Evidence for prehistoric mining activity has now been reported in many areas throughout the world , particularly in Europe, the Near East, China and parts of North and South America. This study does not propose to list every site, but rather to report on those sites which have been described in some detail, particularly in Europe. Summaries will be included relating primarily to those sites exploiting copper-bearing ores within a period from the Chalcolithic (Copper Age) to the Late Bronze Age, with reference, as appropriate, to activity during the Roman period. Particular reference will be made where possible to geological and mineralogical conditions of the ore deposit, also to evidence for ore processing and smelting technologies.
Evidence for the first use of metals in Europe originates from Bulgaria (Coles & Harding 1979), where excavation of a ‘tell’ (artificial mound) at Karanova revealed occupational evidence beginning about 8000BP and continuing for at least 4000 years. Tells are particularly informative as it is possible to discern clear advancement in pottery design and manufacture through successive layers that constitute the mound. The tell at Karanova produced a small awl, consisting of a copper pin set in a bone handle, estimated to date on ceramic evidence from 6500BP. It is suggested that one possible source for the metal could be the site of Aibunar where native copper and associated ore are reported to have been mined from at least 6000BP (Cernyck 1978). A similar awl is recorded from Balomir in Romania, dating to the late Neolithic to early Vinca period (4800-4500BC) of this area (Jovanovic 1976). Near Eastern sites such as Cayönü and Ali Kosh are also considered to date from, and even predate this period (Tylecote 1986).
It is suggested that native copper together with gold and silver was one of the earliest metals to have been fashioned into usable objects such as simple tools and ceremonial objects (Tylecote 1986). These native metals could be found either in alluvial forms or associated with other ores occurring in weathered gossans or exposed oxidised sections of the mineral deposit. Small amounts of native copper have been recorded from some of the Balkan mines (Jovanovic 1976). Typically any ores of this type are likely to have been the first to be removed and therefore the chances of identifying such ores at known early mine sites is greatly reduced. Copper ores which occur in the near surface zone are typically brightly coloured and formed by the oxidation of primary sulphide ores such as chalcopyrite and bornite to red copper oxides - cuprite and the copper carbonates - green malachite and blue azurite. These primary sulphides may also undergo enrichment below the oxidised zone to produce additional sulphide ores such as chalcocite and covellite. Native copper typically occurs at the junction of the oxidised and enriched zones which corresponds to the level of the water table. Other copper ores known as the ‘grey coppers’ or ‘fahlerz’ ores such as tetrahedrite and enargite may also occur as primary or replacement deposits with any of those above.
Native copper is assumed to have been shaped initially by cold hammering, closely followed by the realisation that heated metal (at 200-2500 C) could be worked (forged) more easily into different shapes. Recorded items include: awls, ornamental discs of sheet bronze, hammered flat and perforated for attachment to clothing, tanged daggers, spectacle-spiral pendants, spiral beads and simple flat axes (Tylecote 1986). Large numbers of these artefacts were associated with burials from this early period spanning the Neolithic to Copper Ages. Co-lateral advancements in pottery manufacture observed during the same period suggest that air- draughted kilns/furnaces were capable of producing the required temperatures to melt copper (11000C). The technology to melt native copper and cast it into simple open moulds may therefore have been a simpler process than otherwise expected (Tylecote 1986). As the demand for copper increased, prospection and gathering of ore from surface exposures may not have been sufficient to supply that which was required. This may have led to the first true mining of material from the weathered and rotted parts of the exposed mineral deposit. These earliest workings are likely to have been nothing more than shallow pits, excavated with simple tools of bone, wood and stone.
Tylecote (1986) speculates that during the Neolithic Period the vividly coloured carbonate copper ores malachite (green) and azurite (blue) were already being used for jewellery and possible pottery glazes (Tylecote 1986). It may therefore have been a simple step through experimentation at furnace temperatures and reducing conditions, to sufficiently heat these ores to produce smelted metal . The poor distribution of native copper and lack of scrap metal for reprocessing would also have been a major contributory factor in the developing technology to produce the much needed material. The realisation that these coloured ores could be smelted also led to the need for greater quantities of material to be removed and new sources to be located. These requirements may therefore have marked the transition from simple surface collection of ore to the development of mining proper.
Tylecote (1986) proposes that the early copper smiths may have been revered, and were allowed many privileges by the community, as they possessed the unique ability to turn coloured stone into a useful and visually attractive material - metal. These advances during the Chalcolithic or Copper Age have been considered to progress from east to west across Europe (Tylecote 1986), a view supported by radiocarbon dates centred around 4000BC at Aibunar - Bulgaria, 3800BC at Rudna Glava - Serbia, 3000BC in S.W. Spain and 2500BC at Mitterberg - Austria, also reaching Ireland about 2500BC. However much of this interpretation is based on the typology of ceramic remains and tool artefacts, typically stone hammers, found at the sites. The quoted C14 dates are often infrequent and may represent only one phase of activity at the sites, with other phases not yet realised or simply destroyed by subsequent activity (Craddock 1992). The increasing use of copper during this early period is likely to have brought many social, economic and technological changes, leading to further exploitation of the ores through advancements in mining methods, for example the use and perfection of firesetting for the removal of hard rock.
It is accepted (Tylecote 1986) that during the earliest period of metallurgy the most likely means of improving the hardness of the cutting edge of cast copper axes was by cold hammering. Artefactual evidence indicates that these copper axes soon began to be replaced by implements containing small quantities of arsenic that had the effect of improving the durability of the cutting edge and overall hardness of the metal (Tylecote 1986). The source of this arsenic is presently under debate, formerly it was considered to have been a deliberate addition, developed through experimentation to produce the optimum percentage that would have the greatest affect on the hardness of the metal, with typical values around 1-5% (Charles 1967). However Northover (1989) demonstrated that Irish axes possess an arsenic content that is not related to the hardness of either cutting edges or complete artefact, suggesting that prehistoric smiths were unaware of any improvements due to alloying. Other researchers (Budd et al. 1992, Budd 1992) argue that the presence of arsenic is simply a product of the type of ores being exploited and smelting technique. Experiments have indicated it is very difficult to obtain concentrations of arsenic in copper to match those of known artefacts, from the smelting of copper sulphide and arsenical ores from the oxidised zone of mineralisation. Their results do, however, indicate that appropriate compositions can be obtained through smelting of arsenical copper ores known as the ‘grey coppers’ or ‘fahlerz’ from the enriched zone of mineralisation, including tetrahedrite, famatinite, bournonite, enargite and tennanite.
In eastern Europe it is suggested ( Coles & Harding 1979) that the smelting of local arsenical ores gave rise to increasing availability of tools and particularly weapons, such as the battle axe. They suggest that ores from the Balkan region and technologies of the Near East met in the Caucasus, providing copper smiths with skills to produce some of the finest weapons of the period, inferring that there must have been active trading between the conveniently sited ores and technological skills of these two regions. Towards the end of the Copper or Chalcolithic period the addition of tin to copper to produce a greatly improved alloy, bronze, was first practised. Possible known sources of the tin include Cornwall, Afghanistan and Turkey, with recent research (Yener & Vandiver1993) in the Anatolia region of Turkey proposing the beginning of both tin mining and smelting around 2870BC.
