Archaeometallurgy of the Early Bronze Age

Andrew T Young 2009

The first metal implements used in Britain and Ireland were probably manufactured from native copper (as deposits in veins and therefore not smelted from ores), whilst the earliest known copper-mining took place at Ross Island (Ireland) c2400 BC. Some of the earliest evidence for the use of metal tools is derived from the timbers of the Corlea Trackway in Ireland where cut-marks from a metal axe have been identified on the timbers of the trackway across a peat-bog (stone and metal axes leave very different cut-marks), dated to between 2268 and 2251 BC – excellent archaeological evidence that they were being used as tools (an important distinction – they were not just prestige objects, despite their rarity at that time). The earliest dated copper axes from the British Isles were found in a cache at Castletown Roche in Ireland dated to c2700 BC and included imported types (from Europe), as well as locally derived examples. The earliest known example of a bronze axe was found at Mount Pleasant near Dorchester and is dated to c2500 BC.

Why is it that copper artefacts do not appear on archaeological sites at an early date in Britain and Ireland? The answer lies partly in the mechanical properties of copper itself. It is a non-ferrous metal which can be reheated, melted down and re-used (when a tool becomes unusable through use-wear or damage for example). Tools that were made from copper were superior in many respects to their equivalents made from flint or stone, could easily be re-sharpened, and were less prone to catastrophic failure in use. However copper (in its native form) was far less abundant than flint and other rocks that were being used to make tools. Therefore copper was not only an extremely effective material to make tools and weapons from, it was also much more difficult to obtain than the rocks used to make similar tools. It is partly because of this rarity that archaeologists attribute copper and early bronze artefacts with high-status.

Metallurgy developed very rapidly in Britain and Ireland after its initial inception – there was an extremely short Chalcolithic (Copper Age) in comparison with other areas. This might be because of the ready availability of tin. Britain had the largest tin deposits in north-west temperate Europe. In fact, tin is entirely absent from many areas of Europe and it seems likely that the very considerable deposits in south-west England (especially Cornwall) were of interest to continental communities once binary alloys (bronzes) were being manufactured. Given the archaeological evidence for long-distance exchange (exotic stone types), it seems probable that tin would also be an important trading commodity. There is ample evidence for prehistoric boats capable of making the crossing to and from continental Europe. Indeed, there are examples of sunken boats in British waters from the Bronze Age laden with metals and or metal objects.

Very quickly after copper implements appeared the first binary alloys of copper (bronzes) began to appear. The Early Bronze Age began late in Britain compared with mainland Europe – we might therefore suggest that as there was already contact between the archipelago and the continent, that the knowledge of how to combine copper and tin to make an alloy (superior to its constituent components) was quickly disseminated. Couple that with the fact that tin was so readily available in Britain and the rapid spread of bronze technology can be easily explained. The components were near to hand, and once the know-how was passed-on, it was taken-up rapidly in light of the mechanical advantages offered by bronze over copper. Binary alloys of copper and tin are far stronger than the constituent metals, are less brittle, more flexible and easier to cast. Furthermore, a relatively high-tin bronze (c10% tin) retains its lustre longer than copper and is a brighter ‘golden’ colour in comparison. These qualities no doubt contributed to make it more attractive than copper at this time.

Historically, later prehistory has been subdivided into three periods – the Neolithic, Bronze Age and Iron Age. However, this macro chronological scale is entirely arbitrary and merely reflects the major technological innovations for each period, not necessarily the incidence or abundance of those materials. In the Early Bronze Age for example, many tools were still made from flint or stone. Metal tools appear to have been the exception initially rather than the rule. Certain classes of stone tools continued in use right into the Late Bronze Age and Early Iron Age. Some stone tools might are associated with metal production (mauls for example). The Early Bronze Age can be sub-divided into a series of technological stages associated with specific advances made in the production of bronze objects. One of the commonest implements manufactured from bronze was the axe. It is therefore relevant to consider the evidence obtainable from axes (and accoutrements involved with their production) as they serve as temporal markers. The bronze axe typology itself is closely related to developments in their form associated with technological innovation. These morphological changes relate to advances made by their makers with regard to efficiency (particularly hafting) and manufacturing techniques.

