4 EVIDENCE FOR PREHISTORIC MINING
The purpose of this section is to describe and discuss the evidence for surface and underground mining at the Great Orme. This will include information on the mining features and artefacts that are considered to be representative of early or primitive mining. Comparisons will also be made to equivalent features and artefacts that are known to be of 18-19th century origin.
The beginning chapter will describe the chronology of discovery, size and extent, and associated spoil deposits of both the surface and underground workings. This will be followed by descriptions of artefacts from these areas, including details on the form, distribution, likely use and origin of stone tools, fragmented bone, charcoal from possible firesetting, bronze fragments and other minor objects. All references to surface and underground locations relate to figures 8,9,10,11 and appendix B. The final section will describe and discuss the available evidence for ore processing, smelting and distribution of the finished metal.
4.2 Description of the Mine Workings
This section will describe the extent, form, complexity and chronology of discovery of the Bronze Age workings, and how geological factors govern these conditions. Comparisons will be made to the more recent 18-19th century workings and any obvious intervening ages of mining activity. By contrasting different features from these periods, it has been possible to produce a list of criteria that characterise the prehistoric workings. This will become useful in future identification of early mining at the Great Orme, as well as providing a means of possible characterisation at other potential and known sites. A major proportion of this information will be presented as a series of charts (section 6.2), incorporating detail from all previous work dealing with the site (Lewis 1990a&b, 1993a,1994, Jenkins&Lewis 1990, James 1990, Dutton 1990, 1993, Dutton&Fasham 1994). Sections will deal with morphology and physical features of the workings, type and distribution of artefacts and mining structures, and descriptions of spoil deposits. Individual comment and supplementary information will be included in the main text.
An attempt to differentiate ages of activity within the known span of the Bronze Age workings will be discussed, and proposals made as to how certain features can correspond to a progression and advancement of mining techniques through these periods.
The first discoveries of early workings likely to be of Bronze Age origin were described last century when miners broke into old workings. Two particular accounts are of note, from 1831 and 1849, and these are detailed in section 2.5. Later in the following century Davies (1948) investigated the site but was unable to gain access to any surface or underground workings as all entrances were either capped off or sealed beneath the extensive tips of spoil that existed on the site at this time (refer also to section 2.5).
Our first accounts of the exploration of nineteenth century workings date to the late 1960’s when members of the Shropshire Caving Club gained access to extensive areas of workings beneath Bryniau Poethion via Treweeks and later the Roman shafts. During this period workings to a depth of 70m were explored and partly surveyed, they were found to contain many artefacts from the past two centuries including various tools, clay pipes and pieces of clothing. A number of stone hammers were also located mainly in the workings off Roman shaft. Explorers of the time (P. Appleton - pers. comm.) suggested, like the miners of the last century, that these tools can only have been used during some very early period of mining activity at the site.
Renewed interest in the mines occurred in the late 1970’s when an amateur archaeologist Duncan James and local enthusiasts further explored the workings at a depth of 40m off the Roman shaft. After clearance of backfilled rubble a set of workings (location 5) very different to those typical of the nineteenth century were discovered. Throughout these workings were many of the characteristic indicators that are now associated with Bronze Age mining activity, as detailed through this section and referred to in section 2.5.
In April 1985 the author was able to accompany D. James on an excursion through the workings then C14 dated to the prehistoric period (table 1). This lead to an active period of exploration and discovery by members of the recently formed Great Orme Exploration Society through all the workings from the past two centuries accessible via Roman and Treweeks shafts. At this time it was not possible to access any of the workings with connections to Vivians, Owens or Pyllau shafts. Despite this limitation, many discoveries were made, including further artefacts and features from recent centuries and also additional areas where evidence for prehistoric mining existed (eg. locations 1-4, 13-15 & 25-27).
As mentioned previously in section 2.5, a proposed survey as part of a reclamation scheme at the mine site allowed access to be gained to workings off the three sealed shafts. In November - December 1987 entrances were formed to each of the shafts, beginning with Pyllau, then Vivians and Owens. In Pyllau shaft no workings occurred until a depth of 140ft (43m) and all of these were clearly of 19th century origin.
For Vivians shaft the situation was very different with labyrinthine workings (figures 20-21) from virtually the surface to a depth of 132ft (40m), the majority of these trended northwards and were considered to connect with the same vein systems encountered in the Roman shaft workings, later survey confirming this. To the south the workings were far less extensive and smaller in size indicating a termination of mineralised ground. Eventually using a systematic approach to exploration/survey, links were made to the workings from the base of Treweeks shaft (90-110m AOD level). These were made at reference points Nos. 69 &55, so allowing ‘through access’ between Treweeks and Vivians shafts, but more importantly providing a shorter and safer route to these deeper areas of workings. Throughout the workings indications of early mining activity were found in numerous locations (eg. Nos. 18 & 29-33), some of which appeared to have had very little disturbance during the recent centuries of mining (locations 19 & 21).
At Owens shaft the workings occurred off a series of five individual levels between the depths of 40ft (12m) and 140ft (43m). Again the majority of the passages conformed to those of the 19th century. However at the 60ft (18m) level (133-135m AOD) an extensive area of largely undisturbed early workings were found (locations 8-11). They had been driven along the dip of the ore-bearing mudstone/dolomite, whose likely entrance lay to the south where the inclined strata would have originally broken surface in the hollow below the present Pyllau farm house.
Through the following years to the present, continued explorations revealed further areas of early workings (figure 19) and it is likely many more await discovery. Of the workings found in more recent years the most notable are those off the 140ft (43m) level in Owens shaft (locations 22-23), and those at location 35 from Vivians shaft. Apart from these new discoveries, areas of known workings continue to be studied in more detail, revealing in some cases material which was previously overlooked, eg. further metal fragments from location 21.
Excavations at surface exposed and continued to reveal a combination of true surface workings (trench mining) and sealed entrances to isolated areas of underground workings. Many of these had no obvious direct connections to known workings accessed off the individual shafts. In 1988 during the completion of a reclamation scheme at the site, areas of rock and stratified spoil were exposed around Vivians shaft. Investigation of this area by Gwynedd Archaeological Trust (GAT) prompted the removal of further spoil, leading to the discovery of a small gallery which provided charcoal for the earliest C14 date to the site. Excavation in the following year revealed continuing sequences of early spoil containing charcoal, bone and stone hammers, and exposures of rock that indicated an early mined surface.
In 1990 the proposed tourist mine development by Great Orme Mines Ltd (GOML) necessitated the removal of large quantities of spoil from the site in order to relocate former entrances to the workings so that an underground tourist route could be created. During the summer of 1990 a large quantity of material, of the order of 20,000 m3 was removed by machine excavator directly to the north of Vivians shaft. This exposed what is now identified to be part of a large opencast working dating to the prehistoric period. During this phase of excavation numerous stone hammers and bone fragments were found, together with charcoal fragments from possible firesetting. Shallow grooving and concavities of the exposed rock surfaces suggested they had been worked with these simple implements. A number of entrances were also located, many of which were found to connect with the trench-like surface features where ores had been removed (plates 2 & 4), whilst others connected with true underground workings (location 7b). A second phase of machine and archaeological excavation took place during the winter of 1990-91, and concentrated on the removal of spoil from the present tourist entrances (location 17), and to a lesser extent from the main opencast to the north. Further phases of spoil removal progressed during following winters of 1991-92, 92-93 & 94-95, enabling the the main opencast to be enlarged (exposing locations 7a & 7c), a continuous surface walkway to be created and access to the tourist entrances improved.
Throughout this same period limited archaeological excavations by GAT (Dutton & Fasham 1994) and GOML progressed on site, mainly during the summer months, and concentrated on areas to the west (locations 7b) and north (location 7c) of the main opencast. These excavations revealed further spoil deposits and associated mined rock exposures, consisting of pocked surfaces and trench veins that linked through into a series of passages extending along the lines of the mainly north-south veins. Excavations at location 7c continued to October 1995.
Introduction
This study intends to deal primarily with those workings of a true underground nature partly due to the fact that the surface workings have been fairly comprehensively described and interpreted (Dutton 1993, Dutton & Fasham 1994). However these studies have dealt more with specific areas, notably 7b with excavations centred around Vivians shaft from 1988-92. Therefore this particular section will incorporate information gained from the main excavations by GOML from 1990-95, and also comment on an overall chronology of development of the surface workings.
As described in section 2.6 machine and hand excavations by GOML have proceeded on site to the north and south of Vivians shaft from May 1990 to present. Throughout this period two main areas have been uncovered by the removal of nearly 100,000 tonnes of mine spoil. The main feature is the opencast and its associated surface workings (locations 7a-c) that lie to the north of Vivians Shaft (plates 2 & 3), while to the south east lie the tourist entrances (or Southern Entrances) and obvious trench workings that outcrop along a possible fault scarp (plates 3 & 5).
Opencast
Machine excavation to reveal the opencast proceeded to a large extent between June 1990 and March 1991, with spoil removal from about 138m AOD down to 128m AOD. During this time many of the features that are now seen in the opencast were exposed. The upper layers of spoil were typically 18/19th century in nature, being angular, of varying size with occasional drill holes, and contained iron artefacts, timber stemples and other assorted material of a contemporary age. Remains of a buried masonry building on the east side of the opencast at a level of approximately 135m AOD indicated there must have been various stages of activity through these centuries. Similar finds were made by Dutton & Fasham (1994) who recorded the remains of a masonry floor situated directly to the east of Vivians shaft. From old plans of the site of 1847 it is possible this was a crushing floor associated with an ore washing area. The excavated spoil also contained the occasional stone hammer and items of bone. This spoil material appeared to continue to the rock outcrop on the east side of the opencast.