In the British Isles the introduction of tin bronze was rapid during the end of the third millennium BC (Craddock 1992), while in the Near East it was gradual occurring between 6000-3000BC, this could imply that the manufacture of tin bronzes in western Europe occurred independently to those of the Near East. Likely sources of tin were Cornwall and the Wicklow mountains of Ireland, although no evidence for tin related mining at any location in the British Isles of this period is known. Alluvial or placer tin deposits are generally considered to have been the extracted forms, as indicated by a number of artefacts ascribed to the Bronze Age encountered in some of the placer tin workings last century (Penhallurick 1986).
Towards the later half of the Middle Bronze Age to the Late Bronze Age it is widely accepted that many advancements, such as two part clay moulds and greater exploitation of available ores, meant that production was more efficient (Tylecote 1986). This is also borne out by the many hoards attributable to this period. Two types of hoards are known, one having being deposited into lakes or ponds as part of a ritual offering, the other as founders hoards of scrap metal, cut or bent so enabling ease of transport and remelting. The numbers of objects constituting the hoards were far greater than any hoards of the earlier periods, sometimes exceeding them by a factor of 10. These increased numbers are interpreted as indicating that the metal was less valuable than previous and that the availability of transport had improved (Tylecote 1986). International trade may also have been a product of the increased output, with a range of items on offer, including cauldrons, buckets, general tools and ceremonial objects. Since, mainly weapons and smith’s tools and only occasional ornaments (eg. Migdale, Sutherland. Harbison 1968) were previously produced, it may be that these articles indicate that there were no longer shortages of metal in the Late Bronze Age.
There are numerous sites throughout the world where evidence for early mining dated to the Copper and Bronze Age periods are recognised. The following sections of this study will detail some of those sites across Europe (figure 1) and western Asia. Where possible an emphasis on those details that are the subject of this study will be made. In principal these will include geological/mineralogical conditions of the ore deposit and their relationship with the morphology of the mine workings. The distribution of artefacts/features and their value as indicators of early mining techniques will also be considered and reference made to any evidence for ore processing, smelting, metal distribution and wider perspectives to the surrounding
landscape. Sites will be described across Europe and to the Near-Middle East.
2.2 Prehistoric Mining in Europe and the Near-Middle East
Early mining activity in this region has been identified within the Huelva Province approximately 80km north of Cadiz. There are numerous sites here, notably Rio Tinto and Chinflon. Some were investigated during the late 1920's (Davies 1935) before they were destroyed by recent mining. Davies reported on the extensive nature of the workings, he interpreted some as being of a pre-Roman age, characterised by galleries with radiating tunnels that exploited a seam of ores contained within a body of massive pyrite capped by a thick gossan. These workings contrasted markedly with those of a Roman age where shaft and adit mining systems were the practised method.
Limited archaeological investigation occurred during the late 1960's when an excavation at Cerro Salomon identified pottery, slag, crucible and tuyere fragments (Blanco & Luzon 1969, Blanco et al 1970). A pre-Roman age was ascribed, considered to represent a Phoenician occupation of the region.
During the mid 1970's renewed investigations culminated in a comprehensive mining and metallurgical study of the region, lead by Rothenberg and Blanco (1981). Extensive field work studies followed by laboratory investigation identified a range of mining and smelting activity from Copper Age 1 (Chalcolithic) to the Roman period. In total, about 40 sites were investigated, revealing evidence in the form of stone hammers, pottery and often extensive deposits of smelting slags with associated furnace remains distributed over a wide area. The proliferation of slag and proximity of smelting to the mine sites is perhaps the most remarkable feature of this area.
A complete range of mining activity was recognised, including shafts, galleries, tools and smelting areas. This was dated by characteristic pottery remains and a range of stone hammers, that were considered by Rothenberg to be indicative of early mining. Generally there seems to have been a lack of material for radio-carbon dating. The proposed earliest exploitation of suitable ores falls within Copper Age 1, producing malachite from surface and trench workings, which were largely confined to sequences of tuffs and quartzitic rocks. Activity in Copper Age 2 is also proposed, once more partly on cultural evidence from other studies in the province and partly on the composition of copper-bronze implements from this period. It is argued that these mines were producing appropriate ores with noticeable arsenic content to account for the composition of metal implements during this period (Rothenberg & Blanco 1981). This implies that mining had progressed below the zone of oxidation to where sulphide ores of a suitable composition occur.
At present there is no evidence for any Early-Middle Bronze Age mining or metal production. However there is evidence that a revival in activity occurred during the Late Bronze Age. During this period Rothenberg and Blanco (1981) assumed there was an improvement in extractive techniques and smelting. This was interpreted as resulting from the exploitation of ores from the oxidised zone of mineralisation, generally by trench working, perhaps linked to limited underground workings. Once more, evidence for this later period of activity is speculative, being represented by cultural remains from surrounding districts, rather than direct evidence from the mine area. At Rio Tinto some slag from the smelting of silver ores is reported to be dated to the Late Bronze Age by the occurrence of pottery shards. This appears to be unusual as no copper slags are known, even though mining techniques may have been adequate to work the high grade ores from the harder enriched zone of mineralisation. If silver ores were smelted, then it does indicate a degree of sophistication not previously known, involving the use of fluxing with silica and cupellation via the use of lead as a collector for the silver.
Towards the beginning of the first millennium BC, Phoenician influences are noted, being represented by pottery mixed with local coarse wares. During this period obvious improvements in mining technique derive from the use of iron tools, with mining of copper and silver ores from the enriched zone, including chalcopyrite, covellite together with distinct but minor amounts of gold.
Mining continues through to the Roman period which is the main phase of ore production over the Huelva district. Further improvements in mining produced the shafts, adits and gallery systems that characterise some of the workings seen today. These operations were on a scale previously unknown, many millions of tons of ore and associated silver-copper slags revealing a very organised and developed method of working, which is thought to have required a large labour force with specialist engineers and metallurgists.
A preliminary survey conducted by Barge (1985) in the Montagne Noir and Monts de Lacaune districts of southern France identified twenty three potential early mine sites. All displayed evidence for Gallo-Roman workings, though fewer for prehistoric activity. Of the latter at least two districts now have definite evidence for Bronze Age mining. One is related to a series of natural avens (swallow holes) at Bouche Payrol, near the town of Brusque about 30km west of Lodeve. The other is just north of the village of Cabričres, about 20km south of Lodeve. Mention is also made of the district of Seronais situated between St. Girons and Foix in the south, where a Gallo-Roman site exhibits potential evidence of early mineral exploitation.
Bouche Payrol
This site is certainly one of the more curious if not unique mines exploiting copper in early times. A mineral deposit has been worked through and into an area linking at least four natural avens within the limestone country rock. Workings extend to nearly 100m in depth where there is a connecting system of galleries, shafts and inclines, all excavated by fire-setting or metal pick; together with partly in situ timber bridges, fittings, stairs and handholds (Barge 1985). The deepest workings are considered to belong to the Roman period.
Evidence for prehistoric working is confined to the first 10m below surface, forming an upper system of interconnecting narrow passages, galleries, pillars, shaftways and links to surface. No pick marks are recognised in these areas. At least two hammerstones, one with a pecked groove, are recorded from a point at 8m depth, while surface spoil tips have produced nine hammerstones of quartzite and one of basalt, together with a piece of cordoned-ware pottery of probable Mid-Late Bronze Age age (Constantini 1980).
