Content uploaded by Adel El-Turki
Author content
All content in this area was uploaded by Adel El-Turki on Feb 01, 2018
Content may be subject to copyright.
1
1. Conservation and Restoration Work
1.1 A conservative history and the condition of the
mosaics pavament
In the year 2005, the decision was made in
detaching of the mosaic pavement from its origi-
nal location, this is due to; the absence of the pro-
tection and safety measures as well as the risk of
theft and damage (Figure 1). In order to facilitate
its removal, the mosaic was cut into 5 large panels,
each measuring approximately 4.90 x m 2.40 and
corresponding to the number of parts in which the
picture was articulated. The panels were laid on
new concrete supports, with a metal net, then were
displayed along the external wall of the pavilion
reserved to the exhibit of the seaboard villas “Vil-
las Museum” (Figure 2). The detaching operations,
begun by the spreading the mosaic’s surface with
acetate polyvinyl resin: the rear face of each 5 pan-
els was polished and treated by means of mechan-
ic instruments, the entire laying surfaces and
also the mortar over which the tesserae had been
originally placed. Such operations seriously dam-
aged the mosaic pavement: in the first place the
uneven spreading of the mentioned glue over the
upper side of the mosaic brought about, as a conse-
quence, the loss of several portion of the pavement
in the subsequent detaching operations, especially
of those portions made of vitreous paste, the most
fragile ones (for their chemical composition and
for their state of conservation).
After the displaying of the panels along the
external wall, a close investigation was made on
the state of the surfaces, as well as an adequate
El Turki
THE MOSAICS OF THE ROMAN VILLA AT WADI LEBDA
LEPTIS MAGNA
ATTI CONVEGNO AIEMA 2012
2
photographic screening. The surface revealed sev-
eral scales that were observed, the scale was com-
prised of carbonate and siliceous deposit; these
could be explained by the fact that the pavement
had remained buried for a very long time. A thin
layer of vinilic resin and bits of canvas (the mate-
rials used in the detaching processes) were also
found over the entire surface of each panel. How-
ever, gaps and holes that had been filled with white
concrete and plaster were also noticeable (Figures
3-4). In addition, the pavement showed some of
the tesserae had fallen away in some parts, near
the areas treated with plaster move (Figures 5-6).
Overall, it has been considered that the main
causes for the damages the mosaic had undergone
can be thus summarised.
1. The high temperature hitting the mosaic sur-
face, owing to direct exposure to sunlight – this
occurred after the panels were removed by
their original site and exposed on the outside
walls of the museum.
2. Different parameters of size-variability in the
artefact due to thermo-hygrometric conditions,
owed to the highly heterogeneous supports on
which the mosaic was placed.
3. The high hygroscopic quality of the plaster used.
4. The cleaning of the artefacts, through chemical
and mechanical means, such as a high precision
micro-drill and a high precision micro-sander.
1.2 Techniques and Restorations Method
The rich decorative elements of the pavement
consist of opus tessellatum and polychromatic
geometric and figurative mosaics of stone and
glass tesserae. The Conservation Plan has led to
the following activities:
1. High-resolution photographic documentation.
2. The consolidation of constituents materials
and the (cleaning) removal of unsuitable plas-
ter repairs, re-integrating the gaps with new
tesserae.
3. Relaying of the mosaics.
4. Reintegration and reassembling the pavement.
5. Characterisation of the constituent materials.
1.2.1 Documentation
The primary phase of the work was to set-up
a photographic repertory, illustrating the present
condition of the pavement, the original damages
and the damages caused by the detachment pro-
cess. The scanning of the materials of the pave-
ment and the obtained images, for which the used
programme that was specialised in graphics and
its provide the work with higher flexibility and
would be able to elaborate the data. In addition,
the data from previous intervention and the docu-
ments concerning the excavations of the pave-
ment were also collected.
The typical photographic documentation of the
intervention (Figures 7-10.), and consisted of the
following:
• A mapping of the ancient repairs and integra-
tions.
3
El Turki | THE MOSAICS OF THE ROMAN VILLA AT WADI LEBDA LEPTIS MAGNA
• A chart of the re-integrations made by the local
operators.
