The Geology of Parys Mountain

The Parys Mountain site is protected by legislation. Damage can result in fines or imprisonment.

A field work guide for geologists has been issued by the Geologist association.

Please follow the guide and do not use hammers in the SSSI or SAM

The mineralisation of Mynydd Parys extends roughly 3 km NNE-SSW in a band 1km wide. It is associated with an ancient volcanic event (late Ordovician ca 480 my BP). Involving the extrusion of mostly silica – rhyolites/dacite lavas and the ejection of ashes.

These deposits grade laterally into the shallow water volcanic sediments which include silceous sinter and cherts and also host intrusive rhyolites and later dolerites. This volcanic sequence overlies the Parys shales and is in turn overlain by later silurian shales. These beds appear to have been compressed into a steep trough shaped structure trending NE-SW and tilted over to the SE. The axis being exposed at the ends of the Great opencast.  The region is also traversed by the steep NNW-SSE cross faults and to the north there are older Precambrian schists of the Mona complex,brought up by the Carmel head and Corwas thrust faults.

The primary mineralisation comprised pyrite (FeS2) which can be seen in the slump structures in the exposures at the centre of the Great open cast,indicating formation on the sea floor. This was followed by a phase dominated by Chalcopyrite (CuFeS2) and then by one dominated by the intimate mixture of sphalerite (ZnS) and Galena (PbS) with only minor chalcopyrite.  This is known as “Bluestone”

These are believed to have formed from exhalations on the sea floor analogous to the black smokers seen in other ocean today. The ore deposit is thus thought to be of the “Kuroko” type and as such is unique in Britain,

A secondary phase remobilisation occurred during the later Caledonian metamorphism and has been dated to be 360 my BP.  As a result of this unique geology several Sites of Special Scientific Interest (SSSI ) have been established on the mountain.

Weathering products are dominated by colourful red/yellow Fe3+ hydrous oxides but also include a diverse range of sulphate materials. Amongst these are abundant jarosite (KFe(SO4)(OH) and anglesite (PbSO4) for which Mynydd Parys is the world type locality.

The minerals, Pisanite ( [(Fe2+,Cu)SO4.7H2O] , Antlerite [Cu3(SO4)(OH)4]  Basaluminite  [Al4(SO4)(OH)10.5H2O] and Anglesite a lead sulphate are also present and are generally rare elsewhere in the UK.

The present weathering in intense, due to the very high acidity generated by the oxidation of pyrite and other sulphide minerals to form sulphuric acid. Some pools have been recorded as having a pH of as low as 2.

This extreme chemistry has resulted in a remarkable flora and fauna. Eight sits have been established as SSSI on the basis of their unusual lichen communities and there are also unique liverworts and mosses.


Anglesey Mining PLC

These extracts are taken from a field guide prepared for a visit on 21 April 98 to Parys Mountain as part of the Symposium on Magmatism and Mineralization in Arcs and Ocean Basins Geoscience 98 Biennial Conference at Keele University, U.K.


The Parys Mountain deposit is located about 3 kilometres from the north coast of Anglesey, in northwestern Wales, within a sequence of Ordovician? to Silurian volcanic rocks and shales (Figure 1). Basement rocks include the Monian Supergroup of late Precambrian age. The deposit was an historic source of copper in Britain over a century (1768-1904), although most production took place during the first 50 years of this period. The ore was recovered from a series of open pits (Figure 2), and also underground workings which extended below the open pits to a maximum of 150 metres below the surface. The originally mined deposits were viewed as vein systems or lodes within Ordovician felsic volcanic and shaley rocks (Greenly, 1919).

No major mining or exploration operations were carried out from 1904 to 1961, when the first drilling programs were begun. From 1961 to 1990, drilling to depths well below the old pits and workings, and in particular to their west, was carried out mainly by CIGOL, Parys Mountain Mines, Intermine, Cominco, Imperial Metals and Anglesey Mining, and an geological picture of the deeper volcanic and sedimentary sequence began to emerge (Hawkins, 1966; Westhead, 1991; Tyler and Charter, 1997). It was during this period of exploration that stratiform lenses of Cu-Zn-Pb-rich massive sulfide mineralization were discovered in the subsurface to the west of the open pits, and also down dip to the north, near the lower contact between a thick sequence of felsic volcanic rocks and underlying shales (e.g. Figures 3 and 4). These zones of sulfide mineralization have been termed the Chapel, Engine, Deep Engine, South Central and North Central Zones.