A number of field monuments in the immediate area of the mine have also been investigated, with the view of detecting any possible associations between them and the mine. Cave sites, burials and dolmens were also considered, many having previously produced artefacts originating in the Bronze Age. It is hoped that continued study of these possible associations, and analyses of collected ores, will demonstrate that the mine may well have been supplying the necessary copper for the production of local metal goods.
Cabričres
Several prehistoric mining works have recently been described at Pioch Farrus and Vallarade in the district of Cabričres (Ambert 1990), many having been disturbed by Gallo-Roman or modern exploitation. A proportion of the workings do, however, still contain evidence of early activity, including pottery, stone mining tools of quartzite, scrapings of malachite and grey copper ore (fahlerz) and fragments of large pottery vessels ascribed to the Early Bronze Age. A series of workings are known, penetrating dolomitic host rock to form small pits up to 5m deep and some larger galleries which are now flooded.
To date there are no reported traces of firesetting; instead ore material appears to have been worked using simple quartz picks, stone wedges and hammerstones. Some of the hammerstones display above average wear, and these are interpreted as having been used for ore crushing as opposed to direct hammering of a worked rock face. A smaller number of cup marked stones occur within surface spoil, none are recorded underground, this suggests a use exclusively for ore treatment by crushing.
Approximately 400m from the mine, at Roque Fenčstre, an ore processing site was identified. Here a number of treatment pits were cut into the shistose country rock. Closely associated with these were a range of stone implements - hammers, grinders and mortars, further pottery fragments, charcoal, shovels made from shoulder blades of oxen and slaggy deposits with globules or prills of copper metal. Collectively this arrangement has been interpreted as a processing site where ores were crushed, sorted, washed and eventually grilled or roasted as a precursor to smelting.
On ceramic evidence a date in the Late Bronze Age is proposed for the workings and processing site, so an initial date in the Chalcolithic period is possible. A single radio-carbon date from Roque Fenestre partly vindicates this, falling within the Early Bronze Age at 1950+/-130BC (Ly-2869).
Seronais
This group of mines is considered to date from Roman times, although earlier workings known locally for some time are thought to occur (Dubois & Guilbaut 1980).
The country rock, mainly dolomite, is cross-cut by a system of hydrothermal veins containing grey copper and some galena, with limited oxidation to produce copper carbonates. Generally the workings, about twenty in number, are confined to mineralised areas, exploiting the ore in a series of surface workings, small opencasts and galleries with supporting pillars left in place. Indications of fire-setting and pickwork are recorded. Many of the workings are infilled with debris, either purposely and/or through weathering and sedimentation over time. No subsequent mining activity appears to have occurred.
Limited excavation of the site has revealed a considerable number of amphorae (type A1), fragments of orange, reddish and brown clay, and also a lamp of a type widely used in the Mediterranean, and small quantities of copper slag. On this evidence an occupation of the site during Roman times is assumed, with activity beginning during second century BC and continuing to first century AD. However, earlier activity, may have also occurred, considering the suitability of the ore and the friable, softened nature of the country rock (Dubois & Guilbaut 1980).
Mining activity from an early period was initially recognised in the Mitterberg, a region south of Salzberg (Pittioni 1951). Here one main site and a number of
related mines were investigated. More recent and continuing research through small scale excavation and exploration has confirmed and elaborated upon the original findings (Eibner 1974, Gale & Ottaway 1990). A number of mine sites have been described, together with associated ore processing and smelting areas.
The ore body occurs as a series of veins through shales, slates and phyllites, comprising chalcopyrite-pyrite-quartz with oxidation to malachite and azurite. The veins, being harder than the country rock, tend to form prominent features, often standing out on ridges and summits. These obvious surface exposures were probably the first to have been exploited by the early miners, initially as surface trenches known locally as "pingen", then progressing below surface to form the inclined and often extensive, galleries described by Pittioni. He infers that many of these workings were driven by the technique of fire-setting, using timber supplied locally from the extensive forests that still cover this region. Renewed mining activity during the 19th and 20th centuries has destroyed many of the earlier "old man workings". Fortunately, however, descriptions of these discoveries were made and have been useful, together with more recent surveys and excavation in providing an insight and corresponding impetus for future study.
A substantial quantity of artefacts is now recorded from the various workings and associated processing-smelting sites. They include a range of stone implements (pounders, hammers and anvils) corresponding to particular modes of utilisation (Gale & Ottaway 1990), also a collection of wooden items - shovels, knives, wedges, supports and lighting sticks. A bronze hammer and remains of a leather jacket are also recorded. Two radio-carbon dates from the "old man workings" indicate an Early-Middle Bronze Age period, confirming previous opinions based on artefactual evidence.
During the late 1950's to early 1970's several studies in Serbia and north-east Bosnia described a number of copper artefacts, including ornaments, implements, weapons and beads, and there was also limited evidence of smelting (Jovanovic & Ottaway 1976, Jovanovic 1979). They are all considered to belong to the late Vinca period, contemporary with the Eneolithic period, indicating that copper was used in all the regions of the Vinca culture. Apart from some limited evidence for mining in the Bor region of eastern Serbia, no other early evidence was available until 1968. Preliminary investigations carried out at Rudna Glava then identified mining activity considered to be Eneolithic (Jovanovic 1979, 1980).
Rudna Glava
This site was discovered during modern exploitation of an open cast iron mine, revealing a row of vertical shafts partly damaged by modern working. Ore veins penetrating the limestone country rock had been worked to a depth of 25m, exploiting rich concentrations of carbonate ores resulting from the oxidation of primary chalcopyrite which was accompanied by some magnetite. Early mining remained within the confines of the mineral veins of 0.5-2.5m width, then branched off at depth to link with adjacent shaftways. Later activity of possible Roman age is thought to have exploited the harder portions of the ore deposit, producing large oblique galleries, some still containing timbers, thought to have been supports.
Finds from the shafts come from three main zones: access platforms, channels cut from the shafts, and bases of the shafts. The majority of the objects were found at the level of the access platforms. Stone, bone, antler, wooden tools and also pottery had been inserted into cracks and fissures, probably weathered out joints in the limestone. Four pottery vessels found in this situation were remarkably complete and rate as the more important finds from the workings. The first find was a conical jug and a fine ware amphora jug; the second two were intact amphorae with handles and medium burnished surfaces. According to their shape, manufacture and decoration the vessels can be dated to the late Vinca period.
Around 100 stone mining hammers were recorded, generally composed of gabbro, often with wear indicative of heavy working. Some were modified with shallow grooving around the middle to facilitate hafting. Three distinct groups were identified : cylindrical, rectangular and triangular, suggesting various functions with specialisation established through working experience. No evidence for the use of copper tools was found such as distinct groovings on the vein walls, only smooth surfaces being present.
Aibunar
This site perhaps more than any other bears direct comparison with the workings at the Great Orme, as both sites share similar geological and mineralogical conditions. These factors are likely to be important in the determination of the shape, size and extent of the workings.
Aibunar is situated in southern Bulgaria, 8km north west of Stara Zagora. Here at least twelve mine sites from the Eneolithic period (Cernych 1978) have worked an ore body of mainly malachite with some azurite associated with a body of polymetallic ore. Veins are generally parallel to one another, varying in width from 0.5-5.0m, sometimes several combining to form a layer or lens up to 10-15m in breadth. The country rocks are interbedded limestone, marls and dolomites.
