[PDF] Technical ceramics for salt production in Western Sahara

View - Download Technical ceramics for salt production in Western Sahara

Previous PDF

Next PDF

Technical ceramics for salt production in Western Sahara

David Larreina-García

a , Andoni Saenz de Buruaga b , Andoni Tarriño Vinagre b and

Belén Notario

a a

Centro Nacional de Investigación sobre la Evolución Humana (CENIEH), Paseo Sierra de Atapuerca 3, 09002

Burgos, Spain;

b Universidad del País Vasco (UPV-EHU), Miguel de Unamuno 3, 01006 Vitoria-Gasteiz, Spain

ABSTRACT

This paper presents evidence of ceramic technology in western Tiris (Western Sahara), dated by thermoluminescence to the third millennium cal. BP. Western Tiris is an arid region mostly covered by desert where recent archaeologicaleldwork has nevertheless revealed a signi cant network of settlements from the Neolithic period inhabited by nomadic people. Domestic pottery and lithic materials are common in the archaeological register of these sites, but three sherds found in the Lejuad XVII rockshelter present features typical of technical ceramics. Laboratory analyses reveal that abundant mineral and organic temper was added to the natural clay which, in addition to the presence of thicker walls than those usually found in domestic pottery, is interpreted as an attempt to increase resistance to thermal shocks. However, the fragments present only mild signals of exposure to high temperatures, up to a maximum of 900 °C. Discussion of these contradictory data leads to the conclusion that the sherds may have been part of a briquetage mould to extract salt by evaporation, a pyrotechnical industry previously unknown in Western Sahara. Its appearance in an arid environment far from production centres is explained as result of sporadic economic activity rather than cross-cultural mobility and trading, which seems to have been intense in the area from Neolithic times. In addition, this paper introduces the use of micro-computed tomography (µ-CT) as a technique for measuring large porosity derived from burned organic materials.RÉSUMÉ Cet article présente des données concernant la technologie céramique dans le Tiris occidental (Sahara Occidental), qui se place, sur la base de datations par thermoluminescence, au troisième millénaire cal. BP. Le Tiris occidental est une région aride principalement couverte de désert où des travaux archéologiques récents ont néanmoins révélé un réseau important de sites datant du Néolithique qui furent habités par des populations nomades. La poterie domestique et les lithiques sont courants dans leur registre archéologique, mais trois tessons retrouvés dans l'abri sous roche de Lejuad XVII présentent des caractéristiques typiques de la céramique technique. Les analyses de laboratoire révèlent qu'un abondant dégraissant minéral et organique fut ajouté à l'argile naturelle. Ceci, ajouté au fait queARTICLE HISTORY

Received 26 May 2020

Accepted 18 February 2021

KEYWORDS

briquetage; Western Sahara; salt; mobility; µ-CT; porosity calculation © 2021 Informa UK Limited, trading as Taylor & Francis Group

CONTACTDavid Larreina-García

-370 les poteries ont des parois plus épaisses que celles habituellement observées dans la poterie domestique, est interprété comme indiquant une tentative d'augmenter la résistance aux chocs thermiques. Cependant, les fragments ne présentent que de faibles signaux d'exposition à des températures élevées, jusqu'à un maximum de 900°C. L 'évaluation de ces données contradictoires amène à la conclusion que ces tessons ont pu avoir fait partie d'un moule de briquetage pour l'extraction de sel par évaporation, une industrie pyrotechnique jusque-là inconnue dans le Sahara Occidental. Son apparition dans un environnement aride loin des centres de production s 'explique par une activité économique sporadique plutôt que par une mobilité et un commerce interculturels, apparemment intenses dans la région après le Néolithique. En outre, cet article présente l'utilisation de la tomographie micro-calculée (μ-CT) comme technique de mesure de la grande porosité dérivant de matières organiques brûlées.

