A variety of copper alloys, employing variable proportions of the elements tin, zinc, and/or lead, were used during the late Iron Age and Roman periods. These elements also occur naturally in copper ores, but usually only in trace quantities. The various alloys produced by the addition of these elements are: bronze (copper and tin), brass (copper and zinc), gunmetal (bronze with small quantities of zinc) and leaded bronze (Bayley and Butcher 2004, 14). The addition of these different elements affects the qualities of the metal, thus alloys that are high in lead are more fluid, which makes it easier to pour the molten metal into the moulds used in casting. However, the addition of lead also makes the structure of the metal weaker and it will split when hammered. Thus, alloys with a low lead and tin content, which are easier to work by hammering, are better for wrought objects (Bayley 1985, 115; Bayley and Butcher 2004, 13; Dungworth 1996, 402).
Another factor that affects the composition of the alloy is the reuse of scrap metal, a common practice among Roman metalworkers. In a study of Roman brass, Craddock (1978, 12) found that alloys containing less than 4% zinc could result from natural impurities in the ores or the addition of some scrap metal, while alloys containing 22-28% zinc were newly manufactured brasses, and those containing 4-20% zinc resulted from mixing new and scrap metal. In particular, objects with 10-15% zinc were produced when using equal quantities of new and scrap metal, and an alloy of 13% zinc appeared to be the preferred product. It also seems that the scrap metal was carefully sorted by the metalworkers, as the distinct alloys were maintained even after mixing (Bayley and Butcher 2004, 12).
Finally, it should be noted that the untarnished colours of these various alloys are different; thus the possibility that certain alloys were chosen for aesthetic reasons cannot be ruled out. Copper has a pinkish-brown colour, bronze is a brown that becomes paler with higher tin content, and brass is a golden yellow (Bayley and Butcher 2004, 16).
Metallurgical analysis might be used in a number of ways in the study of figurines including dating, determining provenance and identifying workshops. It is known that the alloys used in Iron Age and Roman metalwork changed over time. Dungworth (1996) conducted a study of Iron Age metalwork in northern Britain, and compared his results to those of other studies from the south. He found that in middle Iron Age Britain, all of the objects were in bronze, with no deliberately added zinc or lead. This differs from late pre-Roman Iron Age Britain in the 1st centuries BC/AD when zinc was added and the alloys used included brass and gunmetal.
Brass was being made in the Roman world by the 1st century BC (Craddock 1978, 9), although Dungworth (1996, 410) suggests that early objects in brass were probably imported into Britain and that brass production here does not pre-date the 1st century AD. However, he also notes that the conquest itself did not change the metal compositions in use in Britain, as objects were produced in brass throughout the 1st century, both before and after the conquest (Dungworth 1996, 411). In their studies both Craddock and Dungworth found that there was some difference in the alloys used for different objects. Thus it appeared that zinc was never added to alloys used to make mirrors or statuary (Craddock 1978, 13) and, while simpler versions of both dragonesque brooches and spoons were made in bronze, the later and more ornate versions contained more zinc (Dungworth 1997, 906). In addition, Craddock (1978, 13) suggests that while new brass was produced in the 1st century AD, later objects were generally made of mixed new and recycled metals. Bayley (1990, 138) proposes a link between the military and brass finds, for objects other than coins and brooches, when she found that, at Sheepen, Essex, a high proportion of the mid-1st-century finds in brass were military fittings. This contrasted with Gorhambury, Hertfordshire, where few of the (non-military) finds were in brass.
In determining the types of alloys used, one should also consider access to raw materials. Britain was one of only two provinces (the other being Spain) in the Roman Empire with its own tin supplies (Craddock 1978, 11). Thus it might seem that the production of bronze would be easier, and possibly cheaper, than in other provinces. However, not only was zinc readily available throughout the empire, but it was also cheaper than both tin and copper (Craddock 1978, 11). In addition, lead was relatively cheap, and Bayley (1985, 115) suggests that it may have been added to alloys as 'an economy measure'. Certainly a number of factors may have influenced the type of alloys used to produce figurines, including access to materials and the desired finish.