Most workings consist of surface excavations, their form determined by the shape and width of the vein or mineral outcrop. They progressively deepen to form trench-like features with dividing ridges of the limestone country rock and also wedge-shaped opencasts up to 80m long. They extend laterally at depth forming true underground workings, but of limited extent. Some are accessed by vertical shafts at least 20m in depth.
A number of artefacts have been recorded from excavated spoil at surface and underground, including pottery, antler tools and metal implements. Some of the pottery is dated to the Eneolithic period (equivalent to Copper Age 1, 4000BC), while other material is dated to the Late Bronze Age. No material from an intervening period is known, although future work may alter this situation. Many fragments of antler were located within the mine spoil, these are considered to be the damaged sections of simple picks. One of these displayed remnants of a long sleeve thought to have aided the securing of a copper chisel of a type that is frequent among Eneolithic metal finds of the Balkan Carpathian area.
One of the more unusual features in comparison to other early mining sites is the total absence of hammerstones from any underground context. There are, however, a number of stone implements on the surface dumps, one of which is modified with shallow grooving to allow for hafting. Cernych (1978) concludes that hammerstones were utilised exclusively for crushing during ore processing, and that metal tools like the two found served as general purpose implements for rock breakout and processing alike. One of these is a hammer axe, the other an axe adze. Both are of copper and conform to types considered to belong to the Eneolithic period.
The entire surface mine site was studied, with excavation of selected locations. One of these of limited extent (site 4b), revealed remains of a Late Bronze Age hut with an underlying horizon where two human burials were located. The few scattered shards of early pottery and notable absence of Late Bronze Age pottery lead to the conclusion of an Eneolithic date for the burials.
Collective evidence at Aibunar indicates an Eneolithic mining operation, with extensive workings open to surface, associated dwellings and burials. As yet no smelting remains from the immediate neighbourhood are known. A comprehensive geochemical study supported this observation, concluding that there were no obvious associations between the trace element content of ore deposits and copper implements found locally (Cernych 1978). Comparable investigations in other mining districts, for example north Wales (Jenkins 1986) and at Rudna Glava (Pernicka et al 1993) arrive at similar conclusions. Cernych (1978) proposed the idea that the carbonate ores may have been finely crushed for use as pigments rather than for copper smelting. This possibility is suggested by a single ceramic mortar containing traces of powdered azurite. Similar evidence is also known from Selevac, a Neolithic village in Yugoslavia, where malachite powder, fragments of copper objects and slag were found (Tringham & Krstic 1990).
Numerous metallurgical and mining sites probably exist throughout this region, partly due to its position on the route linking the Mediterranean with the ancient civilisations of Mesopotamia and the Indus valley (Willies 1990). Of a number of sites now being studied, one in particular is providing evidence for the first reported tin mine of the Bronze Age period. This is presently under investigation, and is situated near Camardi on the southern edge of the Anatolian plain (Yener & Vandiver1993).
A number of mines with several entrances have been identified, located on a hill called Kestel. The country rocks are limestones and dolomite overlain by conglomerate. They have been metamorphosed to harder marbles by a gabbro and granite intrusion, which is also probably responsible for the mineralisation. Considerable metasomatism has occurred, replacing the marble with haematite and manganese oxides, in places forming deposits up to 3m across. Quantities of cassiterite are found at the boundary between the marble and hematite. Subsequent oxidation of the haematite has produced limonitic deposits, seen infilling joints widened by solution, some 0.3m wide.
The mines comprise surface workings and backfilled opencasts, giving way to entrances, some several metres wide, linking to cavernous workings with supporting pillars. The rounding of the passage surfaces is considered to be characteristic of firesetting, which is also indicated by the presence of charcoal, which has provided a C14 date of 2400-1600BC. This corroborates an Early Bronze Age date ascribed to pottery shards also found in the workings.
A survey of the immediate surface area around the mines produced large numbers of stone implements, estimated at around 25,000. Composed of gabbro from the nearby intrusion, these tools were categorised typologically into three main types. The first, a hammerstone or maul of typical cobble form with obvious wear to each end, of which a small number were pecked or modified to allow for hafting. The second type displayed wear indicative of ore processing, one variety suggesting use as an anvil for rough crushing with a centrally worn concavity or dimple. The third type suggests a use for finer crushing or grinding through to rubbing, not unlike a grind stone. In addition to these implements a number of hollows, worn into rock exposures adjacent to the mines, were also noted which are also thought to have been ore crushing areas.
Yener and Vandiver (1993) propose that the various stone implements were used to crush the fine grade disseminated tin ores from a coarse to fine grade material. The heavier ore particles would then have to be separated from the lighter marble particles. It is suggested that this was achieved by a process of winnowing, where the crushed material was tossed into the air in a light wind, with the more dense cassiterite falling quickly, while the lighter marble was carried away, falling down wind of the tin ore. If this form of separation is correct, then it implies that processing technologies during this early period had advanced beyond the process of selection by hand of the visually distinct ore, such a simple technique not having been suitable for the separation of these fine grade ores. The lack of available water on or at the site is also likely to have prompted the development of ore processing techniques utilising air separation.
A great deal of evidence for mining from prehistoric and later times exists from this region. Sites are identified from the central Negev mountains of Israel, the Sinai desert of Egypt, Feinan on Jordan’s desert fringes (Hauptman 1989), northern Oman and Syria. At many of these places evidence of actual mining from the earliest periods is minimal or absent, and is considered most likely to have been destroyed by subsequent activity during the Iron Age and more recent centuries. There are, however, remarkable remains of mineral processing, smelting, habitation sites and trading. It is not the purpose of this study to detail every site, but rather to give an overall impression of the forms of mining, so contrasting them with western European sites. For this purpose two locations will be described, those at Timna, Israel, and in northern Oman.
Timna
Timna is probably the best known mining district in the Near East, traditionally being considered synonymous with the site of King Solomon’s mines. Investigations began here in the 1950’s and have continued practically to the present day (Rothenberg 1972, Hauptman 1989).
Mining is considered to have been on a large scale, beginning during the Chalcolithic in the fourth millennium BC and continuing through to Early Bronze Age (EBA) I & II, with activity reaching a peak in Late Bronze Age to Early Iron Age times. No evidence as yet exists for mining activity between these two phases, implying a period of inactivity through the Middle Bronze Age. Overall a picture of widespread activity related both directly and indirectly to copper production is presented, reaching levels beyond those initially assumed.
The earliest evidence takes the form of flint implements and stone tools associated with shallow rough pit workings, pottery remains giving a tentative Chalcolithic date. One of the identified sites (No 42) consisted of a round opening 1m deep with a surrounding deep groove similar to the ‘Schramm’ around the openings of ancient salt mining sites at Hallstatt in Austria, no explanation as to the purpose of the groove being given. Fragmented nodules of chalcocite with malachite and some chrysocolla are considered to have been the main ores exploited by this first stage of mining by irregular trenching and pitting.
In the following period, mining (EBA I) is considered to have pursued the ore to greater depths, via a shaft and gallery system, utilising stone tools, some modified with a central hole for hafting. Dating once more has been based on pottery, and also the discovery of a stone lamp characteristic of this period. No mining has been identified at Timna dating to EBA II. Instead it is represented by evidence for metallurgy to the north of the district, where concentrations of mortars, pestles, slags, furnace remains, habitation sites, burial and ceramics occur. All of this EBA II activity is particularly interesting as it is contemporary with the Early Dynastic period of Egypt , representing a time when many new settlement sites developed. Such increases in population have been taken to imply (Rothenberg 1972) that occupations are likely to relate to the widespread copper industry of the time, bringing with it a degree of trade previously not experienced, and evidenced by imported pottery types now excavated at many of these sites.
