This paper is published in Issue 1 of the Journal of the Great Orme Exploration Society 1, 7-12. 2000.

A revised and more complete version under the title 'An assessment of copper mineralization from the Great Orme Mine, Llandudno, North Wales, as ore in the Bronze Age.' Proceedings of the Yorkshire Geological Society 53, 213-219. 2001, is available off-line.

Abstract
Spoil heap and in situ mineralization from the Great Orme Mines have been classified into six mineral associations and by applying mineralogical, mining and beneficiation criteria their potential roles as Bronze Age copper ores are assessed.
Major veins and open voids within dolostones are infilled by coarse-grained, saddle dolomite-chalcopyrite-pyrite-calcite-malachite. The tonnage, estimated copper grade of 10% Cu and continuity of the veins plus the coarse grain-size of the copper-bearing minerals, suggest that they were the main Bronze Age copper ore possibly accompanied by a contribution from adjacent void-infilling mineralization.
Two minor (and very different) types of mineralization, namely the 'copper ddu ore' - a very fine-grained, friable, limonitic vein infilling, and disseminated, approximately one centimetre diameter, azurite nodules in shales, may have been taken as a by-product of other mining operations (for example mine development).
Neither the fine-grained, disseminated, polymetallic sulphides (Cu-, Pb-, Fe-, Co-, Ni-rich) which are only found in dolostones from the spoil, nor the single, copper-poor, galena-rich assemblage found as a vein infilling, can be considered to be copper ores. These two associations belong to the Mississippi Valley-style lead-zinc mineralization of the Northeast Wales Orefield which is older in age and differently sourced from the copper-dolomite association ores that produced Bronze Age copper metal.
Recognition and sampling of ore, rather than of just mineral specimens from a mine site, have implications for geochemical or isotopic provenance studies that lie far beyond the Great Orme. This is because only run of the mill ores should be treated as having any importance to archaeometallurgists.

Introduction
The concept of an ore is a sophisticated one. It has more to do with mining, economics and profit/benefit than with geology and mineralization. Nowadays ore is sufficiently important that it has a legal definition that nicely encompasses all of these aspects, and it is unlikely that the role and value of ore could have been any different in the Bronze Age.
A useful definition of a metalliferous ore is that of a naturally occurring raw material that can routinely, consistently and economically (in terms of time and resources) be mined, beneficiated and smelted into metal. A profit/benefit (however that was/is perceived) has to be achieved either for an individual or for a community.
For successful, repeatable metal smelting a consistent charge is needed with a high metal content and fixed mineralogy. To meet these two requirements most ores have to be beneficiated - have their metal contents increased by the removal of unwanted minerals (called gangue) and/or the separation and concentration of the metal-bearing ones. Beneficiation, in turn, is most efficient when the metal content, grain size and mineral textures of the ore entering the ore-mill are kept constant, so as literally to produce "run-of-the-mill" ore. These days this consistency is achieved by stockpiling different ores from a mine (or mines) and then blending them before they enter the beneficiation mill. Ore therefore must have both grade (extractable metal content) and tonnage, and so it is ore-as-mined/run-of-the-mill ores rather than exotic and/or heavily mineralized but rare, specimens that have archaeological significance (Ixer 1995, 1999, Ixer and Budd, 1998).
The main significance of such ores is that they make possible the matching of mineral to metal i.e. the provenancing of metal and metal artifacts- but in addition they can yield information about ancient technology, as they comprise the major constraints on mining and smelting methods. Provenancing studies were originally performed using trace elements (Coghlan and Case, 1957; Northover, 1989) but more recently this method has been augmented by stable isotope studies, especially those of lead (Deal, 1985). However, often an inability or an unwillingness to distinguish between ore and mineralized specimens has led to the analysis of inappropriate materials. The subsequent discussion and counter discussion of these analyses, summarized in Budd et al., (1996) and Tite, (1996), have led to the legitimizing of the "Magpie School of Provenancing" where the analyses of colourful, attractive specimens are discussed in black or white terms, whilst dull but correct ores have been neglected.
The different mineral association/ore types found at The Great Orme Mines are considered in the light of these and other considerations and the concept of 'ore-triage' is introduced and applied to them.

