This article appeared in a journal published by Elsevier.

The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

Available online at

www.sciencedirect.com

Journal of Cultural Heritage 9 (2008) e21ee26

Original article

Early evidences of vitreous materials in Roman mosaics from Italy:

An archaeological and archaeometric integrated study
Cristina Boschetti a,b,*, Cristina Leonelli b, Michele Macchiarola c,
Paolo Veronesi b, Anna Corradi b, Cinzia Sada d
a
Padova University, PhD School in Science and Technologies for the Conservation of the Archaeological and Architectonic Heritage,
Earth Sciences Department, via Giotto 4, 35137 Padova, Italy
b
Modena and Reggio Emilia University, Department of Materials and Environmental Engineering, via Vignolese 905, 41100 Modena, Italy
c
CNR-ISTEC, via Granarolo 64, 4818 Faenza (RA), Italy
d
Padova University, Physical Science Department ‘‘Galileo Galilei’’, via Marzolo 8, 35131 Padova, Italy
Received 30 March 2008; accepted 27 May 2008

Abstract

This work displays the lines of a project about vitreous materials used in Roman Republican Age opus vermiculatum mosaics from Italy. This
mosaic technique, originated in Egypt and Greece during the Hellenistic Age, testifies the fist evidences of vitreous materials used in mosaics:
faience and glasses. The use of these materials in vermiculata mosaics from Italy was almost unknown when this research was started, so it was
necessary to map and characterize the whole mosaic production. After a survey operated by a portable digital microscope, a reduced number of
samples were collected to perform a laboratory characterization (OM, SEMeEDS, XRD). Data have been completed by an isotope ratio
characterization (SIMS) to investigate the glasses raw materials provenance.
Ó 2008 Elsevier Masson SAS. All rights reserved.

Keywords: Hellenistic mosaic; Roman age Italy; Opus vermiculatum; Sealing-wax red glass; Faience; Vitreous materials microstructure; Oxygen isotope ratio

1. Research aims the characterization of vitreous materials tesserae are

presented.
This research comes from the idea that the mosaic is
a complex system and its correct study needs to consider 2. Experimental section
several aspects: the archaeological context, the style, the
iconography and the mosaic technique. In fact, in the past, 2.1. Introduction
mosaics were considered mostly as artwork pieces and it was
paid attention almost exclusively to the artistic analysis. This During the Republican Roman Age (2nde1st century BC)
is actually true for objects with a high artistic value, as like the a production of luxurious mosaics made in the so-called opus
Hellenistic opus vermiculatum mosaics, object of this research. vermiculatum technique is attested in Italy. This technique
In this paper the lines of a project concerning the study of this uses tiny tesserae (1e4 mm) to obtain a pictorial effect, suit-
mosaic technique in Italy is reported and the first results about able to reproduce paintings as in the taste of this age. Because
of the long times of execution required, these mosaics were
usually assembled in the workshop on a stone or terracotta tray
* Corresponding author.
to create small pictures, called emblemata, suitable to be set
E-mail addresses: [email protected] (C. Boschetti), leonelli@
unimo.it (C. Leonelli), [email protected] (M. Macchiarola), into the centre of the floor as ready-made works. Otherwise,
[email protected] (P. Veronesi), [email protected] (A. Corradi), cinzia. opus vermiculatum is sometimes attested to create great
[email protected] (C. Sada). dimension mosaics, masterpieces assembled in situ.

1296-2074/$ - see front matter Ó 2008 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.culher.2008.05.003
 Author's personal copy