Northern Oman
A number of mining sites for copper are known from this region, with evidence for Bronze Age, Iron Age, medieval and more recent activity (Weisgerber 1987). Some sites are being worked at present, largely by the Oman Mining Co., which has assisted in providing resources towards the study of the earlier mines. Mineralisation takes the form of massive lens shaped ore bodies of cupriferous pyrite, which readily weather producing surface gossans, and acidic groundwaters that are responsible for the preservation of old timbers, but the deterioration of any iron or bone tools that may have existed from an early period (Weisgerber 1987).
The earliest signs of copper production from this region are known from ‘Arja, al-Sayab and al-Wasit where evidence for mining from the third to second millennium BC is recorded, while at Zahra there is evidence for third millennium smelting activity. Some of the other known sites indicate there is likelihood that similar ores were worked around 1000BC, with the main activity dating to 800-1000AD, upon which the majority of the recent studies have been concentrated. Identified processes include ore crushing, roasting and smelting.
At ‘Arja, where mining is continuing at present, a large number of old workings have been intersected by modern tunnels. Many have been judged to belong to a period circa 900AD. At one location though, exploratory drilling through what appeared to be a gossan cap, penetrated about 30m of eroded fill (spoil) in an early open cast considered to date from the first millennium BC. Weisgerber (1987) proposes that activity from this period was on a larger scale than that previously thought, and concluded that mining would have had to have been fairly organised in comparison to smaller scale mines which are likely to have been less organised. Limited evidence for smelting also exists, comprising mortars for crushing, deposits of slag, remains of bowl furnaces and pottery. However, it is not clear to what period this material belongs.
Both mining and metallurgy are known to have been well developed in India at an early period, notably in the Harapan Culture 2300-1750 BC, after which there was a "dark age" until about 1000BC (Bhardwaj 1979). A period of technological improvement then followed through pre-Maurya, Maurya and post Maurya cultures to about 200AD. A number of sites are considered to have originated during these periods. The most comprehensive recent account discussed ancient mining for gold (Allchin 1962) covering many aspects including deep mining, trenching, pitting and quarrying. For the purposes of this study reference is made to three sites in Rajasthan, where comparable mining techniques had been employed in search of lead, silver and zinc.
The mines in question are situated in north-west India, where they are presently being exploited on a limited scale by Hindustan Zinc Ltd. First reports from this area describe the three sites at Zawar, Rajpura Dariba and Rampura Agucha (Craddock et al. 1985, Willies 1984). Further studies (Willies 1987), with radio-carbon dating on wood and charcoal from the underground workings, indicate activity from 3000 BP to the early Iron Age (India) at 1800 BP. Renewed, though intermittent, activity continued between the 13th and 18th centuries.
Mining has exploited a syngenetic ore deposit, formed as steeply dipping lenses in a dolomite host rock of Precambrian age. Sphalerite, marmatite, galena (silver bearing) and pyrite with some native silver were followed through from surface scrapings and opencasts to progressively deeper and often more elaborate workings, attaining depths of 254m below surface. Low, narrow, ramifying and partly infilled workings, probably signify early phases of development. Other areas are markedly larger, some stoping measuring 20-30m across and 2m high, many containing timber supports and equipment. This material was suitable for C14 sampling, and has provided dates in the region of 2100BP and 1900BP.
Rock breakout was apparently achieved by a combination of metal tools, identified by obvious pick marks, and an elaborate system of firesetting attested by quantities of fragmented charcoal and characteristic smooth rounded profiles to worked surfaces.
The dominance of metal tool marks would seem unusual. However iron is known in India from 2900 BP (Hegde 1984), the use of hammer and gad along pre-existing or fireset fractures being the accepted method of working. Firesetting appears to have increased as the works progressed deeper, involving a system of fireholes to produce benches and inclines similar to those described by Davies (1934) for the Kola Goldfield. Certain of the large stopes retain supporting pillars of ore bearing rock, also piles of waste rock faced with stone walling and occasionally pointed with clay. In places recent and modern mining intersects the older workings. Willies (1987) discusses a variety of practices and requirements that need to be considered when assessing a site of this type, including extraction methods, rock stability, drainage, ventilation, transport, lighting, benefication and distribution of the ore.
2.3 Prehistoric Mining in Great Britain and Ireland
Ireland has been recognised as having probably some of the finest Neolithic and Bronze Age monuments and collections of artefacts in Europe. Many of these artefacts were of copper and bronze which Burgess (1979) has classified into a series of metalworking stages. A number of studies (Coghlan & Case 1957, Coghlan 1963) have attempted to relate compositions of early metal implements to known ore deposits, leading to the proposal that copper and arsenical copper metal were a product of the quartz sulphide vein deposits of south west Ireland. A number of fahlerz (grey copper) ores from this region exhibited arsenic - antimony - silver impurity patterns which could be correlated with Coghlans Group 1 metal of the Copper and Early Bronze Ages. Unfortunately, no mine workings of this period were then known, and it was presumed that, if they had existed, they would have been buried or destroyed by more recent mining activity.
These assertions prompted an investigation to identify such sites, leading to the pioneering research of John Jackson at Mount Gabriel (figure 2), credited with having obtained the first radio-carbon date for a prehistoric mine in the British Isles and Ireland (Jackson 1968). The date from an underground context indicated an Early
Bronze Age period. Despite this, and further evidence of early mining , questions arose as to its authenticity as a single sample; the suggestion was made that it only
represented the age of surrounding peat formations that had deposited in the workings (Briggs 1983). However further investigations (O'Brien 1987) established conclusively that these mines were indeed firmly placed in the Early to Middle Bronze Age (Brindly & Lanting 1990)(figure 3).
At least twelve sites in west Cork are potentially similar to the Mt. Gabriel type mines (O'Brien 1990, 1994), notably at Derricarhoon where stone hammers are recorded (O'Brien 1989). Additional sites occur at Ross Island (figure 2) near Killarney, Country Kerry, though they are geologically different. Excavation presently being undertaken here has revealed definite indications of early activity, with seven calibrated dates (figure 3) of predominant Early Bronze Age (O'Brien 1994). Further potential mines are known from Bunhamon, County Waterford and the Wicklow Mountains, although none of these are presently under investigation.
Mount Gabriel
At Mt. Gabriel the hillside is pitted by dozens of small workings, some mere surface scrapings while many penetrate as shallow inclined openings within a 0-12m depth range. They exploit a low grade ore of sedimentary origin, forming a distinct horizon within Devonian mudstones, sandstones and conglomerates, collectively termed ‘red beds’. Mineralisation is confined generally to the coarser grained horizons and comprises mainly the primary disseminated sulphidic ores chalcocite and chalcopyrite, with some tetrahedrite from the enrichment zone. Pigmentation due to malachite and azurite is limited due to the removal of the oxidised zone by glaciation (Snodin 1972). The grade of mineralisation was initially considered to be high (Jackson 1980, 1984), and this lead to a series of estimates for produced copper metal, and comparisons to known numbers of metal artefacts. Much of this is now inappropriate as present ore re-evaluation implies a far lower grade (O’Brien 1994).