On the west side the spoil was interrupted by a dark coloured clayey horizon of 20-350mm thick (plate 7 & figure 18), which contained bone, snail shells and decayed organic material. Below this clay layer which continued to about 132m AOD, the spoil became less angular in form with increasing numbers of stone tools and pieces of bone. Similar clay bands of varying thickness were found in other parts of the opencast during following years, notably in the area excavated by GAT and GOML. At the time it was considered that this clay unit represented an old soil horizon corresponding to a period separating the lower spoil of possible prehistoric origin from the upper of more recent material. This interpretation is still being reviewed and no firm conclusion has yet been reached as to the significance of the clay layer. An investigation (Dr. Pete Marshall at Dept. of Archaeology, University of Sheffield) of organic material from the soil layer identified pollen from the following species : Pteridium, Cyperaceae, Compositae, Graminae, Filicales, Plantago major and Polypodium. The fact that these species presently occur on the Great Orme, together with the lack of sample material, would suggest that at this stage it is not possible to use pollen studies as a means of characterising the soil horizon.
Below the clay unit the numbers of stone hammers and fragments of bone appeared to increase as the excavation was deepened. Fragments of charcoal also became more common with increasing depth, sometimes occurring as small ‘nests’. Eventually the rockhead was exposed, revealing further stone hammers, some of which laid directly on the rock surface, suggesting they had been deliberately left in these positions (plate 14). An unusual feature of the exposed rock was the virtually complete absence of any material that could be interpreted as an original soil horizon,which is likely to have accumulated once the ore had been removed from these areas. A few of the exposed rock cavities did contain some dark clayey material, but on closer inspection this was interpreted to be rotted mudstone rather than organically derived soil.
Once the bulk of the material had been removed from the opencast, final cleaning of rock surfaces and more detailed excavation was completed by hand. The resulting surfaces were then examined in detail. The exposed rock surfaces displayed features similar to those from underground, displaying trench-like features where ore had been removed along the line of a vein and also pockets, hollows and openings to underground workings (plates 2 & 10). All of these features exhibited smooth curved surfaces and occasional tool marks similar to those made by bone and stone implements. On the west side of the opencast a cave-like opening was uncovered, and on the floor lay the skeleton of a small sheep or goat of indeterminate age. Directly above and to the south of this a human lower jaw bone was unearthed which, from its close proximity to dated contexts identified by GAT (Dutton & Fasham 1994) would appear to be of Middle Bronze Age. Occasional tool marks comparable to those made by iron picks were also noted in the upper parts of the opencast, these being best displayed in a small oval-shaped shaft sunk through the ‘Pyllau Mudstone’ on the east side (plate 6) at an elevation of about 140m AOD. Drill holes for blasting were completely absent from the lower part of the opencast suggesting an age predating the use of gunpowder.
Throughout the opencast only traces of copper mineralisation were observed. The most notable locations include the malachite stained central rock mass at the base of the opencast, the surface of the main terrace on the east side where altered chalcopyrite and rarer copper carbonates occur, and a shale unit on the same side that contains numerous small (up to 6mm) nodules of azurite. Towards the north east corner of the opencast small pockets (up to 150mm) of galena on a vein surface is also present, much of which has altered to the lead carbonate cerussite.
During machine excavation of the opencast a number of very large boulders, some up to 4m x 3m in size, were encountered amongst the spoil towards the sides of the feature. The majority of these blocks were of dolomitised limestone displayed traces of copper mineralisation, a few limestone boulders also being noted, which displayed a pronounced weathered surface not unlike many of the weathered limestone blocks (possible glacial erratics) seen at various locations over the Great Orme today. Many of the excavated blocks now line the sides of the road to the summit of the headland, while the largest limestone block was positioned on the flat ground near to the entrance of the mine. A number of blocks can still be seen within the opencast, situated in the north east corner. Originally these were thought to be in situ rock, however during 1994-95 anomalous dips of bedding and differing lithologies to surrounding material, indicated that they are in fact detached blocks (plate 6). From the observed lithologies, notably the ‘Pyllau Mudstone’ and rubbly horizon, it is estimated the blocks have been displaced along the line of the near vertical north-south veins by as much as 10m. Displacement along these linear features would also explain the near horizontal attitude of the detached blocks in contrast to normal land slip features that typically describe a rotational movement. Further collapsed blocks occur on the northern side of the opencast, these appear to have originated even higher in the local lithological sequence (Engine House beds) having been displaced over at least 14m. It is difficult to ascertain when collapse occurred to the sides of the opencast, although it is likely to predate the use of gunpowder (pre 1700’s) as no drill holes have been observed in any of the detached blocks (see also section 6.3).
Collective evidence, as described above, would seem to indicate that the opencast was mined out at a very early period, and is very unlikely, as has been suggested (David forthcomming) to be an early chamber which collapsed as a result of blasting after discovery in 1849. The most obvious reason being the total lack of any drill holes which would suggest a recent origin.
Southern Entrances (Tourist Entrances)
The southern entrances, which form part of the tourist mine were first located as part of an underground survey conducted in 1988, and accessed via the recently opened Vivians shaft. When initially located, the tunnels leading up to these entrances were extensively backfilled with a combination of early and 18/19th century spoil (plate 4, figure 23). Through excavation and further surveying it was realised that these passages must have originally linked to surface, but were subsequently buried beneath about 20m of mine spoil. As part of the tourist mine development by GOML it was decided to remove this spoil in order to expose the entrances which today form part of the public route underground.
Machine excavation of 18/19th material commenced in the winter of 1990-91, with hand excavation proceeding once rockhead was met. The majority of the spoil was
clearly of recent origin, with occasional iron tools or parts of equipment, coal and cinder from the Engine House and/or associated sources, clay pipe fragments and pieces of timber. The spoil itself was also clearly of the same period, consisting of varying sequences of mainly angular material; sometimes this was well sorted indicating that processed rock/ore had been graded into defined sizes ranging between silt, sand, and fine to coarse gravel components. This material was typical of a mechanical ore-crushing and separation process which had been tipped into the hollow where the present Visitor centre and mine entrances are situated. Documentary evidence (Williams 1979) also indicates the presence of an ore-washing pond or ‘slime pit’ in this area. A number of mussel shells were observed in one of the finer grained units, and it would seem likely these were food remains discarded on the tips. Underlying this horizon a floor of red brick at an elevation of 128m AOD was noted, which was thought to be an 18th century surface, as it occurred far lower than many of the known 19th century floor levels.
A few stone hammers were found amongst the 18/19th spoil, these increasing in numbers as the rockhead was approached accompanied by spoil more typical of an early origin (ie with charcoal fragments and bone). At one location, in vein East 9-10 a cache of four hammers were found, to all intents and purposes they appeared to have been deliberately left there. Another interesting feature of the spoil deposits on the southern side was the presence of stone mortars, about 10 complete and damaged ones being collected. They were on average 300mm across with central depressions on each side. None of these mortars are known from any underground contexts, and very few were located from excavations in the opencast. This suggests that the mortars may have indicated some sort of early processing operation in this part of the site or may simply indicate they had been purposely collected in more recent times for use in the processing operations that are known to have occurred here. This second interpretation would seem to be the less likely as any 18/19th century processing would have been by machine operation, alternatively the presence of mortars here may represent some as yet unidentified period of processing (and mining) at the site.
As increasing quantities of spoil were removed from the southern side and more rockhead was exposed, it became apparent that this area was traversed by a number of mineral veins, most of which had been exploited by trench mining (figure 23 & plate 4). Limited excavations were conducted within these trenches, the spoil they contained bearing many of the characteristics to suggest an early origin for the deposit. The complete lack of drill holes or iron pick marks, but presence of bone tool marks, also strengthened the interpretation that the exposed rock represented an ancient mined landscape, considered to be of Bronze Age origin. A total of 14 mineralised fractures or veins were recorded east of Vivians shaft, each likely to have witnessed exploitation at an early period with trench working continuing in some cases to where dolomitised limestone gives way to a sandstone unit (Craig Rofft Sandstone). Below the sandstone the dolomitic rock in the vicinity of the fractures was again mineralised, and this had resulted in further extraction but this time by the excavation of tunnels following the copper ores into the hillside. From the form of these tunnels and contained material they were considered to be of prehistoric origin.Today they form part of the underground tourist route in the mine.
Closer inspection of particular trench veins, notably vein East 3, indicated they had at some time been closed off to surface by the insertion of stone blocks to form stemples. Further examples of these same features were noted in the north west corner of the opencast during excavation work in 1991-93, once more consisting of deliberately placed blocks that prevented spoil from entering the trench working (plate 10). Similar stone blocks were also noted in some of the underground workings. It may be possible that these stone stemples were emplaced during the early period of mining in order to allow spoil to be stored above these locations, at the same time maintaining access through the trench workings and to surface. Such a system would maximise the quantity of spoil that could be stored at surface and also minimise the distance over which excavated spoil would have to be transported from new areas of working. A number of small revetment walls or walls of ‘deads’ were also exposed during excavation, these have been described by Dutton & Fasham (1994), and are interpreted as being structures to retain mine waste while maintaining access to both surface and underground workings.
