute of Hungary, Budapest. Válek, J., Skružná, O., Kozlovcev, P., Frankeová, D., Mácová, P., Viani, A., Kumpová, I.,
Green, J.R., Barker, C., Stennett, G. 2015, The hellenistic phases of the theatre at Nea 2020. Composition and technology of the 17th century stucco decorations at
Paphos in Cyprus: the evidence from the Australian Excavations. In: Frederiksen, R., Červená Lhota Castle in Southern Bohemia. Int. J. Architect. Herit. 1042–1047.
Gebhard, E.R., Sokolicek, A. (Eds.), The Architecture of the Ancient Greek Theatre. https://doi.org/10.1080/15583058.2020.1731627.
Monographs of the Danish Institute at Athens (17), pp. 319–334. Válek, J., 2015, Lime Technologies of Historic Buildings, Prague.
Groot, C.J.W.P., Ashall, G.J., Hughes, J.J., Bartos, P.J.M. 2004. Characterization of old Weiner, S., 2010. Microarchaeology. Cambridge University Press, Beyond the Visible
mortars with respect to their repair: A state of the art. In: Groot, C., Ashall, G., Archaeological Record, Cambridge.
Hughes, J. (Eds.), Characterization of Old Mortars with Respect to their Repair, Zomeni, Z. 2012. Quaternary marine terraces on Cyprus: constraints on uplift and
RILEM TC 167-COM, pp. 1–8. pedogenesis, and the geoarchaeology of Palaipafos. Unpublished PhD thesis.
Hayes, J.W. 1991, Nea Paphos III: the Hellenistic and Roman Pottery, Nicosia.
Holmes, S., Wingate, M., 2002. Building with Lime: A Practical Introduction. Practical
Action Publishing.

12