• A chart of the gaps filled with plaster and
cement
• A chart of the re-integrations made during the
restorations
1.2.2 The consolidation of constituents materials
The consolidation procedures of the mosaic in
the first place was proceeded to consolidate the
vitreous paste in very careful way (the original
materials appeared highly damaged); consequent-
ly stabilising product consisting of ethylic esters
deriving from siliceous acid, diluted with mineral
turpentine was spread over the artefact several
times, with a soft brush. Nano-lime mortar (to con-
solidate materials having a carbonate basis) and
an acrylic micro-emulsion were used. The task of
pre-consolidating, previous to finishing and clean-
ing the mosaic, and of re-integrating the gaps.
The Cleaning
The first operation was of removing the con-
siderable remaining of polyvinyl resin that had
been used in order to detach the several portions
of the floor. The removal was done by means of a
ATTI CONVEGNO AIEMA 2012
4
tool generating a steam jet; the resulting dirt was
wiped away with soft nylon brushes of different
softness and sizes. This job of brushing away old
glue remaining was indispensable in view of the
subsequent phase, that of removing the carbon-
ate incrustations from the mosaic’s surface; this
was established by means of micro-sanding (a
180-200 mesh inert aluminium oxide) would have
”bounced” back in striking the rubbery glue, thus
having too soft an impact on the deposit accumu-
lated in the long period in which this Roman man-
sion had remained buried. Through this procedure
we managed to remove the carbonate incrusta-
tions almost completely, except for some very
thick ones; these we had to finish by means of a
vibro-engraver and an ultra-sound remover, which
had proved to be effective before, on thick layers
of deposits. In case of a hard and strongly deposit-
ed to the substratum, EDTA (bi-sodic salt deriving
from ethyl diamminic-tetracetic acid) was used.
In areas in which the organic deposits could
not be removed by a steam jet or by mechani-
cal means, the saturated solution of ammonium
carbonate, combined with fine-grained cellulose
pulp, by the use of sponges, brushes and lancets
was used. Having been thus treated, in order to
remove any salts that may have spread over the
mosaic surface, de-ionised water was used with
great care to avoid any water stagnation on the
mosaic’s surface.
Gaps treatment
This operation presented a very high degree of
difficulty, on account of the large number of gaps
5
El Turki | THE MOSAICS OF THE ROMAN VILLA AT WADI LEBDA LEPTIS MAGNA
in the pavement. The integration tests in the gaps
were made with diversified criteria, depending on
the types of existing decorations and on the nature
of the original pavement, as it could be recon-
structed from the photographical surveys done
before the detaching. Certain gaps, small in size
and of easy interpretation could be reintegrated
by reconstructing the decorations with marble or
vitreous paste tesserae (Figures 11-12); with this
purpose in mind we managed to find stone and
vitreous paste tesserae that were compatible with
the original ones.
Some other gaps, which in Roman days had
been repaired with largish marble fragments of
different types and sizes, we did not alter, except
for slight additions of mortar between sections of
marble fragments (Figures 12-13). The gaps that
we found hard to interpret mostly belonged to
large figure scenes that were considerably dam-
aged. Older gaps, which were already there at the
time of the excavations, were just filled in our
intervention with mortar of a neuter tone.
This work turned out to be more complex than
we expected, this associated to the hardness of
the concrete support, and for the further reason
that the tesserae which were closets to the gaps’
perimeter had begun to fall away. While this was
being done, a restorer began re-integrating some
tesserae in the gaps of the decorated margins. Pre-
vious to this we had made tests to define the right
colour shade for the laying mortar; the one that
we selected in the end was a mixture of hydraulic
mortar, yellow sand from a local pit and golden
yellow marble dust, in the proportion 1:1:1.
The re-integration of the missing parts of the
mosaic was done with small tesserae from dif-
ferent types of marble, which had been selected
during the preceding mission, which turned out
to be quite compatible with the original tesserae.
This work consisted in laying an adequate layer
of mortar on the bottom of the concrete support
and then in inserting the tesserae, so as to form a
narrow and uniform weaving. For this purpose the
tesserae had to be cut one by one, with their edge
blunted, when it proved necessary.