Introduction
For very many years mining enthusiasts explored the Great Orme mine workings with no or little understanding of the extent of evidence for earlier phases of mining, now ascribed to the Bronze Age. This is probably also true for many of the other known sites where mining activity of a similar period has now been identified. At the Great Orme the evidence that now characterises the early workings has taken at least ten years to identify. New information continues to appear, and is likely to do so well in to the future. The evidence for early mining at first consisted largely of artefacts (section 4.3), notably stone hammers, animal bone and charcoal fragments (firesetting). Gradually however, the areas where these artefacts were located were observed to be different in many respects from those workings which were clearly of a 18-19th origin. Over a period of time it became possible to distinguish a range of features that typified the earlier workings, which contrasted with equivalent features from more recent mining, eg tool marks. When these features were considered in combination with the associated artefacts, it then became possible to not only distinguish between them and mining activity from the last two centuries, but also to propose likely phases of activity within the Bronze Age period itself.
Form and Extent
Workings from the 18-19th centuries are relatively simple to identify, as they are characterised by the existence of drill holes from blasting, which became a common practice from the early 18th century. Where these drill holes are present, their orientation is useful in determining in what direction a passage or working was driven. Rates of progress for blasting with gunpowder depended on the characteristics of the rock type, but typically rates of 12-18 inches (300-450mm) a day could be expected for driving a six foot high passage in limestone (Williams 1979). Prior to this time the principal mining method consisted of either using picks or a hammer and chisel (or gad) for scraping and chipping the rock away, with much reduced rates of progress in similar conditions of 1-2.5 inches (25-65mm) a day. This earlier method of rock removal, prior to the use of gunpowder is likely to have been in existence since the development of iron tools. Therefore any mining using this technique could span a period of time from the Iron Age to the use of blasting powder, and is also likely to have extended to recent centuries, where evidence indicates that pickwork was common in those areas where ore bearing rock was sufficiently friable. It is quite possible some of these workings could date from even the Roman period, as has been previously suggested for the Great Orme (Davies 1948, Gardner 1958) and for many other mines that purport to have workings from some earlier period (Davies 1935) . It would be extremely difficult to differentiate between those tool marks dating from a Roman period and those of pre gunpowder mining circa 1700-1750. So without alternative evidence such as pottery or other typological artefacts, it is difficult to propose any definite age for workings of this type.
Apart from the drill holes, the 18-19th century workings display other features that distinguish them from workings of pre gunpowder age. Generally, these passages and worked out areas (stopes) tend to have sharp angular rock surfaces with square or rectangular profiles that are a product of the blasting. Waste rock or spoil produced from these areas is also angular in shape and quite variable in size ranging from sand sized particles to boulders some 0.5m across, but fragment sizes generally range between 25-150mm across. The workings often contain many of the tools, equipment, mining features and personal items of the miners from these centuries (tables 5,6,7,8 & figure 27). In contrast there are other workings that are clearly different, and it is these that are considered to be of an early origin.
Apart from the artefacts (section 4.3) that are contained within these earlier workings , there are particular surface rock exposures, tunnel profiles and associated spoil deposits that are clearly distinguishable from those of recent centuries.These earlier workings are found in both areas of dolomitised limestone/dolomite and of units of mudstone. In each of these situations the form of the passages is quite distinct. In the former rock type the workings have fairly smooth curved surfaces, with irregular lense-shape tunnel profiles following the line of the mineral lode (figures 22 & 24,25,26, & plates 11-12). The largest of these features are better described as stopes, and may vary in width from 250mm to 3m and up to 10m in height. In the mudstones the passage shape is more square to rectangular in profile with less rounded rock surfaces, which is likely to be the result of deterioration and spalling due to hydration and fracturing of the mudstone along joints, as seen at location 7a (figure 8). However, areas do exist where little spalling has occurred in certain mudstones, this is likely to be the result of increasing calcareous material and/or less softening/rotting of the mudstone. Passages displaying this type of profile can be seen in the eastern side of the opencast (plate 6). In practice though, many of the early workings exhibit a combination of these two conditions, primarily due to the fact that mineralisation tends to concentrate beneath mudstones or sandstones particularly where they are transgressed by joints or fractures. This arrangement gives rise to the most common passage shape, being widest at ceiling level defined by the underside of the mudstone or sandstone which has provided a line of weakness, and then tapering to the floor as the vein passes through dolomitic material (plate 13). Carbonate minerals are typically seen within the base of the mudstone, occurring as nodules and infills in cracks and fractures. As described in section 3.0, mineralisation can extend beneath the confining horizons of mudstone or sandstone giving rise to ore bodies known as ‘flats’. These laterally extensive ore masses typically have a flat ceiling defined by the lower surface of the mudstone (or sandstone) with an undulating floor with rounded surfaces of dolomite (plate 20).
Throughout the workings above described, tool marks with curved surfaces and rounded points have often been observed, particularly in the softened dolomites and mudstones towards the ceilings of the passages (plate 17). At times these markings are seen to follow very narrow veins or stringers of mineral, suggesting that even very small quantities of ore had been extracted. These tool marks are ascribed to bone tools and are very different to those made by iron picks or chisels, that in contrast display sharp angular profiles with almost linear points of contact. The bone tool marks have become a useful indicator for the presence of early mining, particularly where other evidence such as bone/stone implements is absent. Their existence at other known early mine sites is also of importance, especially where acidic ground waters are likely to have destroyed any bone that may once have existed. Other evidence of tooling, though far less common than the bone tooling, consists of shallow almost circular depressions of 10-25mm in diameter. From descriptions of similar tool marks at
other known prehistoric mine sites (Craddock 1992), and from firesetting experiments conducted on the Great Orme (Lewis 1990b), these are considered to have been made by stone hammers, and are further described in section 4.3.1.
As described earlier (section 3.0), the main factor in governing the size and shape of the early workings is suggested to be the degree of rotting and deterioration of the dolomites, dolomitised limestones and mudstones and to a lesser extent of the base of the sandstones, where some calcareous material exists. In the case of workings through dolomitic material, it is the rotted component that has been virtually completely removed giving little indication as to the type of material that originally existed at these locations. A few situations do exist however, where remnant rotted dolomite is seen to occur as pockets within the harder unaltered dolomite. The surface between the the rotted and hard dolomite is quite well defined and not transitional, so implying that the hard, smooth, rounded surfaces that define many of the passages correspond to this surface. It is these surfaces that define the morphology of the majority of the known workings that are considered to date from the prehistoric period. When picked at with any blunt instrument or even a finger nail the rotted material is friable enough to be easily removed, so demonstrating that the use of bone or wood tools would be quite appropriate for this type of extraction. Many of the bone tool marks found in these areas further reinforce this interpretation. The same removal techniques would also have been quite sufficient for working the mudstones, and once more the occurrence of bone tool marks support this view. Sizes of the early workings vary greatly, generally the width of the veins becomes less towards the east and west margins of the mineralised area. Particular veins towards the centre of multiple vein system can be up to 3m wide both on surface and underground. The smaller less pronounced veins when excavated form tight tortuous tunnels often with sections that prove very difficult to pass through, in places barely 200mm across. The only means of excavating such narrow ways may have been to use workers of small stature, in some cases perhaps children of an early age.
Some of the early passages still retain pockets of rotted dolomite containing small quantities of altered chalcopyrite. Occasionally appreciable amounts of unaltered chalcopyrite are observed in the side walls of these workings, evidently deliberately left by the early miners. Such evidence may indicate that the principal ores mined in prehistoric times were copper carbonates as well as altered chalcopyrite. This is also supported by various accounts (Humphries 1917) from last century, where miners describe the "old welsh" workings as retaining sufficient quantities of the "yellow ore", likely to be chalcopyrite, to justify exploitation. In another account (Stanley 1850) the miners report “that their predecessors of former times had been unable to work the hardest parts of the rock, in which the richest ore is found, for they have recently obtained many tons of ore of the best quality from these ancient workings”. It is likely that the reworking and enlargement of these early passages has meant that a number of them were destroyed partially or in some cases completely. One nineteenth century account (Humphries 1917) describes one such chamber discovered in 1849 as being "about 40 yards long....."and "the roof being one mass of stalactites reflected back through their lights with dazzling splendour", unfortunately this was completely destroyed by blasting. No direct evidence to locate the position of this chamber has yet been found, although one tentative deduction would place it beneath Bryniau Poethion where a similar size cavern does occur (15m east of location 5). The mine manager’s reports (Vivian 1855-57) also indicate there had been activity in this region over the past few years. The naming of ‘Roman shaft’ at this location would also tend to suggest that older workings had recently been discovered in this area. Other chambers are also known, one of which - 10x15m across and 8m high - presently known as the “big or flat stope’ (location 18) was originally thought to have belonged to the nineteenth century but is now considered on excavated evidence to be largely the product of Bronze Age activity.
