History of Research
Overview of Mineralisation
Skarns and Pegmatites
Greisens, Sheeted Vein Complexes and Breccia Pipes
Main-Stage Lode Mineralisation
The Nature of the Mineralising Fluids
The Cornubian Ore Province - a Review.
The Cornubian Orefield is the most intensely mineralised belt in the British Isles and it has been exploited continuously for over 3000 years (Penhallurick, 1986). Early legends of visits by Phoenician traders remain unsubstantiated, but later Greek accounts of trading for tin at the 'White Hill of Ictis' (St Michael's Mount, which 2500 years ago would have stood out in a flat-lying wooded plain close to the coast) in the 'Cassiterides' (Tin Isles) are generally accepted. Julius Caesar, writing in the first century B.C, speaks of tin production in Britain and it is likely that the mineral wealth of the island (and its strategic position on the Irish gold trade route) was an added incentive for the Roman invasion in 44 A.D.
Most of this early production came from placer deposits (Penhallurick, 1986). Surface exposures (particularly on the coast) of lodes were also worked, principally by opencast methods or by driving on lode into cliffs. These 'coffin' workings are still visible on the coast around St Just [SW370313], close to the workings of Geevor Mine [SW375345] and at Botallack [SW363335] (Noall, 1993; 1999). Underground mining seems to have started around the 12th century. The granting of royal charters in 1201 and 1305 (setting up the Stannary Parliament, with its independent taxation, legal and control systems) was of major importance to the tin trade, granting miners special privileges with regard to land access, prospecting and mineral extraction. The granting of these rights saw in a major phase of prospecting across the south-west, initially from the alluvial workings on the moorlands, out into the lowland valley floors and the discovery of lode outcrops from steam exposures and exploratory trenches. From perhaps the Roman period until the early 13th century, Dartmoor was the principal tin producing area in the orefield (Scrivener, 1982), and during the latter part of the 12th century it became the main source of the metal in Western Europe. During the early part of the 13th century Cornwall took over as the major producer (Cornish tin production rose to double that of Dartmoor, at around 500 tonnes per annum), a position it maintained until the closure of South Crofty Mine [SW668412] in 1998. Dartmoor's production steadily declined (reaching a peak of 285 tonnes in 1515) until the mid 18th century saw a revival of its fortunes (Scrivener, 1982).
Early workings were chiefly of alluvial and eluvial placer deposits, known as 'tin streaming'. The tin-bearing sands and gravels were dug out from the riverbed and banks and the heavy cassiterite separated by sluices and crude sediment traps. Underground or opencast workings, prior to the introduction of gunpowder, were worked by a combination of firesetting and manual extraction with picks and chisels. The cassiterite concentrate or rough ore was then taken to a smelter (known as a 'blowing house') to be further refined, in the case of the rough ore, by crushing (using water-powered sets of stamps) and hydraulic separation, and was then smelted. The molten metal was often poured into granite ingot moulds, some of which still survive (Penhallurick, 1986).
The introduction of gunpowder blasting by Bohemian miners (where they worked the tin deposits of the Erzgebirge) during the reign of Elizabeth I (Penhallurick, 1986), saw the rapid development of underground mining in Cornwall. Initially this was for tin, but the manufacture of brass and the use of copper in the national coinage during the 18th century saw this metal assume prime importance in the orefield. The discovery of large deposits of copper in the Camborne-Redruth and Gwennap districts in the late 1600's spurred on a further phase of exploration throughout the county, the chief focus of which was now the discovery of lodes, as the alluvial deposits were becoming increasingly exhausted. As production increased, the main barrier to extending the mine workings at depth was the position of the water table and the need to pump out excess water to keep the workings dry. This was overcome by the development of horse-driven, water-powered, and later, steam-powered pumping engines (Pryce, 1778). The use of steam power (building on the work of Newcomen, Boulton and Watt and Trevithick) revolutionised the mining industry and allowed deep mining to expand rapidly during the 19th century (Buckley, 1997). Steam engines were used to pump water, haul up ore (and, later, men), transport materials and drive sets of stamps. Cornwall became not only mining centre, but also a test-bed of new industrial and engineering ideas that were exported across the globe.
The 19th century was the heyday of Cornish mining. After a period of closure in the 1790's (when cheap copper ore from Parys Mountain on Anglesey almost wiped out the Cornish copper industry) the mines proliferated and during the century over 2500 mines were operated in the orefield as a whole (Alderton, 1993). Copper and tin were the main products of these mines, but considerable tonnages of other metals and minerals were produced (see Table 1 and Figure 1), particularly iron, lead (Douch, 1964), arsenic (Earl, 1993), manganese, zinc and tungsten. During the 1860's copper mining reached its peak (Dines, 1956) with production reaching 15,500 tons of metal; Britain supplied around 40% of world consumption (and was the largest producer). Production of tin reached a peak in 1870 of a little over 10,000 tons of metal (50% of world production). A ruinous fall in metal prices in 1866 (brought about by the discovery of new copper deposits in Michigan (U.S.A) and Chile; and tin deposits in Malaya) saw the mining industry go into a rapid decline with the closure of many mines and the emigration of thousands of miners and their families to the opening mining fields of Australia, South Africa, Mexico, North and South America and the Far East (Noall, 1999).
Figure 1. A map of the orefield of South-west England (after Dunham et al., 1978).
MINERAL OR METAL TONNES (approximate)
Table 1. Estimated total mineral and metal production from South-west England. After Dines (1956), Alderton (1993) and South Crofty PLC (1988-1998).
World metal prices became increasingly volatile and the industry in Cornwall and Devon became caught in a cycle of 'boom' and 'bust' with fewer mines surviving each crash in prices (Morrison, 1980; 1983). Relatively few mines survived until 1900 (when tin production had fallen to 2000 tons metal per annum; and copper to around 50 tons metal per annum) and after a brief rise in metal prices in the early 1900's saw prospects improve, the First World War and the loss of labour brought many of the surviving mines to the brink of ruin. Casualties in the immediate post-war years included the famous Dolcoath [SW660401] in 1920, and Carn Brea and Tincroft mines [SW667405] in 1921. By the Second World War only South Crofty Mine (Buckley, 1997) and East Pool Mine [SW673415] remained in the Camborne-Redruth District with Geevor Mine at St Just, Castle-An-Dinas wolfram mine [SW946623] north of St Austell and Cligga Mine [SW738538] at Perranporth.
East Pool Mine (Heffer, 1985) and Cligga Mine closed in 1945 and Castle-An-Dinas closed in 1959 (Brooks, 2001). A rise in metal prices during the 1970's (which saw tin eventually reaching over £10,000 per tonne) saw renewed prospecting in the South-West and the reopening of Wheal Jane Mine [SW771427] near Truro and the opening of Wheal Concord [SW723458] at Blackwater and Wheal Pendarves [SW645383] near Camborne. During this period the tin price was stabilised by the International Tin Council (ITC), formed by the main tin-producing nations, buying and selling metal on the London Metal Exchange to keep the price as high as possible. When Brazil and China (not members of the ITC) refused to be bound by any quota agreements and flooded the market with tin metal in October 1985, the ITC was unable to buy all the metal and ran out of money. Its trading was suspended on October 24th and the price fell overnight from £8,140 per tonne to £3,300 per tonne (Down, 1986). Wheal Concord and Wheal Pendarves closed in 1986. Geevor mine managed to survive, in a much reduced form, until 1991 and also in that year Wheal Jane closed. This left South Crofty Mine as the sole surviving mine in the Cornubian Orefield. It was hoped that the tin price would rise, but it fluctuated between £2,900 and £4,300 per tonne (averaging £3,400); with production costs of around £4000 per tonne the mine was continuously losing money, despite every effort to minimise costs. The mine eventually closed on March 6th 1998, bringing to an end some 3000 years of mining history. The mine was purchased and unabandoned in 2001; at the time of writing (2002) mining has not yet resumed and the venture faces an uncertain future.
Back to top of Page.
History of Research.
The early works of Carew (1602), Borlase (1758), Pryce (1778), Phillips (1814), Thomas (1819) and Carne (1822) are excellent first hand accounts of individual mines and deposits. By the early part of the 19th century scientific descriptions of lodes and deposits were beginning to be made. The works of Henwood (1843) and, later, Collins (1912) provide excellent descriptions of lodes, relationships between lodes of different orientations and the nature and variable mineralisation of later faults (crosscourses), in many famous mines of the time that were inaccessible by the early 20th century. MacAlister (1908) recognised that the granite intrusions and their accompanying mineralisation took place in three stages: (1) intrusion of granitic magma with accompanying thermal metamorphism of the country rocks close to the contact; (2) intrusion of quartz porphyry dykes along fractures and cleavage planes in the metasediments; (3) deposition of ores in both sedimentary and igneous host rocks. With various refinements this statement remains valid today.