1.2.3 Relaying of the mosaics
After the entire complex procedures of the res-
toration and conservation that carried out on the
mosaic. It has been planned to place the mosaics
in their permanent place of exhibition, the new
pavilion of the Museum of the Villas, which had
been specially built for that purpose. The Eni
North Africa Company undertook to finance the
planning and the construction of this building,
which was erected in the immediate vicinity of the
“Lepcis Villas Museum”.
In order to protect certain portions of panels
placed on them a layer of cotton gauze, that was
ATTI CONVEGNO AIEMA 2012
6
fixed on their surface with an acrylic resin solu-
tion, spread over with a brush; indeed their situ-
ation presented risks that might have worsened
their status, because of having been moved with
the consequent loss of some of the tesserae. The
external margins of all panels were carefully exam-
ined and all screens were controlled.
In view of the complexity of these operations a
special cart was built, in order to move the panels
to the new pavilion, the metal materials we elimi-
nated which had fixed the mosaic sections to the
wall; then the panels were placed, vertically, on
the cart (Figure 14). Then, gradually the mosaics
were carried to the new wing of the museum, by
following the order of the panels’ numbers, from
n.1 to n.5, so as to keep their original situation,
then marked in advance (on the floor of the new
wing) the spots where the mosaic section were to
be placed; on these soft stone bricks were set, at a
regular distance from one another, so as to form
a base which was both discontinuous and higher
than the floor (Figure 15). Such a device allowed
us to set the various panels on the same level, by
means of lead ”cushions” that had been designed
and calculated specifically for each panel. Empty
space was left between the panels’ concrete base
and the floor of this room, so as to prevent the
migration of humidity upwards from the ground
as well as from the foundations of the building, by
way of infiltration, onto the pavement. The same
space was also provided as a thermo-hygrometric
exchange with the external environment, conse-
quently allowing a steady balance of the moisture
and avoiding any form of stress to the materials of
the mosaic. Each panel was first placed in a lean-
ing position, and then laid over the plinths that
had been prepared in advance; all panels were laid
on the same level. In consideration of the panels’
weight, all this work was done by first assembling
the two sections of the first panel, then those of
the second, of the third and so on, up to the fifth
panel. After this final re-assembling, the space
between the plinths and the gallery wall that had
been set up for the exhibition was re-filled, as far
up as the level of the mosaic floor.
7
El Turki | THE MOSAICS OF THE ROMAN VILLA AT WADI LEBDA LEPTIS MAGNA
1.2.4 Reintegration and reassembling
the pavement
After relayed of the mosaic panels in the per-
meant location, the reintegration and reassem-
bling procedures was performed. As mentioned
above in a conservation history of the mosaic. The
consequences from the detaching operations were
performed with insufficient means of experience;
had led to the loss of an enormous amount of the
original materials, while causing a number of gaps
both serious and large enough as to prevent a cor-
rect perusal of the mosaic in its entirety.
A part of these gaps were reintegrated by local
personnel by means of repairs which did not
reconstruct the original image in any organic man-
ner, but had simply filled the missing spots, with
some of the original deserve, plus various materi-
als, such as glass fragments, stone fragments etc.
Such materials, moreover, were placed mostly in a
messy manner (Figures 16-17)
After cleaning operations, the complex map of
various areas on the mosaic panels became clear;
therefore it seemed necessary to define a set of cri-
teria for the re-integrations. Injections with emul-
sion of acrylic resin and soft brushes were used to
fix the loss of original tesserae. (Figure 18).
In the first place we completed cleaning the
border and the base of each gap, so as to obtain a
good levelling and a good compatibility of the re-
integrated tesserae with the original ones; then we
took away all the concrete that had been used in
the preceding intervention. After wetting the sur-
faces we applied a mortar made with an hydrau-
lic lime characterised by a low salt content; soon
afterwards, while the mortar was still soft, tesserae
were inserted.
All materials were used for filling the gaps,
accurate as the choice of litho-types and neutral
colour hydraulic lime mortar. It also considered
that it was important to confer some colour uni-
ATTI CONVEGNO AIEMA 2012
8
formity to the mosaic surface, where differences
were most evident. This was performed using
rebalancing colour shades by means of water-
colours, wherever colours resulted different, on
account of drastic differences between the original
materials and the replacement materials, mechani-
cal and chemical methods were used, to scrape and
slightly tarnish the new tesserae. Then, the floor
mosaic was finally spread with a low concentra-
tion acrylic resin, which we sprayed on with a
pneumatic atomizer (Figure 19).