e22 C. Boschetti et al. / Journal of Cultural Heritage 9 (2008) e21ee26

One of the main archaeological problems linked to these 2.2.3. Analytical techniques
findings is, for the emblemata, the production site identifica- Firstly it was performed an on site surface observation by
tion and, in general, the identification of the workshops a portable optical microscope (Olympus Mic-D, magnification
provenance. In fact, there are similar mosaics, the oldest 20e225): this instrument gave the chance to collect a great
dating back to the end of the 3rd century BC, in other Medi- number of information about the surface materials micro-
terranean sites, such as Delos, Rhodes and Samos islands in structure and to identify the presence of non-original parts.
Greece, Pergamon in Asia Minor, Alexandria in Egypt, Malta The characterization was completed collecting a group of
island and, in Spain, Empuries. Vermiculata mosaics from samples prepared in polished sections to be characterized by
Egypt and Greece have been already studied [1] and here it different techniques: optical microscopy (Olympus SZX16),
was recognized the earliest use of vitreous tesserae, made in electron microscopy (SEM, Philips XL 40) with microanalysis
Egyptian faience, glass and Egyptian blue. Otherwise, an EDS, XRD (Panalyitcal X’Pert, radiation Cu Ka, 2q 5e70 ,
archaeometric characterization was performed only on a small on polished sections with focused beam, acquisition time 3 h),
group of glass tesserae from Delos [2], while faience used in Secondary Ion Mass Spectrometry measurements (IMS 4f
mosaics has never been characterized. mass spectrometer, Cameca, Padova, Italy, using a 14,5 KeV
The solving of the provenance problem was not possible Csþ primary beam and by negative secondary ion detection).
with the existing literature data, because there is a total lack in
the study of the opus vermiculatum technique in Italy, not only 2.2.4. Experimental data and results
as it concerns the technical aspects, but also in the basis of the
archaeological investigation. For example, it is noticeable that 2.2.4.1. Faience tesserae. Faience tesserae have been detected
the exact number of evidences and their distribution is not in a variety of green and blue shades in eighty-two mosaics
clearly known. In this perspective this project [3] has been i.e., with one exception, in the whole production. It was
divided in two steps: the first step is the archaeological study, possible to sample ten tesserae from different sites (Table 1)
aimed to build a background suitable to carry on the second which, after the optical microscope observations, appeared to
step, i.e. an archaeometric characterization aimed to reach be divided in two groups: the first one, attested only in light
a good knowledge of the materials used. Mosaics have been green colour, is opaque and with few pores, while the second,
considered in all their constitutive elements: tesserae, mortars, turquoise, green or blue, is richer in pores and stones, visible
trays and the painted surface finishing, but in this paper only as gray and dark brown grains (Fig. 2). Electron microscopy
the vitreous materials will be discussed. confirmed the division in two groups: the first one presents
a non-homogeneous structure, with high silica, containing
crystals not very well distinct from the matrix. The EDS
2.2. Vermiculata mosaics from Italy: the project
chemical composition (Table 2) reveals the use of iron as
colouring agent and a high concentration in magnesium and
2.2.1. The archaeological investigation
alumina indicating the use of sands rich in impurities.
The archaeological study started with the creation of the
The second type presents isolated silica crystals with sharp
evidences catalogue: the mosaics resulted to be in total eighty-
edges embedded into a glass matrix and, occasionally, it was also
three, mainly located in Pompeii and in Latium, but also in
possible to analyze traces of the colouring agents: in a yellowish
Sicily and in the North of the country. Observing the whole
green sample, it was detected the presence of an unmolten spot
production it was possible to draw up its technical features and
containing lead and antimonies, testifying the use, as well
to find peculiar elements, index of a local production. In fact,
documented in Ptolemaic faience [5,6], of lead antimoniate to
in the entire production it was detected a frequent use of
obtain this particular green shade, resulting from the mixing of
faience tesserae in the shades of blue and green, as like in
the green colour, due to the iron and copper content, and the
mosaics from the Southern and Eastern Mediterranean region,
yellow of lead antimoniate. Moreover, in a turquoise tessera, it
but in Italy glass presence is lower attested and it is mainly of
was detected a spot containing cobalt, testifying the mixing
a bright homogeneous red colour. Moreover, in Alexandria and
between the green of iron and copper and the blue of cobalt. In
Delos, tesserae are cut from glass rods, while in Italy they are
general, comparing the overall compositions with the samples of
obtained from glass cakes. As it concerns other elements, thin
the first group, here it is possible to detect values higher in silica
lead stripes, used in Greece and Egypt to draw up the contour
and lower in alumina. In the whole sampling, analyzing the
lines, are not attested in Italy, while, as like in the other
overall composition, low alkalis values were detected: this is due
Mediterranean sites, it was detected the practice to finish the
to the low quantity of glass phase in fact, performing the analysis
surface with a painting layer [4], minimizing the perception of
on the vitreous matrix, higher values were detected (Na2O 7.4e
the mortar joints (Fig. 1).
12.5 Wt%; K2O 0.8e2.9 Wt%).
In both the structures there is no evidence of a vitrified
2.2.2. The archaeometric study layer on the surface, so the samples can be classified as glassy
After the acquisition of a good archaeological knowledge faience, a non-glazed faience variant [7].
of the mosaics, it was performed the archaeometric charac- The mineralogical analyses show mainly the high temper-
terization, aimed to investigate the microstructure and the ature phases of SiO2: cristobalite and trydimite. Low quartz,
chemical composition of the vitreous materials. and an amorphous phase is common to both microstructures.
 Author's personal copy

C. Boschetti et al. / Journal of Cultural Heritage 9 (2008) e21ee26 e23

Fig. 1. An emblema mosaic seen in section (a): starting from the bottom it is possible to observe the tray, the mortar layers and the tesserae. Below, a sequence of
images taken by the portable optical microscope on the tesserae surface: the faience colour variation (beg), a group of sealing-wax red glass tesserae covered, as
usual, by a green-whitish deterioration layer (h) and the painting on the surface in orange (i) and light blue (j) (For interpretation of the references to colour in this
figure legend, the reader is referred to the web version of this article).