The 1920's and 1930's saw the Cornubian Orefield used to test new theories of ore genesis and zonation. Dewey (1925) and Davison (1921, 1925a, 1927) developed a model of zonation (with tin/tungsten/arsenic mineralisation passing out into copper, then lead/zinc, and finally antimony/manganese/iron mineralisation) based on mono-ascendant single-pass hydrothermal fluids emanating from the Cornubian Batholith, giving rise to a concentric arrangement of mineralisation radiating out from the exposed granite cupolas (Figure 2).
Figure 2. A comparison of the zonation models of Davison and Dines (from Hosking, 1979).
Davison's model suggested that the focal points of this mineralisation were grouped around original high spots in the batholith roof and that the mineral zones formed a series of 'shells' parallel to the granite contact with a copper zone overlying and overlapping a central tin zone and itself overlain and overlapped by a lead/zinc zone. Davison cited the concentric zonal pattern of mineralisation around St Agnes [SW723507] as an example (though this is not as simple as it first appears, Sn lodes in the district dip mostly to the north, while Cu lodes dip south and are of a later phase of mineralisation), modified by denudation, of this original pattern (Figure 3).
Figure 3. The pattern of mineral zoning in the St Agnes District, used by Davison in the formulation of his theory of zonation (after Bromley and Holl, 1986).
Dines (1934) provided field evidence against Davison's model. He held the view that the various zones were considerably flatter than the granite contact (Geevor mine remains the classic example of this phenomenon) and that the higher the zone, the greater its lateral extent. Dines also suggested that the appearance of certain minerals in particular zones was temperature dependent and determined by the temperature gradient between the granite and the surface. He also noted the irregular distribution (particularly of tin) of mineralisation related to the position of cusps in the granite roof and coined the term 'emanative centres' to describe these focal points of mineralisation. His model, however, was still based on a 'single pulse' mono-ascendant premise, using the cooling granite as the only source of heat.
In 1956 Dines' exhaustive memoir The Metalliferous Mining Region Of South West England was published, it remains perhaps the single most important account of the mining industry in South-West England published to date. Hosking (1964, 1979) refined and expanded on Dines' model and took into account earlier phases of mineralisation that straddled the magmatic/hydrothermal boundary and pre-dated the main stage lodes. He also saw localised temperature gradients and wallrock interactions as more important than a regional temperature gradient related to magmatic emplacement. Hosking recognised seven depth/temperature zones characterised by distinctive assemblages of ore and gangue minerals and also noted characteristic wallrock assemblages associated with each zone (Table 2). In applying a temperature framework to his model he was greatly aided by the work of Sawkins (1966) and Bradshaw and Stoyel (1968) who, through the study of fluid inclusions, found that not only were there significant differences in the depositional ranges of tin, copper and lead/zinc mineralisation, but that each mineral species has its own, fairly restricted, temperature zone. He also tried to apply a time frame to the span of mineralisation, envisaging a 200 million year protracted episode running from the Permian to the Eocene.
Table 2. Hosking's model of the paragenetic sequence within the Cornubian Orefield (from Hosking, 1964).
During the 1960's structural studies at Geevor Mine (Garnett, 1962) and South Crofty Mine (Taylor, 1965) made important contributions to our understanding of the mechanisms of lode formation, as did a landmark paper by Moore (1975) on the origin of the lode-bearing fracture systems across the orefield. Further studies at Mount Wellington Mine (Cotton, 1972), Wheal Jane (Walters, 1988; Holl, 1990), Wheal Pendarves (Alderton, 1976), the St Just District (Jackson, 1977), the Tavistock District (Bull, 1982), Dartmoor (Scrivener, 1982), the Wadebridge District (Clayton, 1992) and South Crofty Mine (Farmer, 1991) looked at the individual deposits from mineralogical, geochemical and structural perspectives.
Moore (1982) argued that the pattern of W-Sn-Cu-Zn zoning above emanative centres could be explained in terms of a pattern of 'hot spot' geothermal circulation similar to that seen at Wairaki in New Zealand. While this model explained some of the pervasive alteration patterns seen in parts of the batholith (e.g., the St Austell Granite), it failed to take into account the textural evidence used later by Halls (1987, 1994) in his mechanistic approach to the formation of the lode system.
Complex models of polyascendant fluid phases and structural reactivation had now replaced the earlier models of Davison and Dewey. The focus of research shifted to the geochemistry of the ore fluids (Rankin and Alderton, 1983, 1985; Jackson et al., 1982) and the building of a geochronological framework for the timing and duration of mineralisation. The works of Halliday (1980), Darbyshire and Shepherd (1985, 1987, 1994) and Chesley et al. (1991,1993) favoured the view of a protracted history of mineralisation, spanning over 200 million years, that was advanced in major reviews of the orefield by Dunham et al. (1978), Bromley and Holl (1986), Bromley (1989), Jackson et al. (1989) and Willis-Richards and Jackson (1989). Similar studies by Chen et al. (1993) and Clark et al. (1993) point to a much narrower timeframe and a diachronous pattern of mineralisation across the batholith, reflected in reviews by Alderton (1993), and Scrivener and Shepherd (1998).
Back to top of Page.
Overview of Mineralisation.
Most of the mineralisation present in the orefield can be directly linked to the granite batholith in some way, although some deposits clearly pre-date the granite and a variety of syngenetic sedimentary and SEDEX origins have been ascribed to these (Clayton et al., 1990; Clayton, 1992). Deposits falling into this category include the manganese deposits of East Cornwall and West Devon and the stratiform Pb-Sb-Cu deposits of the Wadebridge district in North Cornwall. The Sn-W-As-Cu mineralisation for which the region is famous occurs in a variety of forms, but principally in high-angle fissure veins (lodes) in or close to the granites (Garnett, 1962; Farmer, 1991).
Mineralisation across the Cornubian Orefield can be divided into the following, chronologically arranged, groups: (1) pre-granite orebodies of sedimentary/sedimentary-exhalative type; (2) syn-granite intrusion orebodies - skarns and pegmatites; (3) early post-granite intrusion orebodies - greisens and sheeted vein complexes; (4) main stage polymetallic orebodies - Sn-Cu-As-Zn-Pb lodes and carbonas, etc; (5) late post-granite mineralised (Zn-Pb-Ag-Co) and unmineralised fissure veins - crosscourses.
Back to top of Page.
Deposits of this type fall into two main groups: (1) manganese/iron oxide and hydroxide deposits; (2) massive and vein sulphide deposits.
The manganese deposits occur primarily in the Lower Carboniferous chert beds to the west of Dartmoor, which extend into the Callington area of Cornwall, close to Bodmin Moor. A large belt of similar chert-bearing rocks extends across the north side of the St Austell pluton. The ore occurs as irregular, ramifying lensoid masses, intimately associated with chert and, sometimes, jasper; and consists primarily of pyrolusite, psilomelane and wad, often with a central core of the Mn silicate, rhodonite. Rhodochrosite and tephroite are also noted from some deposits, particularly that of Treburland Mine [SX218814] near Altarnun (Golley and Williams, 1995). Deposits within the metamorphic aureoles of the granites are dominated by silicates, suggesting silicification/metasomatism of the original oxides and hydroxides. The deposits, spatially and genetically related to lavas and tuffs and occurring as replacement bodies in the cherts (as at Greystones Quarry [SX344784] near Launceston), are irregular and patchy and none of them are very large. Between 1800 and 1907 58,660 tonnes of Mn ore was produced, of which 90% was raised by Chillaton and Hogstor mines [SX429812] northwest of Tavistock (Dunham et al., 1978).
A number of small ochre deposits have been exploited to the southeast of Bodmin Moor, and also in South Devon. These deposits appear to be chemical exhalative horizons within basic lavas and tuffs that have undergone hydrothermal alteration (Jackson et al., 1989).
Various other stratiform exhalative bodies have recently been discovered, particularly in South Devon and eastern Cornwall (R. Scrivener, pers. comm.). These include quartz-barite-ankerite bodies carrying disseminated and massive sulphides of Zn, Fe, Cu, Sb, Co, Ni and Hg. They are of interest as potential carriers of precious metals (Alderton, 1993) and may be more common than previously thought, acting as potential sulphur reservoirs for later stages of mineralisation. Quartz-barite bodies are also known to occur.
Vein deposits of argentiferous lead are known from the Combe Martin [SS564466] area of north Devon. The source of the lead and associated metals (Zn, Ni, Cu, Sb) is unknown, but U-Pb ages of 360 ± 30 m.y (Dunham et al., 1978) confirm a pre-granite origin for the deposits. Copper veins found in the slates to the south around North Molton [SS742295] may be of a similar age.