Present evidence from surface and underground indicates that the early workings cover an area of at least 240m by 130m stretching from the Pyllau valley to Bryniau Poethion, with vertical depths to 70m. The survey plans (figures 8,9,10,11) give a good impression of their extent, but not of the complexity and varied shapes of the worked areas, mainly because of limitations of the survey. However a few areas have been surveyed in more detail to give a true impression of the labyrinthine nature of the workings. This information has been plotted to give 3-dimensional representations of two main areas centred around locations 19 (figure 21) and from 17-18 (figure 20). These views indicate how the early workings and later 19th century tunnels have followed the interconnecting north-south, east-west and oblique trending mineral channels or veins.
A limited but important series of detailed excavations has been conducted in a few of the known areas of prehistoric workings. These have concentrated on the removal of material (wood charcoal or bone) for C14 dating, and include locations 7a, 7c, 11, 18, 20, 21, 26 and 35, and also for Uranium series dating of calcite flowstone from locations 5, 7a, 22, 23, 34 and between 17 to 21 (ref. pt. 219). More detailed excavations have been completed at locations 7c, 18 and 21 (figures 24,25,26) by GOML (David 1992-95 & Lewis 1990-92). This work consisted of the removal of sequences of both 19th century and prehistoric spoil, revealing stone tools, bone, charcoal and further fragments of metal from location 21 ( see section 4.3.3).
Spoil and Associated Deposits
The recognition of at least eight varieties of spoil have been proposed (Dutton & Fasham 1994) from the surface and underground of both prehistoric and recent workings. These are considered to reflect the source rock type and the mining technique employed for its removal. However, these differences are often subtle and care is needed in distinguishing similar types especially where the context appears to be the same. Generally, distinctions are based on degrees of angularity/rounding, sorting and size. For the purpose of this study, descriptions of spoil type will concentrate on material from underground contexts.
Typically, spoil from Bronze Age contexts tends to be fairly evenly sorted having average sizes of 50-125mm. Fragments are sub-angular to rounded in shape, with a matrix of a sandy component derived from the rotted dolomite. Occasionally more angular dolomite fragments mixed with charcoal are found. These deposits are considered to have been produced by firesetting, and quite often display a reddening indicating the oxidation state of iron oxides in the normally pink-brown coloured dolomite has been altered by heating. Complete or spalled stone hammers occur throughout these spoil types, while bone has a tendency to occur to a greater extent in the more sandy material with rounded fragments.
Many of the early workings appear to have been backfilled with spoil. This would seem to be a common practice with the technique commonly used last century as well as in modern day mining. All types of mining generate large quantities of waste rock which requires convenient disposal. The most practical sites are therefore workings where all ore has been removed, which can be readily filled with material from those productive workings in close proximity. Many of the early workings at the Great Orme still contain part of their original backfill, much of this material having been removed during the 18/19th centuries when miners prospected these areas. Evidence for this backfill exists as remnant spoil adhering to the walls and ceilings and infilling small pockets in the earlier tunnels. Occasionally, almost completely backfilled workings which do not appear to have seen any disturbance by recent mining are found. The positions of these sites are presently being recorded for future study. It is also possible that backfilled passages may have assisted with controlling ventilation through the complex workings at the site. It could be argued that if firesetting was being used in the deeper areas of the mine, then some means of ventilation would be essential for safe and efficient working.
In contrast to the ‘early’ spoils, 18-19th century spoils are composed mainly of dolomite and dolomitised limestone unaffected by rotting. This material displays a greater degree of angularity and size variation (20-300mm), which is considered to be indicative of extraction by blasting and/or pick and chisel work. The degree of sorting in this material is generally poor with varying size components of derived rock type. Quite often, fragments of limestones are noted, which is considered to be indicative of non productive mining through barren rock. Sometimes complete or fragmentary iron tools are associated with these spoils, so helping to further characterise the material as being 18-19th century in origin. A minor proportion of underground and surface spoils contain a clay component which is likely to be the product of rotting and hydration of mudstones/shales. Similar, but more reddened material with a noticeably clayey consistency is encountered in some surface spoils, this has probably been superficially derived from glacial drift.
Spoil deposits can easily be overlooked as an indicator for early mining, as they are often conveniently taken to be all of the one type. There is therefore as the above descriptions indicate, an opportunity to investigate these seemingly uninteresting materials further, to perhaps record quantitatively, identify and interpret many more varieties. These may not only categorise prehistoric and recently derived material, but also subdivide Bronze Age material on the basis of mining technique as related to the controlling geology, and in turn lead to deducing the relative ages of locations where the material originated.
Microscopic examination of spoil may also prove worthwhile for paleoenvironmental material, particularly where there is noticeable organic material content. One such unit identified during excavation of the main opencast was sampled for pollen analysis (section 4.2.3). Unfortunately the low count and poor state of preservation of the pollen due to the alkaline conditions meant that individual sub-species could not be identified. Never the less, seven species were identified (section 4.2.3), although these were no different to species presently found on the headland.
Careful examination of certain underground spoil of an assumed early origin has revealed a number of bronze fragments, these are described in section 4.3.3. Occasionally, small mollusc shells are found in areas of undisturbed spoil, predominantly where entrance passages have collapsed allowing surface derived material to be washed in and to accumulate (location 11 - Owens shaft). The morphology of mollusc shells can be a useful indicator of former climatic conditions at the time of the animals death. This affinity of climate sensitivity can therefore be related to known patterns of climatic change in prehistory and so help with the provision of additional ways in deducing relative ages of material surrounding the molluscs.
Calcite Speleothems
Another interesting feature of the early workings is the deposition of calcite, taking the form of stalagmites, stalactites and more typically of flowstone accumulations on the walls and floors of the passages (plate 20). These speleothems are known from several areas in the workings, notably locations 7a, 5, 22, 23, 34 and the route between 17 and 21. Occasionally they form impressive features coloured green-blue by copper-iron mineralisation. At two of the known locations (5 & 34) flowstone deposits consisting of a lower layer of calcite and and upper layer of aragonite are known. Normally aragonite, which is chemically the same as calcite (CaCO3) but with a different crystal form (calcite: trigonal, aragonite: orthorhombic) is metastable and therefore its mode of formation in the workings is considered to be unusual. This change in crystal form may be due to the presence of other ions that have influenced the formation of the flowstone.
In places the calcite deposits have plainly taken some time to form, attaining thickness up to 250mm (location 5, figure 8), and are seen capping spoil containing bone and stone tools and charcoal fragments. Calcite speleothems are also found in some of the typical 19th century workings where thicknesses up to 25mm are known to occur. A notable example is found in the deep workings accessed off the Penmorfa drainage level where an appropriately named feature known as the ‘ice bridge’ can be seen. The existence of speleothems can not be considered to always date from an early period, and therefore they must be considered with other evidence if an early age is to be ascribed to them.
Observation indicates that some of the earlier calcite formations were damaged last century as prospecting miners deliberately blasted them in search of ore, and other stalactites were broken as miners explored the areas of old workings. In some of these locations new calcite growth forming off the the broken points of stalactites can clearly be seen, and detached fragments are scattered over the passage floors. It is possible that study of calcite growth rates at these locations could help to indicate when this damage occurred. A few places have recently been discovered where formations are undisturbed, indicating that recent miners were not able to access such areas (location 7a). In one particular part of the mine between locations 33-34, a suspended floor of flowstone has been observed near the roof of what appears to be an early stope, and above it the walls are coated in similar flowstone. This is interpreted as being a deposit of flowstone that accumulated over the open walls and backfilled material in an early stope. Later prospecting by recent miners then removed the backfill leaving the suspended floor (situation 6, figure 27). There are indications of similar floors in other parts of this central vein towards Vivians shaft, but unfortunately many of these were destroyed by recent prospection. All that now remain are faint lines on the upper walls of the stopes and thin coatings of flowstone above this line, while below no flowstone is present, once more indicating that some considerable quantity of spoil must once have existed in these early stopes.
A means to investigate the age of the calcite flowstone using uranium series dating is currently being studied at Liverpool University (King forthcoming). This work has already supplied a number of dates, which in principle are comparable to C14 dates on material from nearby areas. At one particular location (7a) sample material has been submitted for both of the dating methods, the results being detailed in section 5. Further study within this and other fields of speleothem dating, such as laser luminescent analysis and paleomagnetic methods are also intended as part of a future programme of investigation.
Introduction
Details included here are supplementary to previous papers (Lewis 1990a,1994) , with updated information and interpretations from personal observations and reference to reports by Dutton&Fasham (1994) and Gale (1995).
In excess of 2000 stone implements of various sizes and shapes (plates 14-15, figures 29,30) are now recognised from the surface spoil and under ground workings. Attempts to devise systems of classification that accord with those already proposed for implements throughout the British Isles ( Picken 1988) have proved to be difficult. The main reason is that these systems have relied largely on degrees and types of modification. Unfortunately at the Great Orme only an estimated 5% of the implements display signs of such modification. This figure, however, could be increased if all the known tools from the site were cleaned and inspected more thoroughly, as was the case at Copa Hill where Timberlake (1990) was able to identify increasing numbers of modified tools after more detailed inspection. Another study (Gale 1995) on stone implements considers a greater range of samples from both European (including the Great Orme) and Near East sites, and attempts to classify these according to a wide range of parameters. In the case of the stone tools from the Great Orme this system is considered to be unnecessarily complicated, and a localised classification is therefore proposed, incorporating previous attempts to define the best means of differentiating between types (Dutton 1990 & Lewis 1988, 1990a). The distribution of the tools throughout the workings will also be commented on, with particular reference to the possibility of recognising distinctions in the use of the tools underground.
