1Remains of possible pottery workshops were observed for the first time during the construction of the Epinay-sur-Seine-Le Tréport railway line in 1874. The site was then completely forgotten and, a century later, systematic field-walking by the JPGF association in the Rosne and Ysieux river valleys, in the region of ‘Pays de France’ north of Paris, brought finds that demonstrated continuous pottery production from at least the Gallo-Roman period through medieval times. In 2009, an in-depth study of the material resulting from systematic surface prospecting, conducted within the frame of a research program on local ceramic production, highlighted in situ production of common pottery and terra sigillata (Samian ware) of the Argonne type during the Lower Roman Empire. The existence of this specialized artisanal activity being clearly established, it remained to locate the workshop(s): the presence of characteristic debris (baking supports, pipes, silt nodules with vitrified face, etc.), more or less concentrated in the ploughing horizon on several plots, pinpointed four zones probably corresponding to production units with one or more kilns, distributed on each side of the ancient road from Paris (Lutecia) to Amiens (Samarobriva), on the border of the territories of Ecouen to the east and Saint-Brice to the west, 17 km north of Paris.
2Following magnetic and EMI prospection in the winter of 2013, the first archaeological excavation was carried out in Saint-Brice in the summer of 2013, in the western sector of the site, at La Plaine du Moulin (Guadagnin, 2016) where the presence of a Lower Empire workshop was expected due to surface remains: a levelled kiln, a small dump and layers of occupation containing evidence of Samian ware production dated to the second half of the 4th century, with a terminus post quem provided by a coin issued in Amiens in 353. Production of 3rd century AD common wares can be expected, that of Samian ware in the 4th and at the beginning of the 5th centuries, and common ware again until the 6th/7th centuries.
3After a second series of geophysical surveys (2014), other diagnostic excavations were carried out in 2014 and 2016 in the eastern sector of the site, Les Reserves de Chauffour (Calfur = Calcis furnus) in Ecouen (Guadagnin, 2016); two other production units existed on the fringes of a marsh situated east of the ancient road. They made it possible to study two kilns of the ‘Argonne’ type (Chenet, 1941) belonging to a workshop making Samian-ware, well preserved under thick colluvium.
4The geophysical surveys carried out in both sectors of the site (Ecouen and Saint-Brice) in 2013, 2014 and 2021 were aimed at detecting traces of pottery workshops and, more specifically, tentative kilns. Magnetic and EMI surveys (2013 and 2014) revealed several anomalies, which could correspond to kilns: four in the eastern part of the site at Ecouen, between the road and the river (Rosne), and two in the western part at Saint-Brice (Tabbagh & Thiesson, 2015). The eastern two of these anomalies were verified by excavation to be kilns. The last survey (2021) has yet to be field-proofed.
5The survey at Plaine du Moulin/Saint-Brice in 2013 was carried out manually over an area of 47 m by 80 m, using two caesium magnetometer sensors in vertical gradient mode (Geometrics G858) and an EMI prototype (CS150, perpendicular configuration, 1.5 m inter-coil spacing). We found that it was difficult to use the data coming from the bottom magnetic sensor (h = 0.3 m) because of the disturbance produced by a multitude of small superficial ferrous objects (iron litter), probably as a result of the spread of urban sludge (coming from the nearby city of Paris) during the 19th and 20th centuries. Also, temporal variations several seconds long are recorded every time trains pass on the nearby railway. Some strong EM interference coming from the nearby airport beacons could also explain the linear stripes along the profiles seen on the map (Fig. 1). Nonetheless, the EMI surveys (f = 8 kHz) seem not to be impacted by the interferences and show up as a clear anomaly exhibiting a significant magnetic signal as well as higher electrical resistivity value. A small test excavation on the spot of this anomaly, interpreted as a potential pottery workshop, found a pile of Gallo-roman tegulae and mortar, but no kiln. In a second excavation, located 15 m to the north, based on the position of a well-defined magnetic anomaly (from a reprocessed magnetic map) but with no corresponding susceptibility anomaly, found no noticeable remains. In this area, EMI was found to give the most reliable results.
Figure 1. Saint-Brice – La Plaine du Moulin: map of Earth Magnetic Field (Geometrics G858, h = 0.85 m, local trends), Apparent Magnetic Susceptibility and Electrical Resistivity (CS150).
6A second survey in April 2014, placed east of the ancient road (Ecouen) over an area of 27 m by 100 m (G858) and 27 m by 50 m (CS60 EMI prototype, 0.6 m inter-coil spacing, vertical coplanar configuration), had a lower depth-of-investigation than with the previously used CS150.
7Four anomalies were identified as possibly corresponding to workshop structures, two (A and B in Fig. 2) in the northern part and two (C and D in Fig. 2) in the central part, where apparent magnetic susceptibility reached very high values close to 300 10-5 uSI. Anomalies A and B in the area not covered by the EMI survey corresponded to kilns, which were excavated during the summer of 2014. But archaeological testing revealed nothing at points C and D. Samples taken during the excavations made it possible to observe that heated subsoil (reddish) has a low magnetic susceptibility, significantly lower than that of the surface layer, and that it is the remanent magnetization which allows kilns to be detected. In this area, magnetics proved more reliable for the task of locating kilns.
Figure 2. Écouen – Les Réserves de Chaufour: position of maps from Figs 1 and 2; 3D model of a kiln found on the spot of the B anomaly; mapping the Earth Magnetic Field (Geometrics G858, h = 0.85 m), Apparent Magnetic Susceptibility and Electrical Resistivity (CS60).
8Finally, the EMI survey was carried out again in Saint-Brice in 2021, using a CMD-Mini-Explorer 6L (GF Instruments, Brno) with an RTK GPS (1-m interval between profiles, 7 cm in-line spacing), covered 1.8 ha –more than previously prospected– in search of other kilns that archaeologists were adamant existed there. The output of this instrument are six apparent electrical conductivity dataset corresponding to increasing depths of investigation (inter-coil separation = 0.2; 0.33; 0.5; 0.72; 1.03 and 1.5 m) and six apparent susceptibility dataset. Data were calibrated using a manual zero offset calibration and converted into apparent conductivities and susceptibilities. For inter-coil separation shorter than 0.5 m, conductivity maps show no particular anomalies, except those related to agricultural practice (heavy tractor workload). For deeper levels, clear resistive elongated anomalies that can be interpreted as masonry walls were found, as well as an outline of the 2013 excavations (Fig. 3). Unfortunately, the apparent susceptibilities maps were also heavily distorted by agricultural practice, but surprisingly not at the most superficial levels. This instrument shows the high vertical variability of magnetic susceptibility, which makes it a difficult parameter to use for the detection of kilns or refuse dumps in this situation.
Figure 3. Saint-Brice – La Plaine du Moulin: CMD-Miniexplorer-6L maps (mesh 0.5 m by 0.5 m) for apparent electrical conductivity and magnetic susceptibility.
9In conclusion, even in a very difficult situation like this one (railway interference, sludge), magnetic field measurements, if correctly processed, have good potential for detecting kilns because they are sensitive to the remanent magnetisation in situations where magnetic susceptibility of heated features remains low.
-small120.png){kind=logo}
Flux RSS