Massive and vein sulphides occur in the strongly deformed Middle Devonian to Lower Carboniferous metasediment/volcanic sequences of the Wadebridge-Camelford area (Trevone Basin) of North Cornwall. Two distinct metallogenic events have been identified (Clayton et al., 1990; Clayton, 1992), an early suite of stratiform/stratabound Fe-Cu-Zn deposits occurring in black, carbonaceous slates and Fe-Cu-Zn-Sb-Pb mineralisation occurring in brecciated and carbonatised metabasites. These are augmented by later vein-hosted Sb-As-Au and Pb-Zn-Ag mineralisation. The biogenic sulphides (Clayton and Spiro, 2000) of the stratiform/stratabound deposits were reworked during the Variscan Orogeny and underwent a series of metasomatic events. The later veins formed under conditions of brittle shear and uplift during the Late Variscan utilising locally derived high temperature metamorphic fluids (and using the existing deposits as a sulphur reservoir) and formed at depths of between ~5 km and ~2 km and temperatures of ~380oC and ~280oC.
The copper deposits within the Lizard Ophiolite also pre-date granite intrusion. These consist of irregular discontinuous masses of native copper lining joints and fractures in, sometimes brecciated, serpentinite on the Lizard Peninsula of South Cornwall. The copper is associated with minor chalcocite, malachite and cuprite, sometimes with a steatite/talc gangue (Dines, 1956). The deposits are irregular and small and were worked briefly between 1820 and 1845.
Back to top of Page.
Skarns and Pegmatites.
Contact metamorphosed and metasomatised calcareous sediments are known as 'calc-flintas' in the Cornubian Orefield. Although of considerable areal extent (particularly between the Bodmin Moor and Dartmoor granites) these rocks rarely contain large enough concentrations of metallic elements to be economically worked. Exceptions to this rule are found around the northern flank of Dartmoor, close to Okehampton [SX590950], where Belstone and Ramsley mines worked three sulphide-rich (Cu with As-Co) horizons, interbedded with cherts, up to 30 metres thick, in calc-silicate hornfelses (Dines, 1956). The rocks also carry up to 7 wt% Sn, but this is locked in silicates (Malayite and andradite garnet, etc) rather than as a sulphide or oxide (Alderton, 1993). The nearby quarry at Meldon, where Carboniferous cherts, slates, limestones and dolerite meet the granite contact, has a metasomatic assemblage (internally derived) dominated by garnet, vesuvianite, wollastonite and diopside (Edmonds et al., 1975). A second wave of metasomatism (granite derived) produced axinite and datolite with subordinate hedenbergite, andradite and Fe-wollastonite. Minor amounts of Cu, Bi, As, Co and Ni sulphides occur along with rhodonite-tephroite-spessartine-bustamite assemblages in Mn-rich horizons. Metasomatism and metamorphism of iron-bearing sands has lead to the development of magnetite-bearing ore, on the southeast side of Dartmoor, which was commercially exploited at Haytor Mine [SX772764] (Dines, 1956).
Skarns are also developed on the coast of Penwith around St Just where the granite meets the Devonian cover rocks. These are predominantly doleritic rocks with subordinate metabasic volcanics (Floyd et al., 1993) and interleaved/faulted sections of hornfelsed slates. Metamorphism, shearing and metasomatism, with redistribution of Ca and Fe, has lead to the development of complex garnet-magnetite-dipside-epidote-actinolite-tremolite skarns (Jackson et al., 1982). These are associated with axinite, spinel, tourmaline, hornblende, chlorite, apatite, calcite and sulphides (Van Marcke de Lummen, 1985) and a number of rare tin-bearing silicates such as stokesite (CaSnSi3O9.2H2O), wickmanite (MnSn(OH)6) and Malayite (CaSnSiO5). These sheared belts are found within 100 metres of the granite contact and may reach 15 metres thick and extend downwards beyond 400 metres below sea level (Jackson et al., 1982; Alderton, 1993). A second fissure controlled stage of metasomatism has been recognised giving rise to veins of garnet and other minerals up to 1 metre in thickness. These are particularly prominent in the area around Botallack (where some of the layered metasomatic rocks have been impregnated with cassiterite, e.g., Grylls Bunny Mine above The Crowns) Th data from Grylls Bunny gives temperatures in the range of 280oC - 380oC. An Ar-Ar determination of the age of the skarns at St Just gave a result of 274 Ma (Jackson et al., 1982).
The skarns were formed by the expulsion of Ca, Mg and Fe from the basic volcanics, followed by enrichment in Sn, Fe, B, Be, U, W, Ta, F and Cl, transported by fluids of magmatic departure (Alderton, 1993).
Beer and Ball (1986), examining the Sn and W contents of the killas around the granite plutons found consistent evidence of enrichment in these metals in metasomatic fronts radiating out from the granites. They commonly found a narrow exo-contact zone (some 20-50 metres from the contact) of moderate Sn and W values, succeeded outwards by a zone of high metal values (up to120 ppm+ for Sn) that may reach 500-800 metres from the contact. Beyond this point the values gradually fall away to background (4 ppm for W, 3 ppm for Sn). This metasomatic phase appears to post date extensive alkali metal metasomatism, which was responsible for the growth of feldspar in the hornfelses closest to the granite, with biotite forming further out. The feldspar zone (up to 50 metres from the contact) appears to have resisted the later metal-bearing metasomatism far better than the biotite hornfelses or spotted slates further out, thus explaining the dip in metal values close to the contact. This metasomatic dispersal of Sn and W is unrelated to the vein mineralisation that post-dates (and sometimes modifies) it and marks the expulsion of metallic pneumatolytic vapours and fluids possibly before the granite (carapace, at least) was fully below the solidus. The dispersion haloes formed during this event (and also for the pathfinder element bismuth; Ball et al., 1982) have also been observed above concealed cusps and ridges (e.g. Bosworgy [SW573331], near St Erth; Beer et al., 1975; Rollin et al., 1982) in the granite and have been used as a tool in mineral exploration.
Within the Lands End Granite metasomatic albitisation has produced areas that were extensively replaced by later phases of mineralisation to form large pods of disseminated ore, particularly at junctions between lodes, where the highly fractured ground is extensively replaced by cassiterite with copper sulphides, similar in morphology to the carbonas associated with the main-stage lodes (Thorne and Edwards, 1985; Jackson et al., 1982; Jackson, 1979).
Pegmatites occur in all the major plutons, often as small, discontinuous pods and lenses within the granites. Occasionally they will form larger continuous veins and schlieren or banded pegmatite/aplite/microgranite (Badham, 1980) veins (as at Tremearne [SW613267] and Megilligar Rocks [SW610267] on the margin of the Tregonning-Godolphin Granite). Some show evidence of forming in localised fluid-rich sections of the magma and are contemporaneous with the host rock, while others cut across the host granite and out into the country rock. Other pegmatites of stockschieder type lie along the contacts of adjacent granite intrusions within the main plutons (best displayed in parts of the St Austell granite). Many of the pegmatites carry no metallic minerals, but may occasionally be worked for industrial minerals, as at Trelavour Downs [SW960571], west of St Austell (phlogopite mica) and at Tresayes [SW995585] (a 45 metre wide pegmatite, worked for potash feldspar) north of St Austell (Dunham et al., 1978).
Metalliferous pegmatites (Plate 1) often carry arsenopyrite, lollingite and wolframite with occasional minor cassiterite. They may be associated with tourmaline and can carry an extensive suite of accessory minerals including zinnwaldite, apatite, chlorite, stilbite, stokesite, topaz and triplite (iron phosphate, found at Megilligar Rocks). They may occur as isolated pods, e.g. Rinsey Cove [SW593269] (Hall, 1974), or as larger sets of veins that may be economically worked, e.g. Buttern Hill [SX188822] W-bearing pegmatites on the NE edge of Bodmin Moor and the W-bearing pegmatites at Cligga Head (Dines, 1956).
Plate 1. Wolframite-bearing, quartz-feldspar-tourmaline pegmatite. No:4 lode footwall drive, 400 fathom level, South Crofty Mine (metre rule for scale).
Pegmatites are fairly rare in the Cornubian Orefield and this may be due to extensive fracturing towards the end of crystallisation, which saw fluids escape into the lode fracture system, rather than accumulate within the granite. Th data from mineralised pegmatites falls into the range 250oC to 460oC (Alderton, 1993), whilst the K-Ar date for the Halvosso Pegmatite on Carnmenellis is >285 Ma (Halliday, 1980), which would place it in the latter stages of the cooling of the host granite, post-dating the outer Carnmenellis Granite and its mineralisation (Clark et al., 1993).
Within South Crofty Mine (Plate 2) a large number of sub-horizontal stacked veins (inclined from 0o to 20o and varying from a few centimetres to 1 metre in thickness), termed 'quartz floors', carrying an assemblage of quartz feldspar fluorite wolframite arsenopyrite with rare stannite and occasional later veining by Cu and Fe sulphides (Taylor, 1965; Farmer, 1991; LeBoutillier, 1996), occur in cylindrical zones extending for over two hundred metres in dip height. Although not true pegmatites (they appear to be tensional features infilled with a combination of magmatic and hydrothermal fluids, related to internal shearing within the still-plastic granite) they have a pegmatitic mineralogy and appear to straddle the boundary between purely magmatic and hydrothermal processes.
Plate 2. Stacked 'quartz floor' pegmatite veins. Main Lode drive, 290 fathom level, South Crofty Mine (metre rule for scale).