Firesetting is considered to have been the main method of rock extraction, evidenced by the curved rock profiles and often considerable amounts of charcoal dispersed through the sharp, splintery fragments of spoil. Certain areas of rock wall display ‘peck’ marks from stone hammers, but despite this, these tools appear to have had a limited use for this means of extraction. Instead, many are considered to represent simple crushing devices for ore benefication. Most of the tools, of which about 30% are modified, originated on local beaches, being of rock types similar to cobbles from Schule Bay, 5km away.
A number of pieces of elongated timber have also been recorded, and are thought to be prisers and wedges for forcing off the fractured fireset rock. Separate, but smaller, fragments of pine wood may be a form of tinder for igniting fires. A single short handled shovel of alder, similar to those found at the Mitterberg in Austria (Pittioni 1951), is recorded from mine No 3.
O’Brien’s approach to the investigation of early mining has been a comprehensive one and present information (O'Brien 1994) provides a good basis to what can be achieved by viewing mining and its associated effect on the landscape in a holistic way. As detailed above this has included information on the morphology and distribution of the workings, spoil characteristics, analysis of artefacts, proposed mining techniques and geological conditions of the ore deposit. However, like so many other early known mine sites the exact type and grade of the ore are difficult to quantify. Therefore there will always be a question as to whether ore production was as great as that presently envisaged. O'Brien's approach to viewing the mining landscape in a wider context, is also of importance in attempting to find associations with field monuments (eg. burial sites) of a similar age to the mines. At present he proposes that the mines were operated on a seasonal basis, with labour organised for production during the winter months, so enabling farming during the warmer months. O’Brien has approached the study of other potential early mine sites in a similar way to those at Mount Gabriel, notably Ross Island as described below. Much of this research is presently ongoing and as yet has not therefore achieved a similar level of detail.
Ross Island
Ross Island is situated on the eastern shores of Lough Leane, to the south of Killarney in County Kerry. The mine workings occur in lower Carboniferous limestones, where quartz veins contain sulphide ores of copper, zinc and lead, together with minor amounts of several other ores. Earliest documentary evidence indicates the mines were being worked during the 18th century, with the main phase of production through the early 19th century (Weaver 1838). During this period there are a number of accounts referring to miners discovering older primitive workings containing a large number of stone hammers, many of which were modified by rilling to facilitate hafting.
In 1992 O’Brien (1994) began a series of surface excavations at the site. A number of early workings have now been investigated, many of which have smooth and concave mine walls, these being interpreted as the result of firesetting. Stone hammers and fragments of cattle shoulder-blade scoops are associated with the excavated spoil from these workings. Radiocarbon dates from charcoal, timber and bone from the base of the spoil indicate that these copper mines were in operation prior to 2000BC.
Further excavation in the vicinity of the mines has revealed much evidence for occupation including worked flint, animal bone, early Beaker pottery and several hut circles. A date range of 2400-2200BC for the pottery is supported by five radiocarbon dates (figure 3) on charcoal, which now confirms that Ross Island is at present the oldest known copper mine in north-west Europe and the first Copper Age mine to be discovered in Ireland or Britain.
During the 1930’s the Committee of the British Association commissioned Oliver Davies, an authority on Roman mining, to investigate the likelihood of early mining in Wales. Using documentary evidence from the previous two centuries Davies (1937) was able to identify a number of sites, the most important of which include Mynydd Parys in Anglesey; Nantyreira, Nantyricket and Cwmystwyth in mid Wales and the Great Orme in north Wales. Limited excavations were completed at each site, with the exception of Nantyricket. Unfortunately Davies was not able to make any observations underground, and so his investigations concentrated on surface spoil deposits, recording stone hammers, charcoal fragments and occasional antler together with some bone from the Great Orme. Similar, though not as detailed studies (Roeder 1901) were completed at the Alderley Edge mine in Cheshire, where stone hammers of a rilled type contained in 19th workings were described.
More recent investigations during the past ten years by amateur archaeologists and particularly members of the Early Mines Research Group have been able to reappraise the earlier described sites and discover new sites. Despite extensive re-working in recent centuries, some 22 sites have now produced evidence for early phases of primitive mining (Picken 1988), with seven sites radiocarbon dated to the Bronze Age period (figure 2). The most important of these sites are described in the following section.
Cymystwyth - Copa Hill (SN 816756)
Formerly recognised by Oliver Davies (1937) by the profusion of stone hammers littering the spoil tips , this site has revealed some important information during recent years. A series of excavations completed by Timberlake (1990a&b and pers. comm.) have provided conclusive dated evidence to place activity at Copa Hill firmly in the Early - Middle Bronze Age.
Mining activity within and around Cwmystwyth relates predominantly to the nineteenth century (Hughes 1981), with exploitation of lead and zinc ores through a series of stoped out veins. However, at ‘Copa Hill’ copper ores, principally chalcopyrite with some malachite, are associated with the same mineralisation. Although galena (lead sulphide) has been found within excavated mine spoil, it has been argued (Timberlake 1990a&b) that this was not the ore sought in prehistoric times. Timberlake reasons that traces of copper ores (chalcopyrite) within the same spoil, suggest that this and its oxidised secondary ores were therefore the chief ores exploited in the early period.
Surface investigations have disclosed an open cast working measuring 40x12m with shallow galleries filled with stratified layers of weathered country rock, recent mine spoil and material clearly from an earlier period. Charcoal recovered from these older workings gave the first indications of firesetting, with sampled material providing C14 dates of Early to Middle Bronze Age (figure 3).
Numerous stone hammers and spalled fragments have been discovered from the same deposit as well as from the surrounding ground surface. They are composed of gritstone, sandstone and quartzitic sandstone, and are thought to have been collected in the river valley below the site. However the degree of smoothing and rounding may suggest an origin from local beaches, some 20km away. An analysis of the hammerstones has recorded weights, dimensions, rock types and degrees of wear. Timberlake (1990b) concluded that the average size lay between 150-250mm in length, 90-120mm in width and 0.8-1.2kg in weight with occasional examples up to 3kg and a few to 5kg in weight. The study showed that about 60% had traces of modification to assist with hafting, often very slight with only faint scratchings and edge notchings, no grooved types were noted. A proportion of the spalled fragments displayed wear marks consistent with reuse, possibly as scrapers or for levering away the shattered rock mass subsequent to firesetting. A limited area of blunt pitting within one of the shallow galleries indicated that the stone tools were utilised for rock extraction as well as for crushing the detached ore.
Further excavation is planned at the site, with proposals to deepen the former excavation pit to prove the depth of the opencast, and to gain access to any adjoining galleries. Excavations completed in the summers of 1994-95 have revealed further deposits of early spoil, with a remarkably well preserved timber launder that has been radiocarbon dated to the Mid-Bronze Age (Timberlake 1994). A damaged antler pick has also been recovered from a context just beneath the launder. Paleoenvironmental work initiated by Mighall (1993) is already supportive of radiocarbon dates from the workings, and it is hoped to make collaborative studies on wider cultural aspects of the mining activity (Timberlake pers. comm.).