Mineralization at the Great Orme Mine
The Great Orme Mine (SH772831) lies above and to the west of the Victorian seaside resort of Llandudno. Here, epigenetic copper mineralization belonging to the copper-dolomite class of deposits cuts Late Asbian to Early Brigantian limestones, dolostones and minor siliciclastics (Ixer and Stanley 1996; Ixer and Davies 1996).
There are strong structural and lithological controls on the styles of mineralization and composition of the ores and potential ores found within the Great Orme Mine. Fault and joint planes carry the bulk of the mineralization to give vein-style ores comprising coarse-grained, void-infilling, sulphide-carbonate minerals, or locally a very soft and friable hydroxide-oxide-carbonate mineral association. Sulphide-carbonate minerals are also present within void spaces (vughs) in dolostones and dolomitized limestones. This type of mineralization has a patchy distribution, but is concentrated close to the veins.
Fine-grained lithologies, namely argillaceous dolostones-dolomitic shales-shales carry minor amounts of disseminated, fine-grained sulphides or, very rarely, are azurite- ± malachite-bearing.
The majority of the mineral associations are the result of a primary sulphide mineralization followed by intense, widespread, polyphase, supergene alteration. The primary mineralogy is simple, comprising sulphide veinlets, vugh-infilling or "knots" of saddle dolomite overgrown by chalcopyrite, accompanied by minor pyrite and marcasite. The primary sulphide assemblage has altered to a series of secondary copper sulphides and oxides including digenite, covelline, spionkopite, cuprite, tenorite and native copper but mainly to limonite and malachite. Late generations of calcite are interbanded with botryoidal malachite and limonite and minor amounts of manganese oxide minerals.
General descriptions of the geology and mineralization, including their role in controlling Bronze Age mining, have been given in Lewis (1990, 1994), Jenkins and Lewis (1991) and Dutton et al. (1994). These authors, together with Ixer (1994), have given brief mineralogical descriptions of the ores, ascribing both the lead and the far more abundant copper ores to Mississippi Valley-style mineralization. However, more recently Ixer and Stanley (1996), and especially Ixer and Davies (1996) have shown that, despite the geographical position of the Great Orme - namely lying on the margin of the Northeast Wales Mississippi Valley-style, lead-zinc orefield (Ixer and Vaughan, 1993), the main copper mineralization is later than the minor lead ores, unrelated to them, and that it belongs to the copper-dolomite association. This association is a minor but world-wide class of mineralization found in close proximity to some massive lead-zinc deposits, but is differently and more deeply sourced than they are.

1a Vein copper ores
Historically these were the most important copper ores and this was probably true in Prehistory (Smith, 1991; Lewis, 1994). At least twelve, subparallel, north-south orientated, high angle faults, up to decimetres in width, cross-cut dolostones and dolomitized limestones and were infilled by coarse-grained sulphides and secondary minerals rather than just high-grade malachite as suggested by Joel et al., (1997). Due to the efficient exploitation of these ores in the nineteenth century, very little accessible vein ore remains in situ within the Bronze Age workings of the Great Orme Mine, and hence the detailed mineralogy of these ores has not yet been described.
A thin, one or two centimetre-wide, stringer cutting dolostones beneath the Craig Rofft Sandstone, present at the end of the main Bronze Age chamber, and representing an off-shoot from a main vein, has been collected and is taken to be an example of the copper ores that formed the main veins. The mineralization which comprises primary coarse-grained sulphides, up to 0.5 centimetres across, cross-cuts dolomitized limestones, and the wallrocks next to the vein are green-stained where malachite has pervasively replaced dolomite crystals along their cleavage directions or infills small void spaces in the rock. Although this is surface staining it may account for the incorrect description of the stringer as 'high-grade malachite' by Joel et al. (1997).
Pentagonal dodecahedral pyrite crystals, intergrown with twinned marcasite and chalcopyrite, are extensively oxidized to limonite. Locally, thin (10-30mm wide) chalcocite/digenite rims have formed between the primary sulphides and secondary limonite that carries trace amounts of fine-grained covelline, spionkopite and cuprite. Late stage malachite is present as botryoidal crusts overgrown by fibrous and finally stout malachite crystals growing into void spaces within limonite; it is locally accompanied by calcite and manganese oxides. The mineralogy, ore textures and paragenesis of this ore are identical to the vugh-infilling mineralization, found in the dolostones and dolomitic limestones adjacent to the veins, which, with the removal of the high grade ore, is the most obvious mineralization to be seen today.
The vein copper ores offered tonnage and with the absence of significant cross-faulting a continuity in mining, albeit with variations in vein thickness. In places supergene mineralization is accompanied by dedolomitization and partial dissolution of the dolostone host rocks immediately adjacent to the veins producing a zone of soft, partially friable wallrock - the "weathered rotted wallrock" as figured in Lewis (1994) that was far easier to mine than 'fresh' dolostones. Copper grades for the vein copper ores are not given in the literature but visual inspection suggests that approximately 10% copper metal, in the form of chalcopyrite, malachite but also as copper-bearing limonite, is a reasonable assumption. The grain size of the main copper-bearing minerals was coarse enough that beneficiation by hand-cobbing would produce a copper concentrate suitable for roasting prior to smelting.