They have in the past been considered as potential economic targets (Bratley, 1978) for their tungsten content, but were not worked for such in the latter part of the history of the mine. Steeply-dipping lodes with an identical mineralogy also occur in the mine (Roskear Complex Lode, Complex Lode, Care's Lode and North Pool Quartz Lode, which is faulted and slightly deflected by an elvan dyke) and pre-date the later tourmaline-bearing tin lodes. A similar W-bearing lode occurs at Castle-An-Dinas [SW945623], at Goss Moor, near St Austell, which is cut by a granitic intrusion at depth (Beer et al., 1986; Brooks, 2001).
Back to top of Page.
Greisens, Sheeted Vein Complexes and Breccia Pipes.
Greisen-associated mineralisation is well represented in the Cornubian Orefield and can be broken down into three main groups:
1) Endogranitic sheeted vein systems.
2) Exogranitic vein systems.
3) Pervasive bodies.
The build up of magmato-hydrothermal fluids, in the apical sections and cusps of the individual granite plutons, during the final stages of magmatic crystallisation eventually lead to hydraulic fracturing of the granite overlying these fluid reservoirs. The resulting metasomatic reactions, along the walls of the fractures, with the pneumatolytic fluids resulted in the formation of greisen bordered vein systems within the granite and sheeted vein systems in the overlying and adjacent killas.
Endogranitic vein systems are found throughout the province. The best examples are those of Cligga Head, St Michael's Mount [SW514299], Hemerdon [SX572580] (Dartmoor), Goonbarrow Clay Pit [SX007585] (St Austell) and Bostraze Clay Pit [SW383317] (Lands End). These systems are characterised by presence of numerous sub-parallel quartz veins, typically 1-10 cm in width with strike lengths that may reach in excess of 100 metres (as at St Michael's Mount; Wheeler et al., 2001), with distinct greisen (essentially muscovite, sericite and quartz (both primary and secondary 'globular silica') minor tourmaline) selvedges. These tend to have very sharp boundaries with the host granites and typically reach from 8-30 cm in width. The quartz veins may carry an assemblage consisting of tourmaline, muscovite, orthoclase, chlorite, wolframite, arsenopyrite, stannite (kesterite, Zn-stannite and stannoidite), cassiterite, topaz, apatite and minor amounts of Fe, Cu, Mo and Bi sulphides (Hall, 1971; 1974; Alderton, 1993).
The depositional textures seen in some of these quartz veins show that crystallisation took place in a remarkably static and stable environment (Halls et al., 1999). The quartz crystals form interlocked and bridging arrays with frequent vugs. Within these vugs ore minerals have nucleated without showing any tendency to bilateral symmetry; bunchy aggregates of wolframite (occasionally bridging some of the veins) and prismatic (occasionally twinned) zoned cassiterite are typical. The textures show that the rate of fracture opening was greater than the rate of mineral growth and that nucleation was governed by diffusion in a more or less static fluid at near-equilibrium saturation (Halls et al., 2000). Other veins show quartz arrays bridging and, often, filling the fractures. The associated ore and gangue minerals behave in a similar fashion. Growth may be normal to the walls of the fracture (simple extensional dilation) or oblique to the walls (a combination of shear and extension). In this case mineral growth took place at a greater rate than fracture opening. Occasionally veins of this type also show more than one phase of opening, due to renewed extension after crystallisation of the first phase of infill.
The greisen veins formed as arrays of parallel extensional fractures in the upper parts of their respective cupolas where the pressure of segregated magmatic fluids exceeded the tensile strength (and minimum principal stress) of the rock. Each fracture was propagated laterally and vertically for as long as those conditions were met, and each fracture acted as a separate semi-closed section of the system as a whole. There was no large flow of water through the vein system, the mineralisation in the fractures appears to have been supplied by the immediately surrounding granite; water:rock ratios prevailing during mineralisation were low and physiochemical conditions were stable (Halls et al., 1999, 2000). Fluid inclusion data (Jackson et al., 1989) give a temperature range for the greisenising fluids of 250oC - 500oC, variable salinities (5-40 wt% NaCl) and high pH (3.4 to 4.9).
Of these deposits only that of Cligga Head (see Plate 3) has been commercially worked, producing 300 tons of wolfram concentrates and 200 tons of black tin during World War II (Dines, 1956), though the Hemerdon deposit (SW Dartmoor) was almost brought into production (reserves of 40 million tonnes at 0.18% WO3 and 0.026% Sn; Thorne and Edwards, 1985) prior to the tin crash of 1985.
Plate.3. Parallel greisen veins, Cligga Head. The granite between the veins is now extensively kaolinised. The veins consist primarily of quartz ± tourmaline, with a range of accessory minerals that may include cassiterite, wolframite, scorodite and rare blue topaz (hammer, 40 cm, long, for scale).
Several exogranitic sheeted-vein systems have similar characteristics to those of the granite-hosted deposits, both in terms of their chemistry (both have micas with high Li, Rb and F contents; Alderton, 1976) and mineralogy. These deposits consist of numerous, closely spaced, sub-parallel, narrow (1-2 cm) quartz veins. These veins carry an assemblage of muscovite, chlorite, cassiterite, topaz, tourmaline, stannite, wolframite, chalcopyrite and rare native copper. Wallrock alteration is of limited width and consists primarily of haematisation with the development of muscovite and tourmaline selvedges.
Good examples of this style of mineralisation include Wheal Prosper [SX030642] and Mulberry Openwork [SX020658] (both north of St Austell; Foster, 1878; Bennett et al., 1981) and Redmoor [SX355710] (near Callington). It is thought that these and other, poorly-mineralised (such as Magow Rocks [SW581422] at Godrevey, near Hayle; Bromley and Holl, 1986), deposits overlie endogranitic systems in the granite at depth and point to cuspate areas in the granite roof.
Deposits like Mulberry and Wheal Prosper were worked as bulk tonnage - low grade deposits; Mulberry alone produced over 1 million tonnes of ore, at ~0.4% Sn. Investigation by the BGS has shown that these two deposits still contain significant reserves (Bennett et al., 1981), though in the current economic climate they are very unlikely to resume operation.
The Redmoor deposit, near Callington was investigated in the early 1980's and also came close to being brought into production. It is situated in pelitic sediments and crossed by a number of main-stage lodes that form enrichment zones at the intersection with the sheeted veins. The deposit contains significant amounts of stannite, along with cassiterite, wolframite and arsenopyrite, with a host of accessory minerals including bismuthinite, pyrrhotite, chlorite and fluorite. The sulphide assemblage overprints an earlier Sn-W-As assemblage associated with the emplacement of the complex (Thorne and Edwards, 1985). In common with the endogranitic systems, these deposits have all undergone pervasive tourmalinisation of variable intensity (Bromley and Holl, 1986), affecting the wallrock between the veins.
Pervasive greisenisation is best exemplified by the apical outcrop of the St Agnes granite porphyry at Cameron Quarry [SW704507] (Hosking and Camm, 1985; Bromley, 1989), close to St Agnes Beacon. This exposure shows the roof zone of the granite (marked by aplite-pegmatite and leucogranite 'sheets') and the killas contact. The porphyry (consisting of large phenocrysts of alkali feldspar and quartz; smaller phenocrysts of plagioclase and biotite; set in a fine-grained groundmass of quartz, feldspar, biotite, muscovite and tourmaline) is progressively greisenised away from the contact and converted into a mass of quartz and fine-grained white mica. Voids left by dissolved feldspars were later infilled with pseudomorphs of cassiterite, wolframite and various sulphides. In this instance fluid pressures were not high enough to cause fracturing, the fluids passing along grain boundaries and metasomatising the rock en masse.
Breccia pipes are an expression of this early phase of mineralisation (see Figure 4) formed under unstable conditions, unlike the greisen systems described above. A number of breccia pipes are known to exist (Goode and Taylor, 1980) throughout the orefield, but the best example is that of Wheal Remfrey [SW925576], in the western lobe of the St Austell Granite (Allman-Ward et al., 1982) close to Fraddon Down. The complex is hosted within coarse megacrystic granite and consists of a tourmalinite bordered sheeted vein system and an elongate N-S trending tourmaline-dominated hydrothermal breccia body, measuring 500 metres by 40-100 metres (Bromley, 1989).
Figure 4. Major types of primary tin deposits in Cornwall (after Hosking, 1969).
The breccia body is broadly pipe-like, with steep contacts. It consists of irregular disordered clasts of granite, porphyry, elvan, tourmalinite and pelitic (it is estimated that the current exposure level is only 200 metres below the original granite roof) material that show evidence of transport and rotation. Towards the margins of the body the clasts are not rotated and appear to have travelled very little distance; beyond the breccia body, the host rock is cut by anastomosing vein and veinlets of tourmaline and tourmalinite.