Quantitative analyses are not useful to identify minor crys- blue, was mapped, but the number of evidences is so insig-
talline phase or quartz level to discriminate the raw materials nificant that it was decided to focus the sampling only on red
used in the two cases. (Figs. 1 and 2), the colour, with twelve evidences, highly
attested. Six samples were collected from three mosaics and
2.2.4.2. The glasses. During the on site observation, the use of other six samples have been taken from Italian wall mosaics
glasses in different colours, as like yellow, orange, green and dated from the beginning to the first half of the 1st century
 Author's personal copy

e24 C. Boschetti et al. / Journal of Cultural Heritage 9 (2008) e21ee26

Table 1
Description and provenance of the faience samples analyzed
Sample Colour Mosaic Provenance Conservation place
Roma VN camp2 Green Emblema with Nile landscape Roma, via Nazionale, house Roma, Antiquarium del Celio
Roma VN camp3 Turquoise
Chiusi camp1 Green Emblema with boar and deer Montevenere di Chiusi (Siena) Chiusi (Siena), Museo Archeologico
hunting scene Nazionale
Pollenza camp1 Green Emblema with boar and deer S. Lucia di Pollenza (Ancona), villa Ancona, Museo Archeologico Nazionale
Pollenza camp2 Blue huntig scene delle Marche
Pollenza camp3 Turquoise
Pompei L camp1 Turquoise Emblema with Theseus and the Pompeii, VII, 11, 8e10, House of the On the place
Pompei L camp2 turquoise Minothaurus Labyrinth, triclinium (42)
Pompei L camp3 Yellowish green
Solunto L camp8 Green Emblema with armillary sphere Solunto (Palermo), Leda house, triclinium On the place

Fig. 2. Microstructure observation on faience and sealing-wax red glass tesserae: the optical microscope images (left) are compared to the observation at the
electron microscope (right). From the top: Pollenza_camp1, a first type faience tessera, Pompei_L_camp3, a second type faience tessera and a sealing-wax red
glass tessera from the Navarca house in Segesta (Trapani). Pink rectangles mark the area analyzed by EDS to calculate the chemical composition
(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).
 Author's personal copy

C. Boschetti et al. / Journal of Cultural Heritage 9 (2008) e21ee26 e25

Table 2
The EDS composition of faience tesserae: in green the samples with the first microstructure and in yellow the group with the second one
Chiusi Pollenza Roma_VN Solunto_L Pollenza Pollenza Pompei_L Pompei_L Pompei_L Roma_VN
camp1 camp1 camp2 camp8 camp2 camp3 camp1 camp2 camp3 camp3
(Wt%) (Wt%) (Wt%) (Wt%) (Wt%) (Wt%) (Wt%) (Wt%) (Wt%) (Wt%)
Na2O 0.85 1.07 2.21 e 0.81 3.68 2.42 1.98 2.86 3
MgO 0.97 1.19 4.39 0.37 5.5 0.43 0.57 0.38 0.58 0.39
Al2O3 12.84 13.75 12.45 1.58 11.43 2.73 1.25 1.24 1.19 1.44
SiO2 77.62 73.15 44.37 95.83 31.03 87.2 82.48 86.29 74.04 87.24
Cl 0.39 e e e 0.1 0.31 0.19 0.66 0.43
K2O 1.11 0.99 0.67 0.15 e 0.62 0.36 0.7 0.7 0.65
CaO 4.83 7.79 29.96 0.8 46.05 3.65 4.14 4.08 9.99 3.87
FeO 1.1 1.31 4.42 1.27 5.18 0.47 1.16 1.87 1.7 1.68
CuO e e e e e 1.13 1.83 1.48 1.36 tr
PbO e e e e e e e e 6.91 e
SO3 0.29 e 0.66 e e e 5.46 1.79 e 1.31
TiO2 e 0.75 0.87 e e e e e e e
CoO e e e e e e e e e tr
BaO e tr e e e e e e e e
V2O5 e e e e e e e e e tr
P2O5 tr e e e e e e e e tr
SnO e e e e e e tr e e e
Sb2O5 e e e e e e e e tr e
100 100 100 100 100 100 100 100 100 100