1b Vugh-infilling mineralization within dolostones
This is the most obvious mineralization within the mine area, seen as green-stained knots of sulphide-carbonates infilling void spaces, up to 30cms in diameter, within coarse-grained dolostones and dolomitized limestones (Davies, 1996). However, much of the visible malachite-staining is very thin (only a few tens of microns); this is especially true for joint and bedding surfaces, hence care should be taken when trying to determine the copper content of the rocks by visual inspection.
The petrography of both the primary and secondary mineralization is given in great detail by Ixer and Davies (1996). Epigenetic, multistage, primary fluids followed by supergene mineralizing solutions have formed rhythmical crusts of saddle dolomite overgrown by chalcopyrite plus a little pyrite and marcasite, which in turn are overgrown by successive generations of texturally distinctive, supergene calcite and malachite.
Chalcopyrite carrying rare, micron-sized inclusions of iron-poor sphalerite is intergrown with euhedral pyrite, which locally is nickeliferous/cobaltian, and lesser amounts of twinned marcasite. Chalcopyrite alters to thin (up to 50mm wide) rims of digenite, djurleite, spionkopite and covelline, plus trace amounts of the silver sulphide acanthite, but mainly to limonite and malachite. The supergene mineralization includes at least three generations of calcite, each associated with minor amounts of copper as malachite plus traces of cuprite, tenorite and native copper. In places, late generations of calcite, aragonite, malachite and brochantite are seen to cement Bronze Age spoil material (Lewis, 1990).
The vugh-infilling mineralization is second only to the vein-style ores in terms of potential tonnage. However, the mineralization is stratabound, restricted to certain coarse-grained dolostones and dolomitized limestones that are themselves only found close to the mineralized faults. The host rocks show little sign of being friable and the distribution of the vughs is patchy so that the only vugh-infilling ores that were likely to have been mined are in the softer 'rotted' wallrock-envelope immediately alongside the main veins. This appears to be the explanation for the continued presence of the 'green rock', an outcrop of malachite-rich dolostone and limestone in the 'opencast' (Davies, 1996), situated between two worked out veins. The similarities in ore petrography between the vugh-infilling and main vein ores means that they would also have the same ore grade (up to 10% copper metal) and beneficiation characteristics, namely hand-cobbing to produce a chalcopyrite and/or malachite concentrate.