It is thought that the pipe formed when fluid overpressures initiated a fracture system, similar to that in the formation of greisen veins, however, instead of the fracture ending at a point determined by the stress conditions, it intersected a lower pressure regime and the sudden pressure loss to the system as a whole resulted in explosive decompression. This resulted in violent decrepitation under positive pore fluid pressures and the development of a chaotic fluidised system. It is thought that the trigger for this mechanism was the crystallisation of tourmaline. Boron-rich melts can contain high levels of dissolved water; with the crystallisation of tourmaline the water vapour pressure sharply increased and contributed to the increased stress that lead to fracture (Jackson et al., 1989).
The breccia body shows evidence of very rapid nucleation and crystallisation, with clasts close to the pipe walls trapped in position, adjacent to their point of origin. Some clasts appear to have collapsed into the body from above (slates, etc from the contact), while others have been brought up from depth. The main sense of movement appears to be in from the side walls due to the shock wave on decompression. There is evidence of more than one decompressive event, with rapid crack-sealing of the fracture followed by progressive pressure build-up and repeated failure along the fracture zone once the tensile strength of the rock is exceeded (Allman-Ward et al., 1982; Bromley, 1989).
There is little metallic mineralisation associated with the Wheal Remfrey breccia pipe, although other tourmaline-dominated bodies may carry cassiterite and rutile. They tend to be depleted in chalcophile elements, even with respect to background values (Bromley and Holl, 1986) and appear to be wholly derived from magmatic fluids.
Back to top of Page.
Main-Stage Lode Mineralisation.
The typical lodes of the province are steeply dipping (most >70o) monolithic veins which are concentrated along the axis of the batholith and closely associated with elvan dykes (Jackson et al., 1989). The lode system as a whole has produced almost all the metallic output of the orefield and has produced not only tin and copper, but a range of metals including tungsten, iron, lead, zinc, silver, etc (see Table 1). The origin of these metals is still in debate; the tin (and tungsten) is proposed to have been derived from tin-rich sediments or protolith at the point of anatexis, though some authors point to the possibility of derivation from the mantle (Hutchison and Chakraborty, 1979). Though this seems less likely than a crustal source, Shail et al. (1998) have found traces of mantle helium in fluid inclusions from the orefield, attesting to some mantle involvement in mineralisation. The origin of the Cu-Zn-Pb mineralisation is thought to be due to a combination of xenolith assimilation and hydrothermal leaching of basic rocks (and pelites); it has been calculated that the volume of basic rocks and their copper content could easily supply the amount of copper extracted in the province (Jackson, 1979).
Though on the scale of the orefield, the lode system is extremely complex; within localised areas a number of fairly simple relationships can be established. Mineralogically (as a rule) the lodes show decreasing complexity with depth. Close to surface they are truly polymetallic and may show a mixed oxide/sulphide assemblage that, in many cases, has been modified by supergene activity (see Figure 5) to give a large potential list of secondary minerals. These include secondary sulphides, hydroxides, oxides, sulphates, arsenates, carbonates and native metals; many of these were first described in Cornwall and the area has been the focus of professional and amateur mineral collectors for centuries.
Figure 5. A section through a typical Sn-Cu lode, showing the relative position of the gossan, supergene and primary sections and the zoning seen in some of the major structures in the Camborne-Redruth District (from Hosking, 1988).
Within this near-surface zone the ores of tin, arsenic, copper and zinc were worked (though often in separate stages, depending on the economics of the time); many mines also produced minor amounts of lead and silver (Dolcoath) and occasional U, Fe, Bi, Mn, Ni and Co (Wherry Mine [SW470294], Penzance) ores. In the Camborne-Redruth and Gwennap areas (as elsewhere) the lodes close to surface were dominated by copper mineralisation (Dines, 1956). In the supergene zone original simple sulphides (chalcopyrite, pyrite) were replaced by malachite, azurite, tenorite, cuprite and native copper. Rare secondaries, such as olivenite and liroconite, etc, are known from the Gwennap area and Porthleven area (where Cu and Pb ores occur together). Below the oxidised zone secondary chalcocite, bornite, enargite and covellite were deposited before passing back into primary sulphides below the water table. These very shallow (often < 50 metres) rich deposits were mined at an early date (1700 onwards) and the high financial returns gained were responsible for the proliferation of mining activity across the orefield during the 18th century.
With increasing depth the zinc and lead mineralisation died away leaving a zone of simple sulphides dominated by copper and arsenic (Collins, 1912; Dines, 1956). As the granite/killas contact was approached tungsten became locally important, reaching its greatest development immediately below the contact (e.g. Rogers Lode, East Pool Mine; Dines, 1956). Below the contact copper declined and tin became increasingly important, and at depth (~500 metres from surface) cassiterite is often the sole ore mineral present.
This change from a simple oxide-dominated assemblage at depth, passing into mixed oxide/sulphide assemblages close to the contact and complex polymetallic assemblages at surface was the foundation for the theories of hydrothermal zonation formulated during the early 20th century. However, the relationship between the various phases was not always as clear cut and within a single lode there is often evidence of a protracted history of mineralisation, brecciation, shearing and further mineralisation, that negates the idea of mono-ascendant fluids. Most lodes do not show a single continuum of pressure/temperature-controlled mineralisation, they show a series of punctuated events, with the later sections of the lodes showing lower temperature assemblages. In this way some lodes initially worked for copper may later have had the walls of the existing stopes reworked for their tin or tungsten content (e.g. the North Tincroft Lode of South Crofty Mine; LeBoutillier et al., 2000a; 2000b; 2001).
The gangue minerals associated with the ores also vary with depth and are also temperature dependent (see Table 2). At depth (associated with tin ore) the main gangue minerals are tourmaline (fine-grained, powdery to flinty, Prussian blue to dark blue) and quartz. At higher elevations, lower temperatures and in lower energy environments (in areas reactivated by further fracturing) this gives way to a chlorite-dominated (Taylor, 1965; Farmer, 1991) assemblage (though initially in places still retaining a proportion of tourmaline) with quartz and fluorite. Lower temperature phases are dominated by quartz, siderite, fluorite, marcasite and rare calcite. While this trend to lower temperature mineral assemblages and lower energy environments over time and proximity to the surface is broadly correct, recent studies have shown that some shallow deposits can also be tourmaline-dominated and that some of the chlorite assemblages record violent brecciation events with clasts transported considerable distances ( LeBoutillier et al., 2000a; 2000b; 2001). This again suggests a series of punctuated mineralisation events, utilising fluids from a variety of sources and under a variety of physiochemical conditions.
Many lodes show a complex interplay between tectonically-driven episodes of mineralisation and remobilisation of constituents by convecting hydrothermal fluids. This last particularly applies to copper and uranium mineralisation (pitchblende and coffinite are themselves a late-stage infill in some lodes, e.g., No4 Lode at South Crofty Mine; Cosgrove and Tidy, 1954), which, in many secondary phases, are highly mobile and readily dissolved. An environment in which a cyclical system of pressure/temperature changes occurs may see the deposition of the rare fibrous form of cassiterite known as 'wood tin' (Hosking et al., 1987), if the fluids are supersaturated with respect to tin.
Often in contrast to the mineralogy (particularly at depth) the structural history of many lodes is complex and shows a series of brecciation and shearing events responsible for depositing a variety of individual assemblages in the lode over time (Plate 4).
Plate 4. The NPQ Lode, 390 fathom sublevel east, South Crofty Mine. The lode is one of an en echelon series of quartz-feldspar ± wolframite ± arsenopyrite ± fluorite ± chlorite early 'pegmatitic' lodes (showing dilational 'banded' and mylonitic shear textures) that has been reactivated during main-stage mineralisation with a central (dark) brecciated core of tourmaline, massive cassiterite and quartz (hammer, ~30 cm, for scale).
Brecciation textures are common in lodes in the deeper workings of many mines (Dines, 1956) and occasionally at surface (Dunham et al., 1978; Goode and Taylor, 1980). In the deeper workings of South Crofty Mine many lodes showed an early tourmaline ('blue peach')/quartz cassiterite breccia with a cassiterite/quartz cement (Plates 5 and 6). This was sometimes followed by other brecciation events, but was more often followed by further lower energy dilational episodes, giving the lode a banded appearance (some of these bands were, occasionally, microbreccias, emplaced within the lode), particularly along the hangingwall. Some of the reactivation episodes lead to the deposition of later chlorite-dominated assemblages, while other events lead to fine fracturing across the lode and the deposition of low temperature chalcedony-marcasite-siderite assemblages.
Plate 5. The Dolcoath North Lode, 380 fathom level, South Crofty Mine. The lode is predominantly composed of brecciated blue peach and quartz with minor (<1%) cassiterite. Some later lode-parallel shears carry minor fluorite and haematite. The wallrock adjacent to the lode contacts is irregularly tourmalinised. The width of the lode is ~1 metre.