AD. Because of the complex problems linked to this material, agent: this model of production was confirmed by an experi-
the analytical data will be discussed in detail elsewhere. For mental reproduction in electric kiln and in a wood fired
the moment, it is possible to report that all the red glasses were furnace, at the moment still ongoing.
identified by SEMeEDS with the so-called sealing-wax red
glass [8], a silicaesodaelead glass of composition of SiO2
38e51, PbO 16e35, Na2O 8e12, CuO 7e11, CaO 2e5, 3. Conclusions
Al2O3 1e2(Wt%). Minor elements are: K2O, Cl, MgO, FeO,
MnO, P2O5, Sb2O3. As a conclusion it can be stated that these This study gave the possibility to draw up the main tech-
glasses were coloured and opacified by cuprous oxide crystals, nical features of vermiculata mosaics from Italy, permitting to
clearly detected by the X-ray diffraction. The colouring raw hypothesize a local production. Moreover, the use of glass and
material, as attested by literature data [9], is clearly metallic faience tesserae was never reported before and thanks to the
copper, detectable as unmolten particles. characterization it was possible to enhance the knowledge of
these materials and, in the future, it will be possible to link
2.2.4.3. The isotope signature characterization. To make clear these data to the literature. Otherwise, the frequent use of the
the identification of the production site, a characterization sealing-wax red glass is meaningful, because it testifies the
(SIMS) of the isotope ratio of a group of meaningful elements choice of a refined material difficult to produce and certainly
(O, Sr, Cu, Fe, Pb) in glasses and faiences is in course. For the expensive. The isotope signature suggests the existence of
moment the work has been concluded only in red glasses for a specialized production, but in the future data shall be
oxygen isotope ratio, suitable to trace the provenance of the completed with a more comprehensive characterization.
sand [10]. The samples analysis revealed the existence of four
different groups: the first one is formed by sealing-wax red References
glasses dated 2nde1st century BC and, with small values
variation, mosaic tesserae and vessels of the 1st century AD. [1] A.-M. Guimier-Sorbets, M.-D. Nenna, Réflexsions sur la couleur dans les
These values are well distinct from the ones obtained from mosa€ıques hellénistiques: Délos et Alexandrie, BCH 119 (1995) 529e
a base glass chunk found in Pompeii and other glasses (col- 563.
[2] C. Fiori, Pavement en mosa€ıque de l’Îlot des Bijoux: composition des
ourless and dullish red, with low lead and copper content)
échatillons en matière vitreuse, in: A.-M. Guimier-Sorbets, M.-D. Nenna
forming a second group. A sealing-wax red glass tessera from (Eds.), Réflexsions sur la couleur dans les mosa€ıques hellénistiques:
Sicily (Segesta) has distinct values, so it must be attributed to Délos et Alexandrie, BCH 119 (1995) 553e563.
a different production. The difference between sealing-wax [3] C. Boschetti, A. Corradi, C. Leonelli, P. Veronesi, S. Santoro, Il corpus
reds and dullish red glasses could not be only index of dei mosaici alessandrini in Italia. Un progetto tra archeologia e arche-
ometria, in: C. Angelelli, F. Rinaldi (Eds.), Proceedings AISCOM 12th
a distinct production site, but also of a different technology:
meeting, Padova-Brescia, February 14e17 2006, Scripta Manent, Tivoli,
for the firsts it is true-life to hypothesize the melting of a batch 2007, pp. 547e554.
containing all the components while, for the seconds, the [4] C. Boschetti, A. Corradi, P. Baraldi, Raman characterization of painted
remelting of a base glass with the addition of the coloring mortar in Republican Roman mosaics, Journal of Raman Spectroscopy,
 Author's personal copy

e26 C. Boschetti et al. / Journal of Cultural Heritage 9 (2008) e21ee26

39 (8) (2008), Special issue. In: Baraldi P, Tinti A, editors. Raman [8] N. Brun, M. Pernot, The opaque red glass of Celtic enamels from
spectroscopy in art and archaeology, 1085e1090. continental Europe, Archaeometry 34 (1992) 235e252.
[5] A.J. Shortland, M.S. Tite, Technological study of Ptolemaic-early Roman [9] I.C. Freestone, Composition and microstructure of early opaque red
faience from Memphis, Egypt, Archaeometry 47 (1) (2005) 31e46. glass, in: M. Bimson, I.C. Freestone (Eds.), Early Vitreous Materials, The
[6] Y. Mao, Lead-alkaline glazed Egyptian faience: preliminary technical British Museum Occasional Paper 56 (1987) 173e188.
investigation of ptolemaic period faience vessels in the collection of the [10] K.A. Leslie, I.C. Freestone, D. Lowry, M. Thirlwall, The provenance
Walters Art Gallery, JAIC 32 (2) (2000) 185e204. and technology of Near Eastern glass: oxygen isotopes by laser
[7] M.S. Tite, I.C. Freestone, M. Bimson, Egyptian faience: an investigation fluorination as a complement to strontium, Archaeometry 48 (2)
of the methods of production, Archaeometry 25 (1) (1983) 17e27. (2006) 253e270.