2a Vein Copper ddu
Locally veins cutting dolostones carry black, very fine-grained, friable material rather than mixtures of sulphides and carbonates. In situ samples were collected from the 94m level within the mine. Initial X-ray work on this material, known as copper ddu (black copper), showed it to be a mixture of amorphous phases with a little dolomite and trace amounts of malachite and goethite (Ixer and Davies, 1996). Subsequent XRD analyses failed to find any crystalline phases although trace amounts of chalcopyrite were identified using the SEM (R.Doonan pers. comm.)
Although some tonnes of the material remain in the mine (T. Hammond pers. comm.) and its soft nature would make mining easy, it is difficult to know if this was a copper ore in the Bronze Age. Qualitative chemical analyses (XRF) on two different samples show them to be chemically heterogeneous, one shows the presence of iron, copper and nickel, plus small amounts of cobalt and manganese, whereas the other has major calcium, copper and iron, plus minor amounts of arsenic and trace amounts of manganese, nickel and silver. The friable nature of the material would prevent beneficiation by any physical methods hence the raw copper ddu itself would comprise the potential 'smelter charge'.
In order to test its potential as a source of copper metal the copper ddu was roasted in air. The resultant sinter was X-ray diffracted and studied in reflected light. Initial results, reported by Ixer and Davies (1996), suggested the presence of CaO ± CaFeO2 plus some Ca2MgFe2O6 and a strong peak that matched that of copper metal. However the presence of copper metal was not confirmed by reflected light petrography. Recently, more extensive investigation of the copper ddu showed that although the resultant phases vary with the temperature and length of time of roasting, they included tenorite (CuO) and delafossite (CuFeO) and elemental copper alongside oxides of calcium, magnesium and manganese (R.Doonan, pers comm.) Reflected light petrography also showed the presence of phases with the optical properties of tenorite and delafossite.
Elsewhere in the Northeast Wales orefield similar ores have been mined as sources of cobalt and nickel products rather than of copper metal (Warren et al., 1984).

2b Azurite nodules in shales
Azurite is far less abundant than malachite and is highly sporadic in its occurrence. It has been reported as deep blue crystals infilling vughs in dolostones in the Higher Shaft area (Dutton et al., 1994); within a shale horizon off Owen's Shaft at the 18m level (A. Lewis pers. comm.) and from a separate single, shale horizon in the "opencast" (Ixer and Davies, 1996). The reason for the rarity of the mineral is that the high pH of the ground waters/supergene enrichment fluids has ensured that malachite and not azurite is the stable copper carbonate (Ixer and Davies, 1996). Hence most azurite is present in non-carbonate host lithologies where the pH of the local groundwater would be lower. It is an odd coincidence that azurite at Alderley Edge, another Bronze Age copper mine, is also restricted to shale horizons; however, here it occurs within siliciclastics not carbonates and so is a little more difficult to explain.
Within the 'opencast' a single, non-carbonate-bearing shale horizon, 30cm thick, is unique in carrying small, azurite spheres, up to 1 centimetre in diameter, as the main copper mineral, together with very minor quantities of relict pyrite and chalcopyrite (Ixer and Davies, 1996). The azurite is restricted to this horizon for even within the dolostones directly beneath and in contact with the shale altered pyrite and chalcopyrite is associated with malachite - albeit a blue-green malachite.
The copper grade of the azurite-bearing shale is low, estimated to be less than 2% Cu metal, but the ore is persistent, is soft so would be easy to mine, and the nodules are coarse-grained so would be easy to beneficiate by hand-cobbing to produce a pure copper concentrate. There is evidence that metre-thick shale/mudstone horizons were used as cross-courses to gain access between adjacent vertical veins (Lewis, 1994) and Jenkins and Lewis (1991) have suggested that these horizons carry up to 0.5% copper (as malachite and azurite) which could be recovered as a by-product of mine development. However, there is no evidence that thin, horizontal, copper-bearing strata with low concentrations of copper were mined solely as ore, as the continuing presence of the azurite-bearing shale in the 'opencast' testifies.