This brecciated texture is due to hydraulic fracturing and explosive decompression, similar to the mechanism seen at Wheal Remfry (Halls, 1987; 1994; Halls and Allman-Ward, 1986). The majority of lodes appear to be extensional faults and would have communicated with areas of lower pressure, which were accessible during movement, along their dip-length. With the loss in pressure boron-rich (or silico-cassiterite) fluids, that had previously been building up pressure in the fluid reservoir, were suddenly released and travelled upwards as wallrock pore fluid pressures caused spalling off of fragments along the sides of the lode fracture. It is difficult to ascertain the amount of transport that took place; lode textures are fine-grained and indicative of very rapid nucleation and a number of clasts appear to have moved very little distance before being arrested in the crystallising fluid. Occasionally some clasts show evidence of entrainment (rounded, rolled clasts with 'debris trails' in their wake), but again the distance travelled cannot be quantified.
Plate 6. The NPB2 Lode (east drive) 400 fathom level, South Crofty Mine. The blue peach lode carries a series of internal quartz-filled shears. The lode cuts an elvan dyke and earlier quartz floors. Dip-slip slickenlines and lode orientation suggest that the lode occupies a normal fault with downthrow to the south (right); though in all likelihood the area has undergone a complex series of movements prior to the formation of the final set of slickenlines. Hammer for scale (30 cm).
The same cannot be said of the 'breccia lodes' that outcrop (and subcrop), most commonly, in the Gwinear District southwest of Camborne, along the line of a buried granite ridge that appears to be an extension of the Carn Brea ridge between Camborne and Redruth (Beer et al., 1975; Goode and Taylor, 1980; Taylor and Pollard, 1993). These bodies consist of rounded pebbles and cobbles of a variety of rock types (including granite, up to 1 metre in size, though the contact is some 750 metres below surface) set in a matrix of comminuted rock fragments of various sizes. The bodies are chaotic and disordered and in places the clasts (primarily slate) are so finely ground that the rock appears similar to a sandstone in texture, indicative of violent, high-energy emplacement (Clark, 1990). The walls of these fractures are often scoured and polished by the passage of material and small breccia fragments are found forced into cracks in the walls. Some of these breccia bodies carry (later, infilling) chlorite, cassiterite and chalcopyrite in the matrix and were mined from surface as early as Tudor times (e.g. Relistien Mine [SW601368] at Wall, near Gwinear). The lode at Trevaskis Mine [SW607378], nearby, is in excess of 10 metres wide and lies between intensely brecciated and silicified wallrocks of metadolerite. The ore consists of angular slate clasts cemented by chlorite and fine rock fragments. Within this are veins of quartz carrying a chlorite-arsenopyrite-chalcopyrite-chalcocite-cassiterite assemblage and also a chalcopyrite-sphalerite-galena assemblage that may be later.
Both Relistien and Trevaskis breccias carry clasts of internally brecciated elvan and other examples are also closely associated with elvan dykes. It appears that the elvans and breccia bodies are roughly contemporaneous; clasts of elvan moulded around other clasts found at Trevaskis (Goode and Taylor, 1980) suggest that still-plastic elvan was utilising the same fracture pathways as the breccia lode at close to the same time. In other examples host slates were brecciated before the intrusion of an elvan dyke. These fluidised explosion breccias appear to have formed under conditions of very high pressure where the system had instantaneous connection with areas much closer to surface than other lodes in the district. Rapid boiling and the development of a gas-fluid medium, similar to that seen in volcanic breccia pipes (e.g., Ardsheal Hill, near Kentallen, Scotland) lead to a violent explosive reaction with material entrained and blown up along the fracture. This appears to have been a largely barren episode, as in almost every case the mineralisation that accompanies these structures post-dates their emplacement. Some lodes preserve evidence that they originated as breccia lodes and were later reactivated during later phases of mineralisation, with large-scale replacement of original textures and materials. Rule (1865) notes the occurrence of rounded stones in some of the lodes at Wheal South Frances, near Carnkie and Phillips (1896) records a large block of killas found in the Main Lode of Dolcoath Mine, some 450 metres below the granite contact.
Lodes (or later assemblages in a pre-existing structure) emplaced in a lower-energy environment typically show a banded appearance, due to repeated opening of the lode fracture. Some of these lode sections may be mylonised by fault movements, while others appear to be open-space dilational infillings with vugs and druses of crystals (and occasionally, as at Dolcoath Mine, pockets of carbon dioxide in sealed vugs). They may show a range of assemblages of various temperature/pressure characteristics, or may have been crack-sealed in a single mineralising event.
A second, later, sequence of lodes occur in many districts, which cut across and displace the earlier lodes. These later caunter lodes (Collins, 1912; Dines, 1956) generally carry a lower temperature, mesothermal, assemblage (dominated by copper mineralisation) and strike E-W (in the Camborne-Redruth District). Rotation of the stress field saw these lodes emplaced in a fracture set offset from the dominant lode trend by ~30°. Farmer (1991) saw their origin in terms of the opening of Riedel D shears during conditions of dextral shear; field evidence from South Crofty Mine and other locations around the northern margins of Carnmenellis show that while the main-stage lodes are typically associated with dip-slip or oblique dip-slip movements, caunter-orientation structures are associated with horizontal to sub-horizontal slickenlines formed by shear movements.
Some of the 050°-060° (dominant lode trend) lodes were also reactivated during the deposition of the caunter lodes. At South Crofty Mine segments of caunter lode orientation were opened up within existing lode systems (e.g. Roskear D Lode) and infilled with an assemblage dominated by low temperature quartz, earthy chlorite, haematite, kaolinite, fluorite and chalcedony (LeBoutillier, 1996). These 'caunter jogs' were economically barren and were left as pillar areas along the strike of the lode.
Irregularly shaped sub-horizontal or pip-like replacement bodies sometimes occur at the junction of two, or more, lode structures. These are called carbonas and commonly consist of a fine network of veins, in altered granite, and bunches of ore. They reach their greatest development in the Lands End Granite and the Great Carbona (Collins, 1912; Dines, 1956; Noall, 1993) of St Ives Consols is arguably the most famous. This body (10-20 metres thick by 230 metres long, dipping at 20°) carried cassiterite, copper sulphides and fluorite (a mineral not found in the lodes of this mine) in tourmalinised/chloritised/sericitised granite. The average grade of the ore was 1.5% Sn, which compares very well with the average grade in the lodes; most Cornish mines operated at grades of between 0.70% and 2%; at South Crofty Mine the average R.O.M grade was 1.5% (Owen and LeBoutillier, 1998), but varied up to 2.5%, while individual lode grades over short strike lengths could reach as high as 40% Sn.
The main-stage lodes throughout the province are accompanied by wallrock alteration of varying type and intensity (Scrivener, 1982). The most common types of alteration are tourmalinisation (the progressive replacement of chlorite, micas and feldspars by tourmaline, which may lead to the development of a quartz-tourmaline rock; in pelites replacement of phyllosilicates may be extensive and pervasive), chloritisation (replacement of micas and feldspar by chlorite, due to the influx of Fe and Mg in solution), haematisation (due to the breakdown of existing chlorite, although some textures appear due to primary replacement of feldspars and micas) and sericitisation (the replacement of feldspars by white mica). At South Crofty Mine wallrock alteration haloes could extend in excess of 3 metres from the lode in either direction and sometimes showed overprinting of one type of alteration (particularly haematisation after chloritisation) on another. Such alteration was often barren, but it was not uncommon for Sn grades in the wallrocks (these metasomatic haloes mirror the mobility of Sn, seen in the country rocks around the granite) to exceed that in the lodes and the alteration zone was often included in stoping patterns and formed a significant part of the material extracted. The impregnation of granite with cassiterite is not uncommon in the Camborne-Redruth District; the Great Flat Lode, south of Carn Brea, consisted primarily of tourmalinised/chloritised granite (known to the miners as 'capel') carrying cassiterite over widths of up to 5 metres (Foster, 1878) around a narrow quartz leader vein.
The dating of the main-stage lodes has seen a revolution over the past decade. Halliday (1980) obtained dates for main stage Sn-bearing lodes of between 279±4 and 269±4 Ma. These dates were obtained from muscovite and orthoclase, using Rb-Sr methods, associated with various lodes across the Lands End and Tregonning-Godolphin granites. Using a previously published age of 295-300 Ma for the emplacement of the granite batholith, he envisaged a 20 million year hiatus between the emplacement of the granite and the onset of mineralisation. Between these two events, and intimately associated with the mineralisation, he placed the intrusion of the elvan dykes. This model became refined by later workers (Jackson et al., 1982; Thorne and Edwards, 1985; Bromley and Holl, 1986; Bromley, 1989) with a 'second magmatic event' (the emplacement of the elvan dykes) some 20 million years after granite emplacement, followed by main-stage mineralisation over a protracted period. This model was reinforced by Chesley et al. (1991), who obtained Nd-Sm dates of 259±7 and 266±3 Ma for fluorites from South Crofty Mine and Wheal Jane respectively (though their samples were not paragenetically constrained, and came from a late stage during mineralisation).