The majority of the investigations completed at Copa Hill have been conducted and organised by Timberlake working with members of the Early Mines Research Group. Research, as detailed above, has concentrated on the excavation of spoil from an opencast feature which is thought to have connected with underground workings. Unfortunately, the thickness and volume of spoil have prevented the true extent of these workings being realised. At present only a shallow gallery a few metres in depth has been exposed. Even though the extent of known underground workings are very limited, Timberlake has obtained a great deal of information about the mine from the spoil, artefacts within it and features of the opencast and surrounding landscape. A similar though less detailed approach has been adopted by Timberlake (1994) in the study of other early mine sites in the mid Wales region, notably Llancynfelin (SN 651921) and Nantyreira as detailed below.
Nantyreira (SN 826 874)
Nantyreira is situated near to the source of the River Severn on the eastern slopes of Plynlimon in mid Wales. The mine was worked on a small scale for lead in the late nineteenth century. Although there are no reports of copper ores having been mined here, chalcopyrite is in evidence amongst the present spoil tips (Timberlake 1988). The main lode trends north-south through Ordovician grits and at surface this has been worked out to form a chasm or opencast feature 95m long, 6m wide and 7m deep.
Documentary evidence from the nineteenth century records workings attributed to the ‘old man’, described to be 6-9 feet wide, up to 300yards long and of unknown depth (Bick 1977). Another account (Hamer 1873-6) records the discovery of an iron pick or gad with a wooden handle, a broken piece of antler and a smooth round ball 3.5 inches in diameter, supposedly used for ore crushing.
Oliver Davies, who visited Nantyreira during 1937, noted quantities of burnt rock and charcoal in the vicinity of the opencast. Here two trenches were excavated revealing further similar material, together with hammerstones and muller querns, which Davies (1937) ascribed to the ‘old celtic’ or Roman period.
This former evidence for an early phase of mining at the site prompted further investigations in 1988, when two trenches were excavated through spoil tips near the main lode. Timberlake (1988) describes sequences of material very similar to those of the Davies excavation, including shale, vein rock (quartz), charcoal, some soil and rotted galena and occasional hammerstones. The majority of the hammerstones were unmodified, a few exhibited side picking and one had been used as a mortar. Charcoal (oak) samples provided dates of 3390+/-80 BP (BM-2581) and 3410+/- 80 BP (BM-2583).
Other Mid-Wales Sites
During the past five years a number of potential early mining sites previously mentioned by Picken (1988) have been investigated in this region (Timberlake 1995). At least twenty of these have revealed evidence of stone tools, suggesting the presence of primitive mining. Excavation at two of these sites, Llancynfelin (SN651 921) and Nantyrarian (SN705 814) exposed spoil containing fragments of charcoal which has been radiocarbon dated to 3390+/-35BP (BM-2916) and 3470+/-60BP (BM-2930) respectively (Timberlake 1995).
There is now great potential to investigate these sites in the mid-Wales region, and with this in mind a programme to collate information on past, present and future research in underway (Timberlake pers. comm.). It is hoped that the publication of this report, which will also cover the state of research at other U.K. sites, will encourage funding for a future programme of investigation.
Parys Mountain (SH445905)
Parys Mountain near Amlwch in Anglesey is probably best known for its extensive opencast and underground workings from the late eighteenth and early nineteenth centuries. Documentary accounts from these periods do, however, indicate evidence for mining at a much earlier period. Stanley (1873) and Evans (1873) reported a number of stone hammers, some of them grooved or notched, and considered to be of Roman origin. The existence of twelve Roman copper cakes from Anglesey, three of which originated from very near the mine site, further reinforced this initial assumption.
Oliver Davies, who visited the site in 1937 also ascribed a Roman age to the earliest workings on the hillside. This was partly based on his observations of material and artefacts obtained from two trenches that were excavated through a flat lying spoil tip a short way north of the summit windmill. Davies (1937) records finding twenty four hammerstones, charcoal fragments which he assumed to be from firesetting, and a stone lid. No evidence for any smelting or processing was recorded.
During 1988 members of the Early Mine Research Group proposed to relocate the Oliver Davies excavations (Timberlake 1988). Eventually, after some abortive excavation attempts a site resembling the description given by Davies was located just north of a location known locally as the ‘Oxen Quarry’. Two excavated trenches revealed a series of discernible layers of mine waste, the lower of which contained complete hammerstones, spalled fragments and scattered charcoal. Quartz vein material was also much in evidence, although no pyrite or chalcopyrite was found, probably on account of the acidic conditions. The hammerstones varied in size from 150g to 6.75 Kg, and were generally of unmodified form, although a few showed evidence for side notching and one was faintly pecked around its waist. Identified rock types included dolerite, microgranite, ignimbrite, porphyry, quartzite, serpentine, chert and gritstone. Collected charcoal (species - Oak) has now provided three C14 dates of 3490+/-50BP (BM-2585), 3500+/-50BP (BM-2586) and 3550+/-50BP (BM-2584).
In July 1995 an access was created to an area of underground workings which had been sealed for at least 15 years. A series of workings, assumed to be of 18th and early 19th century were explored with the intention of investigating any evidence for earlier phases of mining. Approximately 160m from the new entrance at a depth of 20m evidence for likely prehistoric activity was found towards the end of a chamber probably worked out in the 18th century. This consisted of stratified, graded and partly cemented spoil deposits which contained complete and fragmented hammerstones, charcoal and organic debris (wood, leaves, acorns and grass). It appeared as though this deposit was in situ and represented a surface infilling of an inclined gallery. A similar deposit was found at the other end of the chamber. Both locations clearly displayed the relationship whereby the 18/19th century workings had cut into the base of these earlier workings. Research at the site is presently ongoing and it is hoped C14 dates will confirm that these workings originated during the Bronze Age.
Alderley Edge (SJ860780)
This location, 20km south of Manchester, is an outlier of Permo-Triassic sandstone, cut by faults with associated copper-lead mineralisation, comprising mainly disseminations of malachite-azurite with some galena (Carlon 1979).
Evidence for early activity was first described by Dawkins (1875) who noted a number of grooved stone hammers in the surface spoil tips. A short while later he returned to excavate with General Pitt Rivers. About one hundred stone hammers were recorded, being divided into three main types : those with a transverse groove, those with a lateral groove and those with a distinct wedge-shape incorporating a groove.
More systematic searches (Roeder 1901, Roeder & Graves 1905) added four more sites with grooved hammers. They also identified a series of pits along the Engine vein opencast. These had clearly been joined by a later stage of mining, by knocking out the sides. From the base of the pits extended smaller diameter shafts, which they considered to be trials. Pieces of charcoal recovered from one pit suggested firesetting may have taken place, but an iron rod was also found, so the exact context could not be established. Clearly activity from some later period, possibly 18th century onwards, had occurred.
More recent descriptions of the Alderley Edge mines by Warrington (1965, 1981) have covered activity over the last few centuries, but noted that no conclusive evidence existed for dating the hammerstones to a time earlier than the medieval period.
Detailed examination of the surface pits and their relationship to historical mining by Gale (1989) has confirmed all previous descriptions of the earlier workings. Gale has identified three phases of mining, an early period of surface pits and shafts with peck marks distinctive of a hammer stone technology, a 17th to 18th century phase of underhand stoping through pickwork, and a drill and blast phase characteristic of the deeper 19th century mining. The first phase of mining is interpreted as Bronze Age in origin, this is being based on the present belief that stone tools are now accepted as indicative of Chalcolithic (Copper Age) to Bronze Age mining.