3a Disseminated polymetallic sulphides
Amongst the specimens obtained from the surface spoil are a small number showing a distinctive ore assemblage. These include the National Museum of Wales specimen NMW 83.41G. M1489 and specimens GO19, GO20 (RAI), which have been described in detail (Ixer and Stanley, 1996), plus an undescribed sample GTO.08 found on the spoil north of Vivian's Shaft (A. Lewis pers.comm.). They all share a polymetallic mineral assemblage, with copper-, lead-, cobalt-, and nickel-bearing minerals within fine-grained, argillaceous dolostones that show widespread dedolomitization. Chalcopyrite, siegenite, galena, pyrite and marcasite are the main sulphides, together with minor amounts of very zoned nickeliferous/cobaltian pyrite, gersdorffite/cobaltite and trace amounts of tennantite and sphalerite. Supergene alteration of the primary phases is minor but the iron-cobalt-nickel sulphides are altered to limonite; chalcopyrite to limonite and malachite; whereas galena is replaced by cerussite. The presence of cobalt- and nickel-bearing sulphides and sulpharsenides suggests that the disseminated polymetallic assemblages do not belong to the copper-dolomite association but rather to the Mississippi Valley-style mineralization of the Northeast Wales Orefield (Ixer and Stanley, 1996).
Specimen GTO.08 is similar to the previously described samples. Mineralization is present as disseminated sulphides within the dolostone but also within veinlets overgrowing saddle dolomite crystals. Minor quantities of detrital quartz (including some metamorphic clasts) and mudstone clasts are present but little argillaceous matter. Chalcopyrite, up to millimetres in diameter, is intergrown with pyrite and associated with small crystals of nickeliferous-cobaltian pyrite and gersdorffite/cobaltite. Chalcopyrite shows alteration to bornite (the only bornite recognised from the Great Orme Mines area) and pale blue copper sulphides and spionkopite. Late stage azurite accompanies malachite cross-cutting limonite after chalcopyrite and pyrite.
The presence of the secondary cobalt arsenate erythrite, as 10-20mm diameter spheres infilling void spaces within in situ shales from Roman Shaft, has been recorded by Jenkins and Johnson (1993). Careful examination of multiple sections from this material (GO 11) (RAI) show the presence of organic matter plus pyrite, nickeliferous cobaltian pyrite, chalcopyrite and trace amounts of galena and marcasite. No primary arsenides or sulpharsenides are seen. Indeed the mineralogy and ore textures display so many differences from the disseminated polygenetic sulphide 'ores' that they cannot represent the same type of mineralization.
As all of the disseminated polymetallic sulphide material comes from surface spoil and does not match any in situ ore or any sampled mineralized wallrock, there is no evidence that the material was mined in the Bronze Age from the Great Orme Mine. Also it is not possible to give any tonnage and difficult to estimate the copper grade. However, the ore would be hard to beneficiate, as much of it comprises fine-grained intergrowths of sulphides so that a separate copper concentrate could not be achieved, making it difficult to smelt due to the presence of nickel and cobalt. This type of mineralization cannot be considered to be a Bronze Age copper ore.

3b Vein lead ores
Galena-rich veins are rare, only one having been found as a void-infilling along a master joint plane within the 'opencast'. Centimetre-diameter crystals of galena are associated with millimetre-diameter chalcopyrite and lesser amounts of nickeliferous pyrite, marcasite and trace amounts of millerite and sphalerite. The galena is free of sulphosalts but carries up to 140 ppm silver (R. Chapman, pers. comm.). Primary galena and chalcopyrite are altered to the secondary copper sulphides, digenite, djurleite, spionkopite and covelline; in addition galena is altered to cerussite. Late malachite and calcite infill small void spaces (Ixer and Stanley, 1996; Ixer and Davies, 1996). The mineralogy and textures of the ore show it to belong to the Mississippi Valley-style mineralization found in the Carboniferous Limestone of Northeast Wales, as described by Ixer and Vaughan (1993), and shown to be earlier than the main copper mineralization (Ixer and Davies, 1996; Joel et al, 1997).
This mineralization cannot be considered to be a Bronze Age copper ore. The copper content is low, less than for any of the other ore types. Indeed both in the field and in hand specimen the material is clearly a potential lead ore with galena being the most abundant ore mineral. The mineralization is rare within the mine, present in the single vein, quite separate from any of the copper ores. This is of significance for at Cwmystwyth, where chalcopyrite-rich and galena-rich ores occur next to one another, Bronze Age miners discarded the galena (Timberlake, 1994) and only took the copper ores. Hence the unintentional incorporation of significant amounts of lead into Bronze Age copper ores intended for smelting at the Great Orme seems impossible.

Discussion and Conclusions
The copper and lead mineralization at the Great Orme Mines, rather than being an unusual example of the Mississippi Valley-style ore deposits of the Northeast Wales Orefield, is more complex and includes a number of genetically distinct mineral associations. In particular, the galena-rich vein, which does belong to the Northeast Wales Mississippi Valley orefield, is earlier and unrelated to the copper-rich mineral assemblages which form Britain's first recognised example of the copper-dolomite association (Ixer and Stanley, 1996; Ixer and Davies, 1996).

To: TABLE 1.