Introduction

Western Tiris (Tiris al-Gharbiyya) is a southern area of barren desert terrain in Western Sahara in the westernmost portion of the Sahara Desert bordering with the Atlantic Ocean and Mauritania (Figure 1). Mainly consisting of desertatlands and strongly con- ditioned by an extreme climate and arid environment, it is currently one of the most sparsely populated regions in the world with barely2.3 inhabitants per km 2 (United Nations2019, citing the number for the whole of Western Sahara). Nowadays, as in pre- vious times in the historical past, Tiris is still a place of reference for the nomadic groups in the western reaches of the Sahara. The research project'Recovery, Conservation and Scientic Study of the Archaeolo- gical, Environmental and Cultural Heritage of Western Sahara'began in 2005 as an initiative with the double aim of establishing a research platform in the region and of engaging in humanitarian co-operation with its inhabitants in order to investigate the cultural roots of an archaeologically and historically much-neglected territory. Between then and 2016eldwork covered an area of30,000 km, locating a total of

1053 prehistoric and pre-Islamic sites that dier from each other functionally, culturally

and chronologically. They include settlements, lithic industrial workshops, funerary and ritual sites and quarries for mining siliceous rocks and range chronologically from the Early and Middle Stone Age (Upper Acheulean to Aterian;350,000-20,000 cal. BP) to the Neolithic (c.7800-3000 cal. BP) and,nally, the Proto-Berber period (c. 3000-

700 cal. BP). The bulk of these sites consist mostly of lithic artefacts that often

number in the thousands (Figures 2-3) (Sáenz de Buruaga2008,2014b,2018; Sáenz de Buruagaet al.2009). Of special interest to this paper is the Lejuad area, which denes one of the most important geographical contexts for our archaeological knowl- edge of Western Sahara and takes the form of a network of settlements of Neolithic tra- dition dating to betweenc.4500 and 1500 cal. BP. More than 150 sites are known, rich in stone artefacts and pottery (both large and small vessels usually decorated with dots and lines that are either impressed and/or incised) and frequently associated with rock art (Sáenz de Buruagaet al.2012; Sáenz de Buruaga and Arruabarrena2015:67-88; Sáenz

AZANIA: ARCHAEOLOGICAL RESEARCH IN AFRICA

de Buruaga2019). Six of the 73 rock art sites have engravings of metallic items - knives, axes and halberds - that stand out since the presence of archaeologically known metal objects in the vast extent of 185,000 km² of Oued ed-Dahab (Río de Oro) is e ectively anecdotical, consisting of just one copper arrowhead found in 1951, plus a copper ring found out of archaeological context in 2009 (Sáenz de Buruaga2015: 365). In 2011, we noticed the presence of a small (fl21.0 ×14.5 cm) horse engraving in the rockshelter of Lejuad XVII (Figure 4) where in 2018 we recovered from the surface the three ceramic sherds that form the case study of this paper (Figure 5). These fragments - LJ01, LJ02 and LJ03, which are analysed in detail below - present unusual features unseen in the domestic pottery recovered thus far, such as thick walls, the addition of extra mineral and organic temper, a very coarse and unnished appearance and confus- ing signals of intensering. They were originally thought to be part of a metallurgical crucible. Signicantly, the shelter of Galabt Lejuad-Gleb 11, which is 4 km from Lejuad XVII, hasve rock engravings of double-edge curved axes that resemble the 'Metgourine'type (Figure 4) (Sáenz de Buruaga2014a). This encouraged us to consider the possibility of a relationship between these images of metal and the possible traces of Figure 1.Map showing the location of the Lejuad area in Western Sahara. 346
metal production. Although the subsequent analyses which we discuss here now suggest that this is not true, the presence of these engravings at Galabt Lejuad-Gleb 11 still holds interesting implications for the Lejuad area.

Figure 2.Typical Palaeolithic site: perspectiveof the northern sector of the Agsumal mountain (Mijek),

in northern Tiris and a series of Upper Acheulean bifaces found near the base of the rock relief. The

scale is in units of 10 cm. Figure 3.Typical Neolithic site in the Emherisat-2sebjaarea (Duguech) in southern Tiris with the characteristic associated assemblage ofaked and polished stone tools and impressed-decorated pottery recovered from the surface. The scale is 5 cm long for the geometric stone artefacts and the ceramics and 10 cm long for the polished stone tools.