Studies on the Continent (e.g. Beck et al. 1985; Condamin and Boucher 1973; Riederer 2002; Tykot et al. 2002) have tried to determine whether there are regional differences in the alloys used, and whether this can be used to identify workshops. The studies in France were conducted on a large number of pieces from sites throughout the country. As in Britain, zinc was not a component of early alloys but was used in varying quantities in later pieces (Condamin and Boucher 1973, 164). Beck et al. (1985, 73 and 76) analysed the composition of 465 copper alloy objects, 196 of which were figurines. They found that there was not enough variation in the quantities of tin or zinc to show regional differences, although some regions used more tin in figurines than others, but there was greater variation in the amount of lead used, particularly among the figurines that were often in a leaded bronze (Beck et al. 1985, 77-103).
The study by Beck et al. (1985) also included an analysis of lead isotopes in 43 objects. This showed a regional use of lead sources in which some groups (Narbonne and Aquitaine) utilised lead from a single source, Bavai (Nord) from two sources and, as one might expect, finds from the Mediterranean coast utilised multiple sources. However, the compositional differences do not seem to relate to the availability of tin or lead sources, but, as appears to be the case in Britain, were the result of the desire to produce objects in particular alloys (Beck et al. 1985, 108-9). These studies also showed that specific alloys were not being used for certain figure types but universally for all figures (Condamin and Boucher 1973, 165). Riederer (2002) also found in his study of c. 1000 figurines in German museums that leaded tin bronze was the most common alloy used, and that while there was great variation within this type, it might be possible to identify particular workshops or regions by their choice of alloy.
One fact that does emerge from these studies is that, while it was possible to determine regional trends, the differences between the alloys were not subtle enough to identify individual workshops. Perhaps further studies combining metallurgical and stylistic analysis could help to do this, but the biggest obstacle is the small number of figurines available. It is rare that large groups of figurines of one particular type are found on one site, as is the case at Bavai where it appears that figurines of Jupiter, Mars and Mercury were being produced (Faider-Feytmans 1964). The French studies show that large numbers of figures need to be analysed in order to see any patterns. For instance, in the study by Tykot et al. (2002), which was composed of a small group of 17 figures from the sanctuary at Collado de los Jardines (Jaén), Spain, the compositions showed no patterns. In addition, groups of objects from a specific site may form homogeneous assemblages with a particular composition that is not representative of that found in the region as a whole (Beck et al. 1985, 76).
The need to sample large numbers of figurines meant that metallurgical analysis was beyond the scope of this project. However, in recent years, small groups of figurines such as those from the temple at Uley, Gloucestershire, have been analysed. Fifteen objects from Uley were subjected to X-ray fluorescence. The results showed that a wide range of alloys was used, and that the three Mercury figurines that were analysed would all have been different colours when unpatinated (Bayley 1993, 101). The Uley figurines are quite different from the French pieces, as fewer are in leaded bronze or gunmetal, but almost twice as many contain significant amounts of zinc (Bayley 1993, 102). Two figurines from Gorhambury, Hertfordshire (Eagle wing 1010 and Satyr/Bacchus 1011) were among 125 objects analysed from the site and both are leaded (Bayley 1990). In addition sometimes authors indicate that, due to their heavy weight, figurines may be in a leaded copper alloy. These include: Vulcan 518 from Cirencester, Gloucestershire; Genius 675 from Earith, Cambridgeshire; Fortuna 776 from Motcomb, Dorset; Vulcan 829 from North Bradley, Wiltshire; Genius 862 from Badbury, Wiltshire; Mercury 863 from Durnford, Wiltshire; Horse and rider 880 from Stow cum Quy, Cambridgeshire; Mars 987 from Beighton, Norfolk; Vulcan 1185 from Richborough, Kent and Lar 1186 from Castle Cary, Somerset. However, the small sample size means that it is not possible to establish the reasons for these differences in metallurgical composition, or whether they are seen throughout Britain. The limited scope of sampling has also not addressed the possibility that different parts of the figurines may have been made from alloys of different composition.
© Internet Archaeology/Author(s)
University of York legal statements | Terms and Conditions
| File last updated: Tue Mar 20 2012