Classification
The most common (90-95%) implement from the site are unmodified generally ovoid shaped pebble, cobble and boulder sized beach derived stones, typically they are described as hammers, mauls, pounders and crushers. They vary greatly in size and shape from some of 50mm long, 0.25 kg in weight to others up to 400mm long and weighing 29kg (64lbs). Many are of a shape that would enable them to be easily held in a single (figures 29c-e, 30d-f) or double-handed fashion (figures 29h, 30a-c). A smaller number of hammers were of a more irregular shape questioning their use as hand-held tools because of their awkward form. These same tools displayed very little wear, supporting the view that shape alone meant that they were perhaps too unwieldy for use. Some of the smaller singly held hammers also have few indications of pronounced wear, and are suggested (Dutton & Fasham 1994) to have been used as percussive devices in conjunction with the broad ended bone chisels or gouges.
Nearly all tools display varying amounts of wear and damage through use, with only about 7% (Gale 1995) having no indications of any apparent wear. These unused tools have been recorded from excavated surface spoil and from the underground workings, with particularly notable occurrences of groups or caches (3-4 tools) of unused tools revealed in surface hollows at the base of the main opencast and another above the present tourist mine entrances. Similar occurrences of tools have been found at Rudna Glava, in former Yugoslavia (Jovanovic 1979), where they are considered to mark the entrances to workings. It has also been suggested (Budd et al 1992) that the occurrence of unused hammers may simply relate to the tools being mislaid prior to use. This is explained as a result of periodic activity, where miners obtained new supplies of hammers rather than seek those tools that were unused and forgotten about. If this was the case, then these tools are likely to have been lost either due to burial or becoming separated deliberately by backfilling or accidentally by collapse. If this argument is to be believed then it does seem unusual to find a number of hammers of this type in workings that are clearly accessible, while fewer are known from areas of backfilled spoil. Another interpretation might be that the tools were left as some ritualistic offering in gratitude for the ores that were removed from these particular locations, not unlike the similar offerings known from many 18th and 19th century mine sites throughout Europe (Bick pers. comm.).
A few (~5%) of the tools exhibited signs of modification to aid with hafting for handles or supportive slings. Where modification was observed it normally consisted of light nicking or pecking on either side of the midriff, or in exceptional cases, shallow grooving to the same positions (figures 29g, 30g). Modified hammers from the surface spoil deposits tend to be around 3kg in weight. No hammers with continuous grooving are known from the site. In the case of the large hammers or pounders, even though no signs of modification are present, it would have been difficult to handle such large tools and therefore some type of sling arrangement is likely to have been fitted, perhaps used in a similar fashion to the modern day ‘demolition hammer’.
The stones used as tools are typical of those found on local beaches, having been derived from glacial moraine of Irish Sea origin and fewer from the north Wales ice sheet. Rock types include predominantly igneous material, basalt, dolerite, gabbro, rhyolite, granite and a conspicuous number of microdiorite (or microtonalite), derived from the intrusion that forms Penmaenmawr headland some 10km distant across Conwy Bay. Here suitable specimens for hammer stones are found below and along neighbouring beaches from the headland. Microdiorite from the fine grained chilled margin of the same intrusion was exploited in Neolithic times for the production of stone axes at Graiglwyd and Dinas (Warren 1923). These axes are known (Clough & Cummins 1988) to have a fairly wide distribution through England and Wales, especially within Wessex, and also north-east Ireland and Scotland. Distribution routes such as these are likely to have remained in use after the Neolithic period, and may have aided with the establishment of potential trading routes for products originating from the mines of the Great Orme.
Sedimentary rock types are represented to a lesser extent, partly due to their lower strength and hardness. However, they do form a noteworthy part of the assemblage, represented by gritstone/sandstones, siltstones, silty mudstones and unusually limestone, once more suggesting that the ore host rock was soft or rotted enough to allow such hammers to be used. Only a small proportion (~10%) of hammer stones are composed of metamorphic rocks, the most common being quartzite.
Approximately 25 mortar stones have been recorded from the excavated surface spoil, but none have been found underground. These stones are generally around 300mm across and up to 100mm deep, with a central depression on both upper and lower surfaces (figures 29a-b). Clearly this hollow is the result of wear through a grinding motion on both surfaces rather than a pounding action. Identified rock types include gritstone, sandstone, dolerite and basalt. A small number of possible pestles that could have been used in conjunction with the mortars have been identified from surface excavations. These are commonly shaped like a flattened sphere (80-100mm diameter) with indications of wear around the circumference of the tool.
Larger dressing stones or anvils also exist within the surface spoil. To date about 12 are known, typically 400-600mm across and up to 250mm deep composed of mainly gabbro with some of dolerite and microdiorite. Wear on the flattened surfaces is consistent with heavy pounding. Stones of these forms have been recognised at various nineteenth century mining sites. However considering the scale of recognised Bronze Age mining at this site stones of this type would certainly have been required as part of the mineral processing. A secondary use of such available dressing stones in more recent times would also therefore be likely.
Distribution and Use
Of the stone tools found throughout the mine site, those derived from the surface tend to display the greater degree of wear and spalling, which agrees with the idea that surface spoil is regarded as waste and therefore any contained material will also fall under this category. Underground, the situation is different, with tools displaying less spalling and lesser wear. In general though, the stone tools from both surface and underground display wear patterns that are not consistent with a high degree of use. This is especially true when comparing those tools from the Great Orme with similar tools from other known sites where prominent wear through use appears to be the norm. It would appear that the Great Orme tools have experienced a more limited or different type of use, and this is considered to be the result of the rotted and softened nature of the ore-bearing rock. Unlike the other early sites where harder ore bearing rock types necessitated the use of stones as direct hammering devices with accompanying firesetting, ore rock at the Great Orme could have been removed with less effort, notably by bone tools. Therefore stone tools would have had a more limited use, partly as direct hammering devices but more likely as pounders and crushers for breaking detached nodules,fragments or blocks of ore-bearing material. The large size of some tools found in workings of restricted height and extent, also indicates such tools can only have been used for primary crushing. The size of the working would have limited the swing of the tool if it were to be used as hammer for striking the side walls, and the large size of the stone would have made it too unwieldy for the same purpose.
Where the practise of firesetting was applied in the workings, it is again thought that stone tools had a restricted use. This suggestion is largely based on personal observation during firesetting experiments conducted on the Great Orme (Lewis 1990b) at a 19th century trial in dolomitised limestone. During the exercise it was quickly realised that bone and antler were quite effective in removing the fractured fireset rock, and that the stone hammers when used to remove the less weakened material, quickly became abraded. The indicated wear from less than a few hours use was of greater extent than the wear apparent on those stone tools from the mine site.
Like the bone tools, hammer-stones produce characteristic pock markings on those surfaces where ore was extracted, consisting of slight almost circular hollows with remnants of powdered dolomite. However these markings are far less common in their distribution when compared with marks from bone implements. This again confirms the belief that the ore body was indeed easily worked, with the requirement for stone hammers only becoming necessary where ore-bearing rock approached a sufficient hardness to prevent removal by bone or antler tools.
Introduction
This section will describe, discuss and interpret the numerous pieces of complete and fragmented bone that characterise the early mine workings and associated spoil deposits at the Great Orme. Existing work (Lewis 1990a, Dutton 1990) will be referenced as well as more detailed studies of the bone material conducted by Hunt (1993) and Hamilton-Dyer (in Dutton & Fasham 1994). The overall purpose will be to consider all available material to cover such aspects as the preservation and origin of the material, anatomical and species compositions, classification and mode of use of the various forms and their distribution through the surface and underground workings.
In excess of 25,000 individual pieces of bone and several fragments of antler (plate 16) have been collected from mainly the surface spoil by the phased excavation over the past five years (1990-95) for the Great Orme Mines tourist development, and to a lesser extent from a programme of safety work at the site between 1988-89 (figures 31-32). The majority of the bone was derived during machine excavation of areas to the north and south-east of Vivians shaft, much of this material was collected by field watchers standing with the machine as excavation progressed. A smaller but more important proportion of the total material has been obtained by a series of detailed excavations conducted by Gwynedd Archaeological Trust (GAT) and Great Orme Mines Ltd (GOML). An equally important but far smaller quantity of material has also been collected from the underground workings, particularly during clearance of spoil from the tourist mine route and the large chamber or stope with which it is associated. A very minor amount of material has also been recovered from other areas of workings, this comprises bone removed for radiocarbon dating and other analytical studies, including those artefacts from the James (1990) excavations of the late 1970’s.
Preservation
Perhaps the most unique feature of discovered bone is its fine state of preservation,
which was also remarked on last century, when prospecting miners came across various bone “ to all appearance, as fresh, though impregnated with copper, as they were when denuded of their fleshy covering” (section 2.6.1, p40). This preservation is a result of the derived spoil from the limestone-dolomite host rocks that gives rise to neutral to slightly alkaline (pH 7-8) conditions, having the tendency to neutralise any acidification from the oxidation of the chalcopyrite ores. At many other known early mine sites the converse is true, with acidic (pH 2-4) conditions arising as a combination of a non-carbonate host rock and primary sulphide mineralisation. Such conditions do not favour the preservation of bone material, and it is likely that such material and even antler would be rapidly taken into solution. Therefore, when considering other early mine sites in non-carbonate host rocks it may only be possible to find evidence for bone or antler tools by searching for the characteristic tool marks made by these implements.