1.2.5 The archaeometrical analysis of the
constituent materials
In order to classify the type of the constitu-
ent materials at the time in which the mosaic was
made. Stone, glass tesserae with different colours
and mortar samples were investigated. The main
objective of this study, in the case of stone this
was involves analysing its colour and appearance
as well as conducting a petrographic analysis. An
archaeometric analysis of the chemical composi-
tion and the colouring agents of the glass tesserae.
Techniques and methods
The thin, cross-section and chemical analysis
were performed at “Materials Testing Laboratory”
of Emmebi Diagnostica Artistica, Rome and ISCR
Istituto Superiore per la Conservazione ed il Res-
tauro, Rome and respectively and consist of:
• Leica polarization microscope for reflected and
transmitted light mod. DM RXP,
• SEM-EDS instrument was a Zeiss EVO 60 used
in extended pressure mode (EP-SEM). Also
attached to the Evo 60 is an INCA X-sight ener-
gy dispersive X-ray spectrometer (EDS Oxford
Instruments Detector 7636 Energy)
• Optical Microscopy with halogen light. Nikon
Eclipse 50i pol, with a digital camera.
• Optical Microscopy with ultraviolet light. Nikon
eclipse E40, 50W mercury vapour lamp, with
digital camera.
• X-ray Fluorescence (XRF). X-ray tube Moxtek
(maximum voltage of 40 kV and maximum cur-
rent of 0.1 mA), solid-state detector SDD (Silicon
Drift Detector - EIS Italy) with energy resolution
of 139 eV at the line of iron and a multichannel
Amptek USA.
Chemical composition of Mosaic tiles
and mortars
The chemical composition (major elements)
of the stone and glass tesserae used in mosaic as
obtained by XRF is presented in Table1. The results
gathered from each sample will be discussed sepa-
rately here after:
Sample 1 (Brown Stone tesserae)
Sample 1 consists of brown stone tessera with
mortar collected from tied deer (Figure 20a). The
photomicrographs of the cross-section was imaged
using optical microscopes with the halogen and
ultraviolet lights are shown in Figures 20b and 20c
respectively. The layers from the innermost to the
outermost are consisted from the following:
a) The plaster layer consists of a matrix including
white and translucent, yellowish-white, grain
size mainly conglomerate and sandstone, and
fluorescent of red-brown colour as shown in
9
El Turki | THE MOSAICS OF THE ROMAN VILLA AT WADI LEBDA LEPTIS MAGNA
the photomicrographs obtained by the ultravio-
let light (Figure 20c).
b) The layer with the thickness between 50 and
100 microns, constituted by a white matrix
piecewise translucent and by including red and
orange colour. The XRF analysis as shown in
Table 1, indicates that the red pigment in the
tessera r is composed of iron oxides) [1], the
element strontium is a characteristic of the ele-
ments of the glass matrix.
Samples 2 and 3 (Blue Glass Tessera and Mortar)
Two blue glass tessera and mortar pigment
have been sampled from the right shoulder of the
dress of the man who keeps his ears from behind
the bear (Figure 21a). Photomicrographs with hal-
ogen and ultraviolet lights obtained from the thin
section are shown in Figures 21c and 21d respec-
tively. XRF analysis shown in Table 1 revealed the
presence of copper, suggests the use of pigments
based on copper type Azurite, Cu3(CO3)2(OH)2
[1-2], the elements calcium, iron are associated to
the mortar and strontium is assigned for the glass
constituent.
The mortar layer consists of a matrix yellow-
ish-white and including translucent and yellow-
ish-white with a particle size predominantly eri-
naceous and coarse; there are also fine and very
fine granules blacks, probably from carbonaceous
(Figure 21b).
