dAf 51
DUNIKOWSKI Christophe, CABBOI Sandra
La sidérurgie chez les Sénons : les ateliers celtiques et gallo-romains des Clérimois (Yonne)
Discovered during the construction of the A5 motorway, the early metalworking site of the Clérinois (Tonne) delivered the remains of some ten workshops which produced iron between the 4th century B.C. and the 5th century A.D.
The published results are many : descriptions and typologies of the structures (18 furnaces studied), anthracology and analyses of the ores and slags. The study of the forged has allowed an estimation of the overall production : 5 000 tonnes of iron from 15 000 tonnes of smelted ore. Four fifths of this iron was produced between the end of the La Tène period and the 2nd century A.D. by large furnaces equipped with systems for removing the slag. These data can already be considered as a reference in the study of ancient metallurgy.
Abstract
Abstract
The excavation of the early metal-working Complex at Clérimois (Yonne) was part of the programme of rescue digs conducted along the line of the A5 motorway between Sens and Troyes (fig. 1). The deposits, which include three forges at intervals of 30 m, were placed on the north slope of a talweg on the edge of the Senonais plateau. Extensive topsoil Stripping and six months of excavation Discovered the remains of some ten Workshops and eighteen exceptionally well preserved furnaces (fig. 9).
The main activity recognized on the site Comprised the reduction of local iron ores (oxides and hydroxides).This happened in three phases between the fourth century a.c. and the fifth century AD. A particular type of furnace is associated with each phase. The post-reduction processing (purification of the blooms and the Manufacture of objects) was practically absent. This explains why it is impossible to determine the quality of the metal produced with any precision.
The start of the iron-working activity is Placed around the 4th century a.c. (14C dating). It was not possible to define the characteristics of this phase of production in detail: the structures of the only workshop observed in the northern part of the site were too poorly preserved (fig. 10: zone1, ensemble A). Nevertheless, the three "pit-furnaces" observed are distinguished by the absence of any means of removing the slag: this built up at the bottom of the depression, either on the floor of the furnace (fig. 59: type I.1) or in a small depression dug out to this effect (fig. 60 : type1.2).
The second phase starts at the end of the La Tène periode and continues after the Roman Conquest up until the 2nd century AD. It is marked by the appearance of very large furnaces using sophisticated techniques which include slag removal and multiple ventilation systems. The production of iron increases considerably during this period and at least five workshops are installed on the side of the talweg (fig. 9 : forges 1, 2 and 3a).
The nine furnaces (fig. 61: type II) functioned in batteries of two, embeded into the subsoil at the point of inflection of the slope. They present a large "crown" of sandstone which supports the sand-clay walls of the furnaces. These latter take the form of a basin, tronconic in form, some 0.80 m high and capped by a funnel-shaped chimney. The walls are thick and had undergone a number of repairs (relining, reconstruction). The floor, circular or oval (between 0.90 m and 1.20 m in diameter), is preceded by a canal for the evacuation of the slag. Three methods of ventilation have been discovered for this group of furnaces :
- two side openings placed about 20 cm above the floor. These communicate with alcoves built between the sandstone crown and the furnace;
- three orifices pierced through the rear wall, all at the same level;
- tuyères (fig. 62) made out of blocks of clay. These are pierced either by a single channel or by a main channel which divides into two diverging canals (bifurcating tuyère).
Two rather different structures are used alongside the type II furnaces at the end of the La Tène period or at the start of the Gallo-Roman period (fig. 66 : type III). Smaller than the others, this pair of kilns forms a battery to the north of zone 2 (fig. 41: forge 4). They stand directly on the geological subsoil. The walls form a horseshoe-shaped crown of stone of which the inner face is lined with clay. The oval floors (0.80 m2) open out onto a shallow depression designed to receive the molten slag. The poor state of conservation of these features does not allow the reconstruction of the ventilation system.
At the end of the 2nd century AD., the production of iron stops. A new type of furnace marks the revival of production from around the 3rd century AD. (fig. 41 : zone 2, forge 3b). These later structures are built of tile and brick bonded by a sandy mortar and stand on the slagheaps of the previous phase of activity. The oval basins posses vertical walls whilst the floors, also oval, cover only 0.30 m2. The ventilation is assured by a double air-flow: a funnel shaped tuyère placed at a height of 50 cm above the floor provides a lateral blast, whilst a frontal blast is provided by a semi-cylindrical tuyère placed obliquely in one half of the mouth of the furnace (fig. 52). The first arrangement is preceded by a trapeze shaped construction which would correspond with the emplacement of the bellows (fig. 67).
The reasons for the implantation of these reduction furnaces at Clérimois seem to stem from the nearby presence of raw materials. The supply of iron ore to the site is not yet fully understood. It is supposed that the small concentrations of surface ores were exploited by simple gathering or by small-scale excavations. The mining and initial treatment of ore was partially carried out on the site during the second phase of occupation. The only fuel used during this phase was the charcoal obtained from oak or beech. During the later Empire, the main species used is still the oak, in conjunction with smaller proportions of beech and willow.
Apart from the furnaces, the other witness to the smelting of iron ores is the mass of waste products. These have been subjected to a series of chemical analysis. The objectives were to characterise the direct reduction processes and to look for the variations in the compositions of the wastes which might correlate with the reduction techniques observed on the site. The homogeneity of the two principal techniques was well illustrated.
The residues piled up into three slagheaps also allowed an estimation of the quantity of iron produced. The approximate volume of these heaps (about 7800 m3) represents about 12000 tonnes of slag. Some 16000 tonnes of ore would thus have been smelted in the course of the various phases of activity, producing some 5000 tonnes of iron. 80 % of this total would have been produced during the second phase of activity, in other words between the end of the La Tène period and the start of the Gallo-Roman period. This allows us to advance the hypothesis of a continuous production during these three centuries. This was made possible by the technology employed from the La Tène period onwards. However, the production increased considerably after the Roman Conquest. The question of a participation or a control by the public authorities (civitas or the Roman State) remains to be answered. From the 3rd century AD. onwards, the organisation of the work and the production itself take on an “artisanal” rather than an “industrial" nature. This could be tied in with fluctuations in the demand for iron, or with the reorganisations of the economic structures. Whatever the case may be, important technological changes occur during the later Empire. These changes carry the mark of a certain know-how foreign to the previous traditions and would seem to anticipate the traditions known throughout the early Middle Age.