No material suitable for C14 dating has yet been found on site; however an oak shovel has recently been reported, originating from the site last century (Garner et al. 1994). This shovel was located by miners in 1878, who also discovered a number of grooved hammers or mauls in a old surface working “from three to four yards in depth”. In 1992 the shovel was rediscovered and subsequently submitted for C14 dating, which indicated an age of 3470+/-90BP (OxA-4050) placing the wood in the first half of the second millennium cal BC (Middle Bronze Age).
The previous descriptions have to some extent attempted to summarise the many aspects of individual sites where evidence for mining in the Copper and Bronze Ages is now recognised. Perhaps the first obvious difference that is apparent is the variability in how individual authors describe the sites on the basis of geology, mineralogy, artefacts and mining features, shape and extent of the workings, processing and smelting, archaeological interpretation and associations on a wider perspective, for example relationships with trading and distribution of ores and metal products.
In dealing with geological and mineralogical conditions the main contrasts would appear to be the composition of the ore type and the nature of the host or county rock. Typically ores tend to be represented by sulphides of copper with associated ores of zinc and lead. Where ores of this type have been exposed to surface weathering they would have undergone oxidation to form gossans or ‘iron hats’ with underlying concentrations of sulphide ores occurring in an enriched zone immediately beneath. At Rio Tinto in Spain such an arrangement occurs with assumed exploitation from the oxidised zone in the Early Bronze Age, while during the Late Bronze Age improvement in mining technology allowed the ores from the deeper enriched zone to be worked (Rothenberg 1972). At many of the other early mine sites any evidence for gossan material is absent, and is presumed to have been removed by glaciation, for example the Mitterberg in Austria, Mount Gabriel in Ireland and Cwmystwyth in Wales. Some of the remaining sites do not record any traces of gossan either, but they do record the presence of ores from the oxidised zone, for example Rudna Glava in Serbia where rich concentrations of copper carbonate ores from oxidised chalcopyrite veins occur in limestone host rocks. Similar mineralogical/geological conditions are also found at Aibunar in Bulgaria and Seronais in France. These particular sites also have many similarities with the early mines of the Great Orme.
Country rock geology is also an important consideration in the appraisal of early mines, particularly resistance to surface weathering and/or combined rotting associated with the oxidation of sulphidic ores. It is likely that these factors alone could govern the ease of ore extraction and corresponding development of early mine workings. For example, ore deposits contained within a friable rock will be more easily removed than those from harder material. It would appear from the described sites that two particular host rock types can be considered to relate to the extent of the known workings. Calcareous rock types are the most common, such as limestone and dolomite, as found at Aibunar, Rudna Glava, Cabriere, Bouche Payrol, India and Turkey. Non calcareous rock which cover the remaining rock types include mudstones/slates (Cwmystwyth & Mitterberg), sandstones/gritstones (Mount Gabriel, Alderley Edge & Timna) and igneous types such as felsite (Mynydd Parys). These two host rock types produce very different ground water chemistry, which is known (Lewis 1994) to influence the preservation of artefacts and organic material contained in the mine spoils. The higher pH of calcareous spoil aiding the preservation of bone and metal objects, while the lower pH of the non calcareous material aids the preservation of organic material such as timber and plant remains (pollen, leaves and seeds).
It is generally observed that early workings in calcareous rock types display tendencies to be more extensive than those in other rock types. This implies that ore bearing limestones-dolomites must have been more easily removed than equivalently mineralised non calcareous rock types. In contrast, post Bronze Age mining tended to exploit the harder host rocks, as evidenced by the numerous Gallo-Roman mines of Europe (Davies 1935), where iron picks were the accepted mining tool. However, many calcareous rocks can be relatively resistant to iron tools and may therefore have proved to be uneconomic to mine during the Bronze Age, even with the use of stone hammers and firesetting. At the Great Orme it has been observed (Lewis 1994) that rotting has occurred to dolomite in the vicinity of the mineral veins, enabling the material to be removed with little effort and simple tools (section 4.2). It would therefore seem likely that similar rotting may be responsible for, or to have contributed to the removal of ore in the previously described sites.
At this point it is also interesting to note that a few of the mine sites described indicate distinct phases of activity, primarily to two periods, notably in S.W. Spain, the Balkan region and Timna in Israel. It is suggested that copper ores were exploited on a limited scale during the Copper to Early Bronze Age periods (4000-2500BC), this being primarily based on ceramic evidence from the spoil tips rather than radiocarbon dating methods. This complete phase of early working is therefore somewhat speculative and may require some re-interpretation if provenanced C14 dates are not forthcoming. The second phase of mining attributed to the mentioned sites falls within the Late Bronze Age period, which sees more active and larger scale mining activity. Similarities with these two phases of mining also occur in India where Bhardwaj (1979) describes a “dark age” from the Harapan Culture (2300-1750BC) until 1000BC.
The two phases, however, may also represent not only time differences but also technological improvements in the exploitation of a mineral deposit, and its subsequent processing and smelting. It is not unreasonable to consider that some of the earliest exposures of copper ores would have attracted early miners by colour alone. These ores are likely to have existed as residual gossans or from the oxidised zone immediately beneath, and would have consisted of oxides, carbonates, secondary sulphides and native copper. Where the native ores occurred in appreciable quantities they are likely to have been removed by simple scraping and hammering, with the other ores left in situ or perhaps removed for use as pigments. Eventually through experimentation with melting the native copper, the possibility of smelting the other ores, notably carbonates and oxides, may have been realised. Even though present evidence suggests these advances in smelting technology had occurred early on, the evidence of small scale mining seems to suggest that actual mining methods, or an understanding that further ores existed below the rock surface, had not improved. The necessary advancements to exploit these ores (carbonates, oxides and sulphides), do not then appear to have taken place until the Late Bronze Age. Without further research at the mentioned sites it is at present difficult to explain why there is this gap in mining activity between the Early and Late Bronzes Age periods. In view of the research that is continually advancing at some western European sites, it is perhaps now time to reassess and investigate further some of the sites where Middle Bronze Age activity is at present absent. The most essential factor would be to obtain C14 dates as a means of provenancing rather than by or in addition to dating by ceramic material alone.
While mining activity at the sites mentioned above appears limited to the Chalcolithic-Early Bronze Age and the Late Bronze Age periods, the remaining sites through Europe, seem to indicate continuous mining from the end of the Early Bronze Age to the Late Bronze Age periods, notably the Mitterberg, Austria and Feinan, Jordan. This continuity is particularly true at the Great Orme, as will be described in the following chapters.
In contrast to many of the European and Near Eastern sites notably S.W. Spain, Mitterberg, Cabričres and Timna, evidence for smelting of copper ores in Britain and Ireland is virtually absent. The exception being Ross Island, where recent excavations (O’Brien 1995) have revealed deposits of roasted ore and minor slag, unfortunately though, C14 dating now indicates a possible ‘Dark Age’ origin. Some processed ores have also been recorded from Mt. Gabriel, but it is not known where they were smelted. Craddock (1990) suggests that the absence of smelting remains indicates the involved processes may have been ephemeral and quite undetectable and were a result of small scale smelt charges and slag free technique. If this is so, then the question arises as to how the sometimes large quantities of ore material, as proposed at the Great Orme, could be smelted without leaving any residual evidence. A following chapter will discuss and attempt to explain the absence of smelting evidence at the Great Orme.