Various aspects of the massive sulfide deposits and the stratigraphy and structure of their host rocks have been studied over the last two decades (Thanusuthipitak, 1974; Pointon, 1979; Pointon and Ixer, 1980; Southwood, 1982, 1984; Westhead, 1993). During this period, it was generally accepted that the disposition of the main geological units on the property was the result of folding of an initial sequence of mudstone overlain by rhyolite and then shales, into a large anticline-syncline structure with an east-west strike and a moderate dip to the north, with the north limb overturned to the south. The occurrence of Silurian shales flanked by volcanic rocks which were assumed to be Ordovican, based on limited graptolite evidence, was taken as providing support for a major synclinal axis (Westhead, 1991). The age of the Silurian shales, which are exposed in the open pits, has been determined from graptolite studies (Greenly, 1919).

From 1985 to 1990, Anglesey Mining plc carried out further drilling (22 holes totalling over 8000 m) in the western part of the property, and sunk the Morris shaft to 300 m depth in the Engine Zone. Underground drilling in 1990 from the 280 m level intersected several zones of massive sulfides near the lower contact between a thick sequence of felsic volcanic rocks and underlying shales of presumed Ordovician age (e.g. Figures 3 and 4). These are known as the Engine and Chapel Zones (first discovered by Cominco in the 1977-81 period). The White Rock Zone, which is dominantly silica with occurrences of semi-massive sulfides, also is present in this area but its stratigraphic relations with the other Zones are uncertain due to underground faulting. In 1990, minable reserves in the Engine and Chapel Zones were estimated as 1.41 Mt grading 1.99% Cu, 3.42% Pb, 6.65% Zn, 99 g/t Ag and 0.79 g/t Au; with reserves in the White Rock Zone of 0.84 Mt grading 0.49% Cu, 3.43% Pb, 6.72% Zn, 78 g/t Ag and 0.66 g/t Au (Charter, 1995). These form part of an overall estimated geological reserve of 6.45 Mt at similar grades.

In 1995, Anglesey Mining initiated a new phase of work involving relogging of available core, remapping of surface geology, and a major lithogeochemical program. This work has led to a revision in our understanding of the overall stratigraphic and structural setting of the massive sulfide mineralization, and in particular, an abandonment of the idea of strong folding of the sequence, with an emphasis instead on the importance of volcanic facies relationships and their geometry and stratigraphic contacts in the subsurface (Barrett, 1995; Tyler and Charter, 1997; Barrett and MacLean, 1997; Tennant, PhD thesis in progress). These studies support the idea that the distribution of mineralization is determined mainly by its relation to various felsic eruptive centres along a more or less homoclinal sequence which dips to the north. Different volcanic units can be defined and traced, even where altered, using recently developed lithogeochemical methods.



In order to identify different felsic and mafic rock types, particularly in altered sequences, binary plots of immobile element pairs are commonly used, as outlined by MacLean and Kranidiotis (1987), MacLean (1990), and Barrett and MacLean (1991; 1994a,b). The most useful plots involve combinations of the elements Al, Ti and Zr, but Y-Nb-Th-Yb are also helpful. On the Y-axis, a compatible element such as Al2O3 or TiO2 is plotted, whereas on the X-axis an incompatible element such as Zr is used (as a monitor of fractionation). In such plots, primary fractionation trends can be defined using a suite of least altered samples that covers the compositional range involved in the particular volcanic terrane under study. In terranes where several homogeneous, but distinct precursor lithologies are present, rather than a continuous fractionation series, altered samples can be treated by relating them back along one of several alteration lines to the appropriate least-altered precursor rock type. This is the case at Parys Mountain, where several distinct rhyolite types have been identified.

The dispersion of sample points along a given alteration line is due to the effects of gain or loss of mobile elements. Net mass gain (e.g. quartz or carbonate or pyrite addition) causes dilution in the concentration of immobile elements relative to their initial values in a rock, whereas net mass loss (which typically accompanies strong sericitization or chloritization) causes residual concentration of immobile elements. Each altered rock type will have its own alteration line; these lines are generally distinct (without overlap) is most of the immobile element plots. Altered rocks accordingly can be identified using simple immobile element plots. This allows even very altered volcanic units to be readily identified, which in turn greatly improves stratigraphic correlations between drill holes.