During the Early Bronze Age, axes were generally made in open one-piece moulds of stone, such as those from Ballyglisheen, Ireland. The use of open one-piece moulds for certain types continued after the introduction of two piece (or lost wax) moulding methods.

The table shown here summarises the date(s) and form(s) of axes.

Table 1 Types of early flat axes and their dating (After Needham et al 1989, 386)

These differences are easily discernible macroscopically and represent attempts by their makers to create more efficient tools using a minimum of raw material and improving hafting (which evidently was problematic). It should be noted that decorative elements (stamped) were present in some examples throughout the period (whilst most were undecorated) and merely an aesthetic refinement on the part of individual manufacturers (or perhaps at the behest of their recipients). Types 1-3G and 5 could be made in open one-piece stone moulds. Flanges were created by cold-working with hammers. Types 4A-4E could be made in this type of mould, though examples with very pronounced stop-ridges may have been cast in two-piece moulds or by the lost-wax method. Statistical analysis of the composition of axes demonstrates that types 1-2 were composed of copper, and that during MA III (c.2200-2000BC), a binary alloy with tin was introduced in ratios varying between 8-14%, and (generally acknowledged as the most beneficial alloy ratio). What is particularly interesting is that comparative analyses show that in Britain and Ireland the consistent levels of tin in alloys is at odds with bronzes in Europe and the near East, where there was a much greater fluctuation in the percentage of tin used in bronzes (within periods and types). This indicates that tin was less widely available in those areas and more readily available in Britain and Ireland. This can only be attributed to the fact that tin was a native ore in Britain but had to be imported in north-west temperate Europe and the near East, leading to the expedient use of more or less tin depending upon local availability, which appears at times to have been sporadic.

In general therefore, the mechanical properties of British and Irish bronze tools at this time were significantly superior to continental equivalents pro-rata. There is sufficient anecdotal evidence therefore to suggest that the export of tin from Britain was an important exchange mechanism at this time. McKerrell (1978) has suggested on the basis of this and other compositional analyses of early bronzes, that Britain was an important supplier of tin to Europe and the Mediterranean. Needham (1989, 393) suggests that in Britain a metalworking tradition evolved which, benefitting from the unique availability of tin, insisted on the most functionally advantageous ratios of tin in binary alloys used for tool and weapon production, a luxury not afforded elsewhere as a corollary of the sporadic supply of tin as a raw material.

Compositional analysis has identified a number of recurring impurities in British and Irish bronzes. These can be summarised as being (proportionally): - antimony, arsenic, bismuth, cobalt, iron, lead, nickel, silver and zinc (Needham 1993, 394). In general the quantities of these impurities is such that they can be considered as unintentional and probably present due to being resident in the ores used to smelt copper from as they confer no mechanically advantageous properties in the quantities present, with a few notable exceptions. One such example is the axe from Lydd which has a 4% lead content (advantageous for the casting process), but may in itself represent an axe made from metal imported from the continent (Needham 1989, 397). A small group of type 4 axes (the Brithdir hoard) have a high arsenic content (over 1.5%), and it has been shown that a source in south-west Wales must have been the source – the high arsenic content in copper ore from that location is well recorded (ibid).

Flat axes were cast in matrices carved into portable stone moulds, usually made from a rock type that was relatively soft (sandstone, limestone and serpentine). Sometimes the moulds for more than one tool were carved into the same piece of stone (see illustration below).