Present day observations indicates that a high proportion of the bone, approximately 95%, displays vivid colourations of blue-green due to copper (0.9%) and iron (0.5%) mineralisation and occasionally a darker speckled appearance due to manganese (1.6%) (Jenkins&Lewis 1989). These colourations only affect the surface of the bone to a depth of about 0.5mm. Analysis by XRD suggests this mineralisation occurs as impregnations rather than by a replacement process.
Occasionally paler coloured, less hardened and often deteriorating fragments of bone are found, notably in spoil which is encountered nearer to surface, as has been recorded from the trench workings excavated to the south-west of Vivians shaft. It would seem that this material has not undergone any surface mineralisation, and consequently is more prone to damage than the harder stained bone. This could be due to reduced amounts of copper (or iron, manganese) available for impregnation from the surrounding spoil, which itself is likely to be influenced by the covering thickness of spoil with available minerals, and also any localised changes in pH that will reduce the solution of these minerals. Differences in overlying lithologies could affect these pH values, for example more clayey material with associated organic remains could acidify groundwaters in the vicinity of the bone preventing impregnation mineralisation. The existence of oxidising sulphide ores in the same area may also add to this effect. It may also be reasonable to consider that the unmineralised bone may have undergone a process of initial impregnation, but following leaching conditions have caused resolution of this material.
Species and Anatomical Distribution
Previous investigations have considered whether the bone material has been derived from discarded food debris or whether they represent specific tools that have been deliberately fashioned to aid removal of the copper ores. Previous studies (Hunt 1993 and Hamilton-Dyer in Dutton & Fasham 1994) and present thinking agree that a combination of the two is more likely and that those fragments which display polishing and rounding of tips indicate wear through use as tools.
Animal species that comprise the bone assemblages include notably ox, domestic and wild pig and to a lesser extent sheep and goat, also antler of red and roe deer. The majority of the bone appears to be from selective parts of the animals, predominantly long bone (figures 31a,c-d,f; 32c-f) with fewer and decreasing numbers of rib (figures 31e, 32g-h), collar (figure 31b) and shoulder blade. Many show rounding and obvious wear at the tapered end while the opposite end has a shape that would enable it to be held in the hand. From this it can only be implied that the wear is consistent with these bones having been used as simple tools for gouging, chiselling and scraping. Most of the remaining bone have fractured ends, suggesting damage through use or deliberate fracturing to produce pointed and sharpened ends more suited for scraping.
Several pieces of human bone and teeth are known from the site, mainly from surface spoil of Bronze Age contexts. The most notable find is a lower jawbone of an adult male, located near the junction of west 3 and north 2 veins just north of Vivians shaft. The three remaining teeth in the jaw consisting of an incisor and two pre molars all displayed pronounced wear suggestive of a diet of coarse food. The jaw presently resides at the Great Orme Mines centre and as yet has not undergone any specialist examination. However, stray finds of two teeth found at the north side of the opencast have been inspected at the Dental Department Cardiff Hospital, who suggested they were likely to be from a male of between 20-25 years old.
Hamilton-Dyer (in Dutton & Fasham 1994) studied two samples of material, one from underground workings investigated by James between 1976-1981 and consisting of 45 components, and a second of 2444 components excavated by GAT from surface workings and associated spoil deposits excavated between 1987-1989. The predominant species identified were cattle (unspecified), with also a fairly high proportion of unidentified large ungulate fragments, particularly tibia, which are thought also to be of cattle. Other species were represented, notably pig, sheep, goat and red deer. Over 50 bones showed definite signs of modification, with wear consistent with having been used as tools. These consisted of cattle or large ungulate components, of which two general types were identified, rib gouges and limb bone chisels. The ribs are thought to have been used as scrapers or gouges as it is unlikely they were strong enough for being used as chisels for removing the harder material. The limb bones are however ideally suited to a chiselling action because of their increased strength, and this is partly confirmed by wear marks at the broader end of some tools suggesting percussion by a stone hammer.
The study recognises that there is a definite inequality in the distribution of the cattle and large ungulate components, with a concentration of ribs and major leg bones, particularly tibia, and only a small number of skull, foot and vertebral fragments. This distribution is interpreted as a disposal of major meat bearing elements, but with fewer fragments of the normal high meat bearing areas such as the pelvis and scapula. Skeletal remains from pig displays a wider distribution, with teeth and skull components implying the casual disposal of meal remains with some reuse as tools. Hamilton-Dyer infers that it is difficult to predict ages of any of the animals because of the fragmented nature of the material, but does conclude that the distribution and sizes are similar to other known Bronze Age sites such as that at Runnymede, Berkshire (Done 1980).
A second study by Hunt (1993) considered three samples of bone. These consisted of two of 150 and 131 fragments each from surface derived contexts and a third of 174 fragments from the large underground stope (location 18) accessed off the tourist route. Hunt also identified cattle as the predominant species, with a similar proportion of unidentified elements, mainly ribs and leg derived bone. These are suggested to be red deer or horse, or more likely cow, on the basis that this is the predominate species. Other species identified include pig, sheep, goat, red/roe deer, horse and dog. These identified proportions of cattle and sheep are broadly similar to other Bronze Age sites in Britain, for example Brean Down (Levitan 1990).
The same study identified butchery markings on a number of components of the three samples, however it seems that fragmentation is assumed to be related to tool use, rather than to patterns of fracture induced by marrow extraction or trampling. Hunt also proposed that certain carcass elements related to food use, for example, scapula, humerus, pelvis and femur; while tibia, ulna, radius and rib, particularly those of cattle were more important for tool fabrication. The noticeably large amounts of fragmented tibia and radius material also related to the process of modification for such fabrication. Other species such as pig, sheep, deer , goat and horse are assumed to have served as food material. Hunt suggested that the under representation of metapodials and limb extremities, if not caused by sample bias, may indicate a certain degree of butchery of the carcasses before bringing them to the mine site. This would seem logical if the assemblages are to be regarded as partly food debris.
A limited study by Dutton & Fasham(1994) identified similar bone assemblages to those above, and described two tools of cattle bone which appear to have been adapted for specific uses. One of these , a rib, has been modified at one end to produce a saw-like forked arrangement, the other, a long bone, is similarly notched with surface marks along one side resembling a ‘Y’ and ‘I’ (figures 31g-f), which could have identified a particular individual. Such notching may have aided the removal of ore bearing material or may represent some other, as yet unknown, purpose. If the notching was indeed to aid ore extraction then it would be reasonable to assume that evidence of similar modifications would be found on other tools, which they are not.
Distribution and Classification
The studies mentioned above indicate that specific elements of bone, particularly those of cattle were selected for use as tools. It was observed as part of this study that collected and in situ material from both surface and underground contexts agrees with this, although this is not based on any detailed investigation. What is obvious, however, is the disparity in the percentages of unmodified bone termed food debris and those bones that have clearly been used as tools. As would be expected on surface the proportion of actual tools is very small, particularly within general spoil material. This does, however, increase where spoil of a primary context is encountered as opposed to a redeposited context. For example at the base or filling small workings extending from the trench workings where locally derived spoil has been generated, the spoil is found to contain higher numbers of complete and damaged bone implements. Dutton & Fasham (1994) considered the distribution of bone throughout the surface spoil, and have suggested there may be a relationship between the various tool forms and the type of spoil material. It is proposed that bone is most common in extraction spoil as opposed to material from processing or prospecting, which is suggested to contain higher proportions of hammer spalls and correspondingly less bone.
Underground, depending on the size of the early working, a similar relationship can be identified, for example in the large stope (location 18) north of Vivians shaft, where fragmented bone is more common than tool form bone. At this location a large quantity of spoil has been removed for part of the tourist route. During excavation it was realised that the lower sections of spoil were in a primary context, while the more disturbed upper layers suggested redeposition of spoil that had entered the stope from the northern base of the opencast. The percentages of tool form bone from this area are therefore likely to be similar to those from the surface, as has been confirmed by observation. At other locations underground where workings of a more restricted size are known and where no direct connections to the surface exist, a different situation occurs. At these sites lower proportions of fragmented bone and higher proportions of tool form bone are known, although generally these areas have on average far less bone than the surface spoil.
Sometimes individual bone tools are located very near to matching tool marks that were identified as having presumably been made by these implements. These features have concave profiles with rounded points, consistent with the shape of the bone itself. Tool marks of this type contrast strongly with those made by metal tools, which have sharper tapered points and straight angled sides. In certain places almost an exact fit can be made between the tool and the markings, indicating that these areas have seen very little disturbance from later periods of mining activity. Typically the bone tool marks are seen in areas of rotted dolomite and the softened mudstones, and are generally most obvious at the junction of these rock types coinciding with the passage sidewall to ceiling. At one location, a small shaft to the west of Vivians shaft a collection of finely preserved tool marks (plate 17) occur together with finger marks in the very rotted dolomite.