The photograph of the second glass tessera is
shown in Figure 22. XRF analysis obtained from
this material is represented in Table, the result
identified the calcium, iron, manganese, copper,
lead, antimony and strontium, the blue coloration
is obtained by the addition of copper oxides and
manganese dioxide, with melting in an oxidiz-
ing atmosphere; antimony, when associated with
the lead, is present as a matting agent. Moreover,
the presence of antimony is also, probably in the
form of calcium antimoniate is due to its effect as
an opacifier. Calcium, which is associable stron-
tium, is a characteristic of the elements of the
glass matrix, together with the manganese often
present in varying amounts in the glass paste [1].
Regardless of the colour; iron, probably present
as oxide is associated with residual red pigment
(Figure 22).
Sample 4 (Pink Stone Tessera and Mortar)
Sample 4 represents the pink stone tessera with
the mortar sample (Figure 23a). The thin cross-sec-
tion as shown in the photomicrographs obtained
with halogen and ultraviolet lights are presented
in Figures 23(b) and 23(c) respectively. The plas-
ter layer present at the residual level, consists of
a matrix and including translucent white and
yellowish-white colours. In addition, yellow layer
thickness of about 200 microns constituted includ-
ing yellow and translucent in ultraviolet light is
almost non-fluorescent; the XRF analysis (Table 1)
shows the presence of iron oxide, the main constit-
uent of the material [1]. The presence of strontium
and zirconium indicative that the Roman might be
used to produce of the glass.
ATTI CONVEGNO AIEMA 2012
10
Sample 5 and 6 (Light and Dark Green
Glass Tesserae)
Light and dark green coloured glass tesserae
were analysed (Figures 24 and 25). The green col-
oration from the light tessera is obtained by the
addition in the process of melting of copper com-
pounds (which are associated with the impurities
of zinc) and manganese dioxide; antimony and
tin, when associated with the lead, are present
as opacifiers, which is associable strontium, is a
characteristic of the elements of the glass matrix,
together with the manganese often present in
varying amounts in the glass paste. Regardless of
colour, the presence of iron is instead correlated to
the mixture in the visible yellow tessera. However,
The elements identified from the dark green tes-
sera as obtained by SEM-EDS analysis is presented
in Table 2, the results are in a good agreement with
the XRF result (Table 1). The green colour is due to
the presence of an iron compound, probably com-
bined with lead antimoniate, and terracotta frag-
ments (Table 2). A few particles composed of tin
are present. The brown multilayer portion is again
a typical alteration of glass [1].
Samples 7 and 8 (Red Glass Tessera)
Two fragments of red glass tesserae from the
blood of the gladiator’s right leg after torn down
are shown Figure 26a and Figure 26b. The second-
ary electron micrograph is shown in Figure 26c.
The results of XRF and EDS performed on thin sec-
tions are shown in Table 1 and Table 3, the results
showed that the red colouring is obtained by means
of compounds of copper and iron oxides; antimo-
ny, when associated with the lead, is present as a
matting agent; sodium, silicon, aluminium, mag-
nesium, calcium which is associable strontium, is
a characteristic of the siliceous glass with sodium
flux, the main glass constituents of the mosaic.
11
El Turki | THE MOSAICS OF THE ROMAN VILLA AT WADI LEBDA LEPTIS MAGNA
The white multilayer portion is a typical altera-
tion of glass due to the leaching of lead, copper
and sodium, and then to the formation of silica gel
(see Figure 26b).