The copper-dolomite association is made up of a number of copper-bearing mineral assemblages, each with its own mineralogy, mining and beneficiation characteristics, and each assemblage may be regarded as a potential copper ore. Table 1 summarizes the characteristics of these assemblages and shows the results of a triage, which, based on grade, tonnage and beneficiation criteria, has classified them into realizable, by-product (potential) and non- Bronze Age copper ores. It strongly suggests that the copper veins supplied the "run-of-the-mill" ore perhaps with addition of adjacent vugh-infilling copper ores; that the ore was hand-cobbed to produce a mixed chalcopyrite, malachite and copper-bearing limonite concentrate ready for roasting. Other than iron, this concentrate would carry few metals except in the low ppm range. It is possible that azurite from shaly horizons and malachite from vughs within the dolostones were collected in the Bronze Age and used for pigment, or were the initial copper ores from the Great Orme (Jenkins and Lewis, 1991). However, their routine use to provide a pure copper carbonate concentrate, during much of the Bronze Age as suggested by Joel et al. (1997), seems unlikely. Whilst a copper concentrate would have been simple to produce from these sources, as potential ores they lack continuity and, if the grades of the present day examples at the Great Orme are representative, metal content. They may only have been exploited as a by-product of other mining activity, for example the cutting of cross-courses within thick (approximately 1 metre) mudstones to travel from vein to vein. The presence of the malachite-bearing vughy dolostones and the thin azurite-bearing shale in the Bronze Age 'opencast' supports this view, namely that they were not considered to be important, independent copper ores.
The importance of the limonite copper ddu ore is more problematical. Although it would have been easy to mine it is local in its distribution and its fine grain size prevents any physical beneficiation from producing a copper concentrate. However, if the identification of copper phases in the sinters are correct, then it suggests that simple heating of copper ddu produces copper - whether that copper is extractable and how much copper is produced are still unknown.
Neither the polymetallic nor the galena-rich mineral assemblages should be considered to be Bronze Age copper ores from the Great Orme. No in situ polymetallic-type ores have been collected from the Bronze Age workings - or elsewhere in the mine. Their rarity in the spoil suggests that their tonnage was very low and their petrography shows that any copper concentrate made from them would be contaminated by significant amounts of cobalt, lead, nickel, iron and some arsenic. The copper content of the galena veins is too low for it to provide a copper concentrate.
Having established the relative importance of the mineral associations and, in particular, what can and cannot be regarded as Bronze Age copper ore, it is now possible, briefly, to indicate their potential for metal provenancing. The copper ores (1a and 1b) can only produce a simple copper concentrate of chalcopyrite-malachite-copper-bearing limonite, plus minor pyrite/marcasite and trace amounts of copper sulphides and oxides, ± azurite, plus a carbonate gangue. This assemblage carries few metals other than copper; iron, found in chalcopyrite, pyrite, marcasite and limonite, is present in major amounts, but other metals are less than 0.5% and probably only in the low ppm range. These metals include cobalt and nickel present in cobaltian and nickeliferous pyrite; silver in chalcopyrite and acanthite and manganese as manganese oxides. Jenkins and Lewis (1991) have reported the trace element concentrations from three roasted samples of malachite and two of chalcopyrite including an example of chalcopyrite-dolomite ore. Their data confirm the mineralogical data presented here, namely that a chalcopyrite concentrate has cobalt, nickel and manganese all below 0.02wt% (200 ppm) and that for void-infilling malachite the values are lower. However, they report arsenic concentration in chalcopyrite of 0.32 to 0.42wt% and high nickel (0.13wt%) in malachite taken from a mudstone host. However, even these relatively enhanced values are unlikely to be reflected in any copper metal produced from these minerals.