AZANIA: ARCHAEOLOGICAL RESEARCH IN AFRICA

The Metgourine axe typology is typical of the High Atlas Mountains of Morocco. In simplied terms, the engravings represent the blade of the axe as a marked crescent, resembling a new moon, mounted on a shaft. Following publication of a paper by Sáenz de Buruaga (2014a) debate ensued on the engravings'meaning, distribution and typology since the Lejuad panels show dierences from the typical Metgourine categor- isation (Auclair2015; Sáenz de Buruaga and Arruabarrena2015; Rodrigue2016; Sea- right-Martinet2016a,2016b). These discussions concerning the typological classication of the engravings are beyond the scope of this paper, but two important conclusions should be noted. First, the engravings depicting metal artefacts in the High Atlas are of objects recognisably inuenced by the Argaric Culture of the southern Iberian Peninsula, with the notable exception of the crescent moon double-edged axe (Metgourine), which is considered as genuinely Saharan in origin. Second, variability in the artefacts'typology may correspond to intercultural exchange between dierent groups moving from the High Atlas to Mauritania and other parts of the Sahara. Chrono- logically the expansion of this typology concentrates in the second millennium BC, i.e.c.

3800-3200 cal. BP.

Leaving out the debate and restricting o

urselves to western Tiris, it is signicant that the closest metal production site to Lejuad is approximately 300 km distant as the crow ies at Akjoujt in west-central Mauritania, which was a major producer of copper metallurgy that reached the peak of production during thersthalfofthethirdmillen- nium cal. BP (rst millennium cal. BC) (Vernet2012,2014). The Lejuad engravings are much closer typologically to the considerably more distant High Atlas region, which is Figure 4.a) Lejuad XVII rockshelter; b) the main mountain complex of Lejuad in southern Tiris I; c) engraving of an equinegure on the right wall of Lejuad XVII rockshelter with a 10 cm scale; d) engravings of two curved double-edged metal axes and a square handle at the Galabt Lejuad-Gleb

11 site with a 10 cm scale; e) detail of a"en éventail"axe at Galabt Lejuad-Gleb 11 with a 5 cm scale.

348
some 1500 km away in a straight line. Metal production is evident there by the fourth millennium cal. BP (second millennium cal. BC), which is in agreement with the esti- mated chronology of some of the engravings (Sáenz de Buruaga2014a:175).Tosum up, today's nomadic way of life based on constant mobility is hardly a novelty since long distance contacts are documented in the region from at least the Neolithic. In spite of unfavourable climatic and geological conditions, the environment of Lejuad and its surroundings in the southern section of the Tiris must thus have o ered sucient resources to guarantee the stability of human groups whose material culture documents contacts with and inuences from other populations elsewhere (Sáenz de Buruaga2013). The distribution of engravings of the Metgourine-type Figure 5.Ceramicsfrom LejuadXVII: aandb) external surfaceandsection of LJ01;candd) surfaceand section of LJ02; e and f) external surface and section of LJ03.

AZANIA: ARCHAEOLOGICAL RESEARCH IN AFRICA

Saharan axe, in particular, demonstrates wide human mobility and cultural interaction (Auclair2015). The casual discovery at Lejuad XVII of the three ceramic fragments mentioned above may reinforce this scenario of mobility and migration for eking out a living in hostile environments by adding another ingredient dierent to the usual lithic, pottery, metal, and art arguments, namely salt.

Methodology

Evidence of the presence of technical ceramics constitutes an‘exotic"material within nds comprise lithic artefacts or domestic pottery framed within a well-explained Neolithic horizon. A set of labora- tory-based analytical methods was launchedinordertogatherasmuchinformationas possible to determine if the three shreds from Lejuad XVII were indeed fragments of the same container, to identify the specic activity practised with it and to provide a relevant chronological framework. The primary tools selected were a range of micro- scopic analyses—Optical Microscopy (OM), Scanning Electron Microscopy (SEM- EDS), X-ray Fluorescence (WD-XRF) and X-ray-Diraction (XRD)—capable of describing and analysing mineral phases and internal microstructures and of obtain- ingbulkchemicalcompositions.Inaddition, we employed X-ray micro-computed tomography (μ-CT) to better explain the construction of the ceramic fabric, thermo- gravimetric analysis (TGA) to estimate thering temperatures, mercury porosimetry to calculate the ceramic porosity and Th ermoluminescence to determine the frag- ments"age. Each sample was utilised in dierent analyses, but due to the limited avail- ability of material not all the samples weresubmitted to the same analytical method (Table 1). Following macroscopic investigation, to obtainat sections observable by OM and SEM microscopy three polished blocks were prepared at the Wolfson Laboratory of the UCL Institute of Archaeology (London, United Kingdom) following the standard procedure of mounting the samples in epoxy resin. They were prepared by grinding them on successivelyner abrasive paper before polished them to a 1 μmnish. In addition, two thin sections for petrographic examination were also prepared for observation under Optical Microscope (using an Olympus BX51). A FEI Quanta