Sample 9 and 10 (Yellow Glass Tessera)
Sample 9 and 10 of yellow glass tesserae were
collected from the gladiator’s shield torn down are
shown in Figure 27a. Typical structure of Samples
9 and 10 is shown in Figure 27b. XRF identified
manganese, iron, tin, lead, antimony and strontium
(Table 1). Figure 27c showing the optical image
which identified the red inclusions which could be
assigned to the presence of iron that not melted at
low temperature. EDS analysis was carried out on
the thin section surface of the tessera to obtain its
elemental composition is shown in Table 3. Figure
27d show the x-ray maps performed on the surface
of the tessera, which identified a high levels of sili-
con, lead, antimony sodium, aluminium calcium,
iron, chloride, potassium and manganese. XRF
analysis revealed that yellow colour is attributed
Table 1: Chemical analysis of stone, glass and mortar materials
Sample Description Elements Wt (%)
Ca Mn Fe Cu Zn Ti Sn Pb Sb Sr Zr Au Hg
Brown stone tessera with lime
mortar 57 371 122
Light blue glass and mortar 43 60 17 65 607
Blue glass tessera:
lime mortar is also present 21 28 300 92 1079 12155 1288
Pink stone tessera:
lime mortar is also present 33 202 117
Light green glass tessera 41 167 429 57 403 9507 1201 381
Red glass tessera from a drop
of blood (the right leg of
gladiator torn down)
211 525 9095 434 222
Yellow glass tessera from
the Shield of the Gladiator
torn down
26 179 484 11897 1900 500
Red stone tessera from the
decorated of the helmet with
mortar patterned
59 26 238 115 97 126
Yellow tessera with mortar
from the bottom right of the
gladiator’s helmet
39 296 143 228 41
Golden Glass tessera 32 82 365 1549 605 124 1079
Table 2: EDS analysis of dark green glass tessera
Spectrum Elemental Composition (wt %)
Na Al Si K Ca Mn Fe Cu Sb Pb Total
1 13.20 2.16 55.97 0.63 5.70 0.42 1.47 2.78 7.22 9.61 99.11
Table 3: EDS analysis of red glass tessera
Spectrum Elemental Composition (wt %)
Na Mg Al Si K Ca Ti Mn Fe Cu Pb Total
1 11.58 2.32 1.84 59.12 3.14 10.06 0.16 0.38 1.28 2.63 6.31 98.83
Table 4: EDS analysis of yellow glass tessera
Spectrum Elemental Composition (wt %)
Na Al Si Cl K Ca Mn Fe Sb Pb Total
1 9.02 3.47 39.77 0.53 0.49 3.93 0.21 1.49 11.55 29.28 99.74
ATTI CONVEGNO AIEMA 2012
12
by the antimony and tin, when associated with the
lead. The yellow colour of the glass tessera as iden-
tified by EDS identified lead antimonate (with the
formula Pb2Sb2O7 corresponding to the mineral
bindheimite, Pb2Sb2O6 [O, OH]) [3].
On the basis of previous study, researchers
drew two main conclusions about the technology
of ancient opaque glass: (1) in general, the same
element (antimony or tin) was used to make white
(calcium antimonate or tin oxide) and yellow (lead
antimonate or lead stannate) opaque glass in a giv-
en period and place; and (2) antimony-based opac-
ifiers were used as an alternative to tin-based ones,
and vice versa. On this basis, the prevailing opin-
ion is that antimony-based opacifiers were used by
Roman glassmakers until the fourth century A.D,
and were thereafter progressively replaced by tin
based opacifiers [3]. Regardless, colour constituent,
calcium, silicon, sodium which is associable stron-
tium, is a characteristic of the elements of the glass
matrix, together with the manganese often present
in varying amounts in the glass paste.
Sample 10 (Red Stone Tessera)
Sample 10 represents the red stone tessera with
the mortar collected from the right bottom of the
decoration of the helmet (Figure 28a). The thin-
section as shown in the photomicrographs with
halogen and ultraviolet lights presented in Figures
28b and 28c are assigned that the layers from the
13
El Turki | THE MOSAICS OF THE ROMAN VILLA AT WADI LEBDA LEPTIS MAGNA
innermost to the outermost are consisted from the
following:
a) The plaster layer consists of a matrix including
white and translucent and yellow-orange with
a particle size predominantly arenaceous mate-
rial which fluoresce under ultraviolet light (Fig-
ure 28c).
b) The layer of a thickness of about 100 microns,
constituted of a matrix including orange and
Secondary Electron Image
Secondary Electron Image
Secondary Electron Image
Secondary Electron Image
red and orange and other fine white milky in
appearance; is not fluorescent under the ultra-
violet light.
c) The white colour layer of thickness of about 25
microns is poorly fluoresce under the ultravio-
let light.
XRF analysis shows the layer of red colour is
obtained by iron oxides (Table 1), together with a
lead-based pigment; elements strontium and zir-
conium are attributable to the glass constituent
material. The presence of titanium is assigned to
the underlying mortar.