Bone tools appear to be more obvious in the shallower workings, with a tendency to decrease in numbers as workings become of greater extent in both lateral and vertical directions. Distribution also seems to be partly governed by the surrounding lithologies and extent of associated rotting . In Owens shaft, for instance, an area of workings of typically early form (locations 10-11), has exploited material from a mudstone horizon approximately 0.5m thick, and underlain by rotted dolomite. A similar situation is encountered in the deeper workings (location 34) off the base of Treweeks shaft, where a series of partly collapsed workings in rotted dolomite beneath the main sandstone horizon are observed to contain far fewer bones than those workings nearer to surface. It is reasonable to argue that the deepest workings are likely to be the youngest, this is partly indicated by the decreasing age of radiocarbon dates from surface to underground workings, and also from the fact that the near surface workings would have to have been excavated before any deeper areas of activity. The general lack of bone implements in these areas, even though stone hammers and fireset charcoal are present, may then suggest there was a transition from bone implements to the use of metal tools. Such tools would also have been more efficient in dealing with harder ore bearing material encountered during the later stages of mining when all the softest parts of the vein had been removed. At present there is no direct evidence for the use of metal tools during the early period, primarily because it is not possible to differentiate those marks made by a bronze tool from those of an iron tool. However, two small wedges of bronze were reported from ‘ancient’ workings last century and numerous fragments of bronze have recently been discovered in a small working (location 21) off the present tourist route. Both sets of finds are detailed more fully in the following section.
It may even be possible that some bone implements located underground may have had some ritualistic significance, comparable to the offerings in historical mining to the "Knockers" or mine spirits (Williams 1979). Similarities might also be entertained with other Bronze Age sites, such as Irlingworth on the river Nene in Northamptonshire, where excavation of a round mound for cremation burials revealed a mass of cattle skulls, mandibles, scapula and pelvises that had been stacked above the grave pit (Parker-Pearson 1993).
An argument has been put forward (Briggs 1990) that the bones found within the workings simply represent those remains of food debris disposed of within the vicinity of the mine during some early period, and subsequently used by miners in more recent times as implements for removing ore. This would be very unlikely, because if bone had been in the ground for any period of time it would have undergone some deterioration even under alkaline conditions, and consequently become brittle and unsuitable for use as any form of tool. Evidence to support this view, includes bone tools with appreciable wear from C14 dated contexts and from beneath deposits of calcite flowstone dated by Uranium series methods. The suggestion that recent miners had spent time searching for suitable bone for such purposes is also highly improbable, especially when metal tools were available.
All of the various studies largely agree the animal species that constitute the bone assemblages are mainly cattle with decreasing quantities of pig, sheep/goat, red/roe deer, horse and rarer remains of dog. Generally more of the larger bone components such as ribs and limb bones, particularly tibia have undergone forms of modification consistent with use as tools or deliberate shaping for use as tools. In conclusion, it would seem that the samples studied indicate that the bone assemblages represent a considerable amount of food debris as well as the remains of tools and fabrication waste from tool manufacture. The bias of cattle bone for tool modification may only actually relate to decisions involved in the selection of suitable bone as mining tools with little or no modification necessary.
General
Documentary records last century describe at least three metal objects that were discovered by prospecting miners exploring the old workings. These were described by Stanley (1850) as picks of bronze, one about three inches in length and another one inch in length. They were found together with a broken stag’s horn in 1831, while the third item, a semi-globular object, was found during a separate event. Two of these items (figures 32a-b), the smaller wedge shaped object (32x20mm) and the ‘button’ shaped object (32mm diameter) survive today and are presently in the British Museum (acquisition Nos. 185310-31,8&9).
During 1988-89 excavations in a recently opened series of workings (location 21) about 20m below surface some fragments of a fairly dense material with surface coatings of blue azurite were found, which were thought to be pieces of ore. However on analysis (Jenkins & Lewis 1990) the material was found to be tin bronze with a noticeable amount of arsenic (sample MF, table 2). Additional analyses at the British Museum and Bradford University also confirmed this (samples BBA 1089 & BZ02, table 3). Further excavation during 1991-92 at the same location (Ref. No. 21) unearthed about 70 similar fragments generally between 3-8mm in size with irregular profiles and sharp edges reminiscent of brittle fracturing. These were evenly distributed throughout successive layers of dolomitic spoil of generally fine to medium grade, with fragmented charcoal (dated to 3000+/-50BP, BM-2641) and occasional complete and fragmented stone tools. No evidence of any slagging residues were observed, although bluish green corrosion products were noted on certain fragments.
At first the metal fragments were thought to be part of some damaged tool (Lewis 1990a) similar to the burred end of a chisel. This interpretation does not however, seem likely, due to the large numbers of fragments which would suggest that the tool would have to have been of a poor quality, and would have been totally impractical if deterioration was so rapid. A recent observation indicates that it is difficult to obtain the type of fracturing exhibited by the fragments through cold hammering, which tends to distort the metal. It is well known amongst present day bronze smiths (Chapman pers. comm.) that the best method of reprocessing scrap bronze is by heat fracturing, this entails heating the metal to a cherry red colour (500-6000 C) and then hammering the material. Deformation is like that of a soft ceramic, causing the bronze to fragment into smaller particles that can be reduced in size further by continued hammering of the hot metal. Bronze fragments produced in this fashion display shapes and fracture patterns very similar to the fragments from the working in question. It could be argued that the heat fractured fragments may be the result of firesetting, although this is very unlikely as they could have only be produced by hammering of the hot metal. The restrictive size of the working (plate 11) would have prevented any activity such as this, and so it seems more likely the fragments were deliberately placed. Perhaps it is more likely some ritualistic explanation needs to be sought, where the heat fractured bronze was placed here as a form of offering.
It is likely that the large numbers of metal fragments found from the one location described above are not the only instance of this type. Surely other locations await discovery. The main question these fragments raise is how extensive was the use of bronze tools at the mine. Sufficient evidence just does not yet exist to answer this, but it would seem conceivable that metal tools were used as harder ore bearing formations were encountered This would occur with increasing depth of working or with the need to exploit the sulphide ores that were overlooked during earlier stages of mining. Evidence of bronze tools at other known mine sites is limited, although two copper implements, a hammer-axe and an adze-axe, are recorded from Aibunar in Bulgaria (Cernych 1978). One possible explanation for the lack of metal tools may simply be that bronze was too valuable a commodity to be left underground, together with the fact that later miners were likely to have removed any tools for use themselves.
Analyses
The small wedge (BBA 1246) and button shaped object (BBA 1247) found last century, and one of the metal fragments (BBA 1089) described above from location 21 were analysed by atomic absorption spectrometry (AAS) at the British Museum (Craddock & Hook 1991). The results (table 3) indicated that the button had the composition of brass, and its iron content suggests it is very likely to have been made by the cementation process. Brass was unknown in Britain until the eve of the Roman invasion (Bayley 1990), and this piece could date from any time between then and when it was found. The remaining items are of bronze, having a relatively high tin content (10-15%). Both pieces are characterised by rather high arsenic (~1%) , nickel (~0.5%) and lead (0.1-0.5%) contents which might suggest a common source. These compositions are within the range of Bronze Age copper alloy compositions, and correspond to Northover’s Middle Bronze Age M2 group (Northover 1980a&b), although at this time it is still not possible to assign specific provenance on the basis of composition alone.
In 1992 two fragments (BZ01 & BZ02) of metal from location 21 , including the one (BBA 1089 = BZ01) earlier studied by the British Museum were analysed by Energy Dispersive X-ray fluorescence Analysis (EDXA) at Bradford University (Budd 1992). Both compositional analyses and metallographic examinations were conducted. Compositions roughly compared with those previously determined by AAS, even though EDXA is relatively less sensitive. Similar levels of tin and arsenic were recorded, but lead content differed, with a lower value <0.1% by AAS compared to 2.73% by EDXA (table 3). Whilst these figures may not be highly representative of the bulk lead content, metallographic and SEM examination of the samples clearly showed fairly large quantities of lead to be present in both samples. It was suggested (Budd 1992) that the AAS value for lead may be in error. It has been proposed (Staniasek & Northover 1982) that the level of lead is subject to considerable macro segregation in copper, and it is therefore conceivable that this could relate to sampling problems. Alternatively, if the items had been cast from common stock metal, it is possible the lead levels may relate to variations caused by segregation within the stock metal or even to the relative position of the casting in a sequence poured from the single melt.
The metallographic examinations of the later two samples (BZ01 &02) confirmed similar metal dendrite structures, with some corrosion products and numerous metal sulphide inclusions. Budd (1992) identified the absence of recrystallised crystals over a great majority of the samples and the presence of intersecting slip traces within many of the dendrite structures. This was assumed to indicate that the sample had been worked at a temperature lower than the recrystallisation temperature of the copper-tin-arsenic solid solution. The presence of isolated recrystallised grains with annealing twins at the centre of some of the copper-rich dendrites would suggest that a specific part of the structure must have reached its recrystallisation temperature at some point during the history of the artefact. Budd suggests that these descriptions indicate that any annealing responsible for the observed structures is likely to have occurred below 7000 C. The evidence seems to indicate that the metal fragments were originally part of a tool cast from molten metal which was subsequently heated and fractured by the process of heat fracturing as earlier described.
The bulk compositions of the samples, with about 11% tin, 1% arsenic, 0.4% nickel and with significant levels of lead, is fairly typical of the Acton Park tradition conventionally dated to between 1400 and 1250bc uncalibrated (Burgess 1974) and said by Northover (1980a, 1988) to account for the majority of Welsh axes from the Middle Bronze Age. The presence of sulphide inclusions in each sample suggests that they were produced from a sulphide ore, which further confirms the belief (Northover 1989) that such smelting technologies were commonplace during the period. To summarise, it appears the fragments have originated from the same artefact but the questions of why or how the fragments were distributed within the location, however remain , the subject of speculation.