This model was radically altered by Chen et al. (1993), who, in dating the granites and mineralisation, were able to show that each pluton of the Cornubian Batholith had its own discrete history of magmatism and mineralisation (a view supported by Clark et al., 1993 and Chesley et al., 1993) They were able to show that the batholith was made up of discrete bodies, intruded between 293-274 Ma, and that mineralisation was diachronous across the orefield, instead of being related to a series of pan-province events; mineralisation and magmatism also overlapped, with mineralisation related to one magmatic pulse occurring (e.g. Carnmenellis) prior to later renewed granite magmatism.
Clark et al. (1993) give dates of 286 Ma for main stage lodes at South Crofty Mine, 272±4 Ma for the lodes of the St Just area and 278±6 Ma for the Sn lodes of central Dartmoor. This data indicates that mineralisation started in the Carnmenellis area some 3 million years after the emplacement of the early granites (10 million years before the emplacement of the fine-grained Boswyn granite in northern Carnmenellis) and was almost complete before the intrusion of the oldest (Zennor Lobe) of the Lands End granites at 274 Ma. The emplacement of the elvan dykes now appears to be related to later pulses of granite being tapped by extensional fractures (rather than the radiogenically-driven remelting of a single large intrusion), some of which were later utilised by ascending hydrothermal fluids.
Back to top of Page.
A distinctive set of fractures which are orientated at roughly 90° to the main-stage lodes are known by a variety of local names (fluccans, trawns, guides), but are commonly referred to as crosscourses. In mining circles only mineralised structures (usually chalcedony with quartz and minor haematite, etc) are referred to as crosscourses (fluccans are gouge-filled structures), while those carrying economic mineralisation are called lodes. They played an important part in mining as they represented 'easy ground' to work in and were selectively mined (being considerably easier to mine than granite or hornfelsed killas) as crosscuts into the workings on main-stage lode structures. The lodes (which have produced a wide range of metals, including Pb, Zn, Ag, Ba, Sb, Co, Ni, Fe, Mn, Bi, and U) belonging to this group are patchy in their distribution (only occasional structures of this type occur in the Camborne-Redruth District, e.g. Cobalt Lode of Pedn an Drea Mine at Redruth), the most important areas being the Menheniot and Tamar Valley Pb-Zn districts.
The development of the crosscourses was controlled by pre-granite tension joints and wrench faults (many of these fractures appear to have a movement history that is pre-, syn- and post-granite emplacement, only becoming mineralised during the final stages of the development of the orefield), oriented NNW-SSE to N-S. In the St Just area the lodes trend NW-SE and the crosscourses run N-S to NE-SW (Mount, 1985). Crosscourses typically reach a few metres in width (but may range from ~1 cm to >100 metres) and often have dextral throws from a few metres to tens of metres.
At South Crofty Mine a large number of chalcedony-filled crosscourses (Dominy et al., 1993,1994a, 1994b) faulted the main-stage lodes (Plate 7). These were typically sub-vertical structures with a banded appearance, (due to repeated infilling over a protracted period) typically 0.5 metres in width (but commonly ranging from <1 cm to ~1 metre) and infilled with chalcedony, quartz and occasional fluorite, siderite, earthy chlorite and soft haematite; kaolinite and bitumen were rare accessories. They ranged from compact to open vuggy structures (the largest on 315 fm Pryce's Lode drive measured some 4 metres high by 2 metres (maximum) wide) with quartz/ fluorite druses infilling vugs. Wallrock alteration was typically absent or confined to minor kaolinisation, with the exception of the Great Crosscourse which was heavily kaolinised. Movements were sometimes on the order of a few metres, but were generally <3 metres and often a few centimetres; both dextral and sinistral displacements were recorded with horizontal or sub-horizontal slickenlines.
Plate 7. A sub-vertical crosscourse; 290 fathom level, Main Lode drive, South Crofty Mine. The crosscourse is infilled with banded chalcedony, with minor haematite and earth chlorite, and shows evidence of repeated movements and dilation. Though predominantly wrench-style faults, the offset quartz floor is evidence of a vertical displacement component in the movement (metre rule for scale).
In economic term the Pb-Zn ± F, Ba deposits of the Tamar Valley and Menheniot areas are the most important. These lodes carry galena, sphalerite, fluorite and quartz with minor barite, calcite and siderite; occasional Cu and Fe sulphides also occur. The deposits are often small and shallow (with the exception of the Pb-Zn-F lode at Wheal Mary Ann [SX294635] at Menheniot, which was mined to 600 metres from surface), though occasionally very rich (Dines, 1956). The reactivation of the crosscourse fracture system was due to the onset of E-W crustal extension after the deposition of the main-stage lodes (Scrivener and Shepherd, 1998). The fracture system intersected the sedimentary rift basins to the north and south of Cornubia and fluids from these basins (Gleeson et al., 2000), charged with Pb, Zn, U, etc penetrated into the massif. Rb-Sr isotopic dating of fluorites from the Tamar Valley (Scrivener et al., 1994) has demonstrated a mid-late Triassic age of 236 Ma; this date effectively signals the end of mineralisation in the province, with only minor hydrothermal remobilisation of U, etc taking place afterwards in the late Mesozoic and Tertiary.
Back to top of Page.
The Nature of the Mineralising Fluids.
The transport and deposition of tin and other metals is a complex interplay between the chemistry of the host fluids, the nature of the wallrocks through which the fluids pass and the fluids (magmatic, meteoric, basinal and connate) encountered during transport. The nature of the mineralising fluids, which were essentially high-salinity brines, has largely been deduced by the study of fluid inclusions. In minerals from ore-bearing veins these are typically two-phase liquid-vapour varieties (Shepherd and Rankin, 1998) which, when heated, homogenise into the liquid phase. The homogenisation values (Th) for all types of mineralisation show a range from 150oC to 500oC, with Sn-W assemblages in the range 250oC - 450oC; Cu-Fe-As Sn assemblages between 200oC - 350oC and Pb-Zn assemblages 150oC - 250oC (Alderton, 1993; Haapala and Kinnunen, 1982; Bradshaw and Stoyel, 1968). Salinities are also variable (Rankin and Alderton, 1983, 1985; Alderton, 1993; Jackson et al., 1989; Bromley and Holl, 1986) and show considerable overlap in salinity ranges (2.5 - 35 wt% NaCl) for Sn-W deposits, while Cu-Fe-As-Zn deposits show a much more restricted range (5 -18 wt% NaCl). Na:K ratios are highly variable (between 1 and 18) with fluids associated with Sn mineralisation having values in the range 1-5, while those associated with sulphide mineralisation have higher values (>4). The composition of some of these fluids can identify individual plutons; Carnmenellis and Bodmin Moor share K/Na characteristics which separate them from the other granites.
Early W-As Sn deposits were formed by magmatic fluids (Manning, 1985) with some likely contributions from metamorphic fluids (due to thermal dehydration of the aureole rocks) mobilised during granite emplacement (Wilkinson, 1990; Primmer, 1985b); however, over time, meteoric fluids brought into the system by convection became important. The mixing of these cooler, less saline fluids with the magmatic fluids can be traced through a number of deposits (Shepherd et al., 1985; Scrivener et al., 1986; Polya, 1989) and these fluids not only supplied scavenged metals to the system, but also sulphur (from a diagenetic pyrite source, distinct from the magmatic sulphur source in the granitic fluids), utilised in the deposition of sulphide assemblages (Jackson et al., 1989).
Various workers have assigned almost all Sn-W and later mineralisation to meteoric-dominated fluids (Jackson et al., 1982; Bromley and Holl, 1986; Scrivener et al., 1986) based on O-H isotopic studies, but there is some doubt whether the data for magmatic fluids (based on only a few samples - Sheppard, 1977) has been properly constrained. It is likely that magmatic fluids were far more important than previously realised (LeBoutillier et al., 2000a, 2000b, 2001, 2002).
Mid - Late Triassic changes (dated at 236 Ma, Tamar Valley; Shepherd and Scrivener, 1987; Scrivener et al., 1994) in the stress regime resulted in the development of N-S trending faults (and reactivation of existing structures) that became the host to Pb-Zn base metal mineralisation and barren crosscourse mineralisation. The fluids associated with this phase have very different signatures to those responsible for earlier phases of mineralisation. They have a strong CaCl2 component (NaCl:CaCl2 wt% ratio of 1.2:1; 11-15 wt% NaCl; 9-13 wt% CaCl2) and Th of 110oC-170oC (Shepherd and Scrivener, 1987), they have many features in common with oilfield brines (Scrivener and Shepherd, 1998; Gleeson et al., 2000) and are thought to have originated in offshore sedimentary basins to the north and south of the Cornubian Massif (Bristol Channel Basin, Plymouth Bay Basin). The fluids are envisaged to have moved into the orefield along NNW-SSE and N-S trending wrench faults, intersecting both basins and fracture system, by tectonically controlled 'seismic pumping' (Bromley, 1989; Alderton, 1993). These fluids carried Pb, Zn, Ag, U, several minor metals (Co, Ni, etc - though these may have been scavenged from local basic intrusives during hydrothermal mixing and circulation of the basin brines and meteoric waters) and elateritic hydrocarbons, that were emplaced in N-S (and NNW-SSE) trending lodes, particularly in the Tamar Valley, Menheniot and Newquay areas (Dines, 1956). There appears to have been late-stage reactivation of the main-stage lodes around this time, across the province, evidenced by the occurrence of hydrocarbons in a number of mines across the peninsula. At South Crofty Mine these elateric hydrocarbons (associated with a marcasite-quartz-chalcedony-fluorite-haematite-kaolinite assemblage) were fairly common in intralode fracture zones and were found to be chemically identical to kerogen samples recovered from the Plymouth Bay Basin (South Crofty, unpublished data), confirming the link with offshore sedimentary basins.