No Bronze Age copper (made from mixed copper sulphide-carbonate ore) from the Great Orme Mine could produce high arsenic copper nor tin bronze without being alloyed. In fact the mineralogy of the ores shows that they could only produce trace element-poor copper metal like most other Bronze Age ores from the British Isles (Ixer and Budd, 1998), and that their usefulness in provenancing Bronze Age metal work based upon distinctive trace element signatures is very limited.
A more recent method of provenancing metal to ore is to measure the relative isotopic abundances of lead in order to characterize their isotopic compositions. For any ore this isotopic composition will depend on the source rocks and the geological time of extraction of the lead from that source. Vary either of these and the isotopic composition/signature changes. However, primary ores and the supergene/secondary ores derived from them have the same signature.
At the Great Orme the galena vein (and the polymetallic assemblage) are part of the Mississippi Valley-style lead-zinc mineralization of the Northeast Wales Orefield and are probably Permo-Trias in age sourced from local Upper Carboniferous sedimentary basins. In contrast, the copper-dolomite association is younger, probably Mesozoic-Tertiary in age, and sourced from the large, regional, Irish Sea/Cheshire, Permo-Triassic basins (Ixer and Davies, 1996). The galena and main copper ores are neither contemporaneous nor from the same source rocks. Since the accidental incorporation of galena into any copper ores used at the Great Orme is very unlikely, the isotopic signature of galena has no archaeological significance for copper metal produced from the Great Orme. It could have importance in provenancing Late Bronze Age lead-containing bronzes, were galena intentionally exploited to help make this alloy.
The copper-dolomite ores (copper veins ± vugh-infilling ores) should share the same lead isotopic signature as each other and this should be carried through to any copper metal produced from them. If the limonitic copper ddu is oxidized vein copper ore then it too should fall within the same isotopic field. Similarly the azurite nodules from the shales may share the same signature as the rest of the copper-dolomite association but they may plot to one side and be more radiogenic. Shales and especially organic-rich shales have higher uranium contents than carbonates and this lithological contribution may effect the lead isotope signature of the azurite.
The limited amounts of published lead isotope data from ore assemblages at the Great Orme include three galenas two of which are from the spoil and one from the galena-vein (2577), (Rohl, 1996; Joel et al., 1997); three malachite-rich specimens, two of which are from vugh-style copper mineralization in the 'open cast' (2723, 2725) and one azurite from the azurite shale in the 'open cast' (Joel et al., 1997; M. Goodway, pers. comm.). As predicted, and as noted by Rohl (in Joel et al., 1997) and by Joel et al. (1997), the lead isotopes for galena plot in a separate field from that of the copper mineralization, with the copper ores being more radiogenic. An explanation for this more radiogenic signature is provided by data from the Ty Gwyn copper mine, approximately 1-5km east of the Great Orme Mines, where uraninite-bearing bitumens are intergrown with saddle dolomite and chalcopyrite in classical, galena-free, copper-dolomite association ores (Parnell, 1988, Ixer and Davies, 1996). In addition, inspection of figure 3 of Joel et al. (1997) shows that the azurite from the azurite-shales is even more radiogenic than the vugh-infilling copper mineralization as represented by the malachite analyses.
Not until all the lead isotopic data are published, alongside detailed mineralogical descriptions of the ore samples, together with those from relevant metal artifacts, will it be possible to test rigorously the present mineralogical predictions. However, these early data do confirm that 'ore-triage' (the mineralogical assessment of material into ore, potential ore and non-ore) should be an essential and natural first step before any sort of geochemical provenancing is undertaken. This would allow for a proper determination of what should be provenanced, by what methods, and should provide an estimate of the chances for success.
In the past the only significant contribution to the provenance debate is that the use of poorly collected and superficially described material from the supragossan zone (as defined by Ixer, 1999) has convinced many people that metal provenancing is an untameable chimera. It may be, but we cannot know this for certain until a number of rigorous, well-integrated studies tell us so.

