Camulodunum type 262, le type Besançon and Central Gaulish granite-tempered ware
Background to the studyThe 1972 paper by A. & M. Ferdière (Ferdière and Ferdière,1972) describes two forms of moulded-rim jar found on early Roman sites in central France. The first is a plain, round-shouldered jar with a slightly out-turned beaded or thickened rim, moulded on the inner face. The second has a heavy triangular rim, moulded on the upper surface, and (in many examples) a horizontal cordon and band of incised decoration on the shoulder. The `typical' fabrics of these forms was described as coarse, grey-brown or black, and micaceous, although it was evident at this stage that several different wares, and sources, were represented in the style. The distribution covered some 30 sites, principally in the Centre region, but also further east into the Saône valley and beyond. The whole class was named le type Besançon, and this label has been widely used in the archaeological literature ever since.
From an examination of examples of these jars on sites in central France it is evident that although a range of fabrics are represented in assemblages of the 1st c. AD an important element of the earlier material (broadly 1st c. BC) is in a single coarse hand-formed grey or grey-brown ware, heavily micaceous and with large feldspar inclusions. In Britain, there are jars of apparently identical form and fabric at where it is form 262 in Hawkes and Hull's typology (Hawkes and Hull,1947), and at Silchester. A single tripod bowl from Camulodunum (type 45A) is in the same ware.
Description and petrographyThe ware has two striking characteristics in the hand-specimen: the abundant flakes of biotite mica which give a distinct golden-brown or red-brown sheen to the vessels, and the common and large (up to 1cm or more) feldspar inclusions, prominent in both surface and section. The vessels range in colour from orange- or red-brown through to dark brown, or almost black, but colours towards the middle of the range red-brown: Munsell 2.5YR 6/6 or dark red-brown: 2.5YR 4/4) are most common, although there can be a great deal of variation within a single vessel. The vessels are hand-formed, and burnished or smoothed on the rim and upper body.
The principal inclusions are derived from the in situ weathering of a coarse-grained granite. Among the inclusions there are numerous composite rock fragments, but these have usually decomposed into their constituent minerals. Feldspar fragments, commonly with a perthitic texture, are most abundant, and there are also sheaves of biotite crystals and smaller quantities of quartz (Tyers,1992).
Distribution and sourceSpecimens of this ware, confirmed by petrographic analysis, have a wide geographical spread, from Colchester (Britain) to Roanne (Forez), and Alet (Brittany) to Basel (Switzerland), and there are a number of other sites, mostly in central France, where the ware has been identified `in the hand'. Rather more problematic is the recognition of those specimens that undoubtedly lie disguised in the literature on le type Besançon, where there are numerous references to `micaceous' fabrics with undefined `large' or `white' inclusions. It clear from both the petrographic study and the hand-specimens that the Cam. 262 jar form occurs in a number of other wares, so caution must be exercised with identifications based solely on the (generally inadequate) published descriptions. For instance, in the middle Loire valley the form is found in a coarser variant of Rigby and Freestone's Central Gaulish Standard Fabric (Rigby and Freestone,1986). Similarly, a large group of moulded-rim jars published by Ettlinger from Augst (Ettlinger,1948) can also be discounted as they are not in the granitic fabric, and are presumably from some more local source. Before a thorough assessment of the history of this industry can be made it will be necessary to disentangle the underlying distribution of the granite-tempered ware from examples of the same form, but in other fabrics. These are in some cases later developments, in other cases purely local phenomena, but may include some widely traded wares as well.
A combination of arguments point to southern Burgundy and particularly the Morvan as the likely source of the ware. Cam. 262 jars in this fabric are certainly present in large numbers among the material from Bibracte, although none of the published reports allow the relative abundance of the ware there to be assessed. The long distance movement of vessels is attested from the early-mid 1st c. BC, with perhaps 20 examples of the larger triangular-rimmed jars from the great oppida at Basel. The jars from Camulodunum and Silchester suggest movement towards northern Gaul and Britain during the Augustan-Tiberian period, and there are some hints that this route may have been opened earlier.
FunctionIn addition to the usual questions of source, distribution and chronology we should also address the problem of the function of these jars, in particular an explanation for their long-distance movement. They may have been seen merely as an attractive cooking ware, but the possibility that they served as a container for some other product must also be considered. There is little direct evidence for this at present, although a directed program of residue analysis might identify any prepared foodstuff that they might have carried. The archaeological context of the vessels could also be investigated in more detail, to identify regular patterns of association indicating function or status. Only then can we identify the forces that propelled these jars from the heart of Gaul as far as the British capital of Cunobelin.
• Url: http://www.potsherd.org.uk/papers/cgmw/resume.php • Author: • Last updated: 27 July, 2004 •