Recent lead isotope studies (Joel et. al. 1996) on other metal fragments from location 21 also confirm the suggestion of a common source, but disagree with Budd (1992) concluding a strong likelihood that the fragments had been produced instead by copper carbonate ores from the Great Orme.
To date there is only evidence for the use of stone and bone tools and likely use of bronze implements at the Great Orme. Unlike for instance Mount Gabriel, Ireland (O’Brien 1994) where wooden prisers and gouges have been recorded, the Great Orme workings show no immediate signs for the use of wooden tools, even though the nature of the ore deposit would have lent itself ideally to the use of such implements. The most likely cause of this absence are the neutral to slightly alkaline groundwater conditions that prevail at the Great Orme which have not favoured the preservation of timber, although charcoal is well preserved. At Mount Gabriel, however, acidic groundwaters prevail, and have assisted in the preservation of timber in the waterlogged anaerobic conditions encountered here. Similar conditions also occur at a few of the other known early mine sties, notably the Mitterberg in Austria where numerous pieces of wooden tapers or spills were found. These are suggested to have functioned as illumination sticks held in the mouths of the miners (Pittioni 1951). Similar pieces of timber have recently been found at Copa Hill, Cwmystwyth (Timberlake 1994) and like those from the Mitterberg are suggested to have been used for lighting. Excavations at Copa Hill have also revealed a wooden launder over 3m long and 20cms wide which was lying in situ over a bed of sticks and loose stones at the entrance to a prehistoric working (Timberlake 1994).
The only indications of timber that do occur at the Great Orme are the impressions of wood preserved either in clay or calcite flowstone in certain areas of the mine. Particularly good examples of impressions were found during the main machine excavation of the opencast during 1990-1991 (near Location 7b). These consisted of voids up to 30mm long and 15mm wide within a mixture of mainly rotted and compacted mudstone, redeposited clays and some dolomite. The surface of the cavity was clearly textured as wood, with traces of carbonised material still adhering to the surface, particularly in the base where a thin film of material had accumulated. A similar arrangement was seen in some of the thicker deposit of calcite flowstone, notably from ‘Jackdaws Hole’ (location 7a). Here cavities with similar profiles, but of smaller size and devoid of any carbonised material are recorded. Some of these voids clearly indicate the surface texture of the original timber, and are typically situated towards the base of the flowstone where also fragments of charcoal occur. The combination of timber and charcoal would seem to suggest that all of this material is the residue of firesetting operations, with the timber fragments representing unburnt fuel, perhaps due to the wetter conditions encountered at this location.
A more detailed investigation of the bone tool marks and associated features found throughout the workings may reveal that some of these have in fact been made by wooden implements. Methods of in situ casting of these markings are presently being pursued using a cold cure moulding compound (sodium algenate). It may then be possible through microscopic examination of the moulds to differentiate between the features made by bone, wood or even stone implements.
Even though bone and stone tools are very common at the Great Orme mine site, it is remarkable that no pottery ascribable to the Bronze Age is known from either underground or in the immediate vicinity of the surface. This lack of pottery also characterises other early mine sites in Britain and Ireland and contrasts with many European sites, where ceramic evidence appears to be more common and has frequently been used as dating evidence, primarily due to the lack of other suitable datable remains. However, pottery is known from a number of locations in the confines of the headland and the local vicinity (Bibby 1979). The most notable is that from Kendrick’s cave (SH78008284, figure 28), situated on the east facing cliffs of the Great Orme overlooking Llandudno town; where shards of Food Vessel, Beaker Pottery and Peterborough Ware have been recorded together with older material of Upper Palaeolithic to Neolithic origin (Davies 1989). Similar pottery fragments, though of fewer numbers were also recorded (Bibby 1979, Stone pers. comm.) during an excavation of limited extent of the promontory fort of Pen-y-dinas (SH77908285). This hillfort commands a position overlooking Llandudno Bay, and within its perimeter are at least 70 hut circles, which are thought to date from the Iron Age period. A single piece of pottery has also been recorded (RCAHM 1956) from Llety’r Filiast burial chamber or cromlech (SH 77218295) situated 180m south of the mine. Unfortunately the shard and a piece of bone from the excavation earlier this century have now been lost. A number of potential sites from the Neolithic- Bronze Ages have been recorded over the Great Orme (Bibby 1979, Lewis 1993a). It is hoped that future study and excavation of these sites may indicate further evidence for occupation/activity during these periods on the headland.
Dutton & Fasham(1994) suggests that the absence of pottery at the mine may relate to the fact that the deposition and subsequent redeposition of spoil would fragment any ceramic material to a size requiring the use of sieving to distinguish it from the surrounding matrix. This is possible, but unlikely as areas of spoil of a primary undisturbed context do occur. These have been noted within the opencast, where horizons of early spoil still contained remarkably large fragments of charcoal, sometimes occurring as complete cross sectional segments of small branches 40mm in diameter. It can therefore be argued that these whole fragments have lain undisturbed since the time of their original deposition. Similar undisturbed deposits of spoil are also encountered underground.
The evidence for lighting in early mines at present is rather limited, with existing information giving only an indication to the methods available. At the Mitterberg mines in Austria, Pittioni (1951) describes numerous pine wood spills or splints that he suggested were used for lighting, as opposed to fuel for firesetting. Some of the used spills, indicated by their short length and carbonised ends, exhibited a feature of particular interest. These took the form of indentations at the unburnt end of the wood which were interpreted as teeth marks, indicating that these lighting sticks had originally been held in the miners mouths, so leaving their hands free to perform the tasks of ore extraction. Similar elongate chips of pine wood are recorded from the Mount Gabriel Bronze Age mines (O’Brien 1994), with some individual chips up to nearly 300mm in length. It is argued that the length of the chips and the resinous nature of the pine as opposed to hardwood, would enable them to be used either for lighting or as tinder for igniting the timber fuel for firesetting. O’Brien also observed from field evidence that pine may have been less common than the indigenous hardwoods, and therefore was selected specifically for these purposes. The limited quantities of pine derived charcoal in typical fireset charcoal residues also supports this argument. Pine splinters from bog wood were commonly used by poorer people in Ireland for domestic uses in recent times. Lucas (1954) provides several 19th century references to this practice in the Cork-Kerry region, where splints between 0.3-0.5m were used either singly for lighting indoors or tied in bunches for outdoor torches.
Early miners at the Great Orme must have had similar means of illumination to those mentioned above, especially when the extensive form and depth of the workings is considered. Lighting and methods of ignition for supplying lighting must therefore have been advanced to a sufficient level to allow for efficient working in these areas. Unfortunately, as described earlier, the ground water conditions at the mine do not favour the preservation of timber, instead only relict indications remain to give an impression of wood fragments that have long since deteriorated (see section 4.3.4). It is not unreasonable to assume that pine wood splints had a widespread use through the workings, even though no charcoal of this species has yet been identified. Present evidence for lighting is still somewhat circumstantial, and is represented primarily by the discovery of small heaps of elongate charcoal fragments found in pockets and cavities in the dolomite of passage sidewalls. A good example of this feature is seen in the workings (location 7a) off the east face of the opencast, where charcoal fragments up to 35mm in length occur, having been interpreted as burnt twigs or shaved pieces of presently unidentified wood . On site observation of this wood would seem to indicate that the fragments are in fact hardwood rather than softwood, and so suggest they may be more likely associated with firesetting than illumination. Stone or clay oil/fat burning lamps may have been another source of illumination, and could have provided an efficient means of lighting the deeper workings particularly where strong draughts or damp/wet conditions hindered lighting from wooden splints. No traces of any pottery remains are known from the mine workings, and the only candidate for a stone lamp is an object which is reported to have originated from the mine late last century and is presently in the custody of GOML. This consists of a flattened elongate oval shaped siltstone cobble (200x108mm) with a central shallow concavity (147x55mm and 12mm deep), which was found on the surface spoil tips and assumed at the time to be a ‘Roman’ oil lamp (plate 18). It is difficult to ascertain this is the case as the concavity is very shallow and would only have been able tocontain a small quantity of fuel, alternatively it may represent a cobble derived from the local glacial tills and mixed with spoil during clearance work to expose further areas of mineralised ground. One method of determining whether the assumed lamp ever held any oil would be to complete an examination of the concavity and analyse for traces of material it may have once contained.
During explorations of workings off Owens shaft in December 1987 a cylindrical earthen ware vessel (plate 19) was found in a short 19th century working immediately above another level (locations 8-11) with obvious indications of prehistoric mining. For a number of years the vessel was considered to have been a relic from last century, then more recently it was removed for identification purposes. The vessel is 237mm high and 190mm in diameter and on closer inspection was found to have been hand made as opposed to being made on a wheel. In form it displayed some similarities to pottery containers of the prehistoric period and resembled descriptions of earthen ware vessels used in prehistoric salt mines at Hallstatt in Austria (Jowett pers. comm.). These vessels were held in rope slings and used to carry hot charcoals into the salt mines to provide a method of igniting timber for firesetting or lighting.