600 SEMtted with an energy-dispersive spectrometer (EDS) and Oxford Instru-

ments INCA software to interface with the EDS was used to estimate the bulk elemen- tal composition. A third sample was powdered for WD-XRF and XRD analyses. The sample wasrst crumbled in a Retsch BB51 jaw crusher and theresulting fragments were then milled to a grain size of approximately 50 μm using a Retsch RS200 vibratory disc mill equipped with a tungsten carbide coated jar. The powder was dried to remove moisture, mixed Table 1.Lejuad XVII ceramics: laboratory-based methods used per sample.

OM SEM µ-CT XRF XRD Hg porosimetry TL

350
with an alkaliux (LiBO2/Li2B4O7, 34/66% wt.) and fused at 1150°C in a Pt-Au cru- cible using a PANalytical Perl'X3uxer to obtain a glass disc. The bulk chemical ana- lyses were carried out using an Axios PANalytical WD-XRF spectrometer, while the instrument used for XRD was a X´Pert PRO MPD PANalytical diractometer. For the latter, a small quantity of milled sample (fl1 g) was placed on a back-load sample holder and the interpretation of the resulting diraction pattern performed with PANalytical Highscore Plus v4.7 software. The bulk of the analyses were undertaken at the Microscopy and Microtomography,and Archaeometry laboratories at the Centro Nacional de Investigación sobrela Evolución Humana (CENIEH), Burgos, Spain. The exceptions were mercury porosimetry, which was carried out at the INCAR Instituto Nacional del Carbón in Oviedo, Spain, and the thermoluminescence dating, which was carried out by Dr N. Zacharias at the University of the Peloponnese,

Kalamata, Greece.

Characterisation of the Lejuad ceramics

Three small fragments of technical ceramics - the term understoodas referring ceramics that have been used in high-temperature operations rather than as domestic pottery (Martinón-Torres and Rehren2014: 107) - were found on the surface of the Lejuad XVII rockshelter (Figure 5;Table 2). The best-preserved sherd - LJ03 - shows a degree of convexity, suggesting that the vessel was not large, but none of the sherds allows us to reconstruct shape or estimate size. LJ03 has a rough internal surface that is orange in colour, while its exterior is rougher and exhibits light brown ashy colours with occasional darker'smoky'areas that suggest contact withames. LJ01 also has two distinct faces, the external one being uneven and uniformly black rather than smoky, the internal one orange. Lastly, LJ02 has uneven surfaces that display the orange colour of the clay fabric, indicating the loss of material on both of them. Overall and regardless of preservation state, the sherds were manufactured from a crudely baked fabric with multiple coarse inclusions visible to the naked eye. All are undecorated and unpolished, showing predominantly orange tones typical of oxygen- richring atmospheres. In general, the ceramic fabric seems to have been mostly unal- tered by high temperatures since there are no obviously visible traces of discoloration due to contamination by metallic oxides, slag or glaze, or of bloating or cracks. LJ01 and LJ02 both have relatively thick (fl20 mm) walls, whereas LJ03 is thinner (10 mm). Freshly cut sections show that there is a gradient in colour from a black thin external layer that fades into a grey-blue at the core beforenally becoming yellow-orange at the inner surface

Figure 6

). In LJ03, the yellow-orangish layer ends in a very thin (fi0.5 mm)'rust'- coloured internal layer of slightly smoother texture and brittle appearance. Table 2.Main macroscopic characteristics of the Lejuad XVII fragments.

LJ01LJ02LJ03

Dimensions (mm)36×31×2247×36×1950× 36×10

Weight (g)8.2107.7

External surface colourblackorangelight brown

Internal surface colourorange-redorange-light brown orange-light brown Section colour (outer/inner) black-grey/orangish black-grey/orangish black-orangish/rusty

AZANIA: ARCHAEOLOGICAL RESEARCH IN AFRICA

Figure 6.Lejaud XVII: photographs of the ceramics in cross-section (on mounted polished blocks) showing the colour gradient between the surfaces. This is black-grey at the external surface, turning toward orange towards the interior. Note the abundant porosity and oxidising colours on the thin inner layer in LJ03. Figure 7.Lejuad XVII ceramics: a) typical heterometric structure of the ceramic paste mainly showing

the grain textures of quartz, feldspar and plagioclases, as well as abundant porosity. The scale is 1 mm

long (taken in polarised light, XPL); b) detail of a rounded detrital quartz grain (0.75 mm diameter)

showing a reaction crown, fresh-formed phyllosilicate crystallites surrounding the grain and pores. The scale is 0.2 mm (taken in plain polarised light, XPPL). 352