Figure 1 a multiple matrix stone mould for early flat axes (After Tylecote 1986, 108)

Stone moulds have the benefit that that can be re-used and are fairly robust. Clay moulds were used only once and are destroyed when opened and do not survive well (hence the evidence for clay moulds of all types is very spartan). The ceramic accoutrements used in early metal casting were complex – they are not simply clay. The moulds were often comprised of layers of several different mixes of clay, clay with sand, or clay with finely ground charcoal. A single mould may comprise three or more different layers. Compositional analysis shows that the layer immediately surrounding wax-models was often composed of fine clay combined with ground charcoal. A layer of clay mixed with sand was often wrapped around this inner core. The addition of charcoal prevents cracking when the moulds dry out and are fired prior to use. Drying and firing moulds prevents catastrophic failure during casting, when the mould is exposed to a tremendous temperature changes. They are heated prior to casting to mitigate thermal shock. Experiments show that moulds which are not fired prior to casting often crack and break. As metal was so valuable during the earlier period, smiths could not afford to waste time, energy or material and were probably very careful to prepare all their accoutrements properly.

Clay crucibles, like moulds, were carefully manufactured complex ceramics. Like moulds, they are often comprised of multiple layers of different materials. Some of the best examples of early crucibles have been found in Devon. Like moulds, they do not survive well. Used crucibles, having been repeatedly subjected to extreme thermal shock are very friable and are comparatively rare. Despite the abundance of evidence for metalworking in this period – there are very few intact crucibles in the archaeological record (Tylecote 1962, 130 and see Concepción et al 1993, 33; Crew et al 2001, 2); an ‘incense cup’ from a Bronze Age level at Jarlshof in Shetland might represent a small casting crucible. Otherwise they appear to have been generally triangular in shape, and to have survived in this form through to the Early Iron Age (ibid). The illustration here shows some of the best preserved BA crucibles known.

Figure 2 Two small triangular crucibles (left) c10-15cm diameter, and a circular crucible c20cm diameter (right) with three legs which may have assisted with tipping for pouring molten metal, all from Devon (After Tylecote 1986, 130; 136).

The contents of small crucibles of the size and nature noted for the BA cool extremely rapidly by radiating heat, pouring of the molten metal must have taken just seconds (Tylecote 1962, 136). Due to their small capacity – in the region 40-50cc (Tylecote 1986, 137), “it is likely that only enough metal for one object at a time was melted in these crucibles” (ibid and see Tylecote 1986, 136).

Early crucibles appear to have been generally very simple, and made from clay, like the small triangular crucibles from Aghaloghan, Co. Antrim, which were made from very fine light grey clay. Otherwise they seem to have been made from a coarse sand-clay mix (Harding 2000, 222). Cylindrical crucibles from Lagore were more complex – they were two-layered with a thick coarse outer layer and a thin inner layer of very fine clay. Similar crucibles are known from Buston, Ayrshire (Tylecote 1962, 140 and see Tylecote 1986, 118). The thermochemical qualities of copper are such that when melted in crucible metallic elements may be volatised and absorbed by the crucible, allowing us to determine the use to which prehistoric crucibles were put (Dungworth 2000, 84 and see Bayley 1995, 1). Native coppers which did not require smelting may have been used in the EBA and compositional analysis is able to determine this (Coghlan 1951, 92). Usually they survive as just fragments, and this in part might be caused by the extreme temperature changes they were subjected to which may well have taphonomic implications (Concepción et al 1993, 33). Ceramic tuyères are the only evidence for bellows and blowpipes, mostly of a small conical type, or a cylindrical variety such as that from Ewanrigg (Harding 2000, 220-221), though there are examples of EBA wooden pipes from Scandinavia, where wetland sites better preserve organic materials.

The accoutrements used in early metal casting do not survive well and it was only relatively recently that furnaces have been recognised. It is possible that many archaeological features described as hearths in the literature were in fact furnaces. My experiments have shown that bowl-type furnaces break-down very rapidly after use and survive only as heated clay fragments associated with carbonised wood. Recent research has shown that some features previously described as hearths were in fact furnaces (minute copper prills have been found in curated samples). Early furnaces were of the bowl type and we have so little evidence for bellows their form is entirely conjectural.