Lower temperature (<100oC) high salinity hydrothermal fluids also played a major role in the kaolinisation process (Psyrillos et al., 1998; Manning et al., 1996) that generated the huge Cornubian china clay reserves.
The formation of the ore-forming fluids is dependent on a number of factors; firstly there has to be a mechanism that will allow the effective concentration of incompatible elements (Sn, W, Cu, As, B, Li, Pb, and Zn) in solution. Above the solidus this is usually accomplished by immiscibility (providing that water solubility is reduced, e.g. by the loss of boron in the magma as tourmaline crystallises; London and Manning, 1995) and the movement of volatiles to apical positions in the roof of the pluton (Jackson et al., 1989). Under subsolidus conditions this may be achieved by communication between miarolitic cavities (Candela and Blevin, 1995) in granites and pegmatites.
The fluids themselves are often brine/vapour mixtures (Haapala and Kinnunen, 1982; Heinrich et al., 1992) that, under the right conditions (e.g. low magmatic Fe content), can effectively partition the incompatible elements to such an extent that they are almost non-existent in the magmatic host rocks (Lehmann and Harmanto, 1990). The presence of a vapour phase (often dominated by CO2 and H2S; Lowenstern, 2001) has been shown to be of major importance in the separation of phases in the hydrothermal system, particularly copper (Candela, 1989b; Candela and Piccoli, 1998), which may be present in amounts ten times greater than in the accompanying brine (Heinrich et al., 1992). Oxygen and hydrogen fugacity is also known to play a major role in phase separation, with low fO2 (and high CO2) favouring the transport of W (Candela and Bouton, 1990) and high fH2 favouring Sn transport (Taylor, 1979; Taylor and Wall, 1993; Heinrich, 1990).
The fluids themselves are brines and the transport of metals is thought to be in complex form with chlorine (Taylor and Wall, 1993). Boron and fluorine may also contribute to metal complexing compounds, though they are relatively unimportant in comparison to chlorine (as chloride ions) and phosphorous (as phosphate ions), as they are far less effective in partitioning metals into aqueous fluids (Manning, 1984).
Chloride and hydroxychloride complexes are the most efficient transporters of tin as SnCl2, SnOHCl and Sn(OH)2Cl2 (Wood and Samson, 1998; Heinrich, 1990) in Sn (II) form. These complexes are able to transport tin in concentrations of many thousands of ppm at high temperatures (>400°C) and high pH. In hypersaline conditions species such as KSnCl3, K3SnCl5 and Na2SnCl4 become important (Heinrich, 1990) and fluids derived from highly fractionated K-rich granite are known to be capable of carrying elevated concentrations of Sn. The deposition of tin from these fluids takes place in a number of environments. Decreases in fluid acidity exert a major control on cassiterite deposition. The neutralisation of fluids by hydrolysis of feldspars (to muscovite + quartz ± biotite ± chlorite ± Fe sulphides) during greisenisation, or reaction with granitic wallrocks, leads to the deposition of cassiterite, but typically as a disseminated deposit. The ability to deposit further cassiterite is determined by the availability of feldspar and the area of the rock/fluid interface, which may be quite small. This kind of reaction, therefore, does not often create economic deposits, unless the process is repeated on a series of 'reaction fronts' advancing into the wallrocks (Heinrich, 1990). A more effective scenario would involve the formation of a hydrogen-rich vapour phase, the corresponding rise in oxidation state and decrease in fH2 could lead to cassiterite deposition (providing the fluids are in equilibrium with the host rocks) of the order of ten times greater than by hydrolysis alone (Heinrich, 1990).
Above 400°C and 1-2 kbar, Sn deposition is not greatly affected by changes in temperature or pressure (Taylor and Wall, 1993). It is far more sensitive to changes in fluid acidity and alkali chloride concentrations. The most effective method of Sn extraction would be to mix the primary magmatic fluids with cooler, low-salinity, meteoric fluids, particularly if the meteoric fluids are near neutral acidity and contain CO2 and HCO3- (Heinrich, 1990). Under these conditions massive cassiterite deposition would occur, limited only by the available space and concurrent crystallisation of gangue phases.
Tungsten may also be transported as a chloride complex (Manning, 1984), but there is increasing evidence to suggest that polytungstate ions (e.g. HWO4-, WO42-, H7(WO4)65-) are the dominant species in hydrothermal fluids (Wood and Samson, 1998). Tungsten also readily forms silico tungsten acids (e.g. H8Si(W2O7)6, which would precipitate wolframite in low Ca environments) under the right conditions (Horsnail, 1979). Unlike cassiterite, wolframite deposition appears unaffected by acidity and oxygen/hydrogen fugacity and instead (depending on the supply of available iron) appears largely temperature dependent (Heinrich, 1990).
A range of other metals including Zn, Pb, Sb, Fe and Cu may also form chloride and hydroxide complexes, but studies have shown that sulphide and bisulphide complexes are likely to be more important in the transport of these metals, particularly under conditions of high sulphur concentrations (Wood and Samson, 1998). As is also transported as hydroxide (low S concentrations) and bisulphide (high S concentrations) complexes, but rarely (if at all) as a chloride complex (Wood and Samson, 1998). Sulphur is usually present in hydrothermal fluids as hydrogen sulphide, H2S, and reacts with the dissolved metal ions to form a range of sulphide complexes, e.g. Pb(HS)3-, Cu(HS)2- and ZnS(HS)-; dependent on pH conditions. With decreases in temperature and changes in pH and sulphur concentrations these complexes break down to precipitate sulphide minerals and release H2S. Transport via sulphide complexes requires high concentrations of sulphur to be present in the hydrothermal brines; previously the source of this sulphur was assumed to be primarily from the country rocks via mixing with meteoric fluids (Jackson et al., 1989), but there is an increasing body of evidence to show that granites can carry significant volumes of magmatic sulphur that contribute directly to hydrothermal systems (Kontak, 1990). Inherited sulphur (and metals) from xenolith assimilation would also have made a contribution to the total sulphur budget.
The ore-bearing fluids responsible for the mineralisation across the Cornubian Orefield were the product of a series of events involving the mixing and recycling of fluids from a variety of sources: magmatic, metamorphic, meteoric, connate and basinal. The high heat production (due to radiogenic decay of primary U and Th) of the granites (Willis-Richards and Jackson, 1989; Jackson et al., 1989; Lucas and Willis-Richards, 1998) generated a convective system around the granite plutons that scavenged metals from the country rocks and redeposited them in the lode systems; in addition to metals supplied directly from fluids of magmatic departure. As the convective system began to migrate deeper and closer to the plutons (as overall temperatures began to fall), the fluids were responsible for remobilising some of the mineralisation (particularly copper), giving rise to complex overprinted paragenesis in some areas (Dines, 1956; Seccombe and Barnes, 1990).
Circulating thermal brines are still encountered today and have been analysed at a number of localities, including the Rosemanowes [SW736346] borehole and South Crofty Mine. Alderton and Shepherd (1977) analysed a number of springs, finding a range of temperatures from 16 - 52oC, with salinities up to 1.5% (much lower than that recorded from the main stage of mineralisation, where salinities over 20% are frequent) which is half that of modern sea water. The Br/Cl ratios are comparable to seawater and they are typically of neutral pH. As well as Ca, Na and Cl, they carry substantial amounts of Li, Mg, K and HCO3- and are of meteoric origin. There has been interest in the alkali metal content of these brines in the past and they have received some attention due to their lithium content (Beer et al., 1978) which may reach up to 118 ppm in some springs.
The reason for the persistence of thermal springs so long after the magmatic event is due to the high heat flow generated by radiogenic decay within the granite, which in Carnmenellis runs at 3.9 x 10-3 Wm-3 (Burgess et al., 1982; Edmunds et al., 1984). Geothermal gradients vary from 29.8oC/km in the Carnmenellis Granite, 20 - 50oC/km in the surrounding sediments close to the contact, 39oC/km in South Crofty Mine and 45oC/km at Wheal Jane. Current heat flow values for the various plutons are broadly similar with Land's End and St Austell granites running some 15% higher than the other plutons at 127mWm-2.
Back to top of Page.
Forward to Next Page
Back to Introduction
This page last updated 16/04/2006