Acknowledgements
Mr. Tony Hammond of Great Orme Mines Ltd. and Mr. John Davies are thanked for their material, advice and for illuminating discussions that underpin much of this work. Drs. Roger Doonan (English Heritage) and Graham Hendry are thanked for XRF and XRD analyses of the copper ddu ore. The project was partially funded by SERC research grant GR3/9560 and the continued generosity of the Constantine XI Palaeologos Travel Fund.

References
Budd, P., Haggerty, R., Pollard, A.M., Scaife, B. and Thomas R.G. 1996. Rethinking the quest for provenance. Antiquity, 70, 168-174.

Coghlan, H.H. and Case, H.J. (1957). The early metallurgy of copper in Ireland and Britain. Proceedings of the Prehistoric Society 23, 91-123.

Davies, J. (1996). Detailed stratigraphy of the Copper Mines of the Great Orme with emphasis on the occurrence of mineralization. Unpublished undergraduate thesis. University of Birmingham.

Deal, E.C. (1985). Use of lead isotopes in archaeological studies. In: 33rd Annual Conference on Mass Spectrometry and Allied Topics. East Lansing, U.S.A. 66-67.

Dutton, L.A., Fasham, P.J., and Jenkins D.A. et al; (1994). Prehistoric copper mining on the Great Orme, Llandudno, Gwynedd. Proceedings of the Prehistoric Society 60, 245-286.

Ixer, R.A.F. (1994). Bronze Age Ores at the Great Orme Mines. Abstract, Research Colloquium on the Great Orme Copper Mines. Llandudno.

Ixer, R.A.F. (1995). Ancient Mining, Modern Collections. Abstract, Mineralogy in the Field 1995. A one-day symposium organised by the Russell Society in association with the University Museum at Keble College, Oxford.

Ixer, R.A. (1999). The Role of Ore Geology and Ores in Archaeological Provenancing. In. (S.M.Young, A.M.Pollard, P.Budd and R.A.Ixer, Eds) Metals in Antiquity, BAR International Series 792, 43-52.

Ixer, R.A. and Budd P. (1998) The mineralogy of Bronze Age copper ores from the British Isles: implications for the composition of early metalwork. Oxford Journal of Archaeology 17, 15-41.

Ixer, R.A. and Davies, J. (1996). Mineralization at The Great Orme Copper Mine, Llandudno, North Wales. UK Journal of Mines & Minerals 17, 7-14.

Ixer, R.A. and Stanley, C.J. (1996). Siegenite-bearing mineral assemblages found at the Great Orme Mine, Llandudno, North Wales. Mineralogical Magazine 60, 978-982.

Ixer, R.A. and Vaughan, D.J. (1993). Lead-zinc-fluorite-baryte deposits of the Pennines, Mendips and North Wales. In (R.A.D. Pattrick and D.A. Polya, Eds.). Mineralization in the British Isles. Chapman and Hall. 355-418.

Jenkins, D.A. and Lewis, C.A. (1991). Prehistoric Mining for Copper in the Great Orme, Llandudno. In (P. Budd, B. Chapman, C. Jackson, R. Janaway and B. Ottaway, Eds). Archaeological Sciences 1989. Oxbow Archaeological Monograph 9, 151-161.

Jenkins, D.A. and Johnson, P.B. (1993). Abandoned metal mines: a unique mineralogical and microbiological resource. Journal of the Russell Society, 5, 40-44.

Joel, E.C., Taylor, J.J., Ixer, R.A., Goodway, M. (1997). Lead isotope analysis and the Great Orme Mine, In (A. Sinclair, E. Slater and J. Gowlett, Eds.) Archaeological Sciences 1995. Oxbow Archaeological Monograph. 64. 123-131.

Lewis, A. (1990). Underground exploration of the Great Orme Copper Mines. In (P.Crew and S.Crew, Eds.) Early Mining in the British Isles. Plas Tan y Blwch, 5-10.

Lewis, A. (1994). Bronze Age Mines of the Great Orme. In (T.D.Ford and L.Willies, Eds.) Mining before Powder. Peak District Mines Historical Society 12, 31-36.

Parnell, J. 1988. Mineralogy of uraniferous hydrocarbons in Carboniferous-hosted mineral deposits; Great Britain. Uranium. 4, 197-218.

Rohl, B.M. (1996). Lead isotope data from the Isotrace Laboratory, Oxford. Archaeometry data base 2, Galena from Britain and Ireland. Archaeometry 38 165-180.

Smith, W.D.F. (1991). The Great Orme Copper Mines. Creuddyn Publications, Llandudno.

Timberlake, S. (1994). Evidence for Early Mining in Wales. In (T.D.Ford and L.Willies Eds.) Mining before Powder. Peak District Mines Historical Society 12, 133-143.

Tite, M.S. (1996). In defence of lead isotopes. Antiquity 70, 959-962.

Warren, P.T., Price, D., Nutt, M.J.C. and Smith, E.G. (1984). Geology of the country around Rhyl and Denbigh. Memoir of the British Geological Survey.