Sample 11 (Yellow Stone Tessera)
Sample 11 represents the yellow stone tessera
with mortar collected from the bottom right of the
gladiator’s helmet (Figure 29a). The photomicro-
graphs with halogen and ultraviolet lights were
ATTI CONVEGNO AIEMA 2012
14
obtained from the thin section are shown in Fig-
ures 29b and 29c respectively. The description of
the layers from the innermost to the outermost are
shown below:
a) The white layer as shown in the photomicro-
graph of Figure 28 is not fluoresce under the
ultraviolet light.
b) The layer of the constituted by a matrix includ-
ing yellow and yellow and brown colours with
a thickness of about 100 microns had a fine and
coarse angular translucent morphology which
is not fluorescent under the ultraviolet light.
c) white layer is weakly fluoresce under the ultra-
violet light.
The XRF analysis of red layer identified that
the pigment is comprised of iron oxides together
with a lead-based pigment such as red lead; the ele-
ments strontium and zirconium can be attributa-
ble to the glass constituent material while calcium
is from the underlying mortar (Table 1).
Sample 12 (Gold Glass Tessera)
Figure 30a showing the structure of gold glass
tessera sample. The photomicrographs as obtained
from the thin-section with halogen and ultraviolet
lights are shown in Figures 30b and 30c respec-
tively. The layers from the innermost to the outer-
most are consisted of the following:
a) The colour of the vitreous paste is due to small
amounts of iron oxides.
15
El Turki | THE MOSAICS OF THE ROMAN VILLA AT WADI LEBDA LEPTIS MAGNA
b) The white layer appearance of thickness of 5-10
microns is fluoresce under the ultraviolet light
in blue tones.
c) Thin gold leaf (see XRF analysis; Table 1); the
images of Figures 30b and 30c showing the
overlap between two leaves which divided by
the fluorescent milky layer.
d) The white layer of thickness of about 130
microns containing milky appearance traits
including translucent white is fluoresce under
the ultraviolet light.
The colour of the glass tessera is due to small
amounts of oxides of iron; antimony, when associ-
ated with the lead, is present as a matting agent;
calcium, which is associable, zirconium and stron-
tium, is a characteristic of the elements of the glass
matrix, together with manganese, often present in
varying amounts in the glass paste, regardless of
colour, the metal foil is made of gold.
Mortars
Mineralogical and petrographic analysis were
also performed on the mortar samples. The sam-
ples were collected from the original of the mosaic
bedding layer as shown in Figure 31. The results
revealed that the mortar made up of lime-based
mortar prepared with a poor aggregate of very
fine particle size of arenaceous to silty coarse. The
granules, appear to be non-oriented and with a
value in the percentage of granules with respect to
the very low binder around.
The binder in thin section appears to be car-
bonate with homogeneous structure and texture
micritic. There were no interactions (edges reac-
tion) between the clasts of the aggregate and the
matrix. Some fractures shrinkage affecting both
the binder that the contact granules / binder. The
porosity is low (<20%).
On the surface there is a drafting chromatic
well adherent to the substrate, contact with non-
rectilinear, of non-uniform thickness (from 0.15 to
0 mm) are composed of iron oxides that include
two crystals (one carbonate and one quartz) of
smaller than 0.15 mm.
ATTI CONVEGNO AIEMA 2012
16
REFERENCES
SCHIBILLE Nadine and FREESTONE Ian
C., “Composition, Production and Pro-
curement of Glass at San Vincenzo al
Volturno: An Early Medieval Monas-
tic Complex in Southern Italy”. PLoS
Onev.8(10); 2013.
BARBER D.J., FREESTONE I.C., MOULD-
ING K.M. (2009). Ancient copper
red glasses: investigation and analy-
sis by microbeam techniques. In:
SHORTLAND A.J., FREESTONE I.C.,
REHREN T., editors. From Mine to
Microscope: Advances in the Study of
Ancient Technology. Oxford: Oxbow
Books. 115-127.
TURNER W.E.S. and ROOKSBY H.P., “A
Yellow Cubic Lead Tin Oxide Opaci-
fier in Ancient Glass”, Physics and
Chemistry of Glasses, v. 5, no. 1, Feb-
ruary 1964, pp. 20-25.