Results of instrumental analyses

In spite of the colour dierence, SEM-EDS results show that the bulk chemical compo- sition of the fabric is essentially the same in all three sherds, comprisingfl55-62% SiO 2 andfl18-21% Al 2 O 3 and incorporating signicant amounts of CaO (fl5-14%) and Fe 2 O 3 fl

7-8%) as major compounds (Table 3). The composition is completed with

fl2% Na 2

O,fl1% MgO and traces of K

2

O and TiO

2 , both at <1%. Dierences are evident when the fabrics are compared with the natural soil in the area. Bulk compo- sition of the soil clay was analysed by WD-XRF from an excavated funerary deposit fl3.85 km from Lejuad XVII and showed a much lower (3.2%) percentage of CaO com- pared to the ceramics. Our microscopic observation reveals that the sherds'paste has an isotropic matrix showing abundant porosity and textures of non-plastic materials (Figure 7). Overall, the mineral particles identied by petrographic examination and SEM-EDS phase ana- lyses are randomly distributed throughout the fabric and tend to be angular or sub- angular, although the largest ones are rounded. The predominant mineral inclusions have a size range of 100-500 µm - although they are frequently larger, up to 2 mm - and correspond mainly to angular grains of plagioclase feldspar. Quartz particles are less dominant and mostly show morphologies as rounded grains of sand of variable sizes such asne grains measuring 50-200 µm, but also as larger grains measuring 0.5-

2.0 mm. Signicantly, the quartz grain boundaries appear to be surrounded by a crown of

small crystals of new crystallised phyllosilicates produced by thermal alteration of the ceramic (Figure 7). Other minerals commonly observed in the matrix are potassium feld- spars (microclines). A few small (fl50-100 µm) grains of titanium oxides (ilmenite, tita- nite) are occasionally present, as well as small (<50 µm) grains of zirconite (Figure 8). Lastly, other metal-rich particles very rarely occur in the fabric and, if present, are tiny, looking like natural mineral or slightly reacted grains (Figures 8e,8f). Grains of barite (BaSO 4 ) were found in LJ02 and LJ01. Within the latter fragment they appear to be almost exclusively concentrated along a thin line separating the black external and orange internal layers (Figure 8b). Residual grains of copper carbonates (61% CaO,

39% CuO) were uniquely identied in LJ02.

The angular shape of the plagioclase grains suggests that they were not an original component of the clay and that they may therefore have been added as temper, some- thing that is also suggested by the larger appearance of plagioclases within the ceramic fabric compared to the natural clay (Table 4) and that is additionally pointed out by Table 3.Bulk chemical composition of the three ceramic fragments from Lejuad XVII and the soil of Lejuad by SEM-EDS andWD-XRF analyses. SEM-EDS analyses were obtained by averaging the chemical analyses ofve areas (1 by 1.4 mm) per sample. Results of the WD-XRF analyses are given as stoichiometric oxides except Cl (by weight %) and are normalised to 100%. Original data are available as supplementary materials. Na 2

O MgO Al

2 O 3 SiO 2 K 2

O CaO TiO

2 Fe 2 O 3

LJ01 black 2.4 0.9 18.9 55.6 0.7 13.0 0.7 7.6

LJ03 black 2.3 1.4 18.2 55.0 0.7 14.8 0.6 6.9

L01 orange 2.7 0.8 20.2 62.5 0.9 5.0 0.6 7.1

LJ03 orange 2.6 0.8 21.1 54.8 0.8 11.1 0.7 7.9

L02 orange 2.1 1.3 18.2 61.8 0.8 7.0 0.7 8.0

Natural soil* 2.2 1.1 12.1 67.5 1.8 3.2 0.9 6.5

AZANIA: ARCHAEOLOGICAL RESEARCH IN AFRICA

Figure 8.Lejuad XVII ceramics: SEM micrographs showing the ceramic microstructure: a) dierences in texture in LJ03 between the orangish-rusty ceramic (the darker, upper part of the photograph) and thered ceramic (showing a more compacted structure). Note the vacuoles and mineral grains cross-

ing over the two areas; b) the transition line between the normal fabric (left) and the black layer in

LJ01 showing a mostly identical microstructure; the thin bright line is rich in barite particles (white); c) detail of coarse unfused plagioclases (dark grey) and phyllosilicates (light grey) and a

few ilmenite particles (white); d) typical texture with dominant plagioclases (light grey) and quartz

inclusions (dark grey) plus an occasional presence of ilmenite (bright grey) and zirconite (white); e)

detail of copper-rich metal particles (white) in LJ02; f) detail of barite crystals in LJ01 showing their

needle shape (white); the two light grey particles are ilmenite. 354
the compositional data (Table 3). We can thus conclude that the bulk chemical compo- sition of the blackred layer is equivalent to the orange ceramic, whereas the main dier- ence with the natural clay is a decrease in SiO 2 in favour of CaO and Al 2 O 3 .Figure 9 shows a negative correlation between the lime - donated by the plagioclases - and the alumina and silica - the two main ceramic oxides naturally present in the clay. Con- sequently, thesaturation of thefabric with plagioclasesrich in alumina and lime results in an increment of these two oxides in the bulk chemical composition, with a corresponding decrease in the proportion of silica, whereas the higher lime levels in the black fabric are seemingly due to enrichment caused by the more intensering and visible in the appari- tion of incipient phases of phyllosilicates (Figure 7). Note that the rest of the oxides in the composition maintain the same proportions, which suggests that the main step in the preparation of the ceramic paste involved tempering the natural clay with plagioclases. X-ray micro computed tomography slices show that the sherds'fabric has abundant porosity, some of large size (fl2-3 mm) and exhibiting the characteristicbre-shaped voids left by burnt-out organic temper after the ceramic wasred (Martinón-Torres and Rehren2014). These vacuoles are particularly profuse in LJ01-02, but only occasion- ally present in LJ03. In LJ02 there are two areas where there seems to be a concentration of elongated pores, which seem to correspond to small clusters of organic materials not well sorted during clay preparation (Figure 10).

Table 4.Lejuad XVII ceramics: XRD results of mineral inclusions present in ceramic andclays. Note that

the presence of calcite was conrmed by Raman spectroscopy. Quartz Calcite Feldspar Plagioclase Phyllosilicate

Ceramic fabric 21

-40% 5-10% 5-10%>41%<5%

Soil>41%-5-10%21-40%11-20%

Figure 9.Lejuad XVII ceramics: binary plot of the major elements SiO 2 +Al 2 O 3 of the bulk composition against CaO.

AZANIA: ARCHAEOLOGICAL RESEARCH IN AFRICA

Regarding thering, SEM and μ-CT still micrographs reveal that there are but rare signals (e.g. bloating, cracking, etc.) of submission to high temperatures, while the main body of the sherds shows a packed matrix of patched appearance, surfaced by abun- dant angular shape minerals that seem barely sintered and not properly melted. This aspect is better appreciable on SEM micrographs (Figure 8d). However, the tomography also shows a very distinct layer of more thermally altered material corresponding to thequotesdbs_dbs24.pdfusesText_30

[PDF] céramiques

[PDF] Céramiques de Cornouaille Quimper - Anciens Et Réunions

[PDF] Céramiques de femmes De Brueghel à David Teniers

[PDF] Céramiques glaçurées de l`Antiquité tardive et du haut - France

[PDF] ceramiste - Kaolin`e - Festival

[PDF] CERAMITEC 2006 aura lieu à Munich du 16 au 19 mai en même

[PDF] Cerapur Cerapur Comfort

[PDF] CERAPURSMART caldaia BOSCH caldaia JUNKERS a - Italie

[PDF] CeraStyle Mini. - Moser Professional

[PDF] CeraTherm - Mitigeur thermostatique bain - Anciens Et Réunions

[PDF] CeraTherm - Mitigeur thermostatique douche mural (Code article

[PDF] Ceratherm 50 Mitigeur thermostatique bain/douche - Anciens Et Réunions

[PDF] Cerato (Plumbago) - Anciens Et Réunions

[PDF] Cerato - Plumbago - Anciens Et Réunions

[PDF] Ceratodermia palmo-plantar de Buschke-Fischer - Anciens Et Réunions