In Gironde, the archaeological levels dated between the 9th and 13th centuries frequently yield the remains of a very special ceramic. Its appearance earned it the colloquial name of “ nougatine ”. Of a composite nature, it combines large angular fragments of chamotte (crushed fired clay) with a clayey cement of cream to red ocher color (fig. 1). This association confers low cohesive strength which explains a high fragmentation rate. This type of paste attested since the 1980s corresponds to category No. 3 of the standard for Bordeaux ceramics from the 10th to the 15th centuries drawn up by Sylvie Fabre-Dupont Maleret (Fabre-Dupont Maleret 1995 and 1996). The shards collected are of one form, with a frustoconical profile. Its very thick wall is pierced with at least one circular orifice about 2 cm in diameter (fig. 2). There are variants depending on date and / or their place of discovery. The chamotte ceramics listed to date in Gironde cover a large part of the department (fig. 3) where they are present in all types of habitats. On the urban site of place Camille Julian in Bordeaux, where one type was partially reconstructed, they represent around 2% of the vestiges dated 12th century (Fabre Maleret 1995). In Entre-deux-Mers, the castral site of Le Castéra in Langoiran (12th / 13th century) yielded 1,841 fragments. The archaeometric study carried out showed many differences between ceramic with chamotte clay and common ceramic in a light coloured regional paste. The first is from a totally different raw material supply. Contemporary Bordeaux potters use predominantly kaolin clay, while chamotte paste incorporates marly clays rich in calcium carbonate and iron as a binder (fig. 4). Local pottery productions do not use chamotte as a degreaser but sand. The high percentage of chamotte, almost equivalent to that of clay, technically implies a paste that is difficult to use because at the limit of plasticity. The numerous vacuoles of elongated shapes with varied orientations and the high porosity of the paste implies high porosity, short mixing times and rapid shaping. The presence of water swelling clay minerals such as smectite and detritic calcite supports the hypothesis of low temperature firing (less than 800° C). The clay is only surface fired and the core of the very thick walls remains raw. These vases have no equivalent in the light coloured grainy regional pastes between the 9th and 13th centuries, when open forms tend to disappear from the potter’s repertoire. The manufacturing and cooking processes for chamotte vases are so distinct from contemporary local pottery practices that it seems improbable that potters could have made them. Coupled with the ceramological and archaeological study, experimental archeology has shown that their fairly simple manufacture is within the reach of occasional or domestic practice (fig. 6 and 7). Experimentation across the entire production process has made it possible to verify some of the hypotheses stated (choice of raw materials, paste preparation, shaping techniques, finishes and type of firing). Chamotte, at the origine, seems to come from the crushing of a architectural terracotta whose chemical analyzes have revealed very diverse compositions (limestone or nonlimestone) and where petrography has again highlighted chamotte in the paste of terracotta. For the experiment, we used fragments of ancient tegulae, commonly found on local medieval sites. This step, much longer and laborious than the simple extraction of sand as a degreasing material, must really be essential to justify such an expenditure of time and energy. To obtain the clay matrix of the samples analyzed, a locally abundant carbonaceous common clay. It contains impurities which make it unsuitable for making fine ceramics but does not pose a problem for chamotte clay pottery. From our observations it was possible to reproduce the shaping process. The bottom is first formed of a thick plate of paste pressed on a flat surface and leveled with the finger. Large coils are used to mount the walls to the desired height and inclination. After the time for hardening, the finish was made by smoothing and a hole was pierced from the outside to the inside. The vases thus obtained are thick-walled which complicates the drying phase. Firing was carried out in the open area which could explain the low temperature reached and the use of chamotte as a degreaser to limit thermal shock, but this firing method still proved to be still too hot. To date, the exact function of these specific vessels remains unclear. The characterization of the paste, systematically under-fired, makes it possible to exclude the use of this ceramic as a container for liquids and therefore its use for washing, cooking pot or storage vase for un-dried preparations. The Gironde examples are very often associated with domestic hearth discharges or discovered near these hearths. This association with house-holds, also observed elsewhere, encourages the description of these vases as overturned cloches used to cover fires (Husi 2014). Ceramic fire-covers were produced by potters in the late Middle Ages, but their walls were much thinner, their paste did not need to be “ chamotted” and their shape was different. The use of chamotte vases therefore seems justified by the need to meet very specific needs related to cooking food. The reassembly of two vases discovered at Aubie-et-Espessas highlighted the presence near the bottom of two diametrically opposite holes (fig. 8). Their arrangement suggests that these openings are not linked to the evacuation of content but rather to a handling system. In all the specimens observed, only the edges are fired. The chamotte was probably added as a prefireded degreaser in order to compensate for insufficient firing of the paste, or even to allow its use raw, while preserving its very refractory properties. These indices allow us to suppose that they are inverted vases which can be shaped in the same way as domestic hearth plates and fired on the surface according to the frequency of their use. This mobile bell-shaped cover could be used for baking leavened bread, sub testu (“ under a cover”) as described by the Latin agronomist Caton (Caton, Agr., 74.1). In Italy, the tradition of baking bread under a bell-shaped cover (testo da pane) continued until the 13th century and seems to be identified in southern France at the end of the 12th-13th centuries (Lécuyer 1992). In Romania, where several archaeological discoveries, attested between the 7th and the 10th century, show the use of baking cloches with comparable characteristics (Stanciu 2018), and ethnology shows how they were used. During the Ropotin festival (Mesnil & Popova 2002), raw baking cloches, placed on the floor, are used to bake breads. In the Romanian countryside, cooking under ţest was also used for vegetable or meat dishes. The ţest is easy to use and heats quickly. An iron rod passed through the two holes serves as a lever and keeps the cover open during preheating (fig. 9). Appearing in the 9th century, the chamotte clay pottery cloche is omnipresent from the 10th to the 12th century, before disappearing in the 13th century from the Gironde medieval ware. This constitutes not only a good chronological marker but also the expression of a production adapted to the cooking, under baking cloches, of foods like bread on the domestic hearth. If chamotte paste is linked to such culinary practices, its abandonment could reveal changes in the regional domestic cooking practices., En Gironde, les niveaux archéologiques datés entre le ixe et le xiiie s. livrent fréquemment les restes d’une céramique très singulière qui contraste avec les productions céramiques médiévales régionales. L’aspect de sa pâte lui a valu l’appellation familière de “ nougatine”. Cette céramique de grande dimension, au profil tronconique, à la paroi très épaisse, entretient un lien étroit avec les foyers domestiques. L’étude pluridisciplinaire de ces céramiques a permis de déterminer leur mode de fabrication et de définir leur fonction qui est la cuisson des aliments sous cloche., Marache Valérie, Faravel Sylvie, Ben Amara Ayed, Cantin Nadia, Sireix Christophe, Charpentier Xavier. "La nougatine" : étude régionale d’une céramique à pâte chamottée médiévale. In: Aquitania : une revue inter-régionale d'archéologie, tome 35, 2019. pp. 175-189.