The new Parys Mountain lithogeochemical data set currently comprises about 400 drill-core samples and 100 surface samples. Al2O3-Zr relations are shown in Figure 5a, and TiO2-Zr relations in Figure 5b, although for simplicity, only rhyolites A, B and C-1 are shown. Based on these and related plots, five distinct rhyolite units can be identified on the property; these are termed rhyolites A, B, C-1, C-2, and D. In addition, there are two volumetrically very minor mafic groups, probably thin sills, the main occurrence of which is a thin body near the lower contact of the southern rhyolites. Rhyolites A and C-1 are volumetrically dominant in the eastern and central portions of the property, with rhyolite D more aboundant in the eastern portion. Rhyolites B and C-2 occur close to the contact between hangingwall rhyolites and footwall shales in some areas.

Plots of one incompatible and immobile trace element (e.g. Zr, Nb, Th, Y, Yb) against another have also been used to help subdivide the rhyolite and mafic types. In such plots, a linear fractionation trend extending away from the origin is expected for magmas derived from a common source but linked through processes such as fractionation. Alteration effects will move samples along lines which are parallel to, and superimposed upon, any primary fractionation trend. In a plot of Nb versus Zr (not shown), rhyolites A, C-1, C-2 and D lie on essentially the same trend, whereas rhyolite B defines a separate trend (with a higher Zr/Nb ratio). The dispersion of samples along each of these two main trends is due mainly to alteration, although the different slope for rhyolite B reflects a different magma source. Basalts (not shown) plot close to the rhyolite B trend, with Zr contents of 150-200 ppm. A plot of Y versus Zr (not shown) indicates that most rhyolite A samples are of tholeiitic affinity (Zr/Y = 2.0-4.5), whereas the other rhyolite types are mainly of transitional affinity.

Chondrite-normalized REE plots are shown in Figure 6 for representative samples of the main rhyolite and mafic rock types (the rhyolite C samples are C-1 type). The REE patterns also discriminate well between the different rhyolite groups, with the slopes of the patterns (Lan/Ybn) reflecting the differences noted above in other immobile element ratios such as Zr/Y. Samples within a given group retain near-parallel REE patterns, although they show ëapparentí variations in their absolute REE contents. It should be noted, though, that much of the vertical variation within a given group is the result of mass change effects in mobile elements, rather than primary magmatic variations in REE contents.

Downhole Lithogeochemical Variations

Downhole variations in lithological type can be effectively monitored by plotting ratios such as Al2O3/TiO2 or Zr/TiO2, and variations in magmatic affinity by using ratios such as Zr/Nb and Zr/Y. In most VMS hydrothermal systems, these ratios are generally insensitive to alteration, although local Y mobility has been noted in areas of extreme chloritic alteration. Where a change in primary lithology occurs, immobile element ratios such as Al2O3/TiO2 and Zr/TiO2 should change to a new value (although not necessarily by the same absolute factor). At Parys Mountain, Al2O3/TiO2 ratios are generally >50 in rhyolites, about 15-25 in mudstones, and 5-10 in basalts. The corresponding ranges for the Zr/TiO2 ratio are about 800-4000 in rhyolites, 100-200 in mudstones, and 50-100 in mafic rocks.

In hole CZ-4, in the Chapel Zone, footwall mudstones are overlain by rhyolite C, followed by a thin mud interval, and then rhyolite A (Figure 7). In this figure, the Al2O3/TiO2 ratio is shown on the lower X-axis, and the Zr/TiO2 ratio, on the upper X-axis. Within each of the main rhyolite units in hole CZ-4, and in all other holes as well, these ratios remain nearly constant, which indicates that each eruptive unit had a characteristic and near-homogeneous composition from base to top, which in turn forms the basis for defining chemostratigraphic units. It is also worth emphasizing that within a given chemostratigraphic unit, the immobile element ratios remain closely parallel despite wide ranges in downhole alteration effects.

Alteration Geochemistry

Relative to the main southern rhyolite-shale contact on the western part of the property, the intensity of alteration generally decreases stratigraphically downward (in the shales), and also decreases stratigraphically upward (in the rhyolites). The main alteration minerals are sericite, chlorite, quartz and pyrite. In areas of extreme alteration, as noted earlier, it is not possible to visually identify the original rock types (e.g. rhyolite versus mudstone), let alone distinguish between rhyolite types, although this can be done effectively using the immobile element methods discussed above.

Alteration in the rhyolites includes significant additions of silica in some samples, but losses in others. Samples with silica loss show variable enrichments in Fe, Mg and K, reflecting chloritization and sericitization. Only a few rhyolites are chemically relatively unaltered, and even these have possibly been affected by some alkali exchange, making it difficult to estimate the precursor composition of the rhyolites (this is one reason why subdivision of the felsic lithologies is risky without using immobile element ratios). Many rhyolites display strong to near-total Na depletion, even those far removed from obvious mineralized zones. This raises the possibility that some of the alkali alteration could have taken place prior to the development of mineralizing systems, e.g. during high-temperature emplacement of the rhyolites, or during a period of relatively low-temperature, low intensity interaction with seawater within the volcanic pile.

As a guide to general alteration effects, one can plot mobile components such as SiO2 and MgO against the Zr/TiO2 ratio (Fig. 8). The Zr/TiO2 ratio provides a monitor of primary lithological variation (as discussed above). Examination of Figure 8 indicates that rhyolite B has undergone a wide range of silica addition and depletion, and also moderate to extreme additions of Mg. Rhyolite C commonly shows some silica depletion and Mg addition. Some of the mudstones are extremely altered, with major additions of Mg or K (and silica loss). In order to assess the alteration on an absolute quantitative basis, however, it is necessary to calculate mass changes for each element (e.g. MacLean and Kranidiotis, 1987; MacLean and Barrett, 1993), using a separate precursor composition for each rhyolite type (and for the mudstone). In the next stage of the Parys Mountain lithogeochemical program, calculated mass changes will be used to establish the geometry of the hydrothermal systems, and to define alteration gradients (vectors) in both the rhyolites and mudstones which can help to guide exploration.

Geological Setting

The Engine and Chapel Zones, located in the western portion of this sequence, contain reserves plus resources of 3.09 Mt grading 2.1% Cu, 3.6% Pb, 6.8% Zn, 109 g/t Ag, 0.7 g/t Au. The sulfides occur at or near the top of a mudstone footwall, which is overlain by some 200 metres of rhyolites (massive to flow-banded to fiamme-bearing to autobrecciated to rubbly), with a few thin intercalations of mudstone or volcaniclastic sediments. In this area, a widespread upper interval of rhyolite A commonly overlies domes and lenses of rhyolite C and locally thin tongues of rhyolite B, which in turn overlie footwall mudstones. In places, however, rhyolite A rests almost directly on mudstone, or on mineralization. Rhyolite A in the Engine and Chapel Zones contains some orthoclase + albite phenocrysts, whereas Rhyolite C-1 is aphyric and in part medium-grained. The presence of some peperitic contacts with mudstone suggests that rhyolite C-1 was partly injected below the surface, but the occurrence of fiamme-bearing rhyolite C-1 elsewhere in this area suggests that it was extruded as well. Rhyolite B forms a minor volumetric type east of the Pen-y-mynydd fault, where it occurs as both massive units and altered volcaniclastic intervals near the level of mineralization. West of the Pen-y-mynydd fault, in an area which has been little explored, massive rhyolite B is more common. In places, thin basaltic units occur near the rhyolite-mudstone contact, and display peperitic contacts with the mudstone.

Within the Engine and Chapel Zones, abrupt lateral variations in the thickness of rhyolite C-1 over distances of only 50-100 metres probably reflect the presence of local domes of rhyolite C that gradually “inflated” via extrusions and partly subsurface injections of magma. Hydrothermal systems which were discharging through mudstones at the time of rhyolite C emplacement appear to have been largely suffocated, although some alteration and veining penetrated into the lower parts of the domes. Mudstone areas immediately lateral to rhyolite domes appear to be most favorable for hosting sulfide lenses, which in principle could occur either as in-situ accumulations or as sulfide-bearing debris shed by slumping from dome flanks. Eventually, widespread eruptions of rhyolite A buried the accumulations of sulfides, mudstones and volcaniclastic debris in the flanking lows, and also covered the domes of rhyolite C. Work in progress suggests that a center of rhyolite A eruption existed to the north of the rhyolite C domes.

One question which needs to be resolved is whether the rhyolite eruptive centres are related to regional faulting of the basement rocks. Northwestern Wales was the site of rifting of a continental margin in the lower Paleozoic (Kokelaar, 1988; Howells et al., 1991). Syn-mineralization faults are difficult to identify at Parys Mountain because the footwall generally comprises a monotonous shale sequence, while hangingwall rhyolites show lateral thickness variations related at least partly to volcanic constructional processes. With further work, trends in the subsurface disposition of rhyolite eruptive enters may become evident. Most rhyolites are subalkaline with low TiO2 contents and low Zr/Y ratios (2-4) but high Nb and relatively high REE contents, features typical of rifted continental margin rhyolites. Basalts are enriched relative to MORB, with high contents of Ti, Zr, Nb and light REE, and are of transitional-to-alkaline affinity. Together with regional geological data, the affinities of the volcanic rocks and their bimodal distribution suggest that the tectonic setting of the deposit was a rifted